downloaded as a PDF - Next Digital Decade

THE NEXT DIGITAL DECADE 

ESSAYS ON THE FUTURE OF THE INTERNET 

Edited by Berin Szoka & Adam Marcus

THE NEXT DIGITAL DECADE 

ESSAYS ON THE FUTURE OF THE INTERNET 

Edited by Berin Szoka & Adam Marcus

NextDigitalDecade.com 

TechFreedom 

techfreedom.org 

Washington, D.C. 

This work was published by TechFreedom (TechFreedom.org), a non-profit 

public policy think tank based in Washington, D.C. TechFreedom’s mission is 

to unleash the progress of technology that improves the human condition and 

expands individual capacity to choose. We gratefully acknowledge the generous 

and unconditional support for this project provided by VeriSign, Inc. 

More information about this book is available at NextDigitalDecade.com 

ISBN 978-1-4357-6786-7 

© 2010 by TechFreedom, Washington, D.C. 

This work is licensed under the Creative Commons Attribution- 

NonCommercial-ShareAlike 3.0 Unported License. To view a copy of this 

license, visit http://creativecommons.org/licenses/by-nc-sa/3.0/ or send 

a letter to Creative Commons, 171 Second Street, Suite 300, San Francisco, 

California, 94105, USA. 

Cover Designed by Jeff Fielding.

THE NEXT DIGITAL DECADE: ESSAYS ON THE FUTURE OF THE INTERNET 3 

TABLE OF CONTENTS 

Foreword 7 

Berin Szoka 

25 Years After .COM: Ten Questions 9 

Berin Szoka 

Contributors 29 

Part I: The Big Picture & New Frameworks 

CHAPTER 1: The Internet’s Impact on 

Culture & Society: Good or Bad? 49 

Why We Must Resist the Temptation of Web 2.0 51 

Andrew Keen 

The Case for Internet Optimism, Part 1: 

Saving the Net from Its Detractors 57 

Adam Thierer 

CHAPTER 2: Is the Generative Internet at Risk? 89 

Protecting the Internet Without Wrecking It: 

How to Meet the Security Threat 91 

Jonathan Zittrain 

A Portrait of the Internet as a Young Man 113 

Ann Bartow 


Saving the Net from Its Supporters 139 

Adam Thierer 

CHAPTER 3: Is Internet Exceptionalism Dead? 163 

The Third Wave of Internet Exceptionalism 165 

Eric Goldman 

A Declaration of the Dependence of Cyberspace 169 

Alex Kozinski and Josh Goldfoot 

Is Internet Exceptionalism Dead? 179 

Tim Wu

4 TABLE OF CONTENTS 

Section 230 of the CDA: 

Internet Exceptionalism as a Statutory Construct 189 

H. Brian Holland 

Internet Exceptionalism Revisited 209 

Mark MacCarthy 

CHAPTER 4: Has the Internet Fundamentally 

Changed Economics? 237 

Computer-Mediated Transactions 239 

Hal R. Varian 

Decentralization, Freedom to Operate & Human Sociality 257 

Yochai Benkler 

The Economics of Information: 

From Dismal Science to Strange Tales 273 

Larry Downes 

The Regulation of Reputational Information 293 

Eric Goldman 

CHAPTER 5: Who Will Govern the Net in 2020? 305 

Imagining the Future of Global Internet Governance 307 

Milton Mueller 

Democracy in Cyberspace: Self-Governing Netizens 

& a New, Global Form of Civic Virtue, Online 315 

David R. Johnson 

Who’s Who in Internet Politics: 

A Taxonomy of Information Technology Policy & Politics 327 

Robert D. Atkinson 

Part II: Issues & Applications 

CHAPTER 6: Should Online Intermediaries 

Be Required to Police More? 345 

Trusting (and Verifying) Online Intermediaries’ Policing 347 

Frank Pasquale 

Online Liability for Payment Systems 365 

Mark MacCarthy


Fuzzy Boundaries: The Potential Impact of Vague 

Secondary Liability Doctrines on Technology Innovation 393 

Paul Szynol 

CHAPTER 7: Is Search Now an “Essential Facility?” 399 

Dominant Search Engines: 

An Essential Cultural & Political Facility 401 


The Problem of Search Engines as Essential Facilities: 

An Economic & Legal Assessment 419 

Geoffrey A. Manne 

Some Skepticism About Search Neutrality 435 

James Grimmelmann 

Search Engine Bias & 

the Demise of Search Engine Utopianism 461 

Eric Goldman 

CHAPTER 8: What Future for Privacy? 475 

Privacy Protection in the Next Digital Decade: 

“Trading Up” or a “Race to the Bottom”? 477 

Michael Zimmer 

The Privacy Problem: What’s Wrong with Privacy? 483 

Stewart Baker 

A Market Approach to Privacy Policy 509 

Larry Downes 

CHAPTER 9: Can Speech Be Policed 

in a Borderless World? 529 

The Global Problem of State Censorship 

& the Need to Confront It 531 

John G. Palfrey, Jr. 

The Role of the Internet Community 

in Combating Hate Speech 547 

Christopher Wolf

6 TABLE OF CONTENTS 

CHAPTER 10: Will the Net Liberate the World? 555 

Can the Internet Liberate the World? 557 

Evgeny Morozov 

Internet Freedom: Beyond Circumvention 565 

Ethan Zuckerman

Foreword 


Berin Szoka 

This book is both a beginning and an end. Its publication marks the beginning 

of TechFreedom, a new non-profit think tank that will launch alongside this 

book in January 2011. Our mission is simple: to unleash the progress of 

technology that improves the human condition and expands individual capacity 

to choose. This book also marks an end, having been conceived while I was 

Director of the Center for Internet Freedom at The Progress & Freedom 

Foundation—before PFF ceased operations in October 2010, after seventeen 

years. 

Yet this book is just as much a continuation of the theme behind both PFF and 

TechFreedom: “progress as freedom.” As the historian Robert Nisbet so 

elegantly put it: “the condition as well as the ultimate purpose of progress is the 

greatest possible degree of freedom of the individual.” 1 This book’s twenty-six 

contributors explore this theme and its interaction with relentless technological 

change from a wide variety of perspectives. 

Personally, this book is the perfect synthesis of the themes and topics that set 

me down the path of studying Internet policy in the late 1990s, and weaves 

together most of the major books and authors that have influenced the 

evolution of my own thinking on cyberlaw and policy. I hope this collection of 

essays will offer students of the field the kind of authoritative survey that would 

have greatly accelerated my own studies. Even more, I hope this volume excites 

and inspires those who may someday produce similar scholarship of their 

own—perhaps to be collected in a similar volume celebrating another major 

Internet milestone. 

I am deeply grateful to Shane Tews, Vice President for Global Public Policy and 

Government Relations at VeriSign, who first suggested publishing this sort of a 

collection to commemorate the 25th anniversary of the first .COM domain 

name (registered in 1985) by asking what the future might bring for the 

Internet. Just as I hope readers of this book will be, she had been inspired by 

reading Who Rules the Net? Internet Governance & Jurisdiction, a collection of 

cyberlaw essays edited by Adam Thierer and Clyde Wayne Crews, and published 

by the Cato Institute in 2003. This book would not exist without the 

unconditional and generous support of VeriSign, the company that currently 

operates the .COM registry. 

1 ROBERT NISBET, HISTORY OF THE IDEA OF PROGRESS 215 (1980).

8 FOREWORD 

Nor would the book exist without the superb intellectual contributions and 

patience of our twenty-six authors, and all those who assisted them. I must also 

thank PFF Summer Fellows Alexis Zayas, Jeff Levy and Zach Brieg for their 

invaluable assistance with editing and organization, and Jeff Fielding for the 

book’s stunning cover artwork and design. 

Most of all, I must thank Adam Thierer and co-editor Adam Marcus. The two 

and a half years I spent working closely with them on a wide range of 

technology policy topics at PFF were the highlight of my career thus far. 

I look forward to helping, in some small way, to discover the uncertain future of 

progress, freedom, and technology in the next digital decade—and beyond. 

Berin Szoka 

December 16, 2010


25 Years After .COM: Ten Questions 

Berin Szoka 

While historians quibble over the Internet’s birth date, one date stands out as 

the day the Internet ceased being a niche for a limited number of universities, 

governments and military organizations, and began its transformation into a 

medium that would connect billions: On March 15, 1985, Symbolics, a 

Massachusetts computer company, registered symbolics.com, the Internet’s first 

commercial domain name. 2 This book celebrates that highly “symbolic” 

anniversary by looking not to the Internet’s past, but to its future. We have 

asked twenty-six thought leaders on Internet law, philosophy, policy and 

economics to consider what the next digital decade might bring for the Internet 

and digital policy. 

Our ten questions are all essentially variations on the theme at the heart of 

TechFreedom’s mission: Will the Internet, on its own, “improve the human 

condition and expand individual capacity to choose?” If not, what is required to 

assure that technological change does serve mankind? Do the benefits of 

government intervention outweigh the risks? Or will digital technology itself 

make digital markets work better? Indeed, what would “better” mean? Can 

“We the Netizens,” acting through the digital equivalent of what Alexis de 

Tocqueville called the “intermediate institutions” of “civic society,” discipline 

both the Internet’s corporate intermediaries (access providers, hosting 

providers, payment systems, social networking sites, search engines, and even 

the Domain Name System operators) and our governments? 

Part I focuses on five “Big Picture & New Frameworks” questions: 

1. Has the Internet been good for our culture and society? 

2. Is the open Internet at risk from the drive to build more secure, but less 

“generative” systems and devices? Will the Internet ultimately hinder 

innovation absent government intervention? 

3. Is the Internet really so exceptional after all, or will—and should—the 

Internet be regulated more like traditional communications media? 

4. To focus on one aspect of the Internet exceptionalism, has the Internet 

fundamentally changed economics? What benefits and risks does this 

change create? 

5. Who—and what ideas—will govern the Net in 2020—at the end of the 

next digital decade? 

2 John C Abell, Dot-Com Revolution Starts With a Whimper, WIRED MAGAZINE, March 15, 2010, 

http://www.wired.com/thisdayintech/2010/03/0315-symbolics-first-dotcom/

10 25 YEARS AFTER .COM: TEN QUESTIONS 

Part II tackles five “Issues & Applications” questions: 

6. Should intermediaries be required to police more—or be disciplined in how 

they police their networks, systems and services? Whether one thinks the 

Internet is truly exceptional, and whether it has changed economics largely 

determines one’s answer to these questions. 

7. While debates about the role of online intermediaries and the adequacy of 

their self-regulation focused on net neutrality in the last digital decade, the 

battle over “search neutrality” may be just as heated in the next digital 

decade. Are search engines now the “essential facilities” of the speech 

industry that can be tamed only by regulation? Or are they engines of 

empowerment that will address the very concerns they raise by ongoing 

innovation? 

8. As the Internet accelerates the flow of information, what future is there for 

privacy, both from governments and private companies? Is privacy a right? 

How should it be protected—from both government and private 

companies? 

9. The book concludes with two Chapters regarding the Internet in a 

borderless world. The first focuses on governments’ regulation of speech. 

10. The second focuses on the potential for governments’ “disruption” by 

speech—by unfettered communication and collaboration among the 

citizenry. In both cases, our authors explore the consequences—and 

limits—of the Internet’s empowerment of users for democracy, dissent and 

pluralism. 

Part I: Big Picture & New Frameworks 

The Internet's Impact on Culture & Society: 

Good or Bad? 

Andrew Keen, the self-declared “Anti-Christ of Silicon Valley” 3 is scathing in 

his criticism of the Internet, especially “Web 2.0.” Keen declares we must avoid 

the siren song of “democratized media,” citizen journalism, and, as the title of 

his first book puts it, the Cult of the Amateur. He laments the “technology that 

arms every citizen with the means to be an opinionated artist or writer” as 

producing a techno-utopian delusion little different from Karl Marx’s fantasies 

of a communist society—“where nobody has one exclusive sphere of activity 

but each can become accomplished in any branch he wishes.” 

Keen recognizes the reality of Moore’s Law—the doubling of computing 

capability every two years—but refuses to accept the idea that “each advance in 

3 Tim Dowling, I don't think bloggers read, THE GUARDIAN, July 20, 2007, 

http://www.guardian.co.uk/technology/2007/jul/20/computingandthenet.books


technology is accompanied by an equivalent improvement in the condition of 

man.” Information technology is leading us into an oblivion of cultural 

amnesia, narcissism, and a childish rejection of the expertise, wisdom and 

quality of creative elites. For Keen, a “flatter” world is one in which genius can 

no longer rise above a sea of mediocrity, noise and triviality. His message on 

the verge of the next digital decade might as well be: “Abandon all hope, ye 

who enter here!” Keen’s pessimism is as strident as a certain Pollyannaish 

utopianism on the other side. 

Is there a middle ground? Adam Thierer, Senior Research Fellow at George 

Mason University’s Mercatus Center, insists there must be. In two related 

essays, Thierer describes two schools of Internet pessimism: net skeptics 

generally pessimistic about technology and “net lovers” who think the “good ol’ 

days” of the Internet were truly great but are nonetheless pessimistic about the 

future. This first essay responds to Net skeptics like Keen—putting him in the 

context of centuries of techno-pessimism, beginning with the tale from Plato’s 

Phaedrus of Theuth and Thamus. Thierer’s response is Pragmatic Optimism: 

“We should embrace the amazing technological changes at work in today’s 

Information Age but with a healthy dose of humility and appreciation for the 

disruptive impact and pace of that change. We need to think about how to 

mitigate the negative impacts associated with technological change without 

adopting the paranoid tone or Luddite-ish recommendations of the pessimists.” 

Is the Generative Internet at Risk? 

Harvard Law Professor Jonathan Zittrain summarizes the themes from his 

influential 2008 book, The Future of the Internet—And How to Stop It. Zittrain is 

Thierer’s prototypical Net-loving pessimist who worries how technology will 

evolve absent intervention by those capable of steering technology in better 

directions. Zittrain worries that consumer demand for security will drive the 

developers and operators of computer networks, services and devices to reduce 

what he calls the “generativity” of their offerings. Thus, unregulated markets 

will tend to produce closed systems that limit experimentation, creativity and 

innovation. In particular, Zittrain decries the trend towards “appliancized” 

devices and services—which, unlike the traditional personal computer, can load 

only those applications or media authorized by the developer. Not only does 

this diminish user control in the immediate sense, greater “regulability” also 

creates the potential for the Internet’s “gatekeepers” to abuse their power. 

Thus, Zittrain echoes the prediction made by Larry Lessig in Code—without a 

doubt the most influential Internet policy book ever—that “Left to itself, 

cyberspace will become a perfect tool of control.” 4 

4 Lawrence Lessig, CODE AND OTHER LAWS OF CYBERSPACE 5-6 (1999).


In the end, he proposes essentially two kinds of solutions for “Protecting the 

Internet without Wrecking It.” The first is essentially an appeal to the civic 

virtues of “netizenship.” Second, regulation may be required to force 

companies to “provide basic tools of transparency that empower users to 

understand exactly what their machines are doing,” as well as “data portability 

policies.” More radically, he proposes to impose liability on device 

manufacturers who do not respond to takedown requests regarding 

vulnerabilities in their code that could harm users. And, returning to his core 

fear of appliancized devices, he proposes that “network neutrality-style 

mandates” be imposed on “that subset of appliancized systems that seeks to 

gain the generative benefits of third-party contribution at one point in time 

while reserving the right to exclude it later.” 

Ann Bartow, Professor at the University of South Carolina School of Law, 

offers a stinging rebuke of Zittrain’s The Future of the Internet. She summarizes the book as 

follows: “We have to regulate the Internet to preserve its open, unregulated 

nature.” Her essay draws an analogy to James Joyce’s 1916 novel, A Portrait of 

the Artist as a Young Man—emphasizing Zittrain’s desire for the independence of 

his digital homeland, much as Joyce wrote about Ireland. But as a leading 

cyber-feminist, she is especially critical of what she characterizes as Zittrain’s 

call for “an elite circle of people with computer skills and free time who share 

his policy perspective” to rule his preferred future (which she calls the 

“Zittrainet”) as “Overlords of Good Faith.” 

As Bartow characterizes Zittrain’s philosophy, “The technologies should be 

generative, but also monitored to ensure that generativity is not abused by either 

the government or by scoundrels; elite Internet users with, as one might say 

today, ‘mad programming skilz’ should be the supervisors of the Internet, 

scrutinizing new technological developments and establishing and modeling 

productive social norms online; and average, non–technically proficient Internet 

users should follow these norms, and should not demand security measures that 

unduly burden generativity.” In the end, she finds Zittrain’s book lacking in 

clear definitions of “generativity” and in specific proposals for “how to avoid a bad 

future for people whose interests may not be recognized or addressed by what 

is likely to be a very homogeneous group of elites” composed primary by male 

elites like Zittrain. 

Like Bartow, Adam Thierer rejects Zittrain’s call for rule by a Platonic elite of 

philosopher/programmer kings in the “Case for Internet Optimism, Part 2: 

Saving the Net from Its Supporters.” Thierer connects the work of Larry Lessig, 

Jonathan Zittrain and Tim Wu as the dominant forces in cyberlaw, all united by 

an over-riding fear: “The wide-open Internet experience of the past decade is 

giving way to a new regime of corporate control, closed platforms, and walled 

gardens.” Thierer argues that they overstate the threats to openness and


generativity. Because “companies have strong incentives to strike the right 

openness/closedness balance…. things are getting more open all the time 

anyway, even though the Internet was never quite so open or generative as the 

“Openness Evangelicals” imagine. In the end, he concludes it is “significantly 

more likely that the [regulated] ‘openness’ they advocate will devolve into 

expanded government control of cyberspace and digital systems than that 

unregulated systems will, as the Openness Evangelicals fear, become subject to 

‘perfect control’ by the private sector.” Thus, Thierer rejects what Virginia 

Postrel called, in her 1998 book The Future and its Enemies, the “stasis 

mentality.” 5 Instead, he embraces Postrel’s evolutionary dynamism: “the 

continuum [between openness and closedness] is constantly evolving and … 

this evolution is taking place at a much faster clip in this arena than it does in 

other markets.” In the end, he argues for the freedom to experiment—a 

recurring theme of this collection. 

Is Internet Exceptionalism Dead? 

Eric Goldman, professor at Santa Clara University School of Law, provides a 

three-part historical framework for understanding the Internet Exceptionalism 

debate. In the mid-1990s, Internet Utopianism reigned triumphant, exemplified 

in the 1996 “Declaration of the Independence of Cyberspace” by John Perry 

Barlow, lyricist for the Grateful Dead. 6 Despite its radicalism, this First Wave 

of Internet Exceptionalism succeeded in getting Congress to add the only 

section of the Communications Decency Act that would survive when the 

Supreme Court struck down the rest of the Act on First Amendment grounds: 

Section 230, which “categorically immunizes online providers from liability for 

publishing most types of third party content” and thus “is clearly exceptionalist 

because it treats online providers more favorably than offline publishers—even 

when they publish identical content.” That law lies at the heart of the 

philosophical debate in this Chapter and Chapter 6: “Should Online 

Intermediaries Be Required to Police More?” The Second Wave (“Internet 

Paranoia”) led regulators to treat the Internet more harshly than analogous 

offline activity. The Third Wave (“Exceptionalism Proliferation”) proposed 

laws treating specific sites and services differently, especially social networks. 

The Deadhead Barlow was dead wrong, declare—essentially—the Hon. Alex 

Kozinski, Chief Judge of the Ninth Circuit Court of Appeals, and Josh 

Goldfoot, Department of Justice litigator—each writing only in their private 

capacity—in “A Declaration of the Dependence of Cyberspace.” While they agree 

5 VIRGINIA POSTREL, THE FUTURE AND ITS ENEMIES (1998). 

6 Declaration of John P. Barlow, Cognitive Dissident, Co-Founder, Elec. Frontier Found., A 

Declaration of the Independence of Cyberspace (Feb. 8, 1996), available at 

http://w2.eff.org/Censorship/Internet_censorship_bills/barlow_0296.declaration.


that online anonymity and long-distance communications indeed make it harder 

for governments to punish law-breakers, governments are not helpless: “By 

placing pressure on [intermediaries like hosting companies, banks and credit 

card companies] to cut off service to customers who break the law, we can 

indirectly place pressure on Internet wrong-doers.” They illustrate their point 

with the examples of secondary liability for copyright infringement and Judge 

Kozinski’s Roommates.com decision. Indeed. they reject “the conceit that 

[cyberspace] exists at all” as a distinct, let alone exceptional place, as well as 

arguments that the costs to Internet companies of handling traditional 

regulations are too high. 

Columbia Law Professor Tim Wu concurs that governments can, and do, 

regulate the Internet because of what he and Jack Goldsmith called, in their 

2006 book Who Controls the Internet?, the “persistence of physicality.” This is not 

necessarily something to be celebrated, as he notes, pointing to China’s very 

innovativeness in finding ways to repress its citizens online—a subject 

addressed in this collection’s final Chapter. Another of Thierer’s “Net-Loving 

Pessimists,” Wu professes Internet optimism but insists we must be “realistic 

about the role of government.” 

Wu summarizes the lengthy account in his 2010 book The Master Switch of how 

government is both responsible for creating information monopolists and yet 

also the only force ultimately capable of dethroning them. For Wu, the Internet 

is not exceptional—from “The Cycle” of alternation between 

centralization/closedness and decentralization/openness. Yet Wu agrees the 

Internet is indeed an exception to the general trend of traditional media: 

“[t]echnologically, and in its effects on business, culture and politics.” Thus, he 

compares the “ideology as expressed in its technology” and the American 

exceptionalism of Alexis de Tocqueville. Yet such exceptionalism, Wu warns, 

“cannot be assumed, but must be defended.” Wu closes with a very useful 

bibliography of leading works in this ongoing debate. 

H. Brian Holland, Professor at Texas Wesleyan School of Law, responds with a 

full-bore defense of what he calls the “modified Internet Exceptionalism” 

encapsulated in Section 230—“modified” to be less audacious than Goldman’s 

First Wave (“the Internet is inherently unregulable”), but still bold in its 

insistence that granting broad immunity to online intermediaries for the conduct 

of their users is vital to the flourishing of “cyber-libertarian” Web 2.0 

communities—such as wikis and social networks, capable of evolving their own 

norms and enforcement mechanisms for policing behavior. Holland provides a 

history of Section 230 and the debate over Internet exceptionalism that frames 

the discussion of intermediary deputization in Chapter 6. He explains how 

Larry Lessig’s conviction that private power leads to perfect control, as 

mentioned above, ultimately split the Internet Exceptionalist consensus against


regulation of the 1990s into two camps. Both camps carried the banner of 

Internet freedom but reached opposite conclusions about whether the real 

threat comes from government or the private sector—most notably, regarding 

Net Neutrality. Despite this fracturing, Holland notes that the exceptional 

deregulation made possible by Section 230 has grown, not contracted, in its 

interpretation by the courts since 1996. 

Similarly, Mark MacCarthy, Adjunct Professor in the Communications Culture 

and Technology Program at Georgetown University, explains how “[t]he initial 

demand from Internet exceptionalists that the online world be left alone by 

governments has morphed into the idea that governments should create a 

global framework to protect and spur the growth of the Internet.” Once the 

exaggerated claims about the impossibility of regulating the Net made by First 

Wave Internet Exceptionalists proved false, the question became not whether 

“[i]ntermediaries can control illegal behavior on the Internet and governments 

can control intermediaries, but should they?” 

Based on his first-hand experience at Visa (described in Chapter 6), MacCarthy 

seems willing to accept more intermediary deputization than Holland but insists 

that “[t]he establishment of these laws needs to follow all the rules of good 

policymaking, including imposing an obligation only when the social benefits 

exceed the social costs.” Furthermore, he warns that “a bordered Internet in 

which each country attempts to use global intermediaries to enforce its local 

laws will not scale. This is the fundamentally correct insight of the Internet 

exceptionalists.” Thus, MacCarthy concludes, “If governments are going to use 

intermediaries to enforce local laws, they are going to have to harmonize the 

local laws they want intermediaries to enforce.” 

Has the Internet Fundamentally 

Changed Economics? 

Google’s chief economist Hal Varian provides a coda to the 1998 book 

Information Rules: A Strategic Guide to the Network Economy (with Carl Shapiro of 

the University of California at Berkeley). That book pioneered the exploration 

of the unique aspects of information economics, and their implications for both 

business and policy. Here, Varian argues that the Internet’s most 

underappreciated impact on our economy lies in the obvious yet underappreciated 

ubiquity of computers in our economic transactions, facilitating 

four broad categories of “combinatorial innovation”: new forms of contract; 

data extraction and analysis; controlled experimentation; and personalization 

and customization. Varian celebrates the transformative potential of cloud 

computing technology to allow even tiny companies working internationally to 

launch innovative new applications and services that, in turn, “can serve as


building blocks for new sorts of combinatorial innovation in business processes 

that will offer a huge boost to knowledge worker productivity in the future.” 

Harvard Law Professor Yochai Benkler is best known for his book The Wealth of 

Networks—a clear allusion to Adam Smith’s 1776 classic The Wealth of Nations. 7 

Those familiar with this part of Smith’s work view him narrowly as an 

economist focused solely on what has traditionally been characterized as 

economic exchange. But Smith in fact was equal parts economist, moral 

philosopher, and jurisprudentialist—and so is Benkler. Benkler’s essay, 

“Decentralization, Freedom to Operate, and Human Sociality,” harkens back to 

Smith’s other key work, The Theory of Moral Sentiments (1759). For both Smith 

and Benkler, man’s natural sociability means that our distributed interactions 

tend to benefit society from the bottom-up—as if by Smith’s “invisible hand.” 

For Benkler, the Internet is “a global network of communications and exchange 

that allows much greater flow and conversation, so that many new connections 

are possible on scales never before seen.” Like Varian, Benkler celebrates the 

potential for cloud computing to facilitate accelerating and unprecedented 

collaboration. 

But the keys to Benkler’s future are sociality, voluntarism, widespread 

experimentation, and the freedom to experiment. The latter insistence makes 

him highly critical of is intellectual property—copyright, patent, etc. Yet he does 

not address the dangers of propertizing personal data as another form of 

intellectual property. What does privacy-property mean for data-driven 

experimentation and the freedom to experiment? This question, unanswered 

here, offers perhaps the most tantalizing organizing theme for a future 

successor to this collection of essays. 

Larry Downes closes this Chapter with an expanded version of the discussion 

of digital economics from his 2009 book The Laws of Disruption—a book in the 

same tradition as Varian and Shapiro’s Information Rules (1998), Postrel’s The 

Future and its Enemies (1998), and Clayton Christensen’s The Innovator’s Dilemma 

(1997). Here, Downes proposes five principles of information economics that 

make the digital economy different: (1) Renewability: “information cannot be 

used up”; (2) Universality: “everyone has the ability to use the same information 

simultaneously;” (3) Magnetism: “Information value grows exponentially as new 

users absorb it;” (4) Friction-free: “the more easily information flows, the more 

quickly its value increases;” and (5) Vulnerability: The value of information can 

be destroyed through misuse or even its own success—information overload. 

7 ADAM SMITH, AN INQUIRY INTO THE NATURE AND CAUSES OF THE WEALTH OF NATIONS 18- 

21 (Edwin Cannan, ed., Methuen & Co., Ltd. 1904) (1776), 

http://www.econlib.org/library/Smith/smWN.html.


For Downes, the Internet has changed economics in a second sense: by 

relentlessly and ruthlessly cutting transaction costs—e.g., the costs of search, 

information, bargaining, decision, policing and enforcement. Thus, Varian’s 

computer-mediation promises to dramatically flatten our economy: “As 

transaction costs in the open market approach zero, so does the size of the 

firm—if transaction costs are nonexistent, then there is no reason to have large 

companies”—what Downes calls “The Law of Diminishing Firms.” 

Downes echoes Postrel’s critique of the stasis mentality: “the old rules do little 

more than hold back innovation for the benefit of those who cannot or do not 

know how to adapt to the economics of digital life.” Like Benkler, Downes 

particularly worries about copyright law’s ability to keep pace, but also explores 

the implications of lower transactions costs for privacy, asking: “What happens 

when the cost of deleting information is higher than the cost of retaining it? 

The answer is that nothing gets deleted.” In Chapter 7, both Downes and 

Stewart Baker explore the costs and benefits of privacy regulation. 

Finally, Eric Goldman offers another three-part conceptual framework—this 

time, for understanding how the Internet has revolutionized markets for 

reputational information. Goldman argues that “well-functioning marketplaces 

depend on the vibrant flow of accurate reputational information.” The Internet 

may allow markets to regulate themselves better: If reputational information 

that was previously “locked in consumers’ heads” can flow freely, it can “play 

an essential role in rewarding good producers and punishing poor ones.” 

Smith’s invisible hand alone is not enough, but “reputational information acts 

like an invisible hand guiding the invisible hand”—the “secondary invisible 

hand.” A “tertiary invisible hand” allows “the reputation system to earn 

consumer trust as a credible source… or to be drummed out of the market for 

lack of credibility…. ” 

Goldman cautions against interventions that suppress reputational information, 

but also highlights the potential unintended consequences of interventions 

intended to make reputation markets work better—like anti-gaming rules and a 

right-of-reply. Like Holland, Goldman emphasizes the central importance of 

Section 230’s immunity in allowing reputation systems to flourish without being 

crushed by intermediary liability or policing obligations. 

Who Will Govern the Net in 2020? 

Each of the three authors in this Chapter wisely resists the temptation to make 

overly specific prophesies and instead considers the broad themes likely to 

shape the policy debate over the Internet’s future. New York School of Law 

Professor David Johnson and Syracuse Information Studies Professor Milton 

Mueller focus on who should govern the Net in 2020—and could just as easily 

have responded to our question about Internet Exceptionalism—while Rob


Atkinson, President of the Information Technology and Innovation 

Foundation, provides a “field guide” to the eight major camps in Internet 

policy. 

Echoing Postrel’s dynamist/stasist theme, like Thierer, Mueller predicts “The 

future of Internet governance will be driven by the clash between its raw 

technical potential and the desire of various incumbent interests—most notably 

nation-states—to assert control over that potential].” He hopes the Internet 

will be governed by a “denationalized liberalism” based on “a universal right to 

receive and impart information regardless of frontiers, and sees freedom to 

communicate and exchange information as fundamental, primary elements of 

human choice and political and social activity.” This will require the authority 

of national and subnational governments must be contained to “domains of law 

and policy suited to localized or territorialized authority,” while Internet 

governance institutions must be completely detached from nation-state 

institutions. Defenders of free speech will ultimately have to use global free 

trade institutions to strike down censorship. 

Mueller finds strong grounds for optimism in the Internet’s empowering and 

democratizing nature, and in the rise of new access technologies like unlicensed 

wireless broadband capable of disrupting existing Internet access bottlenecks. 

But he worries about the growing technological capabilities of broadband 

providers to manage and potentially censor traffic on their networks, and admits 

a darker future of strife, industrial consolidation, censorship and cyber-warfare 

is possible. Like Zittrain, Mueller fears a splintering of the Internet driven by 

conflicts over the Internet’s “Root Server,” and that such conflicts are bound to 

intensify as the drive to secure the Internet against cyber-threats and cyberwarfare 

intensifies. 

Like Wu, David Johnson, reaches back to Tocqueville’s Democracy in America 

(1835). While Mueller proposes a new liberalism, Johnson proposes 

“Democracy in Cyberspace: Self-Governing Netizens and a New, Global Form 

of Civic Virtue, Online.” Paraphrasing Tocqueville, Johnson argues: “The 

Internet establishes a new equality of condition and enables us to exercise 

liberty to form associations to pursue new civic, social, and cultural goals.” 

Thus, the Internet is “inherently democratic”—in ways well beyond politics. 

But the Internet’s nature as an “engine of democratic civic virtue” must be 

defended daily by “netizens—the global polity of those who collaborate online, 

seek to use the new affordances of the Internet to improve the world, and care 

about protecting an Internet architecture that facilitates new forms of civic 

virtue.” Johnson argues against Wu’s apparent resignation to some degree of 

government meddling online: “A world in which every local sovereign seeks to 

control the activities of netizens beyond its borders violates the true meaning of 

self-governance and democratic sovereignty.” Johnson predicts that technology


will empower users to sidestep the traditional controls imposed by 

governments—not perfectly, but well enough. Thus, the Internet can fulfill the 

more modest ambitions of First Wave Internet exceptionalists: by making the 

Internet exceptionally democratic and pluralistic. 

Johnson’s approach resembles Thierer’s Pragmatic Optimism staked out by 

Adam Thierer: "the trajectory of freedom and even civic virtue has been, in 

broad terms, over time, constantly upward—because everyone who gets a 

chance to experience an increased level of democratic self-government—a new 

‘equality of condition.’” Like Varian, Benkler and Downes, Johnson sees the 

Internet’s facilitation of collaboration and communication as the keys to 

democratic empowerment. 

As a think tank veteran, Rob Atkinson offers a “Taxonomy of Information 

Technology Policy and Politics,” describing eight camps and their positions 

along four key issues. First is perhaps the strongest, yet also the hardest to 

define: the Internet Exceptionalists, the “Netizens” who “believe that they 

launched the Internet revolution,” prefer informal Internet governance, and 

generally oppose government intervention online—especially copyright. By 

contrast, Social Engineers distrust large corporations even more than 

government, thus leading them to advocate regulatory solutions. Though 

Atkinson doesn’t draw the connection, this camp might well be unified by 

Lessig’s concept of “code as law”—updated as “choice architecture,” in the 

highly influential 2008 book Nudge: Improving Decisions about Health, Wealth, and 

Happiness by Cass Sunstein and Richard Thaler. Free Marketers are those who 

believe the “Internet empowers people, liberates entrepreneurs, and enables 

markets”—especially by reducing transactions costs. Atkinson’s proposed tent 

may be rather too large, potentially encompassing some who advocate 

regulations like net neutrality or antitrust intervention they believe are the key to 

freeing markets. The term cyber-libertarian, seems both narrower and broader 

than Atkinson’s conception of “free-marketeers.” 8 Indeed, it was originally the 

term Atkinson used for the “Internet Exceptionalist” camp, focused primarily 

on cyber-libertinism and a fanatic rejection of copyright. 

Moral Conservatives, on the other hand, “have no qualms about enlisting 

governments to regulate the Internet” to stamp out sin and sedition. Old 

Economy Regulators reject Internet exceptionalism absolutely and insist on 

continuing to regulate the Internet like all media in the “public interest.” Tech 

Companies & Trade Associations are united not by philosophical approach 

but by their ultimate duty to shareholders, while Bricks-and-Mortars 

8 See Adam Thierer & Berin Szoka, Cyber-Libertarianism: The Case for Real Internet Freedom, THE 

TECHNOLOGY LIBERATION FRONT, Aug. 12, 2009, http://techliberation.com/2009/08/ 

12/cyber-libertarianism-the-case-for-real-internet-freedom/


companies, professional groups, and unions generally work to thwart the 

Internet’s disruption of their business models—exemplifying Virginia Postrel’s 

“stasis mindset.” Atkinson’s own camp is that of the Moderates, who want 

government to “do no harm” to information technology innovations, but also 

to “actively do good’ by adopting policies to promote digital transformation” of 

the economy. 

Part II: Issues & Applications 

Should Online Intermediaries 

Be Required to Police More? 

Seton Hall Law Professor Frank Pasquale argues that the Internet allows 

intermediaries to shroud their operations in what might be called “perfect 

opaqueness”—to extend Larry Lessig’s feared model of “perfect control.” 

Pasquale uses the example of Google to illustrate the many ways in which 

online intermediaries choose to police the Internet, even when not required to 

do by governments. Given the critical policing role played by intermediaries, 

Pasquale proposes an “Internet Intermediary Regulatory Council” to “help 

courts and agencies adjudicate controversies concerning intermediary practice” 

and assure adequate monitoring—a “prerequisite for assuring a level playing 

field online.” The IIRC “could include a search engine division, an ISP 

division focusing on carriers, and eventually divisions related to social networks 

or auction sites if their practices begin to raise commensurate concerns.” 

While leaving open the possibility that the IIRC could be a private entity, 

Pasquale is unabashed in citing Robert Hale, theoretician of the New Deal’s 

regulatory frenzy: “Hale’s crucial insight was that many of the leading businesses 

of his day were not extraordinary innovators that ‘deserved’ all the profits they 

made; rather, their success was dependent on a network of laws and regulation 

that could easily shift favor from one corporate player to another.” But rather 

than repealing these laws and regulation to allow the “evolutionary dynamism” 

of competition to play out, as Adam Thierer proposes, Pasquale is willing to 

“rely on competition-promotion via markets and antitrust only to the extent 

that (a) the intermediary in question is an economic (as opposed to cultural or 

political) force; (b) the ‘voice’ of the intermediary’s user community is strong; 

and (c) competition is likely to be genuine and not contrived.” Otherwise, 

competition is inadequate. “The bottom line,” Pasquale concludes, “is that 

someone needs to be able to look under the hood” of culturally significant 

automated ranking systems.” Thus, the Internet is not exceptional: Pasquale 

believes only careful regulatory oversight can protect us from shadowy 

corporations, just as in Franklin Delano Roosevelt’s telephone-and-radio era.


While Pasquale seems not to object to intermediaries acting as arms of the 

police state so long as they are properly transparent and regulated, Mark 

MacCarthy cautions against the practical problems raised by intermediary 

policing and offers an analytical model for deciding when intermediary 

deputization is appropriate. Based on his experience as Senior Vice President 

for Public Policy at Visa Inc., MacCarthy explores how payment systems have 

handled Internet gambling and copyright infringement as exemplary case studies 

in intermediary deputization because, unlike most online intermediaries, 

payment systems are subject neither to Section 230’s absolute immunity for 

third-party content or activities nor to the notice-and-take-down conditional 

immunity of the Digital Millennium Copyright Act. 

MacCarthy finds cause for optimism about self-regulation: “regardless of the 

precise legal liabilities, intermediaries have a general responsibility to keep their 

systems free of illegal transactions and they are taking steps to satisfy that 

obligation.” But he insists intermediary liability should be imposed only where 

real market failures exists, where supported by “an analysis of costs, benefits 

and equities,” where spelled out clearly, and to the extent local laws are 

harmonized internationally. 

The most troubling form of intermediary deputization comes from uncertain 

secondary copyright liability, writes independent writer, lawyer and programmer 

Paul Szynol in an expanded version of an essay originally written for the 

Electronic Frontier Foundation. He challenges the anti-exceptionalist 

arguments made by Judge Kozinski and Josh Goldfoot. Szynol argues that the 

failure to clearly define such liability chills innovation and investment in 

innovative start-ups—and that that this problem is unique to the Internet, given 

the vastly larger scale of competition facilitated by digital markets. 

Most intriguingly, Szynol argues that Kozinski and Goldfoot contradict their 

argument against Internet Exceptionalism by insisting on a standard for 

secondary liability online that is not actually applied offline. Szynol asks, 

“should a car company be held liable for drivers who speed? After all, it would 

be easy enough to add a ‘speed limit compliance chip.’ Yet auto manufacturers 

are not forced to pay any portion of a speeding driver's ticket. Offline, in other 

words, bad actors—the users of technology—are punished for their own 

transgressions. Online, however, the law chases the manufacturers—and 

applies ad-hoc, ambiguous standards [of secondary liability] to their products.” 

Thus, for all their denunciation of First Wave Exceptionalists like John Perry 

Barlow, Szynol essentially insists Kozinski and Goldfoot are actually Goldman’s 

“Second Wave” Internet Exceptionalists who want to impose more punitive 

regulations online than offline.


Is Search Now an "Essential Facility?" 

Frank Pasquale brings his theory of intermediary regulation to full fruition with 

his sweeping call for “search neutrality.” Like Tim Wu in The Master Switch, 

Pasquale worries that antitrust law is incapable of protecting innovation and 

adequately addressing the “the cultural and political concerns that dominant 

search engines raise." Thus, he aims to “point the way toward a new concept of 

‘essential cultural and political facility,’ which can help policymakers realize the 

situations where a bottleneck has become important enough that special 

scrutiny is warranted.” In particular, Pasquale sees taming search as inextricably 

intertwined with protecting privacy—“Engaging in a cost-benefit analysis [as in 

antitrust law] diminishes privacy's status as a right”—and Google’s potential 

chokehold on information through the Google Books Settlement. 

The existence of competition in search, especially from Microsoft’s Bing, and 

the potential for competition from Facebook and other services yet to be 

invented, are essentially irrelevant to Pasquale, while the First Amendment’s 

protection of search engine operators are a complication to be addressed down 

the road. He concludes by insisting that regulation should be supplemented by 

a publicly funded alternative to the dominant private sector search engine— 

something the French government has heavily subsidized a European “Quaero” 

search engine. Similarly, in Chapter 6, Pasquale proposed to model his Internet 

Intermediary Regulatory Council on the French Data Protection Authority. 

Thus, Pasquale’s over-arching vision seems to be that of a Digital New Deal—a 

la française. 

Geoffrey Manne, Professor at Lewis & Clark Law and Executive Director of 

the International Center for Law & Economics, explains that search engines are 

not the bottlenecks Pasquale suggests—and thus why even the traditional 

essential facilities doctrine, which he says “has been relegated by most antitrust 

experts to the dustbin of history,” should not apply to them. In essence, he 

argues that “search neutrality” would protect only competitors, not consumers, 

because even a popular search engine like Google cannot foreclose advertisers’ 

access to consumers’ attention. Google, like any company, has no legal duty to 

help its rivals. More to the point, even if Google entirely dominated search, it 

could not block consumers’ access to its competitors. This, argues Manne, is the 

relevant market to analyze—quoting Supreme Court Justice Abe Fortas’s 

famous admonition about excessively narrow market definitions: “This Court 

now approves this strange red-haired, bearded, one-eyed man-with-a-limp 

classification.” 

Like Manne, New York Law School Professor James Grimmelmann expresses 

“Skepticism about Search Neutrality,” and the significant practical problems it


would create. As the author of the definitive law review article, The Structure of 

Search Engine Law, 9 Grimmelmann is keenly aware of the concerns raised by 

search, yet he concludes that “the case for search neutrality is a muddle” 

because its “ends and means don’t match.” Echoing Johnson, Mueller, 

Holland, and Thierer’s view of the Internet as a liberating, democratizing force, 

Grimmelmann is clear that the lodestar of search is user autonomy: “If search 

did not exist, then for the sake of human freedom it would be necessary to 

invent it.” He deconstructs eight search neutrality principles—equality, objectivity, 

bias, traffic, relevance, self-interest, transparency and manipulation—and 

finds each lacking, but cautions that “it doesn’t follow that search engines 

deserve a free pass under antitrust, intellectual property, privacy, or other wellestablished 

bodies of law,” and that some other “form of search-specific legal 

oversight” might be appropriate. 

Eric Goldman once again puts the debate in the context of its intellectual 

history. Always focused on questions of exceptionalism, Goldman concludes 

search engines are neutral only in theory (“Search Engine Utopianism”) but 

must “make editorial judgments just like any other media company.” He 

explains that, while “search engine bias sounds scary, … such bias is both 

necessary and desirable”—and the remedy of “search neutrality” is probably 

worse than whatever adverse consequences come with search engine bias. 

Ultimately, he predicts that “emerging personalization technology will soon 

ameliorate many concerns about search engine bias.” 

What Future for Privacy Online? 

Michael Zimmer, Professor of Information Studies at School of the University 

of Wisconsin-Milwaukee, concedes that “the Internet has become a platform 

for the open flow of personal information—flows that are largely voluntarily 

provided by users.” Yet Zimmer discusses lingering reasons for concern about 

the Internet as a “potent infrastructure for the flow and capture of personal 

information.” 

Zimmer explores the conflicts among privacy laws in the U.S., Europe, Canada 

and elsewhere, but concludes that “Companies are, on the whole, not moving 

around in order to avoid strict privacy regulations… instead, there has been a 

gradual increase in awareness and action on the issue of privacy.” Still, Zimmer 

worries that the “‘trading up’ to an increased level of protection of personal 

information flows on our transnational digital networks has not materialized as 

quickly or clearly as one might expect.” Zimmer’s answer is to demand a 

“renewed commitment to the rights of data subjects embodied in the Canadian 

and European Union approach to data protection.” 

9 James Grimmelmann, The Structure of Search Engine Law, 93 IOWA L. REV. 1 (2007).


Zimmer writes from the perspective that views privacy as a “right.” This is, to 

put it mildly, not a perspective shared by the other two authors in this Chapter: 

Stewart Baker, a Partner at Steptoe & Johnson LLP and former Assistant 

Secretary for Policy at the Department of Homeland Security (DHS), and Larry 

Downes, who has expanded his essay from his 2009 book The Laws of Disruption. 

Baker spent his time at DHS battling privacy advocates over programs he felt 

justified to protect Americans against terrorism—leading him to ask, “What’s 

Wrong with Privacy?” He traces the answer back to the 1890 law review article, 

“The Right to Privacy” by Supreme Court Justice Louis Brandeis and Samuel 

Warren that gave birth to modern privacy law. Baker rejects their “reactionary 

defense of the status quo” as Boston elites who didn’t much like the news 

media reporting on the details of their private parties. In essence, Baker finds in 

the privacy “movement” the same “stasis mentality” defined by Virginia Postrel. 

Like Postrel, Baker argues for dynamism: “Each new privacy kerfuffle inspires 

strong feelings precisely because we are reacting against the effects of a new 

technology. Yet as time goes on, the new technology becomes commonplace. 

Our reaction dwindles away. The raw spot grows a callous. And once the initial 

reaction has passed, so does the sense that our privacy has been invaded. In 

short, we get used to it.” 

Baker rejects the concept of “predicates” for government access to data (e.g., 

requiring “probable cause” for a warrant), the “Brandeisian notion that we 

should all ‘own’ our personal data,” and attempting to limit uses of information. 

Baker has little to say about the private sector’s use of data but proposes a 

system of auditing government employees to rigorously monitor their use of 

private information. 

Larry Downes, too, rejects the concept of intellectual property in personal 

information—but is willing to concede that Warren and Brandeis “weren’t 

entirely wrong” in that “‘private’ information can also be used destructively.” 

He thus leaves open the possibility of narrow laws tailored to limiting specific, 

destructive uses of information—such as anti-discrimination laws. But Downes 

is highly skeptical about governmental enforcement of “privacy rights,” and 

ultimately echoes John Perry Barlow’s optimism about the potential for 

Netizens to solve their own problems: “Where there are real conflicts, where 

there are wrongs, we will identify them and address them by our means.” 10 

Specifically, Downes argues that “the same technologies that create the privacy 

problem are also proving to be the source of its solution. Even without 

government intervention, consumers increasingly have the ability to organize, 

identify their common demands, and enforce their will on enterprises”— 

detailing examples of how reputational pressure can discipline corporate privacy 

10 Barlow, supra note 6.


practices. Three cheers for the sound of Eric Goldman’s three invisible hand 

clapping, perhaps? Ultimately, Downes vests his greatest hope in the Internet’s 

potential to create new markets by lowering transactions costs—this time, a 

market for private data in which an explicit quid pro quo rewards consumers for 

sharing their personal data for beneficial, rather than destructive uses. 

Can Speech Be Policed 

in a Borderless World? 

John Palfrey, Harvard Law Professor and co-director of Harvard’s influential 

Berkman Center for Internet & Society, speaks with unique authority on 

censorship as one of the co-authors of exhaustive surveys of global censorship 

conducted by himself, Jonathan Zittrain and others at Berkman. These studies 

confirm Tim Wu’s conclusion that governments can and do censor speech 

effectively, contrary to the hopes of First Wave Internet Exceptionalists. 

Palfrey provides a beginner’s guide to the techniques used in, goals of, and 

practical problems created by content filtering. Most disturbingly, he notes the 

growing use of “soft controls” through governmental pressure and governmentfostered 

social norms intended to squelch dissent. 

Like Zittrain, Mueller and Johnson, Palfrey fears “we may be headed toward a 

localized version of the Internet, governed in each instance by local laws.” He 

thus demands a greater international debate about speech controls that forces 

states to discuss whether they “actually want their citizens to have full access to 

the Internet or not.” In particular, he echoes Mueller’s call for international free 

trade institutions to strike down censorship barriers to free speech. 

Christopher Wolf, Partner at Hogan Hartson LLP, focuses not on speech that 

governments hate, but on “hate speech” we all—or nearly all—would find 

objectionable. Yet he notes how difficult it can be to distinguish these two 

categories of censorship. Furthermore, he concludes, after much crusading 

against hate speech, that “laws against hate speech have not demonstrably 

reduced hate speech or deterred haters.” Thus, he concludes that “Hate speech 

can be ‘policed’ in a borderless world, but not principally by the traditional 

police of law enforcement. The Internet community must continue to serve as 

a ‘neighborhood watch’ against hate speech online, ‘saying something when it 

sees something,’ and working with online providers to enforce community 

standards.” Thus, like Johnson, Mueller and Barlow, Wolf looks to Netizens to 

combat hate speech.


Can the Net Liberate the World? 

The book closes by discussing the most tragic disappointment of the First Wave 

Internet Exceptionalists’ vision. Where John Perry Barlow insisted, defiantly, 

that governments those “weary giants of flesh and steel… [did not] possess any 

methods of enforcement we have true reason to fear,” the reality is that 

oppressive governments continue to reign, sometimes even using the Internet 

to serve their agenda. Can the Net liberate the world—or will it, too, become 

another tool of “perfect control,” as Larry Lessig feared? Or will imperfect 

controls work well enough to allow tyrants to hang on to power? 

Evgeny Morozov is a leading commentator on foreign affairs, a visiting scholar 

at Stanford University and a Schwartz fellow at the New America Foundation. 

He praises the Internet’s ability to quickly disseminate information and allow 

dissidents to organize. Yet, having grown up in the Soviet Union, he is deeply 

skeptical about the much-hyped potential for Web media to live up to the hype 

about democratization. He rejects two critical assumptions underlying this 

hype. First, he concludes that the legitimacy of undemocratic regimes is derived 

less from “brainwashing” that can be cured by exposure to the alternative views 

online and more from popular support for authoritarian regimes that promise to 

deliver economic growth or play effectively on other concerns, such as 

nationalism or religion. Second, he suggests the Internet can actually facilitate 

surveillance, fuel genuine support for existing regimes, allow government to 

subtly manipulate public opinion, or simply make authoritarianism more 

efficient. 

John Palfrey’s acid observation in the previous Chapter bolsters Morozov’s 

suggestion that much of the world may not actually want to be liberated: “In 

China and in parts of the former Soviet Union, very often the most fearsome 

enforcer of the state's will is the old woman on one's block, who may or may 

not be on the state's payroll.” 

Optimists like Johnson, Mueller, Thierer and Holland would likely differ from 

Morozov—and the U.S. State Department has tended in this direction, too. In 

January 2010, Secretary of State Hillary Clinton gave a bold speech embracing 

this optimism about the liberating potential of the Internet, and announcing a 

commitment to “supporting the development of new tools that enable citizens 

to exercise their rights of free expression by circumventing politically motivated 

censorship.” 11 

11 Hillary Rodham Clinton, Remarks on Internet Freedom, Jan. 21, 2010, 

http://www.state.gov/secretary/rm/2010/01/135519.htm


Internet entrepreneur Ethan Zuckerman is a senior researcher at the Berkman 

Center and founder of Geekcorps, a non-profit dedicated to building computer 

infrastructure in developing countries. He joined John Palfrey in the study of 

censorship circumvention tools mentioned above. 12 Despite his passionate 

commitment to promoting such tools, as Secretary Clinton proposed, he 

concludes that “We can’t circumvent our way around Internet censorship” 

because of the costs and practical challenges of attempting to circumvent 

censorship on a scale sufficient to make a real difference. Thus, he views 

circumvention as just one of many tools required to thwart “soft censorship, 

website blocking, and attacks on dissident sites. But ultimately, what is most 

required is building the right “theory of change” to inform the multi-pronged 

strategy necessary for the Internet to achieve its democratizing potential. 

Conclusion: Discovering the Future 

of the Internet & Digital Policy 

In these thirty-one essays, our authors paint a complex picture of the future of 

the Internet and digital policy: Technological change inevitably creates new 

problems, even as it solves old ones. In the end, one’s perspective ultimately 

depends on whether one thinks the “net” effect of that change is positive or 

negative—depending on how much, and in what ways, government intervenes 

online. 

Personally, this collection brings me back to where I started my study of 

Internet policy—reading John Perry Barlow’s “Declaration of the Independence 

of Cyberspace” in 1996, and Virginia Postrel’s The Future and Its Enemies in 1998. 

Despite its now-obviously excessive utopian naïveté about the Internet’s 

crippling of the State, Barlow’s poetry still resonates deeply with many, 

including myself, as a powerful synthesis of Internet exceptionalism and cyberlibertarianism, 

a vision of progress as empowerment and uplifting of the user. 

Yet like my former colleague Adam Thierer, it is Postrel’s evolutionary 

dynamism that most guides me, with its emphasis not on a “carefully outlined 

future” or “build[ing] a single bridge from here to there, for neither here nor 

there is a single point,” but on the process of discovery by which the future 

evolves. 13 Like Postrel, I do not imagine that the disruption and transformation 

wrought by the Digital Revolution will always be rosy or easy. But we cannot— 

12 HAL ROBERTS, ETHAN ZUCKERMAN & JOHN PALFREY, 2007 CIRCUMVENTION LANDSCAPE 

REPORT: METHODS, USES, AND TOOLS (March 2009), http://dash.harvard.edu/ 

bitstream/handle/1/2794933/2007_Circumvention_Landscape.pdf?sequence=2. 

13 Postrel, supra note 5 at 218.


as the legendary King Canute once tried with the English Channel—command 

the tides of technological change to halt. 

Thierer’s “Pragmatic Optimism” demands much more than a resignation to the 

inevitability of change. At its heart, it is requires a cheery confidence in what 

David Johnson dubs the “Trajectory of Freedom”—“in broad terms, over time, 

constantly upward”—but also a commitment to the process by which that 

trajectory is discovered. This is progress—progress as freedom. 14 But progress 

also requires freedom, the freedom to discover, innovate and experiment, if 

technology is to achieve its full potential to improve the human condition and 

expand individual capacity to choose. 

I leave it to you, the reader, to choose—to discover—your own answers to the 

many questions of law, economics, philosophy and policy explored in this 

unique book. 

14 ROBERT NISBET, HISTORY OF THE IDEA OF PROGRESS 215 (1980).

CONTRIBUTORS 

Robert D. Atkinson 31 

Stewart Baker 31 

Ann Bartow 32 

Yochai Benkler 32 

Larry Downes 33 

Josh Goldfoot 34 

Eric Goldman 34 

James Grimmelmann 35 

H. Brian Holland 35 

David R. Johnson 36 

Andrew Keen 36 

Hon. Alex Kozinski 37 

Mark MacCarthy 37 

Geoffrey Manne 38 

Evgeny Morozov 39 

Milton Mueller 39 

29

30 CONTRIBUTORS 

John Palfrey 40 

Frank Pasquale 40 

Berin Szoka 41 

Paul Szynol 41 

Adam Thierer 42 

Hal Varian 42 

Christopher Wolf 43 

Tim Wu 44 

Michael Zimmer 44 

Jonathan Zittrain 45 

Ethan Zuckerman 46


Robert D. Atkinson 

Robert Atkinson is the founder and president of the Information Technology 

and Innovation Foundation, a Washington, D.C.-based technology policy think 

tank. He is also author of the State New Economy Index series and the book, The 

Past And Future Of America’s Economy: Long Waves Of Innovation That Power Cycles 

Of Growth (Edward Elgar, 2005). He has an extensive background in technology 

policy, he has conducted ground-breaking research projects on technology and 

innovation, is a valued adviser to state and national policy makers, and a popular 

speaker on innovation policy nationally and internationally. 

Before coming to ITIF, Dr. Atkinson was Vice President of the Progressive 

Policy Institute and Director of PPI’s Technology & New Economy Project. 

While at PPI he wrote numerous research reports on technology and innovation 

policy, including on issues such as broadband telecommunications, Internet 

telephony, universal service, e-commerce, e-government, middleman opposition 

to e-commerce, privacy, copyright, RFID and smart cards, the role of IT in 

homeland security, the R&D tax credit, offshoring, and growth economics. 

Previously Dr. Atkinson served as the first Executive Director of the Rhode 

Island Economic Policy Council, a public-private partnership including as 

members the Governor, legislative leaders, and corporate and labor leaders. As 

head of RIEPC, he was responsible for drafting a comprehensive economic 

strategic development plan for the state, developing a ten-point economic 

development plan, and working to successfully implement all ten proposals 

through the legislative and administrative branches. Prior to that he was Project 

Director at the former Congressional Office of Technology Assessment. While 

at OTA, he directed The Technological Reshaping of Metropolitan America, a seminal 

report examining the impact of the information technology revolution on 

America’s urban areas. 

Stewart Baker 

Stewart A. Baker is a partner in the Washington office of Steptoe & Johnson 

LLP. He returned to the firm following 3½ years at the Department of 

Homeland Security as its first Assistant Secretary for Policy. 

At Homeland Security, Mr. Baker created and staffed the 250-person DHS 

Policy Directorate. He was responsible for policy analysis across the 

Department, as well as for the Department’s international affairs, strategic 

planning and relationships with law enforcement and public advisory 

committees. While at DHS, Mr. Baker led successful negotiations with 

European and Middle Eastern governments over travel data, privacy, visa 

waiver and related issues. He devised a new approach to visa-free travel, forged 

a congressional and interagency consensus on the plan and negotiated 

acceptance with key governments.


Mr. Baker manages one of the nation’s premier technology and security 

practices at Steptoe. Mr. Baker’s practice covers national security, electronic 

surveillance, law enforcement, export control encryption, and related 

technology issues. He has been a key advisor on U.S. export controls and on 

foreign import controls on technology. He has also advised companies on the 

requirements imposed by CFIUS. 

Mr. Baker’s practice includes issues relating to government regulation of 

international trade in high-technology products, and advice and practice under 

the antidumping and countervailing duty laws of United States, European 

Community, Canada, and Australia. He also counsels clients on issues involving 

foreign sovereign immunity, and compliance with the Foreign Corrupt Practices 

Act. 

Mr. Baker has handled the arbitration of claims exceeding a billion dollars, is a 

member of national and international rosters of arbitrators, and is the author of 

articles and a book on the United Nations Commission on International Trade 

Law arbitration rules. 

Mr. Baker has had a number of significant successes in appellate litigation and 

appearances before the United States Supreme Court. He developed—and 

persuaded the Court to adopt—a new theory of constitutional federalism that 

remains the most vibrant 10th Amendment doctrine of the past 30 years. 

Ann Bartow 

Ann Bartow is a Professor of Law at the University of South Carolina School 

of Law in Columbia, South Carolina. She holds a Bachelor of Science from 

Cornell University and a Juris Doctor from the University of Pennsylvania Law 

School. She currently teaches an Intellectual Property Survey, Copyright Law, 

Trademarks and Unfair Competition Law, Patent Law and a seminar entitled 

Pornography, Prostitution, Sex Trafficking and the Law. Her scholarship 

focuses on the intersection between intellectual property laws and public policy 

concerns, privacy and technology law, and feminist legal theory. She also coadministers 

the Feminist Law Professors blog, is a regular blogger at 

Madisonian.net and a contributing editor at Jotwell.com. 


Yochai Benkler is the Berkman Professor of Entrepreneurial Legal Studies at 

Harvard, and faculty co-director of the Berkman Center for Internet and 

Society. Before joining the faculty at Harvard Law School, he was Joseph M. 

Field ‘55 Professor of Law at Yale. He writes about the Internet and the 

emergence of networked economy and society, as well as the organization of 

infrastructure, such as wireless communications.


In the 1990s he played a role in characterizing the centrality of information 

commons to innovation, information production, and freedom in both its 

autonomy and democracy senses. In the 2000s, he worked more on the sources 

and economic and political significance of radically decentralized individual 

action and collaboration in the production of information, knowledge and 

culture. His books include THE WEALTH OF NETWORKS: HOW SOCIAL 

PRODUCTION TRANSFORMS MARKETS AND FREEDOM (2006), which received 

the Don K. Price award from the American Political Science Association for 

best book on science, technology, and politics. 

In civil society, Benkler’s work was recognized by the Electronic Frontier 

Foundation’s Pioneer Award in 2007, and the Public Knowledge IP3 Award in 

2006. His articles include Overcoming Agoraphobia (1997/98, initiating the debate 

over spectrum commons); Commons as Neglected Factor of Information Production 

(1998) and Free as the Air to Common Use (1998, characterizing the role of the 

commons in information production and its relation to freedom); From 

Consumers to Users (2000, characterizing the need to preserve commons as a core 

policy goal, across all layers of the information environment); Coase’s Penguin, or 

Linux and the Nature of the Firm (characterizing peer production as a basic 

phenomenon of the networked economy) and Sharing Nicely (2002, 

characterizing shareable goods and explaining sharing of material resources 

online). His work can be freely accessed at benkler.org. 

Larry Downes 

Larry Downes is an Internet analyst and consultant, helping clients develop 

business strategies in an age of constant disruption caused by information 

technology. 

Downes is author of Unleashing the Killer App: Digital Strategies for Market 

Dominance (Harvard Business School Press, 1998), which was named by the Wall 

Street Journal as one of the five most important books ever published on 

business and technology. 

His new book, The Laws of Disruption: Harnessing the New Forces that Govern Lie and 

Business in the Digital Age (Basic Books 2009) offers nine strategies for success in 

navigating the accident-prone intersection of innovation and the law. 

From 2006-2010, Downes was a nonresident Fellow at the Stanford Law School 

Center for Internet and Society. Before that, he held faculty positions at the 

University of California-Berkeley, Northwestern University School of Law, and 

the University of Chicago Graduate School of Business. Downes is a Partner 

with the Bell-Mason Group, which works with Global 1000 corporations, 

providing corporate venturing methodologies, tools, techniques and support 

that accelerate corporate innovation and venturing programs.


He has written for a variety of publications, including U.S.A Today, Harvard 

Business Review, Inc., Wired, CNet, Strategy & Leadership, CIO, The 

American Scholar and the Harvard Journal of Law and Technology. He was a 

columnist for both The Industry Standard and CIO Insight. He blogs for the 

Technology Liberation Front. 

Josh Goldfoot 

Josh Goldfoot is Senior Counsel with the Computer Crime & Intellectual 

Property Section of the U.S. Department of Justice. He prosecutes hackers and 

other computer criminals, and advises investigators and other prosecutors about 

privacy statutes, the Fourth Amendment, and implications of emerging 

technologies on law enforcement. In 2010, he was awarded the Assistant 

Attorney General’s Meritorious Service Award. He is an accomplished software 

developer and computer technician, and received a United States patent in 2008 

for shape recognition technology. He is a graduate of Yale University and 

earned his law degree from the University of Virginia School of Law. He has 

worked in technology law since 1999, when he advised Internet startups in 

Silicon Valley on intellectual property issues. Prior to joining the Department 

of Justice in 2005, he did appellate and civil litigation, and clerked for judge Alex 

Kozinski on the Ninth Circuit U.S. Court of Appeals. He was a Special 

Assistant United States Attorney in the Eastern District of Virginia for six 

months in 2007 and 2008. 

Eric Goldman 

Eric Goldman is an Associate Professor of Law at Santa Clara University 

School of Law. He also directs the school’s High Tech Law Institute. Before 

joining the SCU faculty in 2006, he was an Assistant Professor at Marquette 

University Law School, General Counsel of Epinions.com, and an Internet 

transactional attorney at Cooley Godward LLP. 

Eric teaches Cyberlaw and Intellectual Property and previously has taught 

courses in Copyrights, Contracts, Software Licensing and Professional 

Responsibility. 

Eric’s research focuses on Internet law, intellectual property, marketing, and the 

legal and social implications of new communication technologies. Recent 

papers have addressed topics such as search engines and online marketing 

practices. 

Eric received his BA, summa cum laude and Phi Beta Kappa, in 

Economics/Business from UCLA in 1988. He received his JD from UCLA in 

1994, where he was a member of the UCLA Law Review, and concurrently 

received his MBA from the Anderson School at UCLA.



James Grimmelmann is Associate Professor at New York Law School and a 

member of its Institute for Information Law and Policy. He received his 

J.D. from Yale Law School, where he was Editor-in-Chief of LawMeme and a 

member of the Yale Law Journal. Prior to law school, he received an A.B. in 

computer science from Harvard College and worked as a programmer for 

Microsoft. He has served as a Resident Fellow of the Information Society 

Project at Yale, as a legal intern for Creative Commons and the Electronic 

Frontier Foundation, and as a law clerk to the Honorable Maryanne Trump 

Barry of the United States Court of Appeals for the Third Circuit. 

He studies how the law governing the creation and use of computer software 

affects individual freedom and the distribution of wealth and power in society. 

As a lawyer and technologist, he aims to help these two groups speak intelligibly 

to each other. He writes about intellectual property, virtual worlds, search 

engines, online privacy, and other topics in computer and Internet law. Recent 

publications include The Internet Is a Semicommons, 78 Fordham L. Rev. 2799 

(2010), Saving Facebook, 94 Iowa L. Rev. 1137 (2009), The Ethical Visions of 

Copyright Law, 77 Fordham L. Rev. 2005 (2009). 

He has been blogging since 2000 at the Laboratorium: www.laboratorium.net. 


Professor H. Brian Holland joined the faculty of Texas Wesleyan School of 

Law in 2009. Prior to his arrival, Professor Holland was a Visiting Associate 

Professor at Penn State University’s Dickinson School of Law. 

After graduating from law school, Professor Holland spent two years as a 

judicial clerk in the U.S. Court of Appeals for the Second Circuit in New York. 

He then joined the Washington, D.C. office of Jones, Day, Reavis & Pogue. 

His work with the firm consisted primarily of appellate work before the U.S. 

Supreme Court and federal courts of appeals, as well as international arbitration 

before the World Bank. Among the significant cases litigated during this period 

were issues of intellectual property and constitutional law (Eldred v. 

Reno/Ashcroft and Luck’s Music Library, Inc. v. Reno/Ashcroft), privacy and 

identity theft (TRW v. Andrews), and federal bankruptcy jurisdiction and venue. 

Professor Holland’s scholarship reflects his interest in technology, governance 

and social change, with a particular emphasis on issues of authority within the 

online environment and the development of social norms in mediated 

communities. He is currently writing on privacy in social networks. His most 

recent work, Social Distortion: Regulating Privacy in Social Networks, has been a 

featured presentation at privacy conferences both in the United States and 

Europe.


Professor Holland received a LL.M., with honors, from Columbia University 

School of Law, completing a self-designed program in technology law. He 

holds a J.D., summa cum laude, from American University’s Washington 

College of Law, and a B.A. from Tufts University. Professor Holland is 

currently pursuing his Ph.D. in Digital Media and Mass Communications at 

Penn State University. His dissertation, now in progress, applies social semiotic 

theories to the concept of fair use in intellectual property law. 


David Johnson joined New York Law School’s faculty in spring 2004 as a 

visiting professor of law. He is a faculty member of the Institute for 

Information Law and Policy. 

Professor Johnson joined Wilmer, Cutler & Pickering in 1973 and became a 

partner in 1980. His practice focused primarily on the emerging area of 

electronic commerce, including counseling on issues relating to privacy, domain 

names and Internet governance issues, jurisdiction, copyright, taxation, 

electronic contracting, encryption, defamation, ISP and OSP liability, regulation, 

and other intellectual property matters. 

Professor Johnson helped to write the Electronic Communications Privacy Act, 

was involved in discussions leading to the Framework for Global Electronic 

Commerce, and has been active in the introduction of personal computers in 

law practice. He co-authored, with David Post, Law and Borders: The Rise of 

Law in Cyberspace, 48 Stanford Law Review 1367 (1996). He is currently 

developing online legal games and law practice simulations. 

Professor Johnson graduated from Yale College with a B.A. summa cum laude 

in 1967. He completed a year of postgraduate study at University College, 

Oxford in 1968, and earned a J.D. from Yale Law School in 1972. Following 

graduation from law school, Professor Johnson clerked for a year for the 

Honorable Malcolm R. Wilkey of the United States Court of Appeals for the 

District of Columbia. 

Andrew Keen 

Andrew Keen is the author of the international hit CULT OF THE AMATEUR: 

HOW THE INTERNET IS KILLING OUR CULTURE (2008). Acclaimed by THE 

NEW YORK TIMES’ Michiko Kakutani as having been written “with acuity and 

passion” and by A.N. Wilson in the DAILY MAIL as “staggering”, CULT OF 

THE AMATEUR has been published in fifteen different language editions and 

was short-listed for the 2008 Higham’s Business Technology Book of the Year 

award.


As a pioneering Internet entrepreneur, Andrew founded Audiocafe.com in 1995 

and built it into a popular first generation Internet music company. He was the 

executive producer of the new media show “MB5 2000” and, between 2001 and 

2007, worked as a senior sales and marketing executive at several Silicon Valleybased 

technology start-ups including Pulse, Santa Cruz Networks and Pure 

Depth. He is currently the host of the popular “Keen On” show on 

Techcrunch.tv as well as the host of the video interview series “The Future of 

Creativity” on the Harvard Business Review website. 

Andrew was educated at London University where he was awarded a First Class 

Honors Degree in Modern History, as a British Council Fellow at the University 

of Sarajevo and at the University of California at Berkeley where he earned a 

Masters Degree in Political Science. 

He is currently writing a second book entitled DIGITAL VERTIGO: ANXIETY, 

LONELINESS AND INEQUALITY IN THE SOCIAL MEDIA AGE which will be 

published by St Martins Press. 

Hon. Alex Kozinski 

Chief Judge Alex Kozinski was appointed United States Circuit Judge for the 

Ninth Circuit on November 7, 1985. He graduated from UCLA, receiving an 

A.B. degree in 1972, and from UCLA Law School, receiving a J.D. degree in 

1975. Prior to his appointment to the appellate bench, Judge Kozinski served 

as Chief Judge of the United States Claims Court, 1982-85; Special Counsel, 

Merit Systems Protection Board, 1981-1982; Assistant Counsel, Office of 

Counsel to the President, 1981; Deputy Legal Counsel, Office of President- 

Elect Reagan, 1980-81; Attorney, Covington & Burling, 1979-81; Attorney, 

Forry Golbert Singer & Gelles, 1977-79; Law Clerk to Chief Justice Warren E. 

Burger, 1976-77; and Law Clerk to Circuit Judge Anthony M. Kennedy, 1975- 

76. 


Mark MacCarthy teaches and conducts research at Georgetown University’s 

Communication, Culture, and Technology Program. He teaches courses on the 

development of the electronic media, technology policy and Internet freedom. 

He is also an adjunct member of the Department of Philosophy where he 

teaches courses in political philosophy and philosophy and privacy. He does 

research and consults in the areas of information privacy and security, Internet 

intermediary liability, global Internet freedom, the future of the online media, 

consumer financial protection, open standards, electronic and mobile commerce 

and other technology policy issues. He is an Affiliate of Georgetown 

University’s McDonough School of Business Center for Business and Public 

Policy, an investigator with the RFID Consortium for Security and Privacy, and 

the appointed expert of the American National Standards Institute on the


International Organization For Standardization (ISO) Technical Management 

Board (TMB) Privacy Steering Committee 

From 2000 to 2008, he was Senior Vice President for Global Public Policy at 

Visa Inc, responsible for policy initiatives affecting electronic commerce, new 

technology and information security and privacy. He regularly represented Visa 

before the U.S. Congress, the U.S. Administration, the U.S. Federal Trade 

Commission, the U.S federal financial regulators and multi-governmental 

groups such as the Organization for Economic Cooperation and Development 

and Asia Pacific Economic Cooperation group. 

Prior to joining Visa, he spent six years as a principal and senior director with 

the Wexler-Walker Group, a Washington public policy consulting firm, where 

he worked with a variety of clients on electronic commerce, financial services, 

privacy and telecommunications. He was Vice President in charge of Capital 

Cities/ABC’s Washington office from 1988 to 1994, representing the 

company’s interests before Congress, the Federal Communications Commission 

and other administrative agencies. From 1981 to 1988 he was a professional 

staff member of the House Committee on Energy and Commerce, where he 

handled communications policy issues. From 1978 to 1981, he worked as an 

economist performing regulatory analyses of safety and health regulations at the 

U.S. Occupational Safety and Health Administration. 

Mr. MacCarthy has a Ph.D in philosophy from Indiana University and an MA in 

economics from the University of Notre Dame. 

Geoffrey Manne 

Currently the Executive Director of the International Center for Law & 

Economics (ICLE), a global think tank, Professor Manne also serves as 

Lecturer in Law for Lewis & Clark Law School. In this capacity he lends his 

expertise to various law school endeavors and teaches the school’s Law and 

Economics course. The ICLE’s website is at www.laweconcenter.org. 

Manne was an Assistant Professor of Law at Lewis & Clark from 2003 to 2008. 

From 2006 to 2008 he took a leave of absence from the school to direct a law 

and economics academic outreach program at Microsoft, and was Director, 

Global Public Policy at LECG, an economic consulting firm, until founding the 

ICLE at the end of 2009. Prior to joining the Lewis & Clark faculty he practiced 

law at Latham & Watkins in Washington, D.C., where he specialized in antitrust 

litigation and counseling. Before private practice Manne was a Bigelow Fellow 

at the University of Chicago Law School, an Olin Fellow at the University of 

Virginia School of Law and a law clerk to Judge Morris S. Arnold of the U.S. 

Court of Appeals for the Eighth Circuit. While clerking he taught a seminar on 

Law & Literature at the University of Arkansas at Little Rock.


During law school Manne was a research assistant to Judge Richard Posner, 

Comment Editor of the University of Chicago Law School Roundtable and a 

Staff Member of the University of Chicago Legal Forum. Among his other 

vocational pursuits was a brief stint at the U.S. Federal Trade Commission. His 

research has focused broadly on the economic implications of legal constraints 

on business organizations, particularly in the contexts of antitrust, nonprofit 

organizations and international law. Manne is a member of the Virginia bar, as 

well as the Bar of the U.S. Bankruptcy Court for the Eastern District of 

Virginia. He is also a member of the American Law and Economics 

Association, the Canadian Law and Economics Association and the 

International Society for New Institutional Economics. 

He blogs for the Technology Liberation Front. 


Evgeny Morozov is the author of THE NET DELUSION: THE DARK SIDE OF 

INTERNET FREEDOM (Public Affairs, 2011). He is also a visiting scholar at 

Stanford University, a fellow at the New America Foundation and a 

contributing editor to Foreign Policy magazine. 


Milton Mueller teaches and does research on the political economy of 

communication and information. He uses the theoretical tools of property 

rights analysis, institutional economics and both historical and quantitative 

social science methods. He has a longstanding interest in the history of 

communication technologies and global governance institutions. Mueller 

received the Ph.D. from the University of Pennsylvania in 1989. 

Mueller’s most recent research projects explore the problem of governing the 

Internet. His new book NETWORKS AND STATES: THE GLOBAL POLITICS OF 

INTERNET GOVERNANCE (MIT Press, 2010) provides a comprehensive 

overview of the political and economic drivers of a new global politics. His 

acclaimed book RULING THE ROOT: INTERNET GOVERNANCE AND THE 

TAMING OF CYBERSPACE (MIT Press, 2002) was the first scholarly account of 

the debates over the governance of the domain name system. His book, 

UNIVERSAL SERVICE: COMPETITION, INTERCONNECTION AND MONOPOLY 

IN THE MAKING OF THE AMERICAN TELEPHONE SYSTEM (MIT Press, 1997) 

set out a dramatic revision of our understanding of the origins of universal 

telephone service and the role of interconnection in industry development. He 

is on the international editorial boards of the journals TELECOMMUNICATIONS 

POLICY, THE INFORMATION SOCIETY, and INFO: THE JOURNAL OF POLICY, 

REGULATION AND STRATEGY FOR TELECOMMUNICATION, INFORMATION 

AND MEDIA.


John Palfrey 

John Palfrey is Henry N. Ess Professor of Law and Vice Dean for Library and 

Information Resources at Harvard Law School. He is the co-author of Born 

Digital: Understanding the First Generation of Digital Natives (Basic Books, 2008) and 

Access Denied: The Practice and Politics of Internet Filtering (MIT Press, 2008). His 

research and teaching is focused on Internet law, intellectual property, and 

international law. He practiced intellectual property and corporate law at the 

law firm of Ropes & Gray. He is a faculty co-director of the Berkman Center 

for Internet & Society at Harvard University. Outside of Harvard Law 

School, he is a Venture Executive at Highland Capital Partners and serves on 

the board of several technology companies and non-profits. John served as a 

special assistant at the U.S. EPA during the Clinton Administration. He is a 

graduate of Harvard College, the University of Cambridge, and Harvard Law 

School. 


Frank Pasquale is a professor of law at Seton Hall Law School and a visiting 

fellow at the Princeton University’s Center for Information Technology Policy. 

He has a JD from Yale, was a Marshall Scholar at Oxford, and graduated from 

Harvard summa cum laude. He has been a visiting professor at Yale and 

Cardozo Law Schools. He has published widely on information law and policy. 

In 2010, he was twice invited by the National Academy of Sciences’s Committee 

on Law, Science, and Technology and its Government-University-Industry 

Roundtable to present on the privacy and security implications of data sensor 

networks. He also was invited by the Department of Health and Human 

Services’ Office of the National Coordinator for Health Information 

Technology to present at a roundtable organized to inform ONC’s 

Congressionally mandated report on privacy and security requirements for 

entities not covered by HIPAA (relating to Section 13424 of the HITECH Act). 

In 2008, he presented Internet Nondiscrimination Principles for Competition 

Policy Online before the Task Force on Competition Policy and Antitrust Laws 

of the House Committee on the Judiciary, appearing with the General Counsels 

of Google, Microsoft, and Yahoo. He is the Chair of the Privacy & Defamation 

section of the American Association of Law Schools for 2010. 

Pasquale has been quoted in the New York Times, San Francisco Chronicle, 

Los Angeles Times, Boston Globe, Financial Times, and many other 

publications. He has appeared on CNN to comment on Google’s China policy. 

He has been interviewed on internet regulation on David Levine’s Hearsay 

Culture poD.C.ast, WNYC’s Brian Lehrer Show, the Canadian 

BroaD.C.asting Corporation’s documentary “Engineering Search,” and on 

National Public Radio’s Talk of the Nation. His recent publications include 

“Beyond Innovation and Competition,” “Network Accountability for the 

Domestic Intelligence Apparatus” (with Danielle Citron), “Restoring


Transparency to Automated Authority,” and “Data and Power.” He is 

presently working on a book titled “The Black Box Society” which examines 

and critiques the rise of secret technology in the internet and finance sectors. 

Berin Szoka 

Berin Szoka is founder of TechFreedom, a non-profit think tank dedicated 

to unleashing the progress of technology that improves the human condition 

and expands individual capacity to choose. 

Previously, he was a Senior Fellow and the Director of the Center for Internet 

Freedom at The Progress & Freedom Foundation. Before joining PFF, he 

was an Associate in the Communications Practice Group at Latham & Watkins 

LLP. Before joining Latham, Szoka practiced at Lawler Metzger Milkman & 

Keeney, LLC in Washington and clerked for the Hon. H. Dale Cook, Senior 

U.S. District Judge for the Northern District of Oklahoma. 

Szoka received his Bachelor's degree in economics from Duke University and 

his juris doctor from the University of Virginia School of Law, where he served 

as Submissions Editor of the Virginia Journal of Law and Technology. He is 

admitted to practice law in the District of Columbia and California (inactive). 

He serves on the Steering Committee for the D.C. Bar's Computer & 

Telecommunications Law Section, and on the FAA's Commercial Space 

Transportation Advisory Committee (COMSTAC). Szoka has chaired, and 

currently serves on, the Board of Directors of the Space Frontier Foundation, 

a non-profit citizens' advocacy group founded in 1988 and dedicated to 

advancing commercial opportunity and expansion of human civilization in 

space. 

Paul Szynol 

Paul Szynol was born in Warsaw, Poland, and moved to the United States in 

1984, the year that New York City’s transit fare rose from 75 cents to 90 cents; 

33 previously unknown Bach pieces were found in an academic library; and 

Canon demoed its first digital still camera. He has lived in New York City, San 

Francisco, Los Angeles, New Haven, Philadelphia, New Jersey, and Warsaw, 

and, during his six drives across the U.S., visited the vast majority of the 

contiguous states. He graduated from Columbia University, where he studied 

history and philosophy, and Yale University, where he studied intellectual 

property law. He has also taken courses at the International Center of 

Photography. In the past, Paul played drums and was a computer programmer, 

and he still tinkers with Pearl drums and Java libraries. He likes dogs, 

documentary photography, music, San Francisco, Linux, and depressing movies. 

He is currently based in New York City.


Adam Thierer 

Adam Thierer is a senior research fellow at the Mercatus Center at George 

Mason University where he works with the Technology Policy Program. 

Thierer covers technology, media, Internet, and free speech policy issues with a 

particular focus in online child safety and digital privacy policy issues. 

Thierer has spent almost two decades in the public policy research community. 

He previously served as the President of The Progress & Freedom Foundation, 

the Director of Telecommunications Studies at the Cato Institute, a Senior 

Fellow at The Heritage Foundation as a Fellow in Economic Policy, and a 

researcher at the Adam Smith Institute in London. 

Thierer is the author or editor of seven books on diverse topics such as media 

regulation and child safety issues, mass media regulation, Internet governance 

and jurisdiction, intellectual property, regulation of network industries, and the 

role of federalism within high-technology markets. He earned his B.A. in 

journalism and political science at Indiana University, and received his M.A. in 

international business management and trade theory at the University of 

Maryland. 

Thierer has served on several distinguished online safety task forces, including 

Harvard Law School’s Internet Safety Technical Task Force, a “Blue Ribbon 

Working Group” on child safety organized by Common Sense Media, the 

iKeepSafe Coalition, and the National Cable & Telecommunications 

Association, and the National Telecommunications and Information 

Administration’s “Online Safety and Technology Working Group.” He is also 

an advisor to the American Legislative Exchange Council’s Telecom & IT Task 

Force. In 2008, Thierer received the Family Online Safety Institute’s “Award 

for Outstanding Achievement.” 

Hal Varian 

Hal R. Varian is the Chief Economist at Google. He started in May 2002 as a 

consultant and has been involved in many aspects of the company, including 

auction design, econometric analysis, finance, corporate strategy and public 

policy. 

He is also an emeritus professor at the University of California, Berkeley in 

three departments: business, economics, and information management. 

He received his SB degree from MIT in 1969 and his MA in mathematics and 

Ph.D. in economics from UC Berkeley in 1973. He has also taught at MIT, 

Stanford, Oxford, Michigan and other universities around the world.


Dr. Varian is a fellow of the Guggenheim Foundation, the Econometric Society, 

and the American Academy of Arts and Sciences. He was Co-Editor of the 

American Economic Review from 1987-1990 and holds honorary doctorates 

from the University of Oulu, Finland and the University of Karlsruhe, 

Germany. 

Professor Varian has published numerous papers in economic theory, industrial 

organization, financial economics, econometrics and information economics. 

He is the author of two major economics textbooks which have been translated 

into 22 languages. He is the co-author of a bestselling book on business 

strategy, INFORMATION RULES: A STRATEGIC GUIDE TO THE NETWORK 

ECONOMY and wrote a monthly column for the NEW YORK TIMES from 

2000 to 2007. 

Christopher Wolf 

Christopher Wolf is a director of Hogan Lovells’ Privacy and Information 

Management practice group. Chris is widely recognized as one of the leading 

American practitioners in the field of privacy and data security law. The 

prestigious Practising Law Institute (PLI) tapped Chris to serve as editor and 

lead author of its first-ever treatise on the subject, and to serve as co-editor of 

its guide to the FACTA Red Flags identity theft regulations. Chris recently was 

heralded for his “lifelong experience as a litigator” by Chambers U.S.A by 

ranking him as one of the nation’s top privacy lawyers. He also was asked to 

form and co-chair The Future of Privacy Forum, a think tank that focuses on 

modern privacy issues with a business practical-consumer friendly perspective, 

collaborating with industry, government, academia, and privacy advocates. 

When MSNBC labeled Chris “a pioneer in Internet law,” it was reflecting on his 

participation in many of the precedent-setting matters that form the framework 

of modern privacy law. 

Chris has deep experience in the entire range of international, federal, and state 

privacy and data security laws as well as the many sectoral and geographic 

regulations. Chris also counsels clients on compliance with self-regulatory 

regimes. 

Chris has appeared as a speaker for the International Association of Privacy 

Professionals and for the Canadian Association of Chief Privacy Officers. He 

appears annually at the PLI Institute on Privacy and Security Law. He also has 

spoken at colleges and universities including Harvard, Stanford, Berkeley, the 

University of Chicago, George Washington University, Georgetown University, 

and the Washington & Lee University School of Law. He is a frequent 

television guest on privacy and related issues, appearing on PBS, NBC, 

MSNBC, CNN, Fox News, and others.


Chris is a fourth-generation Washingtonian who started his career in 

Washington, D.C. as law clerk to The Honorable Aubrey E. Robinson, Jr., of 

the U.S. District Court for the District of Columbia. While in law school, he 

was a member of the Washington & Lee Law Review. 

Tim Wu 

Tim Wu is a professor at Columbia Law School. He teaches copyright and 

communications. 

He is the chair of media reform organization Free Press, and writes for Slate 

magazine on law, media, culture, travel, and dumplings. He has also written 

for some other publications as a pure freelancer, including the NEW YORKER, 

the NEW YORK TIMES, WASHINGTON POST WEEKEND, and FORBES. 

He is also involved in various other projects, usually related to alternative 

channels of content distribution. Many are run through the Columbia 

Program on Law & Technology. One example is Project Posner, another is 

AltLaw, and another is Keep Your Copyrights. 

His first book was WHO CONTROLS THE INTERNET with Jack Goldsmith. 

He is writing a new book on the long patterns of media centralization and 

decentralization; the publisher is Knopf / Random House. 

His topics of study are: Network neutrality, the history and structure of the 

media and communications industries (the book he is currently working on), 

international problems faced by the Internet (see WHO CONTROLS THE 

INTERNET, and copyright and innovation policy (Copyright’s 

Communications Policy). 

His brother is David Wu, author of the Xbox 360 game Full Auto, and his 

mother is Gillian Wu, a scientist. He is married to Kate Judge. His best 

friends are the Famous Five. 


Michael Zimmer, PhD, is an assistant professor in the School of Information 

Studies at the University of Wisconsin-Milwaukee, and an associate at the 

Center for Information Policy Research. 

With a background in new media and Internet studies, the philosophy of 

technology, and information policy, Zimmer studies the social, political, and 

ethical dimensions of new media and information technologies. His research 

and teaching focuses on:


� Ethics and Information Technology 

� Information Policy 

� Web Search Engines 

� Web 2.0 and Library 2.0 

� Privacy and Surveillance Theory 

� Information and Web Literacy 

� Access to Knowledge 

� Internet Research Ethics 

Zimmer received his PhD in 2007 from the Department of Media, Culture, 

and Communication at New York University under the guidance of Profs. 

Helen Nissenbaum, Alex Galloway, and Siva Vaidhyanathan. He was a 

Student Fellow at the Information Law Institute at NYU Law from 2004- 

2007, and was the Microsoft Resident Fellow at the Information Society 

Project at Yale Law School for 2007-2008. Zimmer joined UW-Milwaukee’s 

School of Information Studies in 2008. 

Zimmer earned a B.B.A. in Marketing from the University of Notre Dame in 

1994 and worked for an electronic payment processing company in Milwaukee, 

Wisconsin for several years before moving to New York City to pursue a new 

career in academia. He earned an M.A. in Media Ecology from NYU in 2002, 

and his doctoral studies were supported by the Phyllis and Gerald LeBoff 

Doctoral Fellowship in Media Ecology from the Steinhardt School of 

Education at New York University. His dissertation research was supported by 

an NSF SES Dissertation Improvement Grant. 

Zimmer has published in international journals and delivered talks across North 

America and Europe. He has been interviewed in The New York Times, on 

National Public Radio’s Morning Edition and Science Friday programs, The 

Huffington Post, MSNBC.com, GQ Magazine, The Montreal Gazette, The Boston 

Globe, MIT Technology Review, The Milwaukee Journal Sentinel, and in various 

other national and local print and radio outlets. 

Zimmer was also featured in the “Is My Cellphone Spying on Me?” 

commentary accompanying the 2-disc special edition DVD for the hit 

action/thriller movie Eagle Eye. 


Jonathan Zittrain is Professor of Law at Harvard Law School and the Harvard 

Kennedy School of Government, co-founder of the Berkman Center for 

Internet & Society, and Professor of Computer Science in the Harvard School 

of Engineering and Applied Sciences. He is a member of the Board of Trustees 

of the Internet Society and is on the board of advisors for Scientific American. 

Previously he was Professor of Internet Governance and Regulation at Oxford 

University.


His research interests include battles for control of digital property and content, 

cryptography, electronic privacy, the roles of intermediaries within Internet 

architecture, and the useful and unobtrusive deployment of technology in 

education. 

He performed the first large-scale tests of Internet filtering in China and Saudi 

Arabia in 2002, and now as part of the OpenNet Initiative he has co-edited a 

study of Internet filtering by national governments, ACCESS DENIED: THE 

PRACTICE AND POLICY OF GLOBAL INTERNET FILTERING, and its sequel, 

ACCESS CONTROLLED: THE SHAPING OF POWER, RIGHTS, AND RULE IN 

CYBERSPACE. 

His book THE FUTURE OF THE INTERNET—AND HOW TO STOP IT is available 

from Yale University Press and Penguin UK—and under a Creative Commons 

license. His papers may be found at www.jz.org. 

Ethan Zuckerman 

Ethan Zuckerman served as a fellow of the Berkman Center for Internet and 

Society at Harvard University from 2003 through 2009. Since 2009, he’s been a 

senior researcher at the center, working on projects that focus on the impact of 

technology and media on the developing world and on quantitative analysis of 

media. With Hal Roberts, he is working on comparative studies of tools for 

censorship circumvention, techniques for blocking-resistant publishing for 

human rights sites and on the Media Cloud framework for quantitative study of 

digital media. 

Ethan and Berkman fellow Rebecca MacKinnon founded Global Voices, a 

global citizen media network. Beginning at a Berkman conference in 2004, 

Global Voices has grown into an independent Netherlands-based nonprofit 

with over 200 employees and volunteers in over 100 countries. Global Voices 

maintains an international citizen media newsroom, tracks censorship 

and advocates for freedom of speech online, supports grassroots citizen 

media efforts and is a pioneer in the space of social translation. Global 

Voices’ work has been supported by the MacArthur Foundation, Ford 

Foundation, Knight Foundation, Hivos, Open Society Institute as well as 

Google, Reuters and private donors. Ethan chairs Global Voices’ global board 

of directors. 

In 2000, Ethan founded Geekcorps, a non-profit technology volunteer corps. 

Geekcorps pairs skilled volunteers from U.S. and European high tech 

companies with businesses in emerging nations for one to four month volunteer 

tours. Volunteers have served in 14 nations, including Ghana, Senegal, Mali, 

Rwanda, Armenia and Jordan, completing over a hundred projects. Geekcorps 

became a division of the International Executive Service Corps in 2001, where 

Ethan served as a vice president from 2001 to 2004.

PART I 

THE BIG PICTURE & NEW FRAMEWORKS 

47

CHAPTER 1 

THE INTERNET’S IMPACT ON 

CULTURE & SOCIETY: GOOD OR BAD? 

Why We Must Resist the Temptation of Web 2.0 51 

Andrew Keen 


Saving the Net from Its Detractors 57 

Adam Thierer 

49

50 CHAPTER 1: THE INTERNET’S IMPACT ON CULTURE & SOCIETY: GOOD OR BAD?


Why We Must Resist the 

Temptation of Web 2.0 

By Andrew Keen * 

The ancients were good at resisting seduction. Odysseus fought the seductive 

song of the Sirens by having his men tie him to the mast of his ship as it sailed 

past the Siren’s Isle. Socrates was so intent on protecting citizens from the 

seductive opinions of artists and writers, that he outlawed them from his 

imaginary republic. 

We moderns are less nimble at resisting great seductions, particularly those 

utopian visions that promise grand political or cultural salvation. From the 

French and Russian revolutions to the counter-cultural upheavals of the ‘60s 

and the digital revolution of the ‘90s, we have been seduced, time after time and 

text after text, by the vision of a political or economic utopia. 

Rather than Paris, Moscow, or Berkeley, the grand utopian movement of our 

contemporary age is headquartered in Silicon Valley, whose great seduction is 

actually a fusion of two historical movements: the counter-cultural utopianism 

of the ‘60s and the techno-economic utopianism of the ‘90s. Here in Silicon 

Valley, this seduction has announced itself to the world as the “Web 2.0” 

movement. 

On one occasion, I was treated to lunch at a fashionable Japanese restaurant in 

Palo Alto by a serial Silicon Valley entrepreneur who, back in the dot.com 

boom, had invested in my start-up Audiocafe.com. The entrepreneur, a Silicon 

Valley veteran like me, was pitching me his latest start-up: a technology 

platform that creates easy-to-use software tools for online communities to 

publish weblogs, digital movies, and music. It is technology that enables anyone 

with a computer to become an author, a film director, or a musician. This Web 

2.0 dream is Socrates’s nightmare: technology that arms every citizen with the 

means to be an opinionated artist or writer. 

“This is historic,” my friend promised me. “We are enabling Internet users to 

author their own content. Think of it as empowering citizen media. We can 

help smash the elitism of the Hollywood studios and the big record labels. Our 

technology platform will radically democratize culture, build authentic 

community, and create citizen media.” Welcome to Web 2.0. 

* Andrew Keen is a veteran Silicon Valley entrepreneur and digital media critic. He blogs at 

TheGreatSeduction.com and has recently launched AfterTV, a podcast chat show about 

media, culture, and technology. He is the author of THE CULT OF THE AMATEUR: HOW 

TODAY’S INTERNET IS KILLING OUR CULTURE (Crown 2007).

52 CHAPTER 1: THE INTERNET’S IMPACT ON CULTURE & SOCIETY: GOOD OR BAD? 

Buzzwords from the old dot.com era—like “cool,” “eyeballs,” or “burn-rate”— 

have been replaced in Web 2.0 by language which is simultaneously more 

militant and absurd: empowering citizen media, radically democratize, smash 

elitism, content redistribution, authentic community. This sociological jargon, 

once the preserve of the hippie counterculture, has now become the lexicon of 

new media capitalism. 

Yet this entrepreneur owns a $4 million house a few blocks from Steve Jobs’s 

house. He vacations in the South Pacific. His children attend the most 

exclusive private academy on the peninsula. But for all of this he sounds more 

like a cultural Marxist—a disciple of Antonio Gramsci or Herbert Marcuse— 

than a capitalist with an MBA from Stanford. 

In his mind, “big media”—the Hollywood studios, the major record labels and 

international publishing houses—really did represent the enemy. The promised 

land was user-generated online content. In Marxist terms, the traditional media 

had become the exploitative “bourgeoisie,” and citizen media, those heroic 

bloggers and podcasters, were the “proletariat.” 

This outlook is typical of the Web 2.0 movement, which fuses ‘60s radicalism 

with the utopian eschatology of digital technology. The ideological outcome 

may be trouble for all of us. 

So what, exactly, is the Web 2.0 movement? As an ideology, it is based upon a 

series of ethical assumptions about media, culture, and technology. It worships 

the creative amateur: the self-taught filmmaker, the dorm-room musician, the 

unpublished writer. It suggests that everyone—even the most poorly educated 

and inarticulate amongst us—can and should use digital media to express and 

realize themselves. Web 2.0 “empowers” our creativity, it “democratizes” 

media, it “levels the playing field” between experts and amateurs. The enemy of 

Web 2.0 is “elitist” traditional media. 

Empowered by Web 2.0 technology, we can all become citizen journalists, 

citizen videographers, or citizen musicians. Empowered by this technology, we 

will be able to write in the morning, direct movies in the afternoon, and make 

music in the evening. 

Sounds familiar? It’s eerily similar to Marx’s seductive promise about individual 

self-realization in his German Ideology: 

Whereas in communist society, where nobody has one 

exclusive sphere of activity but each can become accomplished 

in any branch he wishes, society regulates the general 

production and thus makes it possible for me to do one thing 

today and another tomorrow, to hunt in the morning, fish in 

the afternoon, rear cattle in the evening, criticise after dinner,


just as I have a mind, without ever becoming hunter, 

fisherman, shepherd or critic. 1 

Just as Marx seduced a generation of European idealists with his fantasy of selfrealization 

in a communist utopia, so the Web 2.0 cult of creative self-realization 

has seduced everyone in Silicon Valley. The movement bridges counter-cultural 

radicals of the ‘60s such as Steve Jobs with the contemporary geek culture of 

Google’s Larry Page. Between the book-ends of Jobs and Page lies the rest of 

Silicon Valley including radical communitarians like Craig Newmark (of 

Craigslist.com), intellectual property communists such as Stanford Law 

Professor Larry Lessig, economic cornucopians like Wired magazine editor Chris 

“Long Tail” Anderson, journalism professor Jeff Jarvis, and new media moguls 

Tim O’Reilly and John Battelle. 

The ideology of the Web 2.0 movement was perfectly summarized at the 

Technology Education and Design (TED) show in Monterey in 2005 when 

Kevin Kelly, Silicon Valley’s über-idealist and author of the Web 1.0 Internet 

utopia Ten Rules for The New Economy, said: 

Imagine Mozart before the technology of the piano. Imagine 

Van Gogh before the technology of affordable oil paints. 

Imagine Hitchcock before the technology of film. We have a 

moral obligation to develop technology. 2 

But where Kelly sees a moral obligation to develop technology, we should actually 

have—if we really care about Mozart, Van Gogh and Hitchcock—a moral 

obligation to question the development of technology. 

The consequences of Web 2.0 are inherently dangerous for the vitality of 

culture and the arts. Its empowering promises play upon that legacy of the 

‘60s—the creeping narcissism that Christopher Lasch described so presciently, 

with its obsessive focus on the realization of the self. 3 

Another word for narcissism is “personalization.” Web 2.0 technology 

personalizes culture so that it reflects ourselves rather than the world around us. 

Blogs personalize media content so that all we read are our own thoughts. 

1 KARL MARX & FRIEDRICH ENGELS, THE GERMAN IDEOLOGY (1845), text available at 

Marxist Internet Archive, 

http://www.marxists.org/archive/marx/works/1845/german-ideology/ch01a.htm. 

2 See Dan Frost, Meeting of Minds in Monterey, SAN FRANCISCO CHRONICLE, Feb. 27, 2005, 

http://articles.sfgate.com/2005-02-27/business/17361312_1_digital-world-edwardburtynsky-robert-fischell/2 

(quoting Kevin Kelly). 

3 See CHRISTOPHER LASCH, THE CULTURE OF NARCISSISM: AMERICAN LIFE IN AN AGE OF 

DIMINISHING EXPECTATIONS (1978).


Online stores personalize our preferences, thus feeding back to us our own 

taste. Google personalizes searches so that all we see are advertisements for 

products and services we already use. 

Instead of Mozart, Van Gogh, or Hitchcock, all we get with the Web 2.0 

revolution is more of ourselves. 

Still, the idea of inevitable technological progress has become so seductive that 

it has been transformed into “laws.” In Silicon Valley, the most quoted of these 

laws, Moore’s Law, states that the number of transistors on a chip doubles every 

two years, thus doubling the memory capacity of the personal computer every 

two years. On one level, of course, Moore’s Law is real and it has driven the 

Silicon Valley economy. But there is an unspoken ethical dimension to Moore’s 

Law. It presumes that each advance in technology is accompanied by an 

equivalent improvement in the condition of man. 

But as Max Weber so convincingly demonstrated, the only really reliable law of 

history is the Law of Unintended Consequences. 

We know what happened the first time around, in the dot.com boom of the 

‘90s. At first there was irrational exuberance. Then the dot.com bubble 

popped; some people lost a lot of money and a lot of people lost some money. 

But nothing really changed. Big media remained big media and almost 

everything else—with the exception of Amazon.com and eBay—withered away. 

This time, however, the consequences of the digital media revolution are much 

more profound. Apple, Google and Craigslist really are revolutionizing our 

cultural habits, our ways of entertaining ourselves, our ways of defining who we 

are. Traditional “elitist” media is being destroyed by digital technologies. 

Newspapers are in free-fall. Network television, the modern equivalent of the 

dinosaur, is being shaken by TiVo’s overnight annihilation of the 30-second 

commercial and competition from Internet-delivered television and amateur 

video. The iPod is undermining the multibillion dollar music industry. 

Meanwhile, digital piracy, enabled by Silicon Valley hardware and justified by 

intellectual property communists such as Larry Lessig, is draining revenue from 

established artists, movie studios, newspapers, record labels, and song writers. 

Is this a bad thing? The purpose of our media and culture industries—beyond 

the obvious need to make money and entertain people—is to discover, nurture, 

and reward elite talent. Our traditional mainstream media has done this with 

great success over the last century. Consider Alfred Hitchcock’s masterpiece, 

Vertigo and a couple of other brilliantly talented works of the same name: the 

1999 book by Anglo-German writer W.G. Sebald, and the 2004 song by Irish 

rock star Bono. Hitchcock could never have made his expensive, complex 

movies outside the Hollywood studio system. Bono would never have become 

Bono without the music industry’s super-heavyweight marketing muscle. And


W.G. Sebald, the most obscure of this trinity of talent, would have remained an 

unknown university professor, had a high-end publishing house not had the 

good taste to discover and distribute his work. Elite artists and an elite media 

industry are symbiotic. If you democratize media, then you end up 

democratizing talent. The unintended consequence of all this democratization, 

to misquote Web 2.0 apologist Thomas Friedman, is cultural “flattening.” 4 No 

more Hitchcocks, Bonos, or Sebalds. Just the flat noise of opinion—Socrates’s 

nightmare. 

While Socrates correctly gave warning about the dangers of a society infatuated 

by opinion in Plato’s Republic, more modern dystopian writers—Huxley, 

Bradbury, and Orwell—got the Web 2.0 future exactly wrong. Much has been 

made, for example, of the associations between the all-seeing, all-knowing 

qualities of Google’s search engine and the Big Brother in Nineteen Eighty- 

Four. 5 But Orwell’s fear was the disappearance of the individual right to selfexpression. 

Thus Winston Smith’s great act of rebellion in Nineteen Eight- 

Four was his decision to pick up a rusty pen and express his own thoughts: 

The thing that he was about to do was open a diary. This was 

not illegal, but if detected it was reasonably certain that it 

would be punished by death… Winston fitted a nib into the 

penholder and sucked it to get the grease off…. He dipped the 

pen into the ink and then faltered for just a second. A tremor 

had gone through his bowels. To mark the paper was the 

decisive act. 6 

In the Web 2.0 world, however, the nightmare is not the scarcity, but the overabundance 

of authors. Since everyone will use digital media to express 

themselves, the only decisive act will be to not mark the paper. Not writing as 

rebellion sounds bizarre—like a piece of fiction authored by Franz Kafka. But 

one of the unintended consequences of the Web 2.0 future may well be that 

everyone is an author, while there is no longer any audience. 

Speaking of Kafka, on the back cover of the January 2006 issue of Poets & 

Writers magazine, there is a seductive Web 2.0 style advertisement which reads: 

Kafka toiled in obscurity and died penniless. If only he’d had a 

website … . 

4 See THOMAS FRIEDMAN, THE WORLD IS FLAT: A BRIEF HISTORY OF THE TWENTY-FIRST 

CENTURY (2005). 

5 See GEORGE ORWELL, NINETEEN EIGHTY-FOUR (1949). 

6 Id. at 6.


Presumably, if Kafka had had a website, it would be located at kafka.com— 

which is today an address owned by a mad left-wing blog called The Biscuit 

Report. The front page of this site quotes some words written by Kafka in his 

diary: 

I have no memory for things I have learned, nor things I have 

read, nor things experienced or heard, neither for people nor 

events; I feel that I have experienced nothing, learned nothing, 

that I actually know less than the average schoolboy, and that 

what I do know is superficial, and that every second question is 

beyond me. I am incapable of thinking deliberately; my 

thoughts run into a wall. I can grasp the essence of things in 

isolation, but I am quite incapable of coherent, unbroken 

thinking. I can’t even tell a story properly; in fact, I can 

scarcely talk … 7 

One of the unintended consequences of the Web 2.0 movement may well be 

that we fall, collectively, into the amnesia that Kafka describes. Without an elite 

mainstream media, we will lose our memory for things learnt, read, experienced, 

or heard. The cultural consequences of this are dire, requiring the authoritative 

voice of at least an Allan Bloom, 8 if not an Oswald Spengler. 9 But here in 

Silicon Valley, on the brink of the Web 2.0 epoch, there no longer are any 

Blooms or Spenglers. All we have is the great seduction of citizen media, 

democratized content and authentic online communities. And blogs, of course. 

Millions and millions of blogs. 

7 See The Biscuit Report, 

http://web.archive.org/web/20080225015716/http://www.kafka.com/. 

8 See ALLAN BLOOM, THE CLOSING OF THE AMERICAN MIND (1987). 

9 See OSWALD SPENGLER, THE DECLINE OF THE WEST (1918).


The Case for Internet Optimism, 

Part 1: Saving the Net 

from Its Detractors 

By Adam Thierer * 

Introduction: Two Schools 

of Internet Pessimism 

Surveying the prevailing mood surrounding cyberlaw and Internet policy circa 

2010, one is struck by the overwhelming sense of pessimism regarding the longterm 

prospects for a better future. “Internet pessimism,” however, comes in 

two very distinct flavors: 

1. Net Skeptics, Pessimistic about the Internet Improving the Lot of 

Mankind: The first variant of Internet pessimism is rooted in general 

skepticism about the supposed benefits of cyberspace, digital technologies, 

and information abundance. The proponents of this pessimistic view often 

wax nostalgic about some supposed “good ‘ol days” when life was much 

better (although they can’t seem to agree when those were). At a minimum, 

they want us to slow down and think twice about life in the Information 

Age and how it’s personally affecting each of us. Occasionally, however, 

this pessimism borders on neo-Ludditism, with some proponents 

recommending steps to curtail what they feel is the destructive impact of 

the Net or digital technologies on culture or the economy. Leading 

proponents of this variant of Internet pessimism include: Neil Postman 

(Technopoly: The Surrender of Culture to Technology), Andrew Keen, (The Cult of 

the Amateur: How Today’s Internet is Killing our Culture), Lee Siegel, (Against the 

Machine: Being Human in the Age of the Electronic Mob), Mark Helprin, (Digital 

Barbarism) and, to a lesser degree, Jaron Lanier (You Are Not a Gadget) and 

Nicholas Carr (The Big Switch and The Shallows). 

2. Net Lovers, Pessimistic about the Future of Openness: A different 

type of Internet pessimism is on display in the work of many leading 

cyberlaw scholars today. Noted academics such as Lawrence Lessig, (Code 

and Other Laws of Cyberspace), Jonathan Zittrain (The Future of the Internet— 

And How to Stop It), and Tim Wu (The Master Switch: The Rise and Fall of 

Information Empires), embrace the Internet and digital technologies, but argue 

that they are “dying” due to a lack of sufficient care or collective oversight. 

* Adam Thierer is a senior research fellow at the Mercatus Center at George Mason 

University where he works with the Technology Policy Program.


In particular, they fear that the “open” Internet and “generative” digital 

systems are giving way to closed, proprietary systems, typically run by 

villainous corporations out to erect walled gardens and quash our digital 

liberties. Thus, they are pessimistic about the long-term survival of the 

Internet that we currently know and love. 

Despite their different concerns, two things unite these two schools of technopessimism. 

First, there is an elitist air to their pronouncements; a veritable “the 

rest of you just don’t get it” attitude pervades much of their work. In the case 

of the Net skeptics, it’s the supposed decline of culture, tradition, and economy 

that the rest of us are supposedly blind to, but which they see perfectly—and 

know how to rectify. For the Net Lovers, by contrast, we see this attitude on 

display when they imply that a Digital Dark Age of Closed Systems is unfolding 

since nefarious schemers in high-tech corporate America are out to suffocate 

Internet innovation and digital freedom more generally. The Net Lovers 

apparently see this plot unfolding, but paint the rest of us out to be robotic 

sheep being led to the cyber-slaughter: We are unwittingly using services (AOL 

in the old days; Facebook today) or devices (the iPhone and iPad) that play right 

into the hands of the very corporate schemers determined to trap us in high and 

tight walled gardens. 

Unsurprisingly, this elitist attitude leads to the second belief uniting these two 

variants of Net pessimism: Someone or something must intervene to set us on a 

better course or protect those things that they regard as sacred. The critics 

either fancy themselves as the philosopher kings who can set things back on a 

better course, or imagine that such creatures exist in government today and can 

be tapped to save us from our impending digital doom—whatever it may be. 

Dynamism vs. the Stasis Mentality 

In both cases, these two schools of Internet pessimism have (a) over-stated the 

severity of the respective problems they’ve identified and (b) failed to appreciate 

the benefits of evolutionary dynamism. I borrow the term “dynamism” from 

Virginia Postrel, who contrasted the conflicting worldviews of dynamism and 

stasis so eloquently in her 1998 book, The Future and Its Enemies. Postrel argued 

that: 

The future we face at the dawn of the twenty-first century is, 

like all futures left to themselves, “emergent, complex 

messiness.” Its “messiness” lies not in disorder, but in an order 

that is unpredictable, spontaneous, and ever shifting, a pattern 

created by millions of uncoordinated, independent decisions. 1 

1 VIRGINIA POSTREL, THE FUTURE AND ITS ENEMIES, at xv (1998).


“[T]hese actions shape a future no one can see, a future that is dynamic and 

inherently unstable,” Postrel noted. 2 But that inherent instability and the 

uncomfortable realization that the future is, by its very nature, unknowable, 

leads to exactly the sort of anxieties we see on display in the works of both 

varieties of Internet pessimists today. Postrel made the case for embracing 

dynamism as follows: 

How we feel about the evolving future tells us who we are as 

individuals and as a civilization: Do we search for stasis—a 

regulated, engineered world? Or do we embrace dynamism—a 

world of constant creation, discovery, and competition? Do we 

value stability and control, or evolution and learning? Do we 

declare with [Tim] Appelo that “we’re scared of the future” 

and join [Judith] Adams in decrying technology as “a killing 

thing”? Or do we see technology as an expression of human 

creativity and the future as inviting? Do we think that progress 

requires a central blueprint, or do we see it as a decentralized, 

evolutionary process? Do we consider mistakes permanent 

disasters, or the correctable by-products of experimentation? 

Do we crave predictability, or relish surprise? These two poles, 

stasis and dynamism, increasingly define our political, 

intellectual, and cultural landscape. The central question of our 

time is what to do about the future. And that question creates 

a deep divide. 3 

Indeed it does, and that divide is growing deeper as the two schools of Internet 

pessimism—unwittingly, of course—work together to concoct a lugubrious 

narrative of impending techno-apocalypse. It makes little difference whether 

the two schools disagree on the root cause(s) of all our problems; in the end, it’s 

their common call for a more “regulated, engineered world” that makes them 

both embrace the same stasis mindset. Again, the air of elitism rears its ugly 

head, Postrel notes: 

2 Id. 

3 Id. at xiv. 

Stasist social criticism… brings up the specifics of life only to 

sneer at or bash them. Critics assume that readers will share 

their attitudes and will see contemporary life as a problem 

demanding immediate action by the powerful and wise. This 

relentlessly hostile view of how we live, and how we may come 

to live, is distorted and dangerous. It overvalues the tastes of 

an articulate elite, compares the real world of trade-offs to 

fantasies of utopia, omits important details and connections,


and confuses temporary growing pains with permanent 

catastrophes. It demoralizes and devalues the creative minds 

on whom our future depends. And it encourages the coercive 

use of political power to wipe out choice, forbid 

experimentation, short-circuit feedback, and trammel 

progress. 4 

In this essay, I focus on the first variant of Internet pessimism (the Net 

skeptics) and discuss their clash with Internet optimists. I form this narrative 

using the words and themes developed in various books published by Net 

optimists and pessimists in recent years. I make the dynamist case for what I 

call “pragmatic optimism” in that I argue that the Internet and digital 

technologies are reshaping our culture, economy and society—in most ways for 

the better (as the optimists argue), but not without some serious heartburn 

along the way (as the pessimists claim). My bottom line comes down to a 

simple cost-benefit calculus: Were we really better off in the scarcity era when we were 

collectively suffering from information poverty? Generally speaking, I’ll take information 

abundance over information poverty any day! But we should not underestimate 

or belittle the disruptive impacts associated with the Information Revolution. 

We need to find ways to better cope with turbulent change in a dynamist 

fashion instead of embracing the stasis notion that we can roll back the clock on 

progress or recapture “the good ‘ol days”—which actually weren’t all that good. 

In another essay in this book, I address the second variant of Internet 

pessimism (the Net lovers) and argue that reports of the Internet’s death have 

been greatly exaggerated. Although the Net lovers will likely recoil at the 

suggestion that they are not dynamists, closer examination reveals their attitudes 

and recommendations to be deeply stasist. They fret about a cyber-future in 

which the Internet might not as closely resemble its opening epoch. Worse yet, 

many of them agree with what Lawrence Lessig said in his seminal—by highly 

pessimistic—1999 book, Code and Other Laws of Cyberspace, that “we have every 

reason to believe that cyberspace, left to itself, will not fulfill the promise of 

freedom. Left to itself, cyberspace will become a perfect tool of control.” 5 

Lessig and his intellectual disciples—especially Zittrain and Wu—have 

continued to forecast a gloomy digital future unless something is done to address 

the Great Digital Closing we are supposedly experiencing. I will argue that, 

while many of us share their appreciation of the Internet’s current nature and its 

early history, their embrace of the stasis mentality is unfortunate since it 

forecloses the spontaneous evolution of cyberspace and invites government 

4 Id. at xvii-xviii. 

5 LAWRENCE LESSIG, CODE AND OTHER LAWS OF CYBERSPACE 5-6 (1999).


But first let us turn to the Net skeptics, who don’t share such an appreciation of 

the potential benefits of cyberspace. Rather, their pessimism cuts deep and is 

rooted in overt hostility to all things digital. 

The Familiar Cycle of 

Technological Revolutions 

The impact of technological change on culture, learning, and morality has long 

been the subject of intense debate, and every technological revolution brings 

out a fresh crop of both pessimists and Pollyannas. Indeed, a familiar cycle has 

repeat itself throughout history whenever new modes of production (from 

mechanized agriculture to assembly-line production), means of transportation 

(water, rail, road, or air), energy production processes (steam, electric, nuclear), 

medical breakthroughs (vaccination, surgery, cloning), or communications 

techniques (telegraph, telephone, radio, television) have emerged. 

The cycle goes something like this: A new technology appears. Those who fear 

the sweeping changes brought about by this technology see a sky that is about 

to fall. These “techno-pessimists” predict the death of the old order (which, 

ironically, is often a previous generation’s hotly-debated technology that others 

wanted slowed or stopped). Embracing this new technology, they fear, will 

result in the overthrow of traditions, beliefs, values, institutions, business 

models, and much else they hold sacred. As Dennis Baron, author of A Better 

Pencil, has noted, “the shock of the new often brings out critics eager to warn us 

away.” 6 

The Pollyannas, by contrast, look out at the unfolding landscape and see mostly 

rainbows in the air. Theirs is a rose-colored world in which the technological 

revolution du jour improves the general lot of mankind. If something must give, 

then the old ways be damned! For such “techno-optimists,” progress means 

some norms and institutions must adapt—perhaps even disappear—for society 

to continue its march forward. 

Our current Information Revolution is no different. It too has its share of 

techno-pessimists and techno-optimists who continue to debate the impact of 

technology on human existence. 7 Indeed, before most of us had even heard of 

6 DENNIS BARON, A BETTER PENCIL 12 (2009). 

7 William Powers, author of Hamlet’s BlackBerry: A Practical Philosophy for Building a Good 

Life in the Digital Age, reminds us that: 

whenever new devices have emerged, they’ve presented the kinds of 

challenges we face today—busyness, information overload, that sense of life 

being out of control. These challenges were as real two millennia ago as they 

are today, and throughout history, people have been grappling with them and 

looking for creative ways to manage life in the crowd.


the Internet, people were already fighting about it—or at least debating what the 

rise of the Information Age meant for our culture, society, and economy. 

Web 1.0 Fight: Postman vs. Negroponte 

In his 1992 anti-technology manifesto Technopoly: The Surrender of Culture to 

Technology, the late social critic Neil Postman greeted the unfolding Information 

Age with a combination of skepticism and scorn. 8 Indeed, Postman’s book was 

a near-perfect articulation of the techno-pessimist’s creed. ”Information has 

become a form of garbage,” he claimed, “not only incapable of answering the 

most fundamental human questions but barely useful in providing coherent 

direction to the solution of even mundane problems.” 9 If left unchecked, 

Postman argued, America’s new technopoly—”the submission of all forms of 

cultural life to the sovereignty of technique and technology”—would destroy 

“the vital sources of our humanity” and lead to “a culture without a moral 

foundation” by undermining “certain mental processes and social relations that 

make human life worth living.” 10 

Postman opened his polemic with the well-known allegorical tale found in 

Plato’s Phaedrus about the dangers of the written word. Postman reminded us 

how King Thamus responded to the god Theuth, who boasted that his 

invention of writing would improve the wisdom and memory of the masses 

relative to the oral tradition of learning. King Thamus shot back, “the 

discoverer of an art is not the best judge of the good or harm which will accrue 

to those who practice it.” King Thamus then passed judgment himself about 

the impact of writing on society, saying he feared that the people “will receive a 

quantity of information without proper instruction, and in consequence be 

thought very knowledgeable when they are for the most part quite ignorant.” 

And so Postman—fancying himself a modern Thamus—cast judgment on 

today’s comparable technological advances and those who would glorify them: 

being out of control. These challenges were as real two millennia ago as they 

are today, and throughout history, people have been grappling with them and 

looking for creative ways to manage life in the crowd. 

WILLIAM POWERS, HAMLET’S BLACKBERRY: A PRACTICAL PHILOSOPHY FOR BUILDING A 

GOOD LIFE IN THE DIGITAL AGE 5 (2010). Similarly, Baron notes that “from the first days 

of writing to the present, each time a new communication technology appeared, people had 

to learn all over again how to use it, how to respond to it, how to trust the documents it 

produced.” DENNIS BARON, A BETTER PENCIL 5 (2009). 

8 NEIL POSTMAN, TECHNOPOLY: THE SURRENDER OF CULTURE TO TECHNOLOGY (1992). 

9 Id. at 69-70. 

10 Id. at 52, xii.


we are currently surrounded by throngs of zealous Theuths, 

one-eyed prophets who see only what new technologies can do 

and are incapable of imagining what they will undo. We might 

call such people Technophiles. They gaze on technology as a 

lover does on his beloved, seeing it as without blemish and 

entertaining no apprehension for the future. They are therefore 

dangerous and to be approached cautiously. … If one is to err, 

it is better to err on the side of Thamusian skepticism. 11 

Nicholas Negroponte begged to differ. An unapologetic Theuthian technophile, 

the former director of the MIT Media Lab responded on behalf of the technooptimists 

in 1995 with his prescient polemic, Being Digital. 12 It was a paean to 

the Information Age, for which he served as one of the first high prophets— 

with Wired magazine’s back page serving as his pulpit during the many years he 

served as a regular columnist. 

Appropriately enough, the epilogue of Negroponte’s Being Digital was entitled 

“An Age of Optimism” and, like the rest of the book, it stood in stark contrast 

to Postman’s pessimistic worldview. Although Negroponte conceded that 

technology indeed had a “dark side” in that it could destroy much of the old 

order, he believed that destruction was both inevitable and not cause for much 

concern. “Like a force of nature, the digital age cannot be denied or stopped,” 

he insisted, and we must learn to appreciate the ways “digital technology can be 

a natural force drawing people into greater world harmony.” 13 (This sort of 

techno-determinism is a theme found in many of the Internet optimist works 

that followed Negroponte.) 

To Postman’s persistent claim that America’s technopoly lacked a moral 

compass, Negroponte again conceded the point but took the glass-is-half-full 

view: “Computers are not moral; they cannot resolve complex issues like the 

rights to life and to death. But being digital, nevertheless, does give much cause 

for optimism.” 14 His defense of the digital age rested on the “four very 

powerful qualities that will result in its ultimate triumph: decentralizing, 

globalizing, harmonizing, and empowering.” 15 Gazing into his techno-crystal 

ball in 1995, Negroponte forecast the ways in which those qualities would 

revolutionize society: 

11 Id. at 5. 

12 NICHOLAS NEGROPONTE, BEING DIGITAL (1995). 

13 Id. at 229, 230. 

14 Id. at 228-9. 

15 Id. at 229.


The access, the mobility, and the ability to effect change are 

what will make the future so different from the present. The 

information superhighway may be mostly hype today, but it is 

an understatement about tomorrow. It will exist beyond 

people’s wildest predictions. As children appropriate a global 

information resource, and as they discover that only adults 

need learner’s permits, we are bound to find new hope and 

dignity in places where very little existed before. 16 

In many ways, that’s the world we occupy today: one of unprecedented media 

abundance and unlimited communications and connectivity opportunities. 

But the great debate about the impact of digitization and information 

abundance did not end with Postman and Negroponte. Theirs was but Act I in 

a drama that continues to unfold, and grows more heated and complex with 

each new character on the stage. “This conflict between stability and progress, 

security and prosperity, dynamism and stasis, has led to the creation of a major 

political fault line in American politics,” argues Robert D. Atkinson: “On one 

side are those who welcome the future and look at the New Economy as largely 

positive. On the other are those who resist change and see only the risks of new 

technologies and the New Economy.” 17 Atkinson expands on this theme in 

another essay in this collection. 18 

Web War II 

The disciples of Postman and Negroponte are a colorful, diverse lot. The 

players in Act II of this drama occupy many diverse professions: journalists, 

technologists, business consultants, sociologists, economists, lawyers, etc. The 

two camps disagree with each other even more vehemently and vociferously 

about the impact of the Internet and digital technologies than Postman and 

Negroponte did. 

In Exhibit 1, I have listed the Internet optimists and pessimists alongside their 

key works. This very binary treatment obviously cannot do justice to the 

varying shades of optimism or pessimism in in each, but is nonetheless helpful. 

16 Id. at 231. 

17 ROBERT D. ATKINSON, THE PAST AND FUTURE OF AMERICA’S ECONOMY 201 (2004). “As a 

result,” he says, “a political divide is emerging between preservationists who want to hold 

onto the past and modernizers who recognize that new times require new means.” 

18 Robert D. Atkinson, Who’s Who in Internet Politics: A Taxonomy of Information Technology Policy 

& Politics, infra at 162.


Theuthian Technophiles 

( “The Internet Optimists”) 

Nicholas Negroponte, Being Digital 

(1995) 

Exhibit 1 

Kevin Kelly, Out of Control: The New 

Biology of Machines, Social Systems, 

and the Economic World (1995) 

Virginia Postrel, The Future and 

Its Enemies (1998) 

James Surowiecki, The Wisdom of 

Crowds (2004) 

Chris Anderson, The Long Tail: Why the 

Future of Business is Selling Less of 

More (2006) 

Steven Johnson, Everything Bad is Good 

For You (2006) 

Glenn Reynolds, An Army of Davids: 

How Markets and Technology 

Empower Ordinary People to Beat Big 

Media, Big Government, and Other 

Goliaths (2006) 

Yochai Benkler, The Wealth of 

Networks: How Social Production 

Transforms Markets and Freedom 

(2006) 

Clay Shirky, Here Comes Everybody: 

The Power of Organizing without 

Organizations (2008) 

Don Tapscott & Anthony D. Williams, 

Wikinomics: How Mass Collaboration 

Changes Everything (2008) 

Thamusian Technophobes 

( “The Internet Pessimists”) 

Neil Postman, Technopoly: The 

Surrender of Culture to 

Technology (1993) 

Sven Birkerts, The Gutenberg 

Elegies: The Fate of Reading 

in an Electronic Age (1994) 

Clifford Stoll, High-Tech 

Heretic: Reflections of a 

Computer Contrarian (1999) 

Cass Sunstein, Republic.com 

(2001) 

Todd Gitlin, Media Unlimited: 

How the Torment of Images 

and Sounds Overwhelms Our 

Lives (2002) 

Todd Oppenheimer, The 

Flickering Mind: Saving 

Education from the False 

Promise of Technology (2003) 

Andrew Keen, The Cult of the 

Amateur: How Today’s 

Internet is Killing our Culture 

(2007) 

Steve Talbott, Devices of the 

Soul: Battling for Our Selves in 

an Age of Machines (2007) 

Nick Carr, The Big Switch: 

Rewiring the World, from 

Edison to Google (2008)


Theuthian Technophiles 

( “The Internet Optimists”) 

Jeff Howe, Crowdsourcing: Why 

the Power of the Crowd Is 

Driving the Future of Business 

(2008) 

Tyler Cowen, Create Your Own 

Economy: The Path to 

Prosperity in a Disordered World 

(2009) 

Dennis Baron, A Better Pencil: 

Readers, Writers, and the Digital 

Revolution (2009) 

Jeff Jarvis, What Would Google 

Do? (2009) 

Clay Shirky, Cognitive Surplus: 

Creativity and Generosity in a 

Connected Age (2010) 

Nick Bilton, I Live in the Future 

& Here’s How It Works (2010) 

Kevin Kelly, What Technology 

Wants (2010) 

Exhibit 1 Continued 

Thamusian Technophobes 

( “The Internet Pessimists”) 

Lee Siegel, Against the Machine: Being 

Human in the Age of the Electronic 

Mob (2008) 

Mark Bauerlein, The Dumbest 

Generation: How the Digital Age 

Stupefies Young Americans and 

Jeopardizes Our Future (2008) 

Mark Helprin, Digital Barbarism: A 

Writer’s Manifesto (2009) 

Maggie Jackson, Distracted: The 

Erosion of Attention and the Coming 

Dark Age (2009) 

John Freeman, The Tyranny of E-Mail: 

The Four-Thousand-Year Journey to 

Your Inbox (2009) 

Jaron Lanier, You Are Not a Gadget 

(2010) 

Nick Carr, The Shallows: What the 

Internet Is Doing to Our Brains (2010) 

William Powers, Hamlet’s BlackBerry: 

A Practical Philosophy for Building a 

Good Life in the Digital Age (2010) 

In Exhibit 2, I have sketched out the major lines of disagreement between these 

two camps and divided those disagreements into (1) Cultural / Social beliefs 

vs. (2) Economic / Business beliefs.


Exhibit 2 

Optimists Pessimists 

Cultural / Social beliefs 

Net is participatory Net is polarizing 

Net facilitates personalization 

(welcome of “Daily Me” 

that digital tech allows) 

Net facilitates fragmentation 

(fear of the “Daily Me”) 

“a global village” balkanization and 

fears of “mob rule” 

heterogeneity / encourages diversity 

of thought and expression 

homogeneity / Net 

leads to close-mindedness 

allows self-actualization diminishes personhood 

Net a tool of liberation 

& empowerment 

Net a tool of frequent 

misuse & abuse 

Net can help educate the masses dumbs down the masses 

anonymous communication 

encourages vibrant debate + 

whistleblowing (a net good) 

welcome information abundance; 

believe it will create new 

opportunities for learning 

benefits of “Free” (increasing 

importance of “gift economy”) 

mass collaboration is 

generally more important 

Economic / Business beliefs 

anonymity debases culture & 

leads to lack of accountability 

concern about information overload; 

esp. impact on learning & reading 

costs of “Free” (“free” = threat to 

quality & business models) 

individual effort is 

generally more important 

embrace of “amateur” creativity superiority of “professionalism” 

stress importance of “open 

systems” of production 

“wiki” model = wisdom of crowds; 

benefits of crowdsourcing 

stress importance of “proprietary” 

models of production 

“wiki” model = stupidity of crowds; 

collective intelligence is oxymoron; + 

“sharecropper” concern about 

exploitation of free labor


When you boil it all down, there are two major points of contention between 

the Internet optimists and pessimists: 

1. The impact of technology on learning & culture and the role of experts 

vs. amateurs in that process. 

2. The promise—or perils—of personalization, for both individuals and 

society. 

Each dispute is discussed in more detail below. 

Differences Over Learning, 

Culture & “Truth” 

As with Theuth and Thamus, today’s optimists and skeptics differ about who is 

the best judge of what constitutes progress, authority, and “truth” and how 

technological change will impact these things. 

The Pessimists’ Critique 

Consider the heated debates over the role of “amateur” creations, usergeneration 

content, and peer-based forms of production. Pessimists tend to 

fear the impact of the Net and the rise of what Andrew Keen has called “the 

cult of the amateur.” 19 They worry that “professional” media or more 

enlightened voices and viewpoints might be drowned out by a cacophony of 

competing—but less compelling or enlightened—voices and viewpoints. 

Without “enforceable scarcity” and protection for the “enlightened class,” the 

pessimists wonder how “high quality” news or “high art” will be funded and 

disseminated. Some, like Keen, even suggest the need to “re-create media 

scarcity” to save culture or professional content creators. 20 

Some of these pessimists clearly think in zero-sum terms: More “amateur” 

production seems to mean less “professional” content creation will be possible. 

For example, Lee Siegel, author of Against the Machine: Being Human in the Age of 

the Electronic Mob, says that by empowering the masses to have more of a voice, 

“unbiased, rational, intelligent, and comprehensive news … will become less 

19 ANDREW KEEN, THE CULT OF THE AMATEUR: HOW TODAY’S INTERNET IS KILLING OUR 

CULTURE (2007). 

20 Andrew Keen, Art & Commerce: Death by YouTube, ADWEEK, Oct. 15, 2007, 

http://web.archive.org/web/20080107024552/http:/www.adweek.com/aw/magazin 

e/article_display.jsp?vnu_content_id=1003658204. For a response, see Adam Thierer, 

Thoughts on Andrew Keen, Part 2: The Dangers of the Stasis Mentality, TECHNOLOGY LIBERATION 

FRONT, Oct. 18, 2007, http://techliberation.com/2007/10/18/thoughts-on-andrewkeen-part-2-the-dangers-of-the-stasis-mentality.


and less available.” 21 “[G]iving everyone a voice,” he argues, “can also be a way 

to keep the most creative, intelligent, and original voices from being heard.” 22 

The centrality of Wikipedia, the collaborative online encyclopedia, to this 

discussion serves as a microcosm of the broader debate between the optimists 

and the pessimists. Almost every major optimist and pessimist tract includes a 

discussion of Wikipedia; it generally serves as a hero in the works of the former 

and a villain in the latter. For the pessimists, Wikipedia marks the decline of 

authority, the death of objectivity, and the rise of “mobocracy” since it allows 

“anyone with opposable thumbs and a fifth-grade education [to] publish 

anything on any topic.” 23 They fear that “truth” becomes more relativistic 

under models of peer collaboration or crowd-sourced initiatives. 24 

The pessimists also have very little good to say about YouTube, blogs, social 

networks, and almost all user-generated content. They treat them with a 

combination of confusion and contempt. “[S]elf-expression is not the same 

thing as imagination,” or art, Siegel argues. 25 Instead, he regards the explosion 

of online expression as the “narcissistic” bloviation of the masses and argues it 

is destroying true culture and knowledge. Echoing Postman’s assertion that 

“information has become a form of garbage,” Siegel says that the “Under the 

influence of the Internet, knowledge is withering away into information.” 26 Our 

new age of information abundance is not worth celebrating, he says, because 

“information is powerlessness.” 27 

Some pessimists argue that all the new information and media choices are 

largely false choices that don’t benefit society. For example, Siegel disputes 

what he regards as overly-romanticized notions of “online participation” and 

“personal democracy.” Keen goes further referring to them as “the great 

seduction.” He says “the Web 2.0 revolution has peddled the promise of 

21 LEE SIEGEL, AGAINST THE MACHINE: BEING HUMAN IN THE AGE OF THE ELECTRONIC MOB 

165 (2008). For a review of the book, see Adam Thierer, Book Review: Lee Siegel’s Against the 

Machine, TECHNOLOGY LIBERATION FRONT, Oct. 20, 2008, 

http://techliberation.com/2008/10/20/book-review-lee-siegel%E2%80%99sagainst-the-machine. 

22 Id. at 5. 

23 Keen, supra note 19, at 4. 

24 “Wikipedia, with its video-game like mode of participation, and with its mountains of trivial 

factoids, of shifting mounds of gossip, of inane personal details, is knowledge in the process 

of becoming information.” Siegel, supra note 21, at 152. 

25 Id. at 52. 

26 Id. at 152. 

27 Id. at 148.


bringing more truth to more people … but this is all a smokescreen.” 28 “What 

the Web 2.0 revolution is really delivering,” he argues, “is superficial 

observations of the world around us rather than deep analysis, shrill opinion 

rather than considered judgment.” 29 

Occasionally, the pessimists resort to some fairly immature name-calling tactics 

while critiquing Information Age culture. “It would be one thing if such a 

[digital] revolution produced Mozarts, Einsteins, or Raphaels,” says novelist 

Mark Helprin, “but it doesn’t... It produces mouth-breathing morons in 

backward baseball caps and pants that fall down; Slurpee-sucking geeks who 

seldom see daylight; pretentious and earnest hipsters who want you to wear 

bamboo socks so the world won’t end … beer-drinking dufuses who pay to 

watch noisy cars driving around in a circle for eight hours at a stretch.” 30 

Some pessimists also claim that proliferating new media choices are merely 

force-fed commercial propaganda or that digital technologies are spawning 

needless consumerism. “New technologies unquestionably make purchases 

easier and more convenient for consumers. To this extent, they do help,” says 

the prolific University of Chicago law professor Cass Sunstein. “But they help 

far less than we usually think, because they accelerate the consumption treadmill 

without making life much better for consumers of most goods.” 31 

In Siegel’s opinion, everyone is just in it for the money. “Web 2.0 is the 

brainchild of businessmen,” and the “producer public” is really just a “totalized 

‘consumerist’ society.” 32 Countless unpaid bloggers—in it for the love of the 

conversation and debate—are merely brainwashed sheep whom Siegel argues 

just don’t realize the harm they are doing. “[T]he bloggers are playing into the 

hands of political and financial forces that want nothing more than to see the 

critical, scrutinizing media disappear.” 33 He reserves special scorn for Net 

evangelists who believe that something truly exciting is happening with the new 

online conversation. According to Siegel, they are simply “in a mad rush to earn 

profits or push a fervent idealism.” 34 

The pessimists also fear that these new technologies and trends could have 

profound ramifications not just for entertainment culture, but also for the 


29 Id. 

30 MARK HELPRIN, DIGITAL BARBARISM: A WRITER’S MANIFESTO 57 (2009). 

31 CASS SUNSTEIN, REPUBLIC.COM 121 (2010). 

32 Siegel, supra note 21, at 128. 

33 Id. at 141. 

34 Id. at 25-6.


future of news and professional journalism. They worry about the loss of 

trusted intermediaries and traditional authorities. For example, Keen fears that 

Wikipedia, “is almost single-handedly killing the traditional information 

business.” 35 They also argue that “free culture” isn’t free at all; it’s often just 

parasitic copying or blatant piracy. 

Similarly, Nick Carr and Jaron Lanier worry about the rise of “digital 

sharecropping,” where a small group of elites make money off the back of free 

labor. To Carr, many new Web 2.0 sites and services “are essentially 

agglomerations of the creative, unpaid contributions of their members. In a 

twist on the old agricultural practice of sharecropping, the site owners provide 

the digital real estate and tools, let the members do all the work, and then 

harvest the economic riches.” 36 And in opening his book, Lanier says 

“Ultimately these words will contribute to the fortunes of those few who have 

been able to position themselves as lords of the computing clouds.” 37 

Finally, some pessimists worry deeply about the impact of computers and digital 

technologies on learning. They fear these trends will inevitably result in a 

general “dumbing down” of the masses or even the disappearance of reading, 

writing, and other arts. Typifying this view is Mark Bauerlein’s The Dumbest 

Generation: How the Digital Age Stupefies Young Americans and Jeopardizes Our Future 

(2008), but similar concerns are on display in the works of Sven Birkerts, 38 

Clifford Stoll, 39 Todd Gitlin, 40 and Todd Oppenheimer. 41 

The Optimists’ Response 

The optimists’ response is rooted in the belief that, despite their highly 

disruptive nature, the Internet and new digital technologies empower and 

enlighten individuals and, therefore, generally benefit society. 


36 NICHOLAS CARR, THE BIG SWITCH: REWIRING THE WORLD, FROM EDISON TO GOOGLE 137- 

8 (2008). 

37 LANIER, YOU ARE NOT A GADGET at 1 (2010). 

38 SVEN BIRKERTS, THE GUTENBERG ELEGIES: THE FATE OF READING IN AN ELECTRONIC AGE 

(1994). 

39 CLIFFORD STOLL, HIGH-TECH HERETIC: REFLECTIONS OF A COMPUTER CONTRARIAN 

(1999). 

40 TODD GITLIN, MEDIA UNLIMITED: HOW THE TORMENT OF IMAGES AND SOUNDS 

OVERWHELMS OUR LIVES (2002). 

41 TODD OPPENHEIMER, THE FLICKERING MIND: SAVING EDUCATION FROM THE FALSE 

PROMISE OF TECHNOLOGY (2003).


The optimists tend to argue that new modes of production (especially peerbased 

production) will offer an adequate—if not superior—alternative to 

traditional modalities of cultural or artistic production. Despite displacing some 

institutions and cultural norms, they claim digital technologies create more 

opportunities. They speak of “collective intelligence,” 42 the “wisdom of 

crowds,” 43 the importance of peer production, 44 and the rise of what futurist 

Alvin Toffler first referred to as “prosumers.” 45 “There has been a fundamental 

shift in the balance of power between consumers and salesmen over the last 

generation and it points in the direction of consumers,” Tyler Cowen argues in 

his book, Create Your Own Economy: The Path to Prosperity in a Disordered World. 46 

The peer production trend is stressed in works such as The Wealth of Networks: 

How Social Production Transforms Markets and Freedom, by Yochai Benkler, 47 and 

Wikinomics: How Mass Collaboration Changes Everything, by Don Tapscott and 

Anthony D. Williams. 48 “A new economic democracy is emerging in which we 

all have a lead role,” claim Tapscott and Williams, 49 because “the economics of 

production have changed significantly.” 50 

Most optimists also argue that new business models will evolve to support what 

had previously been provided by professional content creators or news 

providers. Glenn Reynolds (An Army of Davids) and Dan Gillmor (We the Media) 

refer of the rise of “we-dia” (user-generated content and citizen journalism) that 

is an increasingly important part of the modern media landscape. Gillmor, a 

former San Jose Mercury News columnist, speaks of “a modern revolution … 

because technology has given us a communications toolkit that allows anyone to 

become a journalist at little cost and, in theory, with global reach. Nothing like 

this has ever been remotely possible before,” he argues. 51 And the optimists 

generally don’t spend much time lamenting the obliteration of large media 

42 HENRY JENKINS, CONVERGENCE CULTURE: WHERE OLD AND NEW MEDIA COLLIDE 4 

(2006). 

43 JAMES SUROWIECKI, THE WISDOM OF CROWDS (2004). 

44 DON TAPSCOTT & ANTHONY D. WILLIAMS, WIKINOMICS: HOW MASS COLLABORATION 

CHANGES EVERYTHING 1, 67 (2008). 

45 ALVIN TOFFLER, THE THIRD WAVE 265 (1980). 

46 TYLER COWEN, CREATE YOUR OWN ECONOMY: THE PATH TO PROSPERITY IN A DISORDERED 

WORLD 117 (2009). 

47 YOCHAI BENKLER, THE WEALTH OF NETWORKS: HOW SOCIAL PRODUCTION TRANSFORMS 

MARKETS AND FREEDOM (2006). 

48 Tapscott & Williams, supra note 44, at 15. 

49 Id. at 15. 

50 Id. at 68. 

51 DAN GILLMOR, WE THE MEDIA at xii (2004).


institutions, either because they think little of their past performance or, 

alternatively, believe that whatever “watchdog” role they played can be filled by 

others. “We are seeing the emergence of new, decentralized approaches to 

fulfilling the watchdog function and to engaging in political debate and 

organization,” Benkler claims. 52 

Optimists also believe that the Information Age offers real choices and genuine 

voices, and they vociferously dispute charges of diminished quality by 

prosumers, amateur creators, new media outlets, and citizen journalists. 

Moreover, they do not fear the impact of these new trends and technologies on 

learning or culture. “Surely the technophobes who romanticize the pencil don’t 

want to return us to the low literacy rates that characterized the good old days 

of writing with pencils and quills,” Baron asks. “Still, a few critics object to the 

new technologies because they enable too many people to join the guild of 

writers, and they might paraphrase Thoreau’s objection to the telegraph: these 

new computer writers, it may be, have nothing to say to one another.” 53 

Finally, in addressing the sharecropper concern raised by Carr and Lanier, the 

optimists insist most people aren’t in it for the money. Shirky notes that 

“Humans intrinsically value a sense of connectedness,” and much of what they 

do in the social media world is a true labor of love. 54 “Amateurs aren’t just pintsized 

professionals; people are sometimes happy to do things for reasons that 

are incompatible with getting paid,” he says. 55 Mostly they do it for love of 

knowledge or a belief in the importance of “free culture,” the optimists claim. 

The Debate Over the Promise— 

or Perils—of Personalization 

Optimists and pessimists tend to agree that the Internet and “Web 2.0” is 

leading to more “personalized” media and information experiences. They 

disagree vehemently, however, on whether this is good or bad. They 

particularly disagree on what increased information customization means for 

participatory democracy and the future of relations among people of diverse 

backgrounds and ideologies. Finally, they differ on how serious of a problem 

“information overload” is for society and individuals. 

52 Benkler, supra note 47, at 11. 


54 CLAY SHIRKY, COGNITIVE SURPLUS: CREATIVITY AND GENEROSITY IN A CONNECTED AGE 

58-9 (2010). 

55 Id.


The Optimists’ Case 

Let’s take the optimists first this time. 

The optimists tend to embrace what Nicholas Negroponte first labeled “The 

Daily Me” (i.e., hyper-personalized news, culture, and information). In 1995, 

Negroponte asked us to: 

Imagine a future in which your interface agent can read every 

newswire and newspaper and catch every TV and radio 

broadcast on the planet, and then construct a personalized 

summary. This kind of newspaper is printed in an edition of 

one.… 

Imagine a computer display of news stories with a knob that, 

like a volume control, allows you to crank personalization up 

or down. You could have many of these controls, including a 

slider that moves both literally and politically from left to right 

to modify stories about public affairs. These controls change 

your window onto the news, both in terms of size and its 

editorial tone. In the distant future, interface agents will read, 

listen to, and look at each story in its entirety. In the near 

future, the filtering process will happen by using headers, those 

bits about bits. 56 

That future came about sooner than even Negroponte could have predicted. 

We all have a “Daily Me” at our disposal today thanks to RSS feeds, Facebook, 

Google Alerts, Twitter, email newsletters, instant messaging, and so on. These 

tools, among others, can provide tailored, automated search results served up 

instantaneously. The optimists argue that this increased tailoring and 

personalization of our media experiences empowers heretofore silenced masses. 

This worldview is typified by the title of Glenn Reynolds’ book: An Army of 

Davids: How Markets and Technology Empower Ordinary People to Beat Big Media, Big 

Government and Other Goliaths. 57 The optimists argue that our “participatory 

culture” promotes greater cultural heterogeneity and gives everyone a better 

chance to be heard. “In a world of media convergence, every important story 

gets told, every brand gets sold, and every consumer gets courted across 

multiple media platforms,” says Henry Jenkins, author of Convergence Culture. 58 

56 Negroponte, supra note 12, at 153-54. 

57 GLENN REYNOLDS, AN ARMY OF DAVIDS: HOW MARKETS AND TECHNOLOGY EMPOWER 

ORDINARY PEOPLE TO BEAT BIG MEDIA, BIG GOVERNMENT AND OTHER GOLIATHS (2006). 

58 HENRY JENKINS, CONVERGENCE CULTURE: WHERE OLD AND NEW MEDIA COLLIDE 3 

(2006). Tapscott & Williams, supra note 44, at 41.


Again, they stress the empowering nature of digital technology as a good in and 

of itself. “The mass amateurization of publishing undoes the limitations 

inherent in having a small number of traditional press outlets,” Shirky claims. 59 

This leads to greater openness, transparency, exposure to new thinking and 

opinions, and a diversity of thought and societal participation. Shirky speaks of 

the “cognitive surplus” unleashed by these changes and its myriad benefits for 

society and culture: 

The harnessing of our cognitive surplus allows people to 

behave in increasingly generous, public, and social ways, 

relative to their old status as consumers and couch potatoes. 

The raw material of this change is the free time available to us, 

time we can commit to projects that range from the amusing to 

the culturally transformative. … Flexible, cheap, and inclusive 

media now offers us opportunities to do all sorts of things we 

once didn’t do. In the world of “the media,” we were like 

children, sitting quietly at the edge of a circle and consuming 

whatever the grown-ups in the center of the circle produced. 

That has given way to a world in which most forms of 

communication, public and private, are available to everyone in 

some form. 60 

Shirky even suggests that “The world’s cognitive surplus is so large that small 

changes can have huge ramifications in aggregate,” and have beneficial impacts 

on politics, advocacy, and “generosity.” 

When it comes to concerns about “information overload,” most optimists see 

little reason for concern. Tyler Cowen argues that using search tools like 

Google and other information gathering and processing technologies actually 

“lengthen our attention spans in another way, namely by allowing greater 

specialization of knowledge:” 61 

We don’t have to spend as much time looking up various facts 

and we can focus on the particular areas of interest, if only 

because general knowledge is so readily available. It’s never 

been easier to wrap yourself up in a long-term intellectual 

project, yet without losing touch with the world around you. 

59 CLAY SHIRKY, HERE COMES EVERYBODY: THE POWER OF ORGANIZING WITHOUT 

ORGANIZATIONS 65 (2008). 

60 CLAY SHIRKY, COGNITIVE SURPLUS, supra note 54, at 63. 

61 TYLER COWEN, CREATE YOUR OWN ECONOMY: THE PATH TO PROSPERITY IN A DISORDERED 

WORLD 55 (2009).


As for information overload, it is you who chooses how much 

“stuff” you want to experience and how many small bits you 

want to put together … . The quantity of information coming 

our way has exploded, but so has the quality of our filters. 62 

Chris Anderson previously made this point in his book, The Long Tail. 

Anderson defined filters as “the catch-all phrase for recommendations and all 

the other tools that help you find quality in the Long Tail” and noted that 

“these technologies and services sift through a vast array of choices to present 

you with the ones that are most right for you.” 63 “The job of filters is to screen 

out [the] noise” or information clutter, Anderson says. 64 Cowen argues that the 

filtering technologies are getting better at this sifting and processing process, but 

so too are humans, he says. The key to this, he argues, is that we are getting better 

at “ordering” information. 

On balance, therefore, the optimists argue that personalization benefits our 

culture and humanity. Dennis Baron concludes, “English survives, conversation 

thrives online as well as off, and on balance, digital communications seems to 

be enhancing human interaction, not detracting from it.” 65 

The Pessimists’ Response 

The pessimists argue that all this Pollyannaish talk about a new age of 

participatory democracy is bunk. Instead of welcoming increased information 

and media personalization, they lament it. They fear that “The Daily Me” that 

the optimists laud will lead to homogenization, close-mindedness, an online 

echo-chamber, information overload, corporate brainwashing, etc. Worst, 

hyper-customization of websites and online technologies will cause extreme 

social “fragmentation,” “polarization,” “balkanization,” “extremism” and even 

the decline of deliberative democracy. 66 

Siegel and Keen are probably the most searing in this critique. To Siegel, for 

example, the “Daily Me” is little more that the creation of a “narcissistic 

culture” in which “exaggeration” and the “loudest, most outrageous, or most 

62 Id. 

63 CHRIS ANDERSON, THE LONG TAIL 108 (2006). 

64 Id. at 115. 


66 Carr worries that every little choice moves us close toward such social isolation: “Every time 

we subscribe to a blog, add a friend to our social network, categorize an email message as 

spam, or even choose a site from a list of search results, we are making a decision that 

defines, in some small way, whom we associate with and what information we pay attention 

to.” NICHOLAS CARR, THE BIG SWITCH: REWIRING THE WORLD, FROM EDISON TO GOOGLE 

160 (2008).


extreme voices sway the crowd of voices this way; the cutest, most self-effacing, 

most ridiculous, or most transparently fraudulent of voices sway the crowd of 

voices that way.” 67 He calls Web 2.0 “democracy’s fatal turn” in that, instead of 

“allowing individuals to create their own cultural and commercial choices,” it 

has instead created “a more potent form of homogenization.” 68 Keen fears the 

rise of “a dangerous form of digital narcissism” and “the degeneration of 

democracy into the rule of the mob and the rumor mill.” 69 

This echoes concerns first raised by Cass Sunstein in his 2001 book 

Republic.com. 70 In that book, Sunstein referred to Negroponte’s “Daily Me” in 

contemptuous terms, saying that the hyper-customization of websites and 

online technologies was causing extreme social fragmentation and isolation that 

could lead to political extremism. “A system of limitless individual choices, with 

respect to communications, is not necessarily in the interest of citizenship and 

self-government,” he wrote. 71 Sunstein was essentially claiming that the 

Internet is breeding a dangerous new creature: Anti-Democratic Man. 72 

“Group polarization is unquestionably occurring on the Internet,” he 

proclaimed, and it is weakening what he called the “social glue” that binds 

society together and provides citizens with a common “group identity.” 73 If that 

continues unabated, Sunstein argued, the potential result could be nothing short 

of the death of deliberative democracy and the breakdown of the American 

system of government. 

Some of the pessimists, like Keen, go further and claim that “the moral fabric of 

our society is being unraveled by Web 2.0. It seduces us into acting on our 

most deviant instincts and allows us to succumb to our most destructive vices. 

And it is corroding the values we share as a nation.” 74 Nick Carr summarizes the 

views of the pessimists when he says: “it’s clear that two of the hopes most dear 

to the Internet optimists—that the Web will create a more bountiful culture and 

that it will promote greater harmony and understanding—should be treated 

67 Siegel, supra note 21, at 79. 

68 Id. at 67. 

69 Keen, supra note 19, at 54-5. 

70 CASS SUNSTEIN, REPUBLIC.COM (2001). 

71 Id. at 123. 

72 See Adam Thierer, Saving Democracy from the Internet, REGULATION (Fall 2001) 78-9, 

http://www.cato.org/pubs/regulation/regv24n3/inreview.pdf. 

73 Sunstein, supra, at 71, 89. 

74 Keen, supra note 19, at 163.


with skepticism. Cultural impoverishment and social fragmentation seem 

equally likely outcomes.” 75 

Another common theme in the works of the pessimists is summarized by the 

title of Siegel’s book (Against the Machine). They fear the “mechanization of the 

soul” 76 or humanity’s “surrender” to “the machine revolution.” 77 In opening of 

You Are Not a Gadget, Lanier fears that “these words will mostly be read by 

nonpersons—automatons or numb mobs composed of people who are no 

longer acting as individuals.” 78 “The trick is not to subject man and nature to 

the laws of the machine,” says Helprin, “but rather to control the machine 

according to the laws and suggestions of nature and human nature. To 

subscribe to this does not make one a Luddite.” 79 

Finally, the pessimists are also concerned about the impact of online anonymity 

on human conduct and language. They argue anonymity leads to less 

accountability or, more simply, just plain bad manners. “If our national 

conversation is carried out by anonymous, self-obsessed people unwilling to 

reveal their real identities, then,” Keen argues, “community denigrates into 

anarchy.” 80 

So Who’s Right? 

On balance, the optimists generally have the better of the argument today. We 

really are better off in an age of information abundance than we were in the 

scarcity era we just exited. Nonetheless, the pessimists make many fair points 

that deserve to be taken seriously. But they need a more reasonable articulation 

of those concerns and a constructive plan for how to move forward without a 

call for extreme reactionary solutions. 

A hybrid approach here might be thought of as “pragmatic optimism,” which 

attempts to rid the optimist paradigm of its kookier, pollyannish thinking while 

also taking into account some of the very legitimate concerns raised by the 

pessimists, but rejecting its caustic, neo-Luddite fringe elements and stasis 

mentality in the process. 

75 Carr, supra note 36, at 167. 

76 Helprin, supra note 30, at 100. 

77 Id. at 9, 100. 

78 Lanier, supra note 37, at 1. 


80 Keen, supra note 30, at 80.


Thoughts on the Pessimists 

First and foremost, if they hope to be taken more seriously, Net skeptics need 

better spokespersons. Or, they at least need a more moderated, less hysterical 

tone when addressing valid concerns raised by technological progress. It’s often 

difficult to take the pessimists seriously when they exude outright hostility to 

most forms of technological progress. Most of them deny being high-tech 

troglodytes, but the tone of some of their writing, and the thrust of some of 

their recommendations, exhibit occasional Luddite tendencies—even if they 

don’t always come out and call for extreme measures to counteract dynamism. 

Moreover, the name-calling they sometimes engage in, and their derision for the 

digital generation can be just as insulting and immature as the online “mob” 

they repeatedly castigate in their works. Too often, their criticism devolves into 

philosophical snobbery and blatant elitism, as in the works of Helprin, Siegel, 

and Keen. Constantly looking down their noses at digital natives and all 

“amateur” production isn’t going to help them win any converts or respect for 

their positions. Moreover, one wonders if they have fingered the right culprit 

for civilization’s supposed decline, since most of the ills they identify predate 

the rise of the Internet. 

The pessimists are often too quick to proclaim the decline of modern 

civilization by looking only to the baser elements of the blogosphere or the 

more caustic voices of cyberspace. The Internet is a cultural and intellectual 

bazaar where one can find both the best and the worst of humanity on display 

at any given moment. True, “brutishness and barbarism,” as Helprin calls it, 81 

can be found on many cyber-corners, but not all of its corners. And, contrary 

to Helprin’s assertion that blogging “begins the mad race to the bottom,” 82 one 

could just as easily cite countless instances of the healthy, unprecedented 

conversations that blogs have enabled about a diverse array of topics. 

Their claim that the “Daily Me” and information specialization will lead to a 

variety of ills is also somewhat overblown. It’s particularly hard to accept 

Sunstein and Carr’s claims that increased personalization is breeding 

“extremism,” “fanaticism” and “radicalization.” A recent study by Matthew 

Gentzkow and Jesse M. Shapiro of the University of Chicago Booth School of 

Business lent credibility to this, finding “no evidence that the Internet is 

becoming more segregated over time” or leading to increased polarization as 

Sunstein and other pessimists fear. 83 Instead, their findings show that the Net 


82 Id. at 42. 

83 Matthew Gentzkow & Jesse M. Shapiro, Ideological Segregation Online and Offline, CHICAGO 

BOOTH WORKING PAPER No. 10-19, April 5, 2010, 

http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1588920.


has encouraged more ideological integration and is actually driving us to 

experience new, unanticipated viewpoints. 84 

While it’s true the Internet has given some extremists a new soapbox to stand 

on and spew their hatred and stupidity, the fact is that such voices and 

viewpoints have always existed. The difference today is that the Internet and 

digital platforms have given us a platform to counter such societal extremism. 

As the old saying goes, the answer to bad speech is more speech—not a 

crackdown on the underlying technologies used to convey speech. It should 

not be forgotten that, throughout history, most extremist, totalitarian 

movements rose to power by taking over the scarce, centralized media 

platforms that existed in their countries. The decentralization of media makes 

such a take-over far less plausible to imagine. 

Sometimes the pessimists seem to just be suffering from a bit of old-fogeyism. 

Lanier, for example, dismisses most modern culture as “retro” and “a petty 

mashup of preweb culture.” 85 “It’s as if culture froze just before it became 

digitally open, and all we can do now is mine the past like salvagers picking over 

a garbage dump.” 86 Many pessimists are guilty of such hyper-nostalgia about 

those mythical “good ‘ol days” when all was supposedly much better. It’s a 

common refrain we’ve heard from many social and cultural critics before. But 

such cultural critiques are profoundly subjective. Many pessimists simply seem 

to be well passed the “adventure window.” 87 The willingness of humans to try 

new things and experiment with new forms of culture—our “adventure 

window”—fades rapidly after certain key points in life, as we gradually settle in 

our ways. Many cultural critics and average folk alike seem convinced the best 

days are behind us and the current good-for-nothing generation and their newfangled 

gadgets and culture are garbage. At times this devolves into a full-blown 

moral panic. 88 “It’s perfectly normal and probably healthy to examine whether 

these changes are good or bad,” says New York Times blogger Nick Bilton, 

author of I Live in the Future & Here’s How It Works. “But we’ll also no doubt 

84 “This study suggests that Internet users are a bunch of ideological Jack Kerouacs. They’re 

not burrowing down into comforting nests. They’re cruising far and wide looking for 

adventure, information, combat and arousal.” David Brooks, Riders on the Storm, NEW YORK 

TIMES, April 19, 2010, http://www.nytimes.com/2010/04/20/opinion/20brooks.html. 

85 Lanier, supra note 37, at 131. 

86 Id. at 133. 

87 Adam Thierer, The “Adventure Window,” Radio Formats and Media Ownership Rules, 

TECHNOLOGY LIBERATION FRONT, Aug. 16, 2006, 

http://techliberation.com/2006/08/16/the-adventure-window-radio-formats-andmedia-ownership-rules. 

88 See Adam Thierer, Parents, Kids & Policymakers in the Digital Age: Safeguarding Against ‘Techno- 

Panics,’ INSIDE ALEC (July 2009) at 16-7, 

http://www.alec.org/am/pdf/Inside_July09.pdf.


look back at many of the debates a generation from now and see that a lot of 

these fears were inflated and maybe a bit ridiculous, too.” 89 

The “sharecropper” concern raised by Carr and Lanier is also over-stated. This 

logic ignores the non-monetary benefits that many of us feel we extract from 

today’s online business models and social production processes. Most of us feel 

we get a lot back as part of this new value exchange. Carr and Lanier are 

certainly correct that Google, Facebook, MySpace, and a lot of other Net 

middlemen are getting big and rich based on all the user-generated content 

flowing across their sites and systems. On the other hand, most cyber-citizens 

extract enormous benefits from the existence of those (mostly free and 

constantly improving) platforms and services. It’s a very different sort of value 

exchange and business model than in the past, but we are adjusting to it. 

Yet for all of Wikipedia’s value as a reference of first (but certainly not final) 

resort, the pessimists have almost nothing good to say about it. Much the same 

goes for open source and other collaborative efforts. They don’t appear willing 

to accept the possibility of any benefits coming from collective efforts. And 

they wrongly treat the rise of collective / collaborative efforts as a zero-sum 

game; imagining it represents a net loss of individual effort & “personhood.” 

That simply doesn’t follow. The masses have been given more of a voice 

thanks to the rise of Web 2.0 collaborative technologies and platforms, but that 

doesn’t mean that media “professionals” don’t still exist. Most bloggers, for 

example, build their narratives around facts and stories found in respected 

“mainstream media” outlets. It’s true that those outlets must now compete in a 

broad sense with many new forms of competition for human attention, but it 

doesn’t mean they still won’t play a lead role in the new information ecosystem. 

Most of all, the pessimists can and must come to terms with the Information 

Revolution while offering more constructive and practical solutions to 

legitimately difficult transitional problems created by disintermediating 

influences of the digital technologies and Net. After all, practically speaking, 

what would the pessimists have us do if we can’t mitigate the problems they 

identify? “Whatever the mix of good and bad,” Notes Wall Street Journal 

columnist Gordon Crovitz, “technology only advances and cannot be put back 

in the bottle.” 90 Would the pessimists have us attempt to put the digital genie 

back in bottle with burdensome restrictions on technology or the creation of a 

permissions-based system of innovation? “[W]hether it’s good for society or 

89 NICK BILTON, I LIVE IN THE FUTURE & HERE’S HOW IT WORKS 63 (2010). 

90 L. Gordon Crovitz, Is Technology Good or Bad? Yes. WALL STREET JOURNAL, Aug. 23, 2010, 

http://online.wsj.com/article/SB10001424052748703579804575441461191438330.html.


bad … is somewhat irrelevant at this point,” argues Nick Bilton. 91 “There’s no 

turning back the clock.” Similarly, Ben Casnocha has correctly noted that: 

the wind at the backs of all techno-optimists … [is] the 

forward momentum of technological development. You 

cannot turn back the clock. It is impossible to envision a future 

where there is less information and fewer people on social 

networks. It is very possible to envision increasing abundance 

along with better filters to manage it. The most constructive 

contributions to the debate, then, heed Moore’s Law in the 

broadest sense and offer specific suggestions for how to 

harness the change for the better. 92 

Regrettably, most of the leading Net pessimists have failed to do this in their 

work. However, good templates for how to accomplish this can be found in 

recent books by William Powers (Hamlet’s BlackBerry: A Practical Philosophy for 

Building a Good Life in the Digital Age) 93 and John Freeman (The Tyranny of E-Mail: 

The Four-Thousand-Year Journey to Your Inbox). 94 These authors, although 

somewhat pessimistic in their view of technology’s impact on life and learning, 

offer outstanding self-help tips and plans of action about how to reasonably 

assimilate new information technologies into our lives. Their key insight: the 

Internet and digital technologies aren’t going away, so we must figure out how 

to deal with them in a responsible manner—both individually and collectively. 

It’s essential other pessimists come to grips with that fact. 

The pessimists are at their best when highlighting the very legitimate concerns 

about the challenges that accompany technological change, including the impact 

of the digital revolution on “professional” media, the decline of authority 

91 Bilton, supra note 89, at 216. 

92 Ben Casnocha, RSSted Development, THE AMERICAN, July 1, 2009, 

http://www.american.com/archive/2009/june/rssted-development. Clay Shirky has 

also noted that “There is never going to be a moment when we as a society ask ourselves, 

‘Do we want this? Do we want the changes that the new flood of production and access and 

spread of information is going to bring about?’” Clay Shirky, HERE COMES EVERYBODY: 

THE POWER OF ORGANIZING WITHOUT ORGANIZATIONS 73 (2008). 

93 WILLIAM POWERS, HAMLET’S BLACKBERRY: A PRACTICAL PHILOSOPHY FOR BUILDING A 

GOOD LIFE IN THE DIGITAL AGE (2010). See also Adam Thierer, Coping with Information 

Overload: Thoughts on Hamlet’s BlackBerry by William Powers, TECHNOLOGY LIBERATION 

FRONT, Sept. 6, 2010, http://techliberation.com/2010/09/06/coping-withinformation-overload-thoughts-on-hamlets-blackberry-by-william-powers. 

94 JOHN FREEMAN, THE TYRANNY OF E-MAIL: THE FOUR-THOUSAND-YEAR JOURNEY TO 

YOUR INBOX (2009). For a review of the book, see Adam Thierer, Can Humans Cope with 

Information Overload? Tyler Cowen & John Freeman Join the Debate, TECHNOLOGY LIBERATION 

FRONT, Aug. 23, 2009, http://techliberation.com/2009/08/23/can-humans-cope-withinformation-overload-tyler-cowen-john-freeman-join-the-debate.


among trusted experts and intermediaries, and the challenge of finding creative 

ways to fund “professional” media and art going forward. 

Thoughts on the Optimists 

Again, the optimists currently have the better of this debate: Web 2.0 is 

generally benefiting culture and society. It is almost impossible to accept that 

society has not benefited from the Internet and new digital technologies 

compared to the past era of information scarcity. The Digital Revolution has 

greatly empowered the masses and offered them more informational inputs. 

But the optimists need to be less pollyannaish and avoid becoming the 

“technopolists” (or digital utopians) that Postman feared were taking over our 

society. There’s often too much Rousseauian romanticism at work in some 

optimist writings. Just as the pessimists are often guilty assuming the Net and 

digital technologies are responsible for far too many ills, the optimists 

occasionally do the opposite by engaging in what Nick Carr labels “the 

Internet’s liberation mythology.” The Internet isn’t remaking man or changing 

human nature in any fundamental way. Nor can it liberate us from all earthly 

constraints or magically solve all of civilization’s problems. Moreover, when it 

comes to economics, all this talk about the Long Tail being “the future of 

business” (Chris Anderson) and of “Wikinomics … changing everything 

through mass collaboration,” (Tapscott and Williams) verges on irrational 

techno-exuberance. 

In particular, optimists often overplay the benefits of collective intelligence, 

collaboration, and the role of amateur production. They are occasionally guilty 

of “the elevation of information to metaphysical status” as Postman lamented. 95 

For example, the optimists could frame “Wiki” and peer-production models as 

a complement to professional media, not a replacement for it. Could the equivalent 

of The New York Times really be cobbled together by amateurs daily? It seems 

highly unlikely. And why aren’t there any compelling open source video games? 

Similarly, free and open source software (FOSS) has produced enormous social 

/ economic benefits, but it would be foolish to believe that FOSS (or “wiki” 

models) will replace all proprietary business models. Each model or mode of 

production has its place and purpose and they will continue to co-exist and 

compete. 

We wouldn’t necessarily be better off if all the “professional” media producers 

and old intermediaries disappeared, even if it is no doubt true that many of 

them will. Some optimists play the “old media just doesn’t get it” card far too 

often and snobbishly dismiss many producers’ valid concerns and efforts to 

reinvent themselves. 

95 Postman, supra note 8, at 61.


There’s also a big difference between “remix culture” and “rip-off culture.” 

Many optimists turn a blind eye to blatant copyright piracy, for example, or 

even defend it as either a positive development or simply inevitable. Remix 

culture generally enhances and extends culture and creativity. But blatant 

content piracy deprives many of society’s most gifted creators of the incentive 

to produce culturally beneficial works. Likewise, hacking, circumvention, and 

reverse-engineering all play an important and legitimate role in our new digital 

economy, but one need not accept the legitimacy of those activities when 

conducted for nefarious purposes (think identity theft or chip-modding to 

facilitate video game piracy.) 

The optimists should be cautious about predicting sweeping positive changes 

from the Internet or Web 2.0 technologies. Consider Shirky’s generally upbeat 

assessment of the impact of “cognitive surplus.” There’s a lot of fluffy talk and 

anecdotal examples in Shirky’s book about how the cognitive surplus spawned 

by cyber-life has affected politics, advocacy, and “generosity,” but I think it’s a 

stretch to imply that the Net is going to upend political systems. In another 

essay in this collection, Evgeny Morozov challenges Shirky on some of these 

points, arguing that “the Internet will not automatically preserve—never mind 

improve—the health of democratic politics.” 96 He’s right. That digital 

technology and the Internet will help reshape society and politics to some 

degree is indisputable. But that doesn’t mean the Net will radically reshape 

political systems or human nature anytime soon. 

Finally, the optimists would be wise to separate themselves from those extreme 

voices in their community who speak of the “noosphere” and “global 

consciousness” and long for the eventual singularity. While he doesn’t go quite 

so far, Wired editor Kevin Kelly often pushes techno-optimism to its extreme. 

In his latest book, What Technology Wants, Kelly speaks of what he calls “the 

technium” as a “force” or even a living organism that has a “vital spirit” and 

which “has its own wants” and “a noticeable measure of autonomy.” 97 

Treating technology as an autonomous force is silly, even dangerous, thinking. 

It is to imbue it with attributes and feelings that simply do not exist and would 

probably not be desirable if they did. Yet, some optimists speak in fatalistic 

terms and make such an outcome seem desirable. They sound like they long for 

life in The Matrix—”Bring on sentient robot masters and the Singularity!” Thus 

does an optimist cross over into the realm of quixotic techno-utopianism. 

Optimists need to place technological progress in context and appreciate that, as 

Postman argued, there are some moral dimensions to technological progress 

that deserve attention. Not all change is good change. The optimists need to be 

96 Evgeny Morozov, Will the Net Liberate the World?, infra at 443. 

97 KEVIN KELLY, WHAT TECHNOLOGY WANTS 198, 41, 15, 13 (2010).


mature enough to understand and address the downsides of digital life without 

dismissing its critics. On the other hand, some of those moral consequences are 

profoundly positive, which the pessimists usually fail to appreciate or even 

acknowledge. 

Conclusion: Toward “Pragmatic Optimism” 

Again, I believe the optimists currently have the better of this debate. It’s 

impossible for me to believe we were better off in an era of information poverty 

and un-empowered masses. I’ll take information overload over information 

poverty any day! As Dennis Baron puts it: “The Internet is a true electronic 

frontier where everyone is on his or her own: all manuscripts are accepted for 

publication, they remain in virtual print forever, and no one can tell writers what 

to do.” 98 

The rise of the Internet and digital technologies has empowered the masses and 

given everyone a soapbox on which to speak to the world. Of course, that 

doesn’t necessarily mean all of them will have something interesting to say! We 

shouldn’t exalt user-generated content as a good in and of itself. It’s quality, not 

volume, that counts. But such human empowerment is worth celebrating, 

despite its occasional downsides. 99 Abundance is better than the old analog 

world of few choices and fewer voices. 

However, the pessimists have some very legitimate concerns regarding how the 

passing of the old order might leave society without some important things. For 

example, one need not endorse bailouts for a dying newspaper industry to 

nonetheless worry about the important public service provided by investigative 

journalists: Who will take up those efforts if large media institutions go under 

because of digital disintermediation? 

The skeptics are also certainly correct that each of us should think about how to 

better balance new technologies and assimilate them into our lives and the lives 

of our families and communities. For example, children need to learn new 

“digital literacy” and “cyber-citizenship” skills to be savvy Netizens. 

To be clear, I am not suggesting that these questions should be answered by 

government. There exist many other ways that society can work to preserve 


99 “Just as well-meaning scientists and consumers feared that trains and comic books and 

television would rot our brains and spoil our minds, I believe many of the skeptics and 

worrywarts today are missing the bigger picture, the greater value that access to new and 

faster information is bringing us.” Nick Bilton, I LIVE IN THE FUTURE & HERE’S HOW IT 

WORKS 136 (2010).


important values and institutions without embracing the stasis mentality and 

using coercion to accomplish that which should be pursued voluntarily. 

As noted, the nostalgia the pessimists typically espouse for the past is a 

common refrain of cultural and technological critics who fear our best days are 

behind us. The truth typically proves less cataclysmic, of course. The great 

thing about humans is that we adapt better than other creatures. When it comes 

to technological change, resiliency is hard-wired into our genes. “The technoapocalypse 

never comes,” notes Slate’s Jack Shafer, because “cultures tend to 

assimilate and normalize new technology in ways the fretful never anticipate.” 100 

We learn how to use the new tools given to us and make them part of our lives 

and culture. Indeed, we have lived through revolutions more radical than the 

Information Revolution. We can adapt and learn to live with some of the 

legitimate difficulties and downsides of the Information Age. 

Generally speaking, the sensible middle ground position is “pragmatic 

optimism”: We should embrace the amazing technological changes at work in 

today’s Information Age but with a healthy dose of humility and appreciation 

for the disruptive impact and pace of that change. We need to think about how 

to mitigate the negative impacts associated with technological change without 

adopting the paranoid tone or Luddite-ish recommendations of the pessimists. 

I’m particularly persuaded by the skeptics’ call for all of us to exercise some 

restraint in terms of the role technology plays in our own lives. While pessimists 

from Plato and Postman certainly went too far at times, there is more than just 

a kernel of truth to their claim that, taken to an extreme, technology can have a 

deleterious impact on life and learning. We need to focus on the Aristotelian 

mean. We must avoid neo-Luddite calls for a return to “the good ‘ol days” on 

the one hand, while also rejecting techno-utopian Pollyannaism on the other. 

We need not go to “all or nothing” extremes. 

In the end, however, I return to the importance of evolutionary dynamism and 

the importance of leaving a broad sphere for continued experimentation by 

individuals and organizations alike. Freedom broadly construed is valuable in its 

own right—even if not all of the outcomes are optimal. As Clay Shirky rightly 

notes: 

This does not mean there will be no difficulties associated with 

our new capabilities—the defenders of freedom have long 

noted that free societies have problems peculiar to them. 

Instead, it assumes that the value of freedom outweighs the 

100 Jack Shafer, Digital Native Calms the Anxious Masses, SLATE, Sept. 13, 2010, 

http://www.slate.com/id/2267161.


problems, not based on calculation of net value but because 

freedom is the right thing to want for society. 101 

Finally, we cannot ignore the practical difficulties of halting or even slowing 

progress—assuming we somehow collectively decided we wanted to do so. 

Turning back the clock seems almost unfathomable at this point absent extreme 

measures that would sacrifice so many of the benefits the Information Age has 

brought us—not to mention the curtailment of freedom that it would demand. 

Regardless, the old Theuth-Thamus debate about the impact of technological 

change on culture and society will continue to rage. There is no chance this 

debate will die down anytime soon. (Just wait till new technologies like virtual 

reality go mainstream!) Despite real challenges in adapting to technological 

change, I remain generally optimistic about the prospects for technology to 

improve the human condition. 

101 Shirky, supra note 59, at 298.

88 CHAPTER 1: THE INTERNET’S IMPACT ON CULTURE & SOCIETY: GOOD OR BAD?

CHAPTER 2 

IS THE GENERATIVE INTERNET AT RISK? 

Protecting the Internet Without Wrecking It: 

How to Meet the Security Threat 91 


A Portrait of the Internet as a Young Man 113 

Ann Bartow 


Saving the Net from Its Supporters 139 

Adam Thierer 

89

90 CHAPTER 2: IS THE GENERATIVE INTERNET AT RISK?


Protecting the Internet Without 

Wrecking It: How to Meet the 

Security Threat 

By Jonathan Zittrain * 

On November 2, 1988, 5-10% of the 60,000 computers hooked up to the 

Internet started acting strangely. Inventories of affected computers revealed 

that rogue programs were demanding processor time. When concerned 

administrators terminated these programs, they reappeared and multiplied. 

They then discovered that renegade code was spreading through the Internet 

from one machine to another. The software—now commonly thought of as 

the first Internet worm—was traced to a twenty-three-year-old Cornell 

University graduate student, Robert Tappan Morris, Jr., who had launched it by 

infecting a machine at MIT from his terminal in Ithaca, New York. 

Morris said he unleashed the worm to count how many machines were 

connected to the Internet, and analysis of his program confirmed his benign 

intentions. But his code turned out to be buggy. If Morris had done it right, his 

program would not have drawn attention to itself. It could have remained 

installed for days or months, and quietly performed a wide array of activities 

other than Morris’s digital headcount. 

The mainstream media had an intense but brief fascination with the incident. A 

government inquiry led to the creation of the Defense Department-funded 

Computer Emergency Response Team Coordination Center at Carnegie Mellon 

University, which serves as a clearinghouse for information about viruses and 

other network threats. A Cornell report on what had gone wrong placed the 

blame solely on Morris, who had engaged in a “juvenile act” that was “selfish 

and inconsiderate.” It rebuked elements of the media that had branded Morris 

a hero for dramatically exposing security flaws, noting that it was well known 

that the computers’ Unix operating systems were imperfect. The report called 

for university-wide committees to provide advice on security and acceptable 

use. It described consensus among computer scientists that Morris’s acts 

warranted some form of punishment, but not “so stern as to damage 

permanently the perpetrator’s career.” 

* Professor of Law, Harvard Law School and Harvard Kennedy School; Professor of 

Computer Science, Harvard School of Engineering and Applied Sciences; Co-Founder, 

Berkman Center for Internet & Society. This chapter originally appeared in the March/April 

2008 BOSTON REVIEW.

92 CHAPTER 2: IS THE GENERATIVE INTERNET AT RISK? 

In the end, Morris apologized, earned three years of criminal probation, 

performed four hundred hours of community service, and was fined $10,050. 

He transferred from Cornell to Harvard, founded a dot-com startup with some 

friends in 1995, and sold it to Yahoo! in 1998 for $49 million. He is now a 

respected, tenured professor at MIT. 

In retrospect, the commission’s recommendations—urging users to patch their 

systems and hackers to grow up—might seem naïve. But there were few 

plausible alternatives. Computing architectures, both then and now, are 

designed for flexibility rather than security. The decentralized ownership and 

non-proprietary nature of the Internet and the computers connected to it made 

it difficult to implement structural improvements. More importantly, it was 

hard to imagine cures that would not entail drastic, wholesale, purpose-altering 

changes to the very fabric of the Internet. Such changes would have been 

wildly out of proportion to the perceived threat, and there is no record of their 

having even been considered. 

Generative systems are powerful—they enable extraordinary 

numbers of people to devise new ways to express themselves 

in speech, art, or code, perhaps because they lack central 

coordination and control. 

By design, the university workstations of 1988 were generative: Their users 

could write new code for them or install code written by others. This generative 

design lives on in today’s personal computers. Networked PCs are able to 

retrieve and install code from each other. We need merely click on an icon or 

link to install new code from afar, whether to watch a video newscast embedded 

within a Web page, update our word processing or spreadsheet software, or 

browse satellite images. 

Generative systems are powerful and valuable, not only because they foster the 

production of useful things like Web browsers, auction sites, and free 

encyclopedias, but also because they enable extraordinary numbers of people to 

devise new ways to express themselves in speech, art, or code and to work with 

other people. These characteristics can make generative systems very successful 

even though—perhaps especially because—they lack central coordination and 

control. That success attracts new participants to the generative system. 

The flexibility and power that make generative systems so attractive are, 

however, not without risks. Such systems are built on the notion that they are 

never fully complete, that they have many uses yet to be conceived of, and that 

the public can be trusted to invent good uses and share them. But multiplying 

breaches of that trust threaten the very foundations of the system. 

Whether through a sneaky vector like the one Morris used, or through the front 

door, when a trusting user elects to install something that looks interesting


without fully understanding it, opportunities for accidents and mischief abound. 

A hobbyist computer that crashes might be a curiosity, but when a home or 

office PC with years’ worth of vital correspondence and papers is compromised, 

it can be a crisis. And when thousands or millions of individual, business, 

research, and government computers are subject to attack, we may find 

ourselves faced with a fundamentally new and harrowing scenario. As the 

unsustainable nature of the current state of affairs becomes more apparent, we 

are left with a dilemma that cannot be ignored: How do we preserve the 

extraordinary benefits of generativity, while addressing the growing 

vulnerabilities that are innate to it? 

* * * 

How profound is today’s security threat? Since 1988, the Internet has suffered 

few truly disruptive security incidents. A network designed for communication 

among academic and government researchers appeared to scale beautifully as 

hundreds of millions of new users signed on during the 1990s, and three types 

of controls seemed adequate to address emerging dangers. 

First, the hacker ethos frowns upon destructive hacking. Most viruses that 

followed Morris’s worm had completely innocuous payloads: In 2004, Mydoom 

spread like wildfire and reputedly cost billions in lost productivity, but the worm 

did not tamper with data, and it was programmed to stop spreading at a set 

time. With rare exceptions like the infamous Lovebug worm, which overwrote 

files with copies of itself, the few highly malicious viruses that run contrary to 

the hacker ethos were so poorly coded that they failed to spread very far. 

Second, network operations centers at universities and other institutions 

became more professionalized between 1988 and the advent of the mainstream 

Internet. For a while, most Internet-connected computers were staffed by 

professionals, administrators who generally heeded admonitions to patch 

regularly and scout for security breaches. Less adept mainstream consumers 

began connecting unsecured PCs to the Internet in earnest only in the mid- 

1990s. Then, transient dial-up connections greatly limited both the amount of 

time during which they were exposed to security threats, and the amount of 

time that, if compromised and hijacked, they would contribute to the problem. 

Finally, bad code lacked a business model. Programs to trick users into 

installing them, or to sneak onto the machines, were written for amusement. 

Bad code was more like graffiti than illegal drugs: There were no economic 

incentives for its creation. 

Today each of these controls has weakened. With the expansion of the 

community of users, the idea of a set of ethics governing activity on the Internet 

has evaporated. Anyone is allowed online if he or she can find a way to a


computer and a connection, and mainstream users are transitioning rapidly to 

always-on broadband connections. 

Moreover, PC user awareness of security issues has not kept pace with 

broadband growth. A December 2005 online safety study found 81% of home 

computers to be lacking first-order protection measures such as current 

antivirus software, spyware protection, and effective firewalls. 1 

Perhaps most significantly, bad code is now a business. What seemed genuinely 

remarkable when first discovered is now commonplace: Viruses that 

compromise PCs to create large zombie “botnets” open to later instructions. 

Such instructions have included directing PCs to become remotely-controlled email 

servers, sending spam by the thousands or millions to e-mail addresses 

harvested from the hard disk of the machines themselves or gleaned from 

Internet searches, with the entire process typically proceeding behind the back 

of the PCs’ owners. At one point, a single botnet occupied fifteen percent of 

Yahoo!’s search capacity, running random searches on Yahoo! to find text that 

could be inserted into spam e-mails to throw off spam filters. 2 Dave Dagon, 

who recently left Georgia Tech University to start a bot-fighting company 

named Damballa, pegs the number of botnet-infected computers at close to 30 

million. 3 Dagon said, “Had you told me five years ago that organized crime 

would control one out of every ten home machines on the Internet, I would not 

have believed that.” 4 So long as spam remains profitable, that crime will persist. 

Botnets can also be used to launch coordinated attacks on a particular Internet 

endpoint. For example, a criminal can attack an Internet gambling Web site and 

then extort payment to make the attacks stop. The going rate for a botnet to 

launch such an attack is reputed to be about $5,000 per day. 5 

Viruses are thus valuable properties. Well-crafted worms and viruses routinely 

infect vast swaths of Internet-connected personal computers. Antivirus vendor 

Eugene Kaspersky of Kaspersky Labs told an industry conference that they 

“may not be able to withstand the onslaught.” 6 IBM’s Internet Security Systems 

1 AOL/NCSA ONLINE SAFETY STUDY 2 (Dec. 2005), 

http://www.bc.edu/content/dam/files/offices/help/pdf/safety_study_2005.pdf 

2 Tim Weber, Criminals ‘May Overwhelm the Web,’ BBC NEWS, Jan. 25, 2007, 

http://news.bbc.co.uk/2/hi/business/6298641.stm. 

3 Bob Sullivan, Is Your Computer a Criminal?, RED TAPE CHRONICLES, Mar. 27, 2007, 

http://redtape.msnbc.com/2007/03/bots_story.html. 

4 Id. 

5 Id. 

6 Id.


reported a 40% increase in software vulnerabilities reported by manufacturers 

and “white hat” hackers between 2005 and 2006. 7 Nearly all of those 

vulnerabilities could be exploited remotely, and over half allowed attackers to 

gain full access to the machine and its contents. 

As the supply of troubles has increased, the capacity to address them has 

steadily diminished. Patch development time increased throughout 2006 for all 

of the top operating system providers. 8 Times shortened modestly across the 

board in the first half of 2007, but, on average, enterprise vendors were still 

exposed to vulnerabilities for 55 days—plenty of time for hazardous code to 

make itself felt. 9 (The patch intervals for browsers tend to be shorter than 

those for operating systems.) What is more, antivirus researchers and firms 

require extensive coordination efforts simply to agree on a naming scheme for 

viruses as they emerge. 10 This is a far cry from a common strategy for battling 

them. 

In addition, the idea of casually cleaning a virus off a PC is gone. When 

computers are compromised, users are now typically advised to reinstall 

everything on them. For example, in 2007, some PCs at the U.S. National 

Defense University fell victim to a virus. The institution shut down its network 

servers for two weeks and distributed new laptops to instructors. 11 In the 

absence of such drastic measures, a truly “mal” piece of malware could be 

programmed to, say, erase hard drives, transpose numbers inside spreadsheets 

randomly, or intersperse nonsense text at arbitrary intervals in Word documents 

found on infected computers—and nothing would stand in the way. 

Recognition of these basic security problems has been slowly growing in 

Internet research communities. Nearly two-thirds of academics, social analysts, 

and industry leaders surveyed by the Pew Internet & American Life Project in 

2004 predicted serious attacks on network infrastructure or the power grid in 

7 IBM INTERNET SECURITY SYSTEMS, X-FORCE 2006 TREND STATISTICS (Jan. 2007), 

http://www.iss.net/documents/whitepapers/X_Force_Exec_Brief.pdf. 

8 SYMANTEC, GLOBAL INTERNET SECURITY THREAT REPORT, TRENDS FOR JULY-DECEMBER 

2007 at 24-28 (April 2008), 

http://eval.symantec.com/mktginfo/enterprise/white_papers/bwhitepaper_internet_security_threat_report_xiii_04-2008.en-us.pdf. 

9 Id. at 6. 

10 See, e.g., Common Malware Enumeration: Reducing Public Confusion During Malware 

Outbreaks, http://cme.mitre.org/ (last visited June 1, 2007). 

11 Bill Gertz & Rowan Scarborough, Inside the Ring—Notes from the Pentagon, WASH. TIMES, Jan. 5, 

2007, at A5, available at http://www.gertzfile.com/gertzfile/ring011207.html.


the coming decade. 12 Security concerns will lead to a fundamental shift in our 

tolerance of the status quo, either by a catastrophic episode, or, more likely, a 

glacial death of a thousand cuts. 

Consider, in the latter scenario, the burgeoning realm of “badware” (or 

“malware”) beyond viruses and worms: Software that is often installed at the 

user’s invitation. The popular file-sharing program KaZaA, though advertised 

as “spyware-free,” contains code that users likely do not want. It adds icons to 

the desktop, modifies Microsoft Internet Explorer, and installs a program that 

cannot be closed by clicking “Quit.” Uninstalling the program does not 

uninstall all these extras, and the average user does not know how to get rid of 

the code itself. What makes such badware “bad” has to do with the level of 

disclosure made to a consumer before he or she installs it. The most common 

responses to the security problem cannot easily address this gray zone of 

software. 

Many technologically savvy people think that bad code is simply a Microsoft 

Windows issue. They believe that the Windows OS and the Internet Explorer 

browser are particularly poorly designed, and that “better” counterparts 

(GNU/Linux and Mac OS, or the Firefox and Opera browsers) can help shield 

a user. But the added protection does not get to the fundamental problem, 

which is that the point of a PC—regardless of its OS—is to enable its users to 

easily reconfigure it to run new software from anywhere. When users make 

poor decisions about what software to run, the results can be devastating to 

their machines and, if they are connected to the Internet, to countless others’ 

machines as well. 

The cybersecurity problem defies easy solution because any of its most obvious 

fixes will undermine the generative essence of the Internet and PC. Bad code is 

an inevitable side effect of generativity, and as PC users are increasingly 

victimized by bad code, consumers are likely to reject generative PCs in favor of 

safe information appliances—digital video recorders, mobile phones, iPods, 

BlackBerrys, and video game consoles—that optimize a particular application 

and cannot be modified by users or third-parties. It is entirely reasonable for 

consumers to factor security and stability into their choice. But it is an 

undesirable choice to have to make. 

* * * 

On January 9, 2007, Steve Jobs introduced the iPhone to an eager audience 

crammed into San Francisco’s Moscone Center. A beautiful and brilliantly 

12 Susannah Fox et al., The Future of the Internet: In a Survey, Technology Experts and Scholars Evaluate 

Where the Network Is Headed in the Next Few Years, Jan. 9, 2005, at i, 

http://www.pewinternet.org/PPF/r/145/report_display.asp.


engineered device, the iPhone blended three products into one: an iPod, with 

the highest-quality screen Apple had ever produced; a phone, with cleverly 

integrated functionality, such as voicemail that came wrapped as separately 

accessible messages; and a device to access the Internet, with a smart and 

elegant browser, and built-in map, weather, stock, and e-mail capabilities. 

Steve Jobs had no clue how the Apple II would be used. The iPhone—for all 

its startling inventiveness—is precisely the opposite. 

This was Steve Jobs’s second revolution. Thirty years earlier, at the First West 

Coast Computer Faire in nearly the same spot, the twenty-one-year-old Jobs, 

wearing his first suit, exhibited the Apple II personal computer to great buzz 

amidst “ten thousand walking, talking computer freaks.” 13 The Apple II was a 

machine for hobbyists who did not want to fuss with soldering irons: all the 

ingredients for a functioning PC were provided in a convenient molded plastic 

case. Instead of puzzling over bits of hardware or typing up punch cards to 

feed into someone else’s mainframe, Apple owners faced only the hurdle of a 

cryptic blinking cursor in the upper left corner of the screen: the PC awaited 

instructions. But the hurdle was not high. Some owners were inspired to 

program the machines themselves, but beginners, too, could load software 

written and then shared or sold by their more skilled counterparts. The Apple 

II was a blank slate, a bold departure from previous technology that had been 

developed and marketed to perform specific tasks. 

The Apple II quickly became popular. And when programmer and 

entrepreneur Dan Bricklin introduced the first killer application for the Apple II 

in 1979—VisiCalc, the world’s first spreadsheet program—sales of the ungainly 

but very cool machine took off. An Apple running VisiCalc helped to convince 

a skeptical world that there was a place for the PC on everyone’s desk. 

The Apple II was quintessentially generative technology. It was a platform. It 

invited people to tinker with it. Hobbyists wrote programs. Businesses began 

to plan on selling software. Jobs (and Apple) had no clue how the machine 

would be used. They had their hunches, but, fortunately for them (and the rest 

of us), nothing constrained the PC to the hunches of the founders. 

The iPhone—for all its startling inventiveness—is precisely the opposite. 

Rather than a platform that invites innovation, the iPhone comes 

preprogrammed. In its first version, you were not allowed to add programs to 

the all-in-one device that Steve Jobs sells you except via the Siberia of its Web 

13 David H. Ahl, The First West Coast Computer Faire, in 3 THE BEST OF CREATIVE COMPUTING 

98 (David Ahl & Burchenal Green eds., 1980), available at 

http://www.atariarchives.org/bcc3/showpage.php?page_98.


browser. Its functionality was locked in, though Apple could change it through 

remote updates. Indeed, those who managed to tinker with the code and enable 

iPhone-support of more or different applications, were on the receiving end of 

Apple’s threat to transform the iPhone into an iBrick. 14 A threat, to be sure, 

that Apple later at least partially disavowed. The machine was not to be 

generative beyond the innovations that Apple (and its exclusive carrier, AT&T) 

wanted. In its second version a year later, the iPhone boasted the App Store. 

Software developers could code for the phone – but the developers, and then 

each piece of software, would require approval from Apple before it could be 

made available to iPhone users. Apple would receive a 30% cut of sales, 

including “in-app” sales of upgrades, and an app could be banned retroactively 

after initial approval. This made the iPhone “contingently generative,” a hybrid 

status that, depending on how you look at it, is either the best or the worst of 

both worlds: a melding of the sterile and the generative. 

Jobs was not shy about these restrictions. As he said at the iPhone launch: “We 

define everything that is on the phone …. You don’t want your phone to be 

like a PC. The last thing you want is to have loaded three apps on your phone 

and then you go to make a call and it doesn’t work anymore.” 15 

In the arc from the Apple II to the iPhone, we learn something important about 

where the Internet has been, and something even more important about where 

it is going. The PC revolution was launched with PCs that invited innovation 

by others. So, too, with the Internet. Both were designed to accept any 

contribution that followed a basic set of rules (either coded for a particular 

operating system, or respecting the protocols of the Internet). Both 

overwhelmed their respective proprietary, non-generative competitors: PCs 

crushed stand-alone word processors and the Internet displaced such 

proprietary online services as CompuServe and AOL. 

But the future is looking very different because of the security situation—not 

generative PCs attached to a generative network, but appliances tethered to a 

network of control. These appliances take the innovations already created by 

Internet users and package them neatly and compellingly, which is good—but 

only if the Internet and PC can remain sufficiently central in the digital 

ecosystem to compete with locked-down appliances and facilitate the next 

round of innovations. The balance between the two spheres is precarious, and 

it is slipping toward the safer appliance. For example, Microsoft’s Xbox 360 

14 Michael, Apple Says It May “Brick” Unlocked iPhones With Next Software Update, Apple Gazette, 

Sep. 24, 2007, http://www.applegazette.com/iphone/apple-says-it-may-brickunlocked-iphones-with-next-software-update/. 

15 See John Markoff, Steve Jobs Walks the Tightrope Again, N.Y. TIMES, Jan. 12, 2007, available at 

http://www.nytimes.com/2007/01/12/technology/12apple.html.


video game console is a powerful computer, but, unlike Microsoft’s Windows 

operating system for PCs, it does not allow just anyone to write software that 

can run on it – games must be licensed by Microsoft. Bill Gates sees the Xbox 

at the center of the future digital ecosystem, rather than its periphery: “It is a 

general purpose computer . . . [W]e wouldn’t have done it if it was just a 

gaming device. We wouldn’t have gotten into the category at all. It was about 

strategically being in the living room.” 16 

Devices like iPhones and Xbox 360s may be safer to use, and they may seem 

capacious in features so long as they offer a simple Web browser. But by 

focusing on security and limiting the damage that users can do through their 

own ignorance or carelessness, these appliances also limit the beneficial tools 

that users can create or receive from others—enhancements they may be 

clueless about when they are purchasing the device. 

If the PC ceases to be at the center of the information 

technology ecosystem, the most restrictive aspects of 

information appliances will come to the fore. 

Security problems related to generative PC platforms may propel people away 

from PCs and toward information appliances controlled by their makers. If we 

eliminate the PC from many dens or living rooms, we eliminate the test bed and 

distribution point of new, useful software from any corner of the globe. We 

also eliminate the safety valve that keeps those information appliances honest. 

If TiVo makes a digital video recorder that has too many limits on what people 

can do with the video they record, people will discover DVR software like 

MythTV that records and plays TV shows on their PCs. If mobile phones are 

too expensive, people will use Skype. But people do not buy PCs as insurance 

policies against appliances that limit their freedoms, even though PCs serve 

exactly this vital function. People buy them to perform certain tasks at the 

moment of acquisition. If PCs cannot reliably perform these tasks, most 

consumers will not see their merit, and the safety valve will be lost. If the PC 

ceases to be at the center of the information technology ecosystem, the most 

restrictive aspects of information appliances will come to the fore. 

In fact, the dangers may be more subtly packaged. PCs need not entirely 

disappear as people buy information appliances in their stead. PCs can 

themselves be made less generative. Users tired of making the wrong choices 

about installing code on their PCs might choose to let someone else decide 

what code should be run. Firewalls can protect against some bad code, but they 

also complicate the installation of new good code. As antivirus, antispyware, 

16 Ryan Block, A Lunchtime Chat with Bill Gates, ENGADGET, Jan. 8, 2007, 

http://www.engadget.com/2007/01/08/a-lunchtime-chat-with-bill-gates-at-ces/.


and anti-badware barriers proliferate, there are new barriers to the deployment 

of new good code from unprivileged sources. And in order to guarantee 

effectiveness, these barriers are becoming increasingly paternalistic, refusing to 

allow users easily to overrule them. Especially in environments where the user 

of the PC does not own it—offices, schools, libraries, and cyber-cafés—barriers 

are being put in place to prevent the running of any code not specifically 

approved by the relevant gatekeeper. Users may find themselves limited to 

using a Web browser. And while “Web 2.0” promises many more uses for a 

browser—consumers can now write papers and use spreadsheets through a 

browser, and software developers now write for Web platforms like Facebook 

instead of PC operating systems —these Web platforms are themselves tethered 

to their makers, their generativity contingent on the continued permission of the 

platform vendors. 

Short of completely banning unfamiliar software, code might be divided into 

first- and second-class status, with second-class, unapproved software allowed 

to perform only certain minimal tasks on the machine, operating within a digital 

sandbox. This technical solution is safer than the status quo but imposes 

serious limits. It places the operating system creator or installer in the position 

of deciding what software will and will not run. The PC will itself have become 

an information appliance, not easily reconfigured or extended by its users. 

The key to avoiding such a future is to give the market a reason not to abandon 

or lock down the PCs that have served it so well, also giving most governments 

reason to refrain from major intervention into Internet architecture in the name 

of public safety. The solutions to the generative dilemma will rest on social and 

legal as much as technical innovation, and the best guideposts can be found in 

other generative successes in those arenas. Mitigating abuses of openness 

without resorting to lockdown will depend on a community ethos embodied in 

responsible groups with shared norms and a sense of public purpose, rather 

than in the hands of a single gatekeeper, whether public or private. 

In the medium term, the battle between generative and sterile will be played out 

between the iPhone and Android, which despite its own version of an App 

Store, also allows outside code to run that doesn’t come from the store; and 

with projects like Boxee and Google TV, which are seeking to bridge the gap 

between the PC and the living room. Each device sets the dial set at a different 

point between complete “open” and completely “closed.” And those dials can 

shift: after a security “spill,” Android could be reprogrammed overnight to be 

more restrictive in the code it runs; and by the same token, Apple could decide 

to loosen its restrictions on iPhone code. 

* * * 

We need a strategy that addresses the emerging security troubles of today’s 

Internet and PCs without killing their openness to innovation. This is easier


said than done, because our familiar legal tools are not particularly attuned to 

maintaining generativity. A simple regulatory intervention—say, banning the 

creation or distribution of deceptive or harmful code—will not work because it 

is hard to track the identities of sophisticated wrongdoers, and, even if found, 

many may not be in cooperative jurisdictions. Moreover, such intervention may 

have a badly chilling effect: Much of the good code we have seen has come 

from unaccredited people sharing what they have made for fun, collaborating in 

ways that would make business-like regulation of their activities burdensome for 

them. They might be dissuaded from sharing at all. 

We can find a balance between needed change and undue restriction if we think 

about how to move generative approaches and solutions that work at one 

“layer” of the Internet—content, code, or technical—to another. Consider 

Wikipedia, the free encyclopedia whose content—the entries and their 

modifications—is fully generated by the Web community. The origins of 

Wikipedia lie in the open architecture of the Internet and Web. This allowed 

Ward Cunningham to invent the wiki, generic software that offers a way of 

editing or organizing information within an article, and spreading this 

information to other articles. Unrelated non-techies then used Wikis to form 

Web sites at the content layer, including Wikipedia. People are free not only to 

edit Wikipedia, but to take all of its contents and experiment with different ways 

of presenting or changing the material, perhaps by placing the information on 

otherwise unrelated Web sites in different formats. When abuses of this 

openness beset Wikipedia with vandalism, copyright infringement, and lies, it 

turned to its community—aided by some important technical tools—as the 

primary line of defense, rather than copyright or defamation law. Most recently, 

this effort has been aided by the introduction of Virgil Griffith’s Wikiscanner, a 

simple tool that uses Wikipedia’s page histories to expose past instances of 

article whitewashing by interested parties. 

Unlike a form of direct regulation that would have locked down the site, the 

Wikipedian response so far appears to have held many of Wikipedia’s problems 

at bay. Why does it work so well? Generative solutions at the content layer 

seem to have two characteristics that suggest broad approaches to lowering the 

risks of the generative Internet while preserving its openness. First, much 

participation in generating Web content—editing Wikipedia entries, blogging, or 

even engaging in transactions on eBay and Amazon that ask for reviews and 

ratings to establish reputations—is understood to be an innately social activity. 

These services solicit and depend upon participation from the public, and their 

participation mechanisms are easily mastered. The same possibility for broad 

participation exists one level down at the technical layer, but it has not yet been 

as fully exploited: Mainstream users have thus far been eager to have someone 

else solve underlying problems, which they perceive as technical rather than 

social. Second, many content-layer enterprises have developed technical tools 

to support collective participation, augmenting an individualistic ethos with


community-facilitating structures. In the Internet and PC security space, on the 

other hand, there have been few tools available to tap the power of groups of 

users to, say, distinguish good code from bad. 

The effectiveness of the social layer in Web successes points to two approaches 

that might save the generative spirit of the Net, or at least keep it alive for 

another interval. The first is to reconfigure and strengthen the Net’s 

experimentalist architecture to make it fit better with the vast expansion in the 

number and types of users. The second is to develop new tools and practices 

that will enable relevant people and institutions to help secure the Net 

themselves instead of waiting for someone else to do it. 

Generative PCs with Easy Reversion 

Wikis are designed so that anyone can edit them. This creates a genuine and 

ongoing risk of bad edits, through either incompetence or malice. The damage 

that can be done, however, is minimized by the wiki technology, because it 

allows bad changes to be quickly reverted. All previous versions of a page are 

kept, and a few clicks by another user can restore a page to the way it was 

before later changes were made. So long as there are more users (and 

automated tools they create) detecting and reverting vandalism than there are 

users vandalizing, the community wins. (Truly, the price of freedom is eternal 

vigilance.) 

Our PCs can be similarly equipped. For years Windows XP (and now Vista) 

has had a system restore feature, where snapshots are taken of the machine at a 

moment in time, allowing later bad changes to be rolled back. The process of 

restoring is tedious, restoration choices can be frustratingly all-or-nothing, and 

the system restoration files themselves can become corrupted, but it represents 

progress. Even better would be the introduction of features that are 

commonplace on wikis: A quick chart of the history of each document, with an 

ability to see date-stamped sets of changes going back to its creation. Because 

our standard PC applications assume a safer environment than really exists, 

these features have never been demanded or implemented. Because wikis are 

deployed in environments prone to vandalism, their contents are designed to be 

easily recovered after a problem. 

The next stage of this technology lies in new virtual machines, which would 

obviate the need for cyber cafés and corporate IT departments to lock down 

their PCs. 

In an effort to satisfy the desire for safety without full lockdown, PCs can be 

designed to pretend to be more than one machine, capable of cycling from one 

personality to the next. In its simplest implementation, we could divide a PC 

into two virtual machines: “Red” and “Green.” The Green PC would house 

reliable software and important data—a stable, mature OS platform and tax


returns, term papers, and business documents. The Red PC would have 

everything else. In this setup, nothing that happens on one PC can easily affect 

the other, and the Red PC could have a simple reset button that restores a 

predetermined safe state. Someone could confidently store important data on 

the Green PC and still use the Red PC for experimentation. This isn’t rocket 

science – there’s already software out there to amount to a Green/Red divide 

on a Windows machine – but it’s not so easy for the average user to deploy and 

use. 

Easy, wiki-style reversion, coupled with virtual PCs, would accommodate the 

experimentalist spirit of the early Internet while acknowledging the important 

uses for those PCs that we do not want to disrupt. Still, this is not a complete 

solution. The Red PC, despite its experimental purpose, might end up 

accumulating data that the user wants to keep, occasioning the need for what 

Internet architect David D. Clark calls a “checkpoint Charlie” to move sensitive 

data from Red to Green without also carrying a virus or anything else 

undesirable. There is also the question of what software can be deemed safe for 

Green—which is just another version of the question of what software to run 

on today’s single-identity PCs. 

For these and related reasons, virtual machines will not be panaceas, but they 

might buy us some more time. And they implement a guiding principle from 

the Net’s history: an experimentalist spirit is best maintained when failures can 

be contained as learning experiences rather than expanding to catastrophes. 

A Generative Solution to Bad Code 

The Internet’s original design relied on few mechanisms of central control. 

This lack of control has the generative benefit of allowing new services to be 

introduced, and new destinations to come online, without any up-front vetting 

or blocking by either private incumbents or public authorities. With this 

absence of central control comes an absence of measurement. The Internet 

itself cannot say how many users it has, because it does not maintain user 

information. There is no awareness at the network level of how much 

bandwidth is being used by whom. From a generative point of view this is 

good because it allows initially whimsical but data-intensive uses of the network 

to thrive (remember goldfish cams?)—and perhaps to become vital (nowroutine 

videoconferencing through Skype, from, unsettlingly, the makers of 

KaZaA). 

Because we cannot easily measure the network and the 

character of the activity on it, we cannot easily assess and deal 

with threats from bad code without laborious and imperfect 

cooperation among a limited group of security software 

vendors.


But limited measurement is starting to have generative drawbacks. Because we 

cannot easily measure the network and the character of the activity on it, we 

cannot easily assess and deal with threats from bad code without laborious and 

imperfect cooperation among a limited group of security software vendors. The 

future of the generative Net depends on a wider circle of users able to grasp the 

basics of what is going on within their machines and between their machines 

and the network. 

What might this system look like? Roughly, it would take the form of toolkits 

to overcome the digital solipsism that each of our PCs experiences when it 

attaches to the Internet at large, unaware of the size and dimension of the 

network to which it connects. These toolkits would run unobtrusively on the 

PCs of participating users, reporting back—to a central source, or perhaps only 

to each other—information about the vital signs and running code of that PC, 

which could help other PCs determine the level of risk posed by new code. 

When someone is deciding whether to run new software, the toolkit’s 

connections to other machines could tell the person how many other machines 

on the Internet are running the code, what proportion of machines belonging to 

self-described experts are running it, whether those experts have vouched for it, 

and how long the code has been in the wild. 

Building on these ideas about measurement and code assessment, Harvard 

University’s Berkman Center and the Oxford Internet Institute— 

multidisciplinary academic enterprises dedicated to charting the future of the 

Net and improving it—have begun a project called StopBadware 

(www.stopbadware.org), designed to assist rank-and-file Internet users in 

identifying and avoiding bad code. The idea is not to replicate the work of 

security vendors like Symantec and McAfee, which, for a fee, seek to bail new 

viruses out of our PCs faster than they pour in. Rather, these academic groups 

are developing a common technical and institutional framework that enables 

users to devote some bandwidth and processing power for better measurement 

of the effect of new code. A first step in the toolkit was developed as “Herdict 

PC.” Herdict PC was a small piece of software that assembles vital signs like 

number of pop-up windows or crashes per hour. [It incorporates that data into 

a dashboard usable by mainstream PC owners. Efforts like Herdict – including 

such ventures as Soluto (www.soluto.com) – will test the idea that solutions 

that have worked for generating content might also be applicable to the 

technical layer. Such a system might also illuminate Internet filtering by 

governments around the world, as people participate in a system where they can 

report when they cannot access a Web site, and such reports can be collated by 

geography. 

A full adoption of the lessons of Wikipedia would give PC users the 

opportunity to have some ownership, some shared stake, in the process of 

evaluating code, especially because they have a stake in getting it right for their


own machines. Sharing useful data from their PCs is one step, but this may 

work best when the data goes to an entity committed to the public interest of 

solving PC security problems and willing to share that data with others. The 

notion of a civic institution here does not necessarily mean cumbersome 

governance structures and formal lines of authority so much as it means a sense 

of shared responsibility and participation. Think of the volunteer fire 

department or neighborhood watch: While not everyone is able to fight fires or 

is interested in watching, a critical mass of people are prepared to contribute, 

and such contributions are known to the community more broadly. 

The success of tools drawing on group generativity depends on participation, 

which helps establish the legitimacy of the project both to those participating 

and those not. Internet users might see themselves only as consumers whose 

purchasing decisions add up to a market force, but, with the right tools, users 

can also see themselves as participants in the shaping of generative space—as 

netizens. 

Along with netizens, hardware and software makers could also get involved. 

OS makers could be asked or required to provide basic tools of transparency 

that empower users to understand exactly what their machines are doing. These 

need not be as sophisticated as Herdict. They could provide basic information 

on what data is going in and out of the box and to whom. Insisting on getting 

better information to users could be as important as providing a speedometer or 

fuel gauge on an automobile—even if users do not think they need one. 

Internet Service Providers (ISPs) can also reasonably be asked or required to 

help. Thus far, ISPs have been on the sidelines regarding network security. 

The justification is that the Internet was rightly designed to be a dumb network, 

with most of its features and complications pushed to the endpoints. The 

Internet’s engineers embraced the simplicity of the end-to-end principle for 

good reasons. It makes the network more flexible, and it puts designers in a 

mindset of making the system work rather than designing against every possible 

thing that could go wrong. Since this early architectural decision, “keep the 

Internet free” advocates have advanced the notion of end-to-end neutrality as 

an ethical ideal, one that leaves the Internet without filtering by any of its 

intermediaries, routing packets of information between sender and recipient 

without anyone looking along the way to see what they contain. Cyberlaw 

scholars have taken up end-to-end as a battle cry for Internet freedom, invoking 

it to buttress arguments about the ideological impropriety of filtering Internet 

traffic or favoring some types or sources of traffic over others. 

End-to-end neutrality has indeed been a crucial touchstone for Internet 

development. But it has limits. End-to-end design preserves users’ freedom 

only because the users can configure their own machines however they like. 

But this depends on the increasingly unreliable presumption that whoever runs


a machine at a given network endpoint can readily choose how the machine 

should work. Consider that in response to a network teeming with viruses and 

spam, network engineers recommend more bandwidth (so the transmission of 

“deadweights” like viruses and spam does not slow down the much smaller 

proportion of legitimate mail being carried by the network) and better 

protection at user endpoints. But users are not well positioned to painstakingly 

maintain their machines against attack, and intentional inaction at the network 

level may be self-defeating, because consumers may demand locked-down 

endpoint environments that promise security and stability with minimum user 

upkeep. 

Strict loyalty to end-to-end neutrality should give way to a new principle asking 

that any modifications to the Internet’s design or the behavior of ISPs be made 

in such a way that they will do the least harm to generative possibilities. Thus, it 

may be preferable in the medium-term to screen-out viruses through ISPoperated 

network gateways rather than through constantly updated PCs. To be 

sure, such network screening theoretically opens the door to undesirable 

filtering. But we need to balance this speculative risk against the growing threat 

to generativity. ISPs are in a good position to help in a way that falls short of 

undesirable perfect enforcement facilitated through endpoint lockdown, by 

providing a stopgap while we develop the kinds of community-based tools that 

can promote salutary endpoint screening. 

Even search engines can help create a community process that has impact. In 

2006, in cooperation with the Harvard and Oxford StopBadware initiative, 

Google began automatically identifying Web sites that had malicious code 

hidden in them, ready to infect browsers. Some of these sites were set up for 

the purpose of spreading viruses, but many more were otherwise-legitimate 

Web sites that had been hacked. For example, visitors to chuckroast.com can 

browse fleece jackets and other offerings and place and pay for orders. 

However, Google found that hackers had subtly changed the chuckroast.com 

code: The basic functionalities were untouched, but code injected on the home 

page would infect many visitors’ browsers. Google tagged the problem, and 

appended to the Google search result: “Warning: This site may harm your 

computer.” Those who clicked on the results link anyway would get an 

additional warning from Google and the suggestion to visit StopBadware or 

pick another page. 

The site’s traffic plummeted, and the owner (along with the thousands of others 

whose sites were listed) was understandably anxious to fix it. But cleaning a 

hacked site takes more than an amateur Web designer. Requests for specialist 

review inundated StopBadware researchers. Until StopBadware could check 

each site and verify it had been cleaned of bad code, the warning pages stayed 

up. Prior to the Google/StopBadware project, no one took responsibility for 

this kind of security. Ad hoc alerts to the hacked sites’ webmasters—and their


ISPs—garnered little reaction. The sites were fulfilling their intended purposes 

even as they were spreading viruses to visitors. With Google/StopBadware, 

Web site owners have experienced a major shift in incentives for keeping their 

sites clean. 

The result is perhaps more powerful than a law that would have directly 

regulated them, and it could in turn generate a market for firms that help 

validate, clean, and secure Web sites. Still, the justice of Google/StopBadware 

and similar efforts remains rough, and market forces alone might not direct the 

desirable level of attention to those wrongly labeled as people or Web sites to be 

avoided, or properly labeled but with no place to seek help. 

The touchstone for judging such efforts is whether they reflect the generative 

principle: Do the solutions arise from and reinforce a system of 

experimentation? Are the users of the system able, so far as they are interested, 

to find out how the resources they control—such as a PC—are participating in 

the environment? Done well, these interventions can encourage even casual 

users to have some part in directing what their machines will do, while securing 

those users’ machines against outsiders who have not been given permission by 

the users to make use of them. Automatic accessibility by outsiders—whether 

by vendors, malware authors, or governments—can deprive a system of its 

generative character as its users are limited in their own control. 

Data Portability 

The generative Internet was founded and cultivated by people and institutions 

acting outside traditional markets, and later carried forward by commercial 

forces. Its success requires an ongoing blend of expertise and contribution 

from multiple models and motivations. Ultimately, a move by the law to 

allocate responsibility to commercial technology players in a position to help 

but without economic incentive to do so, and to those among us, commercially 

inclined or not, who step forward to solve the pressing problems that elude 

simpler solutions may also be in order. How can the law be shaped if one wants 

to reconcile generative experimentation with other policy goals beyond 

continued technical stability? The next few proposals are focused on this 

question about the constructive role of law. 

One important step is making locked-down appliances and Web 2.0 softwareas-a-service 

more palatable. After all, they are here to stay, even if the PC and 

Internet are saved. The crucial issue here is that a move to tethered appliances 

and Web services means that more and more of our experiences in the 

information space will be contingent: A service or product we use at one 

moment could act completely differently the next, since it can be so quickly 

reprogrammed by the provider without our assent. Each time we power up a 

mobile phone, video game console, or BlackBerry, it might have gained some


features and lost others. Each time we visit a Web site offering an ongoing 

service like e-mail access or photo storage, the same is true. 

As various services and applications become more self-contained within 

particular devices, there is a minor intervention the law could make to avoid 

undue lock-in. Online consumer protection law has included attention to 

privacy policies. A Web site without a privacy policy, or one that does not live 

up to whatever policy it posts, is open to charges of unfair or deceptive trade 

practices. Similarly, makers of tethered appliances and Web sites keeping 

customer data ought to be asked to offer portability policies. These policies 

would declare whether users will be allowed to extract their data should they 

wish to move their activities from one appliance or Web site to another. In 

some cases, the law could create a right of data portability, in addition to merely 

insisting on a clear statement of a site’s policies. 

A requirement of data portability is a generative insurance policy applying to 

individual data wherever it might be stored. And the requirement need not be 

onerous. It could apply only to uniquely provided personal data such as photos 

and documents, and mandate only that such data ought to readily be extractable 

by the user in some standardized form. Maintaining data portability will help 

people pass back and forth between the generative and the non-generative, and, 

by permitting third-party backup, it will also help prevent a situation in which a 

non-generative service suddenly goes offline, with no recourse for those who 

have used the service to store their data. 

Appliance Neutrality 

Reasonable people disagree on the value of defining and legally mandating 

network neutrality. But if there is a present worldwide threat to neutrality in the 

movement of bits, it comes from enhancements to traditional and emerging 

“appliancized” services like Google mash-ups and Facebook apps, in which the 

service provider can be pressured to modify or kill others’ applications on the 

fly. Surprisingly, parties to the network neutrality debate—who have focused 

on ISPs—have yet to weigh in on this phenomenon. 

In the late 1990’s, Microsoft was found to possess a monopoly in the market for 

PC operating systems. 17 Indeed, it was found to be abusing that monopoly to 

favor its own applications—such as its Internet Explorer browser—over thirdparty 

software, against the wishes of PC makers who wanted to sell their 

hardware with Windows preinstalled but adjusted to suit the makers’ tastes. 

Microsoft was forced by the law to meet ongoing requirements to maintain a 

17 United States v. Microsoft Corp., 84 F. Supp. 2d 9, 19 (D.D.C. 1999).


level playing field between third-party software and its own by allowing thirdparty 

software to be pre-installed on new Windows computers. 

We have not seen the same requirements arising for appliances that do not 

allow, or strictly control, the ability of third parties to contribute from the start. 

So long as the market’s favorite video game console maker never opens the 

door to generative third-party code, it is hard to see how the firm could be 

found to be violating competition law. A manufacturer is entitled to make an 

appliance and to try to bolt down its inner workings so that they cannot be 

modified by others. So when should we consider network neutrality-style 

mandates for appliancized systems? The answer lies in that subset of 

appliancized systems that seeks to gain the generative benefits of third-party 

contribution at one point in time while reserving the right to exclude it later. 

The common law recognizes vested expectations. For example, the law of 

adverse possession dictates that people who openly occupy another’s private 

property without the owner’s explicit objection (or, for that matter, permission) 

can, after a lengthy period of time, come to legitimately acquire it. More 

commonly, property law can find prescriptive easements—rights-of-way across 

territory that develop by force of habit—if the owner of the territory fails to 

object in a timely fashion as people go back and forth across it. These and 

related doctrines point to a deeply held norm: Certain consistent behaviors can 

give rise to obligations, sometimes despite fine print that tries to prevent those 

obligations from coming about. 

Applied to the idea of application neutrality, this norm of protecting settled 

expectations might suggest the following: If Microsoft wants to make the Xbox 

a general purpose device but still not open to third-party improvement, no 

regulation should prevent it. But if Microsoft does welcome third-party 

contribution, it should not be able to subsequently impose barriers to outside 

software continuing to work. Such behavior is a bait-and-switch that is not easy 

for the market to anticipate and that stands to allow a platform maker to exploit 

habits of generativity to reach a certain plateau, dominate the market, and then 

make the result proprietary—exactly what the Microsoft Web browser case 

rightly was brought to prevent. 

The free software movement has produced some great works, 

but under prevailing copyright law even the slightest bit of 

“poison,” in the form of code from a proprietary source, could 

amount to legal liability for anyone who copies or even uses 

the software. 

Generative Software 

At the code layer, it is not easy for the law to maintain neutrality between the 

two models of software production that have emerged with the Net: Proprietary


software whose source code recipe is nearly always hidden, and free software— 

free not in terms of the price, but the openness of its code to public review and 

modification. The free software movement has produced some great works, 

but under prevailing copyright law even the slightest bit of “poison,” in the 

form of code from a proprietary source, could amount to legal liability for 

anyone who copies or even uses the software. These standards threaten the 

long-term flourishing of the free software movement: The risks are more 

burdensome than need be. 

But there are some changes to the law that would help. The kind of law that 

shields Wikipedia and Web site hosting companies from liability for 

unauthorized copyrighted material contributed by outsiders, at least so long as 

the organization acts expeditiously to remove infringing material once it is 

notified, ought to be extended to the production of code itself. Code that 

incorporates infringing material ought not be given a free pass, but those who 

have promulgated it without knowledge of the infringement would have a 

chance to repair the code or cease copying it before becoming liable. 

Modest changes in patent law could help as well. If those who see value in 

software patents are correct, infringement is rampant. And to those who think 

patents chill innovation, the present regime needs reform. To be sure, amateurs 

who do not have houses to lose to litigation can still contribute to free software 

projects—they are judgment proof. Others can contribute anonymously, 

evading any claims of patent infringement since they simply cannot be found. 

But this turns coding into a gray market activity, eliminating what otherwise 

could be a thriving middle class of contributing firms should patent warfare 

ratchet into high gear. 

The law can help level the playing field. For patent infringement in the United 

States, the statute of limitations is six years; for civil copyright infringement it is 

three. Unfortunately, this limit has little meaning for computer code because 

the statute of limitations starts from the time of the last infringement. Every 

time someone copies (or perhaps even runs) the code, the clock starts ticking 

again on a claim of infringement. This should be changed. The statute of 

limitations could be clarified for software, requiring that anyone who suspects 

or should suspect his or her work is being infringed sue within, for instance, one 

year of becoming aware of the suspect code. For example, the acts of those 

who contribute to free software projects—namely, releasing their code into a 

publicly accessible database like SourceForge—could be enough to start the 

clock ticking on that statute of limitations. In the absence of such a rule, 

lawyers who think their employers’ proprietary interests have been 

compromised can wait to sue until a given piece of code has become wildly 

popular—essentially sandbagging the process in order to let damages rack up.


Generative Licenses 

There is a parallel to how we think about balancing generative and sterile code 

at the content layer: Legal scholars Lawrence Lessig and Yochai Benkler, as well 

as others, have stressed that even the most rudimentary mixing of cultural icons 

and elements, including snippets of songs and video, can accrue thousands of 

dollars in legal liability for copyright infringement without harming the market 

for the original proprietary goods. 18 Benkler believes that the explosion of 

amateur creativity online has occurred despite this system. The high costs of 

copyright enforcement and the widespread availability of tools to produce and 

disseminate what he calls “creative cultural bricolage” currently allow for a 

variety of voices to be heard even when what they are saying is theoretically 

sanctionable by fines up to $30,000 per copy made, $150,000 if the infringement 

is done “willfully.” 19 As with code, the status quo shoehorns otherwise laudable 

activity into a sub-rosa gray zone. 

As tethered appliances begin to take up more of the information space, making 

information that much more regulable, we have to guard against the possibility 

that content produced by citizens who cannot easily clear permissions for all its 

ingredients will be squeezed out. Even the gray zone will constrict. 

* * * 

Regimes of legal liability can be helpful when there is a problem and no one has 

taken ownership of it. No one fully owns today’s problems of copyright 

infringement and defamation online, just as no one fully owns security problems 

on the Net. But the solution is not to conscript intermediaries to become the 

Net police. 

Under prevailing law, Wikipedia could get away with much less stringent 

monitoring of its articles for plagiarized work, and it could leave plainly 

defamatory material in an article but be shielded in the United States by the 

Communications Decency Act provision exempting those hosting material from 

responsibility for what others have provided. Yet Wikipedia polices itself 

according to an ethical code that encourages contributors to do the right thing 

rather than the required thing or the profitable thing. 

To harness Wikipedia’s ethical instinct across the layers of the generative 

Internet, we must figure out how to inspire people to act humanely in digital 

environments. This can be accomplished with tools—some discussed above, 

others yet to be invented. For the generative Internet to come fully into its 

18 LAWRENCE LESSIG, REMIX: MAKING ART AND COMMERCE THRIVE IN A HYBRID ECONOMY 

(2008); YOCHAI BENKLER, THE WEALTH OF NETWORKS (2006). 

19 YOCHAI BENKLER, THE WEALTH OF NETWORKS 275 (2006).


own, it must allow us to exploit the connections we have with each other. Such 

tools allow us to express and live our civic instincts online, trusting that the 

expression of our collective character will be one at least as good as that 

imposed by outside sovereigns—sovereigns who, after all, are only people 

themselves. 

Our generative technologies need technically skilled people of good will to keep 

them going, and the fledgling generative activities—blogging, wikis, social 

networks—need artistically and intellectually skilled people of goodwill to serve 

as true alternatives to a centralized, industrialized information economy that 

asks us to identify only as consumers of meaning rather than as makers of it. 

The deciding factor in whether our current infrastructure can endure will be the 

sum of the perceptions and actions of its users. Traditional state sovereigns, 

pan-state organizations, and formal multi-stakeholder regimes have roles to 

play. They can reinforce conditions necessary for generative blossoming, and 

they can also step in when mere generosity of spirit cannot resolve conflict. But 

that generosity of spirit is a society’s crucial first line of moderation. 

Our fortuitous starting point is a generative device on a neutral Net in tens of 

millions of hands. Against the trend of sterile devices and services that will 

replace the PC and Net stand new architectures like those of Boxee and 

Android. To maintain that openness, the users of those devices must 

experience the Net as something with which they identify and belong. We must 

use the generativity of the Net to engage a constituency that will protect and 

nurture it.


A Portrait of the Internet 

as a Young Man 

By Ann Bartow * 

Introduction 

The core theory of Jonathan Zittrain’s 2008 book The Future of the Internet—And 

How to Stop It is this: Good laws, norms, and code are needed to regulate the 

Internet, to prevent bad laws, norms, and code from compromising its creative 

capabilities and fettering its fecund flexibility. A far snarkier, if less alliterative, 

summary would be “We have to regulate the Internet to preserve its open, 

unregulated nature.” 

Zittrain uses brief, informal accounts of past events to build two main theories 

that dominate the book. First, he claims that open access, which he calls 

generativity, is under threat by a trend toward closure, which he refers to as 

tetheredness, which is counterproductively favored by proprietary entities. 

Though consumers prefer openness and the autonomy it confers, few take 

advantage of the opportunities it provides, and therefore undervalue it and too 

readily cede it in favor of the promise of security that tetheredness brings. 

Second, he argues that if the Internet is to find salvation it will be by the grace 

of “true netizens,” volunteers acting collectively in good faith to cultivate 

positive social norms online. 

One of the themes of the James Joyce novel first published in 1916, A Portrait of 

the Artist as a Young Man 1 is the Irish quest for autonomous rule. Jonathan 

Zittrain’s The Future of the Internet—And How to Stop It is similarly infused with 

the author’s desire for principled, legitimate governance—only of the place 

called cyberspace, rather than the author’s meatspace homeland. 

Portrait’s protagonist, Stephen Dedalus, internally defines himself as an artist 

through a nonlinear process of experiences and epiphanies. He consciously 

decides that it should be his mission to provide a voice for his family, friends, 

and community through his writing. Though Dedalus opts out of the 

* Professor of Law, University of South Carolina School of Law. This essay was adapted 

from A Portrait of the Internet as a Young Man, 108 MICH. L. REV. 1079 (2010), available at 

http://www.michiganlawreview.org/articles/a-portrait-of-the-internet-as-a-youngman. 

The author dedicates this essay to her son Casey, and to the memory of C. Edwin 

Baker. 

1 JAMES JOYCE, A PORTRAIT OF THE ARTIST AS A YOUNG MAN (1916).


traditional forms of participation in society, he envisions his writing as a way to 

productively influence society. Jonathan Zittrain charts the development of the 

Internet as a nonlinear process wrought by both conscious hard work and 

sweeping serendipity. He also strives to provide a voice for technologically elite 

Internet users, and to influence the development of online culture. He paints a 

portrait of the future Internet as chock full of so many enigmas and puzzles that 

it will keep the cyberlaw professors busy for decades, even though according to 

Zittrain, law as traditionally conceptualized will not be important. 

In addition to invoking Joyce, I chose the title of this essay for its decisive 

invocation of maleness. Embedded within Zittrain’s theories of generativity, 

there is also a perplexing gender story, in which men are fertile, crediting 

themselves with helping to “birth” the field of cyberlaw, 2 and engaging in 

stereotypically domestic pursuits such as “baking” restrictions into gadgetry. 3 

Non-generative appliances are deemed “sterile” 4 by Zittrain, sterility being the 

conceptual opposite of generativity. His deployment of reproductive imagery is 

odd. A metaphor equating an author’s creative output to a child is often 

invoked in the context of copyright law by people arguing that authors should 

have extensive control over the works they create. 5 Zittrain’s variation 

characterizes controlled technological innovations as unable to produce progeny 

at all. The metaphor works better if tetheredness is instead envisaged as a form 

of birth control, preventing unwanted offspring only. Certainly the producers 

of closed devices or locked software are able to provide, and generally 

enthusiastic about providing, new and improved versions of their goods and 

services to paying customers. 

2 See, e.g., Lawrence Lessig, Amazon.com Customer Review of THE FUTURE OF THE 

INTERNET—AND HOW TO STOP IT, Cyberlaw 2.0, 

http://www.amazon.com/review/R131R71HS3YJVG/ref=cm_cr_rdp_perm (Dec. 4, 

2008) (“The field of cyberlaw, or the law of the Internet—a field I helped birth … has suffered 

because people like me have spent too much time cheerleading, and not enough time 

focusing the world on the real problems and threats that the Internet has produced.”) 

(emphasis added); see also Daniel J. Solove, Privacy and Power: Computer Databases and Metaphors 

for Information Privacy, 53 STAN. L. REV. 1393, 1416–17 (2001) (noting that Roger Clarke is 

credited with coining the term “dataveillance”). Roger Clarke published suggestions for 

Internet regulations as early as 1988. See Roger A. Clarke, Information Technology and 

Dataveillance, 31 COMM. ACM 498, 508–11 (1988). 

3 JONATHAN ZITTRAIN, THE FUTURE OF THE INTERNET—AND HOW TO STOP IT 2 (2008). 

(“Jobs was not shy about these restrictions baked into the iPhone.”). [hereinafter ZITTRAIN, 

THE FUTURE OF THE INTERNET]. 

4 See, e.g., id. at 2 (“The iPhone is the opposite. It is sterile.”), 73 (“Generative tools are not 

inherently better than their non-generative (‘sterile’) counterparts.”). 

5 See Malla Pollack, Towards a Feminist Theory of the Public Domain, or Rejecting the Gendered Scope of 

the United States Copyrightable and Patentable Subject Matter, 12 WM. & MARY J. WOMEN & L. 603, 

606–07 (2006); see William Patry, Gender and Copyright, THE PATRY COPYRIGHT BLOG, Jun. 20, 

2008, http://williampatry.blogspot.com/2008/06/gender-and-copyright.html.


Zittrain offers a well-executed collection of stories that are intended to anchor 

his global theories about how the Internet should optimally function, and how 

two classes of Internet users should behave: The technologies should be 

generative, but also monitored to ensure that generativity is not abused by either 

the government or by scoundrels; elite Internet users with, as one might say 

today, “mad programming skilz” should be the supervisors of the Internet, 

scrutinizing new technological developments and establishing and modeling 

productive social norms online; and average, non–technically proficient Internet 

users should follow these norms, and should not demand security measures that 

unduly burden generativity. 

The anecdotes are entertaining and educational, but they do not constructively 

cohere into an instruction manual on how to avoid a bad future for people 

whose interests may not be recognized or addressed by what is likely to be a 

very homogeneous group of elites manning (and I do mean man-ning, given the 

masculine dominance of the field) the virtual battlements they voluntarily design 

to defend against online forces of evil. And some of the conclusions Zittrain 

draws from his stories are questionable. So, I question them below. 

Generativity Versus Tetheredness 

Is a False Binary 

Pitting generativity against tetheredness creates a false binary that drives a lot of 

Zittrain’s theorizing. The book was published in May of 2008, but its origins 

can be found in his earlier legal scholarship and mainstream media writings. In 

2006, Jonathan Zittrain published an article entitled The Generative Internet. 6 In it, 

he asserted the following: 

Cyberlaw’s challenge ought to be to find ways of regulating— 

though not necessarily through direct state action—which code 

can and cannot be readily disseminated and run upon the 

generative grid of Internet and PCs, lest consumer sentiment 

and preexisting regulatory pressures prematurely and tragically 

terminate the grand experiment that is the Internet today. 7 

Like the article, the book is useful for provoking thought and discussion, and it 

teaches the reader many disparate facts about the evolution of a number of 

different technologies. But it does not provide much direction for activists, 

especially not those who favor using laws to promote order. Zittrain has come 

6 Jonathan L. Zittrain, The Generative Internet, 119 HARV. L. REV. 1974 (2006) [hereinafter 

Zittrain, The Generative Internet]. 

7 Id. at 1979.


to bury cyberspace law as promulgated by governments, not to praise it. 

“Cyberlaw” as redefined by Zittrain is no longer the science of adapting existing 

real-space legal constructs to the online environment. Instead it is a collection 

of best practices chosen by people with the technological proficiency to impose 

them, top down, on the ignorant folks who are selfishly driven by their shallow 

consumer sentiments (viz., a desire for simplicity and security over openness and 

generativity). 

An abstract for the book, featured at its dedicated website, states: 

The Internet’s current trajectory is one of lost opportunity. Its 

salvation, Zittrain argues, lies in the hands of its millions of 

users. Drawing on generative technologies like Wikipedia that 

have so far survived their own successes, this book shows how 

to develop new technologies and social structures that allow 

users to work creatively and collaboratively, participate in 

solutions, and become true “netizens.” 8 

I will bluntly state (splitting an infinitive in the process) that I did not learn how 

to develop new technologies or new social structures from reading this book. It 

convinced me that new technologies and new social structures could contribute 

productively to the Internet if they develop appropriately, but Zittrain does not 

provide road maps or an instruction manual for developing them. He calls for 

“[c]ivic technologies [that] seek to integrate a respect for individual freedom and 

action with the power of cooperation,” but doesn’t paint a clear picture of 

which precise qualities these technologies or social structures would have, 

beyond cultivating generativity. 9 

Zittrain relentlessly informs the reader that generativity is a very good thing— 

except when it is abused by malefactors. But what, exactly, is generativity? 

Zittrain invokes the terms generative, non-generative, and generativity 

constantly throughout the book (over 500 times), but the definition of 

generative doesn’t remain constant. Sometimes it means creative or innovative, 

while other times it connotes openness, accessibility, or freedom. 10 

8 ZITTRAIN, THE FUTURE OF THE INTERNET, supra note 3. 

9 Jonathan Zittrain, How to Get What We All Want, CATO UNBOUND, May 6, 2009, 

http://www.cato-unbound.org/2009/05/06/jonathan-zittrain/how-to-get-what-weall-want/. 

10 Compare ZITTRAIN, THE FUTURE OF THE INTERNET, supra note 3, at 84 (“Generative systems 

allow users at large to try their hands at implementing and distributing new uses, and to fill a 

crucial gap when innovation is undertaken only in a profit-making model …”), with id. at 113 

(“[T]he PC telephone program Skype is not amenable to third-party changes and is tethered 

to Skype for its updates. Skype’s distribution partner in China has agreed to censor words


Zittrain had written previously that “Generativity denotes a technology’s overall 

capacity to produce unprompted change driven by large, varied, and 

uncoordinated audiences.” 11 Similarly, in the book he says, “Generativity is a 

system’s capacity to produce unanticipated change through unfiltered contributions from broad 

and varied audiences.” 12 He lists five elements of generativity: 

(1) how extensively a system or technology leverages a set of 

possible tasks; (2) how well it can be adapted to a range of 

tasks; (3) how easily new contributors can master it; (4) how 

accessible it is to those ready and able to build on it; and (5) 

how transferable any changes are to others— including (and 

perhaps especially) non-experts. 13 

Generative also seems to mean idiot-resistant. In his article The Generative 

Internet he explains that PCs are highly adaptable machines that are connected to 

a network with little centralized control, resulting in “a grid that is nearly 

completely open to the creation and rapid distribution of the innovations of 

technology-savvy users to a mass audience that can enjoy those innovations 

without having to know how they work.” 14 In the book, he makes the same 

point repeatedly—that most “mainstream” or “rank-and-file” computer users 

are either passive beneficiaries or victims of generativity, rather than generative 

actors. 15 There is a highly influential generative class of individuals who use 

generativity in socially productive ways. There is a nefarious group of 

reprobates who abuse generativity to create online havoc. And then there are 

the rest of the people online, sending and receiving emails, reading and writing 

blogs, participating on social-networking sites, renewing antivirus subscriptions, 

banking, shopping, and reading newspapers online. These users are blithely 

unaware of the generativity that provided this vast electronic bounty and 

complacently believe that, as long as they continue to pay an Internet service 

provider (“ISP”) for Internet access, its delivery will remain relatively smooth 

like ‘Falun Gong’ and ‘Dalai Lama’ in its text messaging for the Chinese version of the 

program. Other services that are not generative at the technical layer have been similarly 

modified …”). 

11 Zittrain, The Generative Internet, supra note 6, at 1980. 

12 JONATHAN ZITTRAIN, THE FUTURE OF THE INTERNET—AND HOW TO STOP IT, supra note 3 

at 70 (emphasis in original). 

13 Id. p. 71. 

14 Zittrain, The Generative Internet, supra note 6. 

15 See, e.g., id. at 3; see also ZITTRAIN, THE FUTURE OF THE INTERNET, supra note 3, at 4, 8, 43, 

44–45, 51, 56, 59, 78, 100, 102, 130, 151–52, 155, 59–60, 198, 243, 245.


and uninterrupted. When they call for more security for electronic devices, they 

themselves are the “damage” that generativity has to “route around.” 16 

The anti-generative concept of tetheredness also does some definitional shapeshifting 

throughout the tome. Sometimes it means unmodifiable, while other 

times it means controlled by proprietary entities, who may or may not facilitate, 

or even tolerate, alterations of their wares by end users. According to Zittrain, 

the dangers of tethers are twofold: Private companies can regulate how 

consumers use their products, and services and governments can use them to 

censor or spy on their citizens. 17 

Tethers can be good things if you are a mountain climber, or if you don’t want 

your horse to run off without you. And far more pertinently, tethers facilitate 

software updating for flaw-fixing and hole-patching purposes. Untethered 

software would require manual updates, a labor-intensive prospect that would 

require a degree of technical proficiency that many Internet users may lack. 

How many people are prepared to give up the advantages of tetheredness in the 

interest of preserving generativity is unclear. Without tethered appliances, the 

functionality of the Internet will be compromised. Try using a program that is 

no longer updated or supported by its vendor. Its obsolescence may render it 

untethered, but unless you have some pretty good programming chops, its 

usefulness will decline rapidly. Zittrain fears people will exchange generativity 

for security in binary fashion, but the relationship between tetheredness and 

convenience needs to be taken into account, as these variables will also affect 

consumer preferences and behaviors. 

The fundamental security most people seek is probably operability. Any threat 

to serviceability, whether from too much generativity or too many tethers, will 

provoke a call for action from users. I couldn’t have accessed the downloadable 

version of Zittrain’s book without a host of tethered utilities, including my 

computer’s operating system, my Internet browser, and Adobe Acrobat, which 

all update automatically with great frequency, as I consented to allow them to 

do when I agreed to the terms of use laid out in the associative end user license 

agreements (“EULAs”). The same with my printer software, my antivirus 

program, my online media players, the online games I play, and every other 

Internet-related utility I use. In a sense, this proves Zittrain’s assertion that we 

have ceded control over the mechanisms of online interface to electronic leash- 

16 This is a sideways reference to the John Gilmore quote, “The Net interprets censorship as 

damage and routes around it.” See Philip Elmer-DeWitt, First Nation in Cyberspace, TIME, 

Dec. 6, 1993, at 62, 64, available at 

http://www.time.com/time/magazine/article/0,9171,979768,00.html. 

17 See, e.g., ZITTRAIN, THE FUTURE OF THE INTERNET, supra note 3, at 56–57, 113 (discussing 

Skype), 109–10, 113 (discussing OnStar), 113 (discussing China’s use of Google.cn), 210–14 

(discussing mobile phones).


wielding tyrants. But, he may have the timing as well as motivation wrong. I 

suspect most of us deferred to tethering commercial enterprises very early in the 

evolution of the mainstream Internet, rather than recently. Zittrain references 

pioneering ISPs CompuServe and AOL as proprietary services that were 

overwhelmed by the generativity of PCs and the Internet. 18 My initial 

nonacademic experiences with the Internet comprised waiting anxiously for 

CompuServe and then AOL to finish installing updates when I needed to check 

my e-mail, and I had to pay for my Internet time by the minute. Things only 

went downhill when AOL went to an “all you can eat” payment structure, 

providing unlimited Internet for a fixed monthly fee. Users surged but AOL’s 

capacity could not meet the demand. 19 Users didn’t want security, they wanted 

performance. Tetheredness, or something similar, may have been linked in 

some way to AOL’s difficulties meeting its customers’ demand, but overselling 

and insufficient server capacity were the true culprits in terms of inhibiting 

operability. In addition, if Zittrain is correct that CompuServe and AOL 

exemplify the evils of tethering, it’s pretty clear the market punished those 

entities pretty harshly without Internet governance-style interventions. 

Software and electronic devices can be simultaneously generative and tethered. 

And it is unfair to criticize people who quite reasonably rely on tetheredness to 

keep their computers and electronic equipment updated and fully functional. 

Many average Internet users might like more transparency about the nature and 

extent of the tethers that connect their computers to large multinational 

corporations, but short of having actual laws that require relevant disclosures, 

this consumer desire is unlikely to be met. For them, generativity is unlikely to 

be helpful or enlightening, as Zittrain correctly notes, because they are not 

skilled enough to take advantage of it. In the absence of helpful laws, they are 

at the mercy of business models. 

Generativity: The Good, 

the Bad & the Ugly 

Zittrain’s stories are intended to show that generative technologies are better 

than tethered ones. But another strand of his narrative illustrates that 

generativity can be used destructively, to support the contention that it cannot 

18 The PC revolution was launched with PCs that invited innovation by others. So too with the 

Internet. Both were generative; they were designed to accept any contribution that followed 

a basic set of rules (either coded for a particular operating system, or respecting the 

protocols of the Internet). Both overwhelmed their respective proprietary, non-generative 

competitors, such as the makers of stand-alone word processors and proprietary online 

services like CompuServe and AOL. ZITTRAIN, THE FUTURE OF THE INTERNET, supra note 

3, at 23–25. 

19 See, e.g., Timothy C. Barmann, Judge to rule this week on AOL service, CYBERTALK, Oct. 26, 1997, 

http://www.cybertalk.com/102697b.htm.


be unfettered. At its worst, he warns, generativity will enable bad actors to 

exploit tethers for nefarious purposes, while tethers will simultaneously restrain 

positive generative responses to these challenges. His accounts of degenerate 

generativity rest uneasily with his exhortation that facilitating generativity should 

be the guiding principle of Internet governance. 

He also suggests deploying the “generative principle to determine whether and 

when it makes sense to violate the end-to-end principle” in the context of 

debates about network neutrality. 20 And the quantum of generativity that is 

promoted becomes the measure for assessing the legitimacy and effectiveness of 

what he characterizes as the intrusions of cyberlaw. He writes: 

The touchstone for judging such efforts should be according 

to the generative principle: do the solutions encourage a system 

of experimentation? Are the users of the system able, so far as 

they are interested, to find out how the resources they 

control—such as a PC—are participating in the environ– 

ment? 21 

Fostering generativity thus becomes the Prime Directive of Internet 

governance. 22 But there are problems he raises elsewhere in the book that 

generativity may not address, or may in fact exacerbate. For example, Zittrain 

references OnStar a number of times, warning that it can be used by law 

enforcement for surveillance purposes because it is tethered, and can be 

accessed remotely. 23 Putting aside questions about whether OnStar is accurately 

described as part of the Internet, one wonders of what practical use OnStar 

would be to its clients if it wasn’t tethered. OnStar seems to be a service that 

caters to people who want higher levels of proactive information and security 

when they are driving than the combination of a GPS unit and mobile phone 

can provide. OnStar customers don’t want generativity; they want someone to 

call the police and an ambulance or tow truck if they have an accident so they 

20 ZITTRAIN, THE FUTURE OF THE INTERNET, supra note 3, at 185. 

21 Id. at 173. 

22 “The Prime Directive is a plot device cooked up by a patently optimistic TV writer (either 

Trek producer Gene L. Coon or writer Theodore Sturgeon, depending on who you ask) in 

the mid-1960s. It’s a freshmen-year philosophy student’s reaction to the Cold War, when 

America and the Soviets were playing out their hostilities by proxy third-world conflicts. 

Effectively, they were interfering in the ‘development’ of underprivileged countries to 

further their own ends with some awful immediate and long-term results. In Roddenberry’s 

vision, humanity had evolved beyond such puppeteering and become an ‘advanced’ race.” 

See Jay Garmon, Why ‘Star Trek’s Prime Directive is stupid’, TECHREPUBLIC.COM, Feb. 12, 2007, 

http://blogs.techrepublic.com.com/geekend/?p=533. 

23 ZITTRAIN, THE FUTURE OF THE INTERNET, supra note 3, at 109–10, 113, 117–18, 187.


don’t have to, or to track down the location of their vehicle if it is stolen. 

Security means more to them than privacy, and if they don’t consciously realize 

they are exchanging one for the other when they sign up with OnStar, it seems 

to me the best solution is to require OnStar to inform them of this trade-off in 

simple and unambiguous terms. The law could also require OnStar to provide 

further information, perhaps including a primer on the search and seizure 

jurisprudence of Fourth Amendment law. Making OnStar generative, so that 

private citizens can readily discern incursions by government actors, would not 

give OnStar customers any more of what they appear to want—a high level of 

security overtly linked to constant, dedicated supervision. Enhanced 

generativity might also provide opportunities for private spying or intentional 

service disruptions by the very villains Zittrain spills so much ink warning 

against. 

Many of his examples of useful online-governance initiatives rely on extensive 

amounts of volunteer labor. But the important technological innovations 

related to the Internet were motivated by some form of self-interest. The U.S. 

Defense Department developed the Internet as a decentralized communications 

system that would be difficult to disrupt during wartime. 24 Tim Berners-Lee 

invented the World Wide Web as a way to facilitate communications with other 

physicists. 25 Pornographers have long used spam, browser hijacking, and 

search-engine manipulation to reach the eyeballs of potential customers. 26 All 

may have relied on generativity (though one might question how open and 

accessible the Defense Department was) but not all are socially beneficial. 27 

Sometimes Internet users may donate their labor involuntarily. Their online 

activities are harvested and bundled into what Zittrain applauds as the mediated 

wisdom of the masses. For example, he notes as follows: 

24 See Joseph D. Schleimer, Protecting Copyrights at the “Backbone” Level of the Internet, 15 UCLA 

ENT. L. REV. 139, 149 (2008); see also JANET ABBATE, INVENTING THE INTERNET 7–41 (1999). 

25 ABBATE, supra; see also Dick Kaser, The Guy Who Did the WWW Thing at the Place Where He Did 

It, INFO. TODAY, Feb. 2004, at 30. 

26 See, e.g., Pornographers Can Fool You With Hi-Tech, FILTERGUIDE.COM, 

http://www.filterguide.com/pornsfool.htm (setting forth various ways in which 

pornographers use technology to fool children) (last visited Oct 21, 2009); PEW INTERNET & 

AMERICAN LIFE PROJECT, SPAM IS STARTING TO HURT EMAIL (2003), 

http://www.pewinternet.org/Press-Releases/2003/Spam-is-starting-to-hurtemail.aspx 

(accounting for pornography-related spams’ impact on email). 

27 See generally Ann Bartow, Pornography, Coercion, and Copyright Law 2.0, 10 VAND. J. ENT. & TECH. 

L. 799, 800 (2008) (“Pornography is a dominant industrial force that has driven the evolution 

of the Internet.”).


The value of aggregating data from individual sources is well 

known. Yochai Benkler approvingly cites Google Pagerank 

algorithms over search engines whose results are auctioned, 

because Google draws on the individual linking decisions of 

millions of Web sites to calculate how to rank its search results. 

If more people are inking to a Web site criticizing Barbie dolls 

than to one selling them, the critical site will, all else equal, 

appear higher in the rankings when a user searches for 

“Barbie.” 28 

But all else is unlikely to be equal. Mattel can hire reputation-defense 

companies like ReputationDefender 29 to bury the critical sites about Barbie 

using search engine-optimization techniques and to surreptitiously edit 

Wikipedia entries. 30 For-profit entities don’t just want to spy on and control 

their customers with tethers. They also want to manipulate as much of the 

Internet as possible to their benefit, and this logically includes taking steps to 

highlight positive information and minimize the visibility of disparagement by 

third parties. 

Additionally, collective actions by the online masses can be oppressive. If more 

people link to websites glorifying sexual violence against women than to 

websites where women are treated as if they are fully human, those sites appear 

higher in the rankings when a user searches for a wide variety of things related 

to sex. The same is potentially true for racist and homophobic sites and other 

content that depict discrete groups in derogatory ways. In this way, negative 

stereotypes can be reinforced and spread virally. 31 

Finally, in the Google PageRank example, the power and input of the masses is 

being harnessed, for profit, by a large corporation. Google is doubtlessly happy 

to use generative tools when they are effective. But contrast the Google search 

28 ZITTRAIN, THE FUTURE OF THE INTERNET, supra note 3 at 160 (footnote omitted). 

29 See id. at 230 (asserting that ReputationDefender uses “moral suasion” as its primary 

technique for manipulating search-engine results). I offer a very different perspective on 

this. See Ann Bartow, Internet Defamation as Profit Center: The Monetization of Online Harassment, 

32 HARV. J.L. & GENDER 383 (2009). 

30 Zittrain himself noted something similar, writing, “If the Wikipedia entry on Wal-Mart is 

one of the first hits in a search for the store, it will be important to Wal-Mart to make sure 

the entry is fair—or even more than fair, omitting true and relevant facts that nonetheless 

reflect poorly on the company.” See ZITTRAIN, THE FUTURE OF THE INTERNET, supra note 3 

at 139. 

31 See ZITTRAIN, THE FUTURE OF THE INTERNET, supra note 3 at 147. Zittrain tacitly 

acknowledges this: “There are plenty of online services whose choices can affect our lives. 

For example, Google’s choices about how to rank and calculate its search results can 

determine which ideas have prominence and which do not.”


engine with Google’s Gmail, and it becomes apparent that the same company 

will keep a service tethered and proprietary when doing so best suits its 

purposes. 32 

The idiosyncratic online juggernaut that is Wikipedia, to which Zittrain devotes 

virtually an entire chapter, also illustrates some of the downsides of excessive 

generativity. 33 Wikipedia is an online encyclopedia that, at least in theory, 

anyone can edit. Zittrain is clearly enamored of it, writing, “Wikipedia stands at 

the apex of amateur endeavor: an undertaking done out of sheer interest in or 

love of a topic, built on collaborative software that enables a breathtakingly 

comprehensive result that is the sum of individual contributions, and one that is 

extraordinarily trusting of them.” 34 Zittrain provides a lot of information about 

Wikipedia, and the vast majority of it skews positive. He writes, “Wikipedia has 

charted a path from crazy idea to stunning worldwide success”; 35 and 

“Wikipedia is the canonical bee that flies despite scientists’ skepticism that the 

aerodynamics add up”; 36 and asserts that the manner in which Wikipedia 

operates “is the essence of law.” 37 Perhaps echoing Zittrain’s enthusiasm, one 

researcher determined Wikipedia has been cited in over 400 U.S. court 

opinions. 38 

Among myriad other facts and anecdotes, Zittrain notes that Wikipedia cofounder 

Larry Sanger is controversial because possibly he is given too much 

credit for his limited contributions to Wikipedia. 39 He also notes that another 

person involved with Wikipedia, former Wikimedia Foundation member Angela 

32 See generally Paul Boutin, Read My Mail, Please, SLATE, Apr. 15, 2004, 

http://slate.msn.com/id/2098946; Deane, Critics Release the Hounds on GMail, 

GADGETOPIA, Apr. 10, 2004, http://gadgetopia.com/post/2254; Google Watch, 

http://www.google-watch.org/gmail.html; Brian Morrissey, An Early Look at How Gmail 

Works, DMNEWS, Apr. 19, 2004, http://www.dmnews.com/an-early-look-at-howgmail-works/article/83946. 

33 ZITTRAIN, THE FUTURE OF THE INTERNET, supra note 3, chapter six. 

34 Id. at 96. 

35 Id. at 136. 

36 Id. at 148. 

37 Id. at 144. 

38 Lee F. Peoples, The Citation of Wikipedia in Judicial Opinions, 12 YALE J.L. & TECH. 

(forthcoming 2009), available at 

http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1272437. 

39 See ZITTRAIN, THE FUTURE OF THE INTERNET, supra note 3, at 143 (“At times—they are 

constantly in flux—Wikipedia’s articles about Wikipedia note that there is controversy over 

the ‘co-founder’ label for Sanger.”); see also ZITTRAIN, THE FUTURE OF THE INTERNET, supra 

note 3 at 142, 145.


Beesley Starling, unsuccessfully fought to have her Wikipedia entry deleted. 40 

That a man who wants undeserved credit and a woman who wants no attention 

at all have likely both been thwarted by Wikipedians is something Zittrain seems 

to view as a positive indicator. Angela Beesley Starling probably feels very 

differently, especially if her reasons for wanting her Wikipedia entry deleted 

included pressing personal safety concerns. The “talk” page of her Wikipedia 

biography quotes her as saying, “I’m sick of this article being trolled. It’s full of 

lies and nonsense.” 41 The forced publicity of Wikipedia entries is something all 

women may encounter under Wikipedia’s “system of self-governance that has 

many indicia of the rule of law without heavy reliance on outside authority or 

boundary.” 42 Research suggests that women, though 51% of the population, 

comprise a mere 13% of Wikipedia contributors, 43 for reasons that probably 

have to do with the culture of this entity, which women may experience more 

negatively than men do. 

Certainly notable living feminists have been on the receiving end of a campaign 

of nasty and untruthful edits to Wikipedia entries they would probably prefer 

not to have. Many entries on feminism have been written or edited by people 

who are actively hostile toward feminists, but they prevail because they seem to 

have a lot of free time and the few feminists who enter the wikifray seem to get 

driven out or edited into oblivion. To take just one example, the entries about 

Melissa Farley, 44 Catharine MacKinnon, 45and Sheila Jeffries 46 have all been 

40 Id. at 143. 

41 See Angela Beesley Starling Talkpage, 

WIKIPEDIA,http://en.wikipedia.org/wiki/Talk:Angela_Beesley_Starling (last visited 

Sept. 4, 2009) (“Angela Beesley has tried to have her biography on Wikipedia deleted, saying 

‘I’m sick of this article being trolled. It’s full of lies and nonsense.’ The Register and 

Wikitruth claim that her objections are ironic in light of the generally liberal policy of 

Wikipedia administrators to the accuracy and notability of biographies in Wikipedia of living 

people. Seth Finkelstein, who tried to have his own entry from Wikipedia removed, called it 

‘a pretty stunning vote of no-confidence. Even at least some high-ups can’t eat the dog 

food.’”) (footnotes omitted). 

42 ZITTRAIN, THE FUTURE OF THE INTERNET, supra note 3 at 143. 

43 See, e.g., Andrew LaVallee, Only 13% of Wikipedia Contributors Are Women, Study Says, WALL ST. 

J., Aug. 31, 2009, http://blogs.wsj.com/digits/2009/08/31/only-13-of-wikipediacontributors-are-women-study-says; 

Jennifer Van Grove, Study: Women and Wikipedia Don’t 

Mix, MASHABLE, Sept. 1, 2009, http://mashable.com/2009/09/01/women-wikipedia; 

Cathy Davidson, Wikipedia and Women, HASTAC, Sept. 2, 2009, 

http://www.hastac.org/blogs/cathy-davidson/wikipedia-and-women. 

44 See Melissa Farley, WIKIPEDIA, http://en.wikipedia.org/wiki/Melissa_Farley (last visited 

July28, 2009). 

45 See Catharine MacKinnon, WIKIPEDIA, 

http://en.wikipedia.org/wiki/Catharine_MacKinnon (last visited July 28, 2009).


heavily edited 47 by a rabid pornography proponent named Peter G. Werner 48 

who sometimes also uses the pseudonym Iamcuriousblue. 49 Each entry is the 

first result returned after a Google search of their names. He has deleted or 

attempted to have deleted entries about other feminists. 50 He shows up under 

one identity or another in virtually every entry in which feminism is mentioned. 

And he successfully convinced the Wikipedia community to ban a feminist 

activist who vigorously contested his edits. 51 Any group that is not well 

represented within the Wikipedia editing community is likely to experience 

similar marginalization. 

Recently, Wikipedia announced that the entries of living people will receive a 

mandatory layer of intermediation. A new feature called “flagged revisions” will 

require that an experienced volunteer editor sign off on any changes before they 

become permanent and publicly accessible. 52 A New York Times report noted 

that this would “divide Wikipedia’s contributors into two classes—experienced, 

trusted editors, and everyone else—altering Wikipedia’s implicit notion that 

everyone has an equal right to edit entries.” 53 This seems to be one realization 

of what Zittrain broadly desires—control over the ignorant wikimasses by a 

designated elite. But the project became significantly less collaborative and 

open when this change was made. 

Wikipedia entries are generated by a massive assemblage of volunteers with 

unknown motivations and agendas. Group behavior is always unpredictable, a 

fact that Zittrain acknowledges but under-appreciates. One somewhat 

organized assemblage that calls itself Anonymous launches cyber-attacks that 

46 See Sheila Jeffreys, WIKIPEDIA, http://en.wikipedia.org/wiki/Sheila_Jeffreys (last visited 

July 28, 2009). 

47 See, e.g., Catharine MacKinnon Talkpage, WIKIPEDIA, 

http://en.wikipedia.org/wiki/Talk:Catharine_MacKinnon (last visited July 28, 2009). 

48 See Peter G Werner Userpage, WIKIPEDIA, 

http://en.wikipedia.org/wiki/User:Peter_G_Werner (last visited July 28, 2009). 

49 See Iamcuriousblue Userpage, WIKIPEDIA, 

http://en.wikipedia.org/wiki/User:Iamcuriousblue (last visited July 28, 2009). 

50 See, e.g., Articles for deletion/Cheryl Lindsey Seelhoff, WIKIPEDIA, 

http://en.wikipedia.org/w/index.php?title=Wikipedia:Articles_for_deletion/Cheryl 

_Lindsey_Seelhoff&oldid=150110815 (last visited Sept. 25, 2009), see also Nikki Craft 

Talkpage, WIKIPEDIA, http://en.wikipedia.org/wiki/Talk:Nikki_Craft (last visited Sept. 

25, 2009). 

51 Telephone interview with Nikki Craft; see also Nikki Craft Talkpage, supra (containing 

conversation in which user Iamcuriousblue discredits Nikki Craft’s Wikipedia article). 

52 Noam Cohen, Wikipedia to Limit Changes to Articles on People, N.Y. TIMES, Aug. 25, 2009, at B1. 

53 Id.


online norms do not seem to have any cognizable role in addressing. 54 As with 

Wikipedians, Anonymous is hostile to others and outsiders. One blogger noted: 

Interestingly … Anon never seems to take down the big sites. 

Walmart.com and the Pentagon are safe from his attentions. 

It’s not that Anon is a big fan of Walmart or the government. 

It’s just so much easier to attack the vulnerable. Big business 

and big government aren’t vulnerable on the Internet. They 

can afford not to be. 

Small discussion boards and blogs, particularly ones that 

advocate unpopular points of view, are often run by individuals 

who put up their own funds, if they can scrape them together, 

and who must be their own IT departments. They can’t afford 

the type of security that requires the big bucks. And since they 

have jobs (unlike Anon, apparently), they have to put their 

desire to maintain an Internet presence in the balance with 

supporting themselves and their families. When the crunch 

comes and time pressures set in, it’s not the Internet presence 

that wins out. 

So the actions of these “apolitical” hackers do have a political 

end: They remove unpopular, radical, fringe viewpoints from 

54 See e.g., Shaun Davies, ‘No Cussing’ Teen Faces Net Hate Campaign, NINEMSN NEWS, Jan. 18, 

2009, http://news.ninemsn.com.au/technology/720115/no-cussing-teen-faces-nethate-campaign 

(stating “McKay Hatch’s No Cussing Club, which encourages teens to ‘chill 

on the profanity’, claims to have over 20,000 members worldwide. Hatch, a 15-year-old 

from South Pasadena in California, garnered wide media coverage for his anti-swearing 

campaign, including an appearance on Dr Phil. But at the beginning of the year, Hatch’s 

email inbox began clogging up with hate mail from an unknown source. Pizza and porn 

deliveries became commonplace for his family, who eventually called in the FBI after 

numerous receiving[sic] death threats and obscene phone calls. Anonymous appears to be 

behind the attacks, with threads on sites such as 4chan.org and 711chan.org identifying their 

members as the culprits. And the pain may not yet be over for the Hatch family— 

Anonymous appears to be planning future raids and has threatened to ‘wipe this cancer [the 

No Cussing Club] from the face of the internet’.[sic] In one 4chan thread, a number of 

users boasted about sending bogus pizza deliveries and even prostitutes to the Hatchs’ 

house, although it was impossible to verify if these claims were genuine. The same thread 

also contained a credit card number purported to be stolen from Hatch’s father, phone 

numbers, the family’s home address and Hatch’s instant messenger address.”); see also Behind 

the Façade of the “Anonymous” Hate Group, RELIGIOUS FREEDOM WATCH, July 6, 2009, 

http://www.religiousfreedomwatch.org/media-newsroom/behind-the-facade-ofthe%E2%80%9Canonymous%E2%80%9D-hate-group/; 

see also Alex Wuori, 

ENCYCLOPAEDIA DRAMATICA, http://encyclopediadramatica.com/Alex_Wuori (last 

visited July 28, 2009).


the web. Big government doesn’t have to eliminate the 

subversive websites; Anon will do it. 55 

The activities of Anonymous have been characterized as domestic terrorism. 56 

And Anonymous certainly takes advantage of generative technologies, just as 

Wikipedians with reprehensible agendas do. Zittrain asserts that bad actors like 

Anonymous are driving the demand for increased security, 57 but he doesn’t 

provide any targeted mechanisms for hindering them, or explain why increasing 

security necessarily compromises productive generativity. 

The Zittrainnet’s Netizens: 

Overlords of Good Faith 

As with a James Joyce novel, there are a variety of transactions that the careful 

reader negotiates with the author. Each section has to be read independently of 

the others, because while it may cohere internally, it may not combine with 

other delineated portions to paint a consistent picture of Zittrain’s preferred 

future for the Internet, which will hereafter be called the “Zittrainnet.” 

Some of the recommendations he makes invite broad democratic participation 

in Zittrainnet governance, while other times he warns against it and suggests 

ways to decrease the threats posed “by outsiders—whether by vendors, malware 

authors, or governments.” 58 One wonders how something as disaggregated as 

the Internet can have outsiders, until recognition dawns about what Zittrain is 

truly suggesting, at least part of the time, in terms of who should control the 

Internet to best ensure its evolution into the Zittrainnet: an elite circle of people 

with computer skills and free time who share his policy perspective. 

Technologists Rule 

Zittrain doesn’t contemplate “anyone” developing serviceable code. Zittrain’s 

view is that only a select few can take productive advantage of generativity, and 

within this elite group are bad actors as well as good. He thinks that cyberlaw is 

the appropriate mechanism to encourage positive uses of generativity while 

55 VeraCity, Dominator Tentacles, http://vera.wordpress.com/2007/08/24/dominatortentacles/ 

(Aug. 24, 2007). 

56 VA. FUSION CTR., VA. DEP’T OF STATE POLICE, 2009 VIRGINIA TERRORISM THREAT 

ASSESSMENT 48 (2009), available at http://www.infowars.com/virginia-fusion-centerreleaseshomegrown-terrorism-document/. 

57 See generally ZITTRAIN, THE FUTURE OF THE INTERNET, supra note 3, at chapter 3. This is one 

of the central claims of the book. 

58 ZITTRAIN, THE FUTURE OF THE INTERNET, supra note 3, at 173.


thwarting the troublesome ones, cyberlaw being computer-code construction 

and norm entrepreneurship within Internet communities, as well as more 

traditionally recognized modes of law formation such as statutes and 

regulations. 59 As far as who exactly will divine good generativity from bad, and 

wield the mighty sword of cyberlaw to defend the former and defeat the latter, 

Zittrain is decidedly vague. In the “Solutions” section of the tome Zittrain lists 

“two approaches that might save the generative spirit of the Net”: 

The first is to reconfigure and strengthen the Net’s 

experimentalist architecture to make it fit better with its nowmainstream 

home. The second is to create and demonstrate 

the tools and practices by which relevant people and 

institutions can help secure the Net themselves instead of 

waiting for someone else to do it. 60 

By “relevant people and institutions” Zittrain seems to mean technologically 

skilled, Internet users of good will. 61 But as far as who it is that will 

“reconfigure and strengthen the Net’s experimentalist architecture” or who will 

“create and demonstrate the tools and practices” on behalf of these relevant 

people and institutions (shall we call them “generativators?”), Zittrain offers few 

specifics. He mentions universities generally, 62 and two organizations he is 

affiliated with specifically, Harvard University’s Berkman Center (where he is 

one of 13 Directors—all male, of course 63) and the Oxford Internet Institute 

(where he is a Research Associate 64), which he describes as “multidisciplinary 

academic enterprises dedicated to charting the future of the Net and improving 

it.” 65 Those who share his visions for the Zittrainnet are supposed to function 

as norm entrepreneurs, guiding lights for the undereducated, inadequately 

skilled online masses to follow, sheep-like. 

Less-relevant people are described as “[r]ank-and-file Internet users [who] enjoy 

its benefits while seeing its operation as a mystery, something they could not 

59 Id. chapter 5. 

60 Id. at 152. 

61 Id. at 246. 

62 Id. at 198, 245. 

63 See People, Berkman Center for Internet and Society at Harvard University, 

http://cyber.law.harvard.edu/people. 

64 See People, OXFORD INTERNET INSTITUTE UNIVERSITY OF OXFORD, 

http://www.oii.ox.ac.uk/people/?status=current&type=&keywords=zittrain. 



possibly hope to affect.” 66 These ignorant non-generativators frighten Zittrain, 

because when he fears that, a crisis comes, they will pressure the government to 

enhance Internet security at the expense of Internet generativity, out of shortsighted, 

ill-informed perceptions of their own self-interest. 67 He knows better 

than they do what’s best for them. 

In a related article he published in Legal Affairs to promote the book, Zittrain 

explains: 

If the Internet does have a September 11 moment, a scared 

and frustrated public is apt to demand sweeping measures to 

protect home and business computers—a metaphorical USA 

Patriot Act for cyberspace. Politicians and vendors will likely 

hasten to respond to these pressures, and the end result will be 

a radical change in the technology landscape. The biggest 

casualty will likely be a fundamental characteristic that has 

made both the Internet and the PC such powerful phenomena: 

their “generativity.” 68 

Many of the stories Zittrain tells in the book are intended to persuade readers 

that unless somebody does something, the Internet will do what the book’s 

cover suggests: derail and drive over a cliff. But after ominously warning his 

audience repeatedly that “Steps Must Be Taken Immediately,” the particulars of 

whom that somebody is and the details of what s/he should be doing are never 

made explicit. 

In addition, the law component of cyberlaw gets surprisingly little attention in 

the book, given that Zittrain is a law professor. According to Larry Lessig, 

“This book will redefine the field we call the law of cyberspace.” 69 This is 

66 Id. at 245. 

67 Id. 

68 See Jonathan Zittrain, Without a Net, LEGAL AFFAIRS, Jan./Feb. 2006, at 34, available at 

http://www.legalaffairs.org/issues/January-February- 

2006/feature_zittrain_janfeb06.msp; see also Lawrence Lessig, Z’s Book Is Out, LESSIG 2.0, 

May 1, 2008, http://lessig.org/b.og/just_plain_brilliant/ [hereinafter Lessig, Z’s Book Is 

Out]; Lawrence Lessig, The state of Cyberlaw, 2005, LESSIG 2.0, Dec. 30, 2005, 

http://lessig.org/b.og/read_this/ (stating “Legal Affairs has a fantastic collection of 

essays about various cyberspace related legal issues by some of my favorite writers about the 

subject. Zittrain’s piece outlines the beginning of his soon to be completed book. It shall be 

called Z-theory.”). 

69 See Lessig, Z’s Book Is Out, supra. Lessig explains his thoughts regarding the importance of 

Zittrain’s book in his blog: 

This book will redefine the field we call the law of cyberspace. That sounds 

like a hokey blurb no doubt. But hokeness [sic] does not mean it is not true.


worrisome to anyone still struggling to ascertain the parameters of cyberlaw in 

the first instance, beyond the macro concerns about top-down versus bottomup 

approaches to governance identified by the scholars mentioned above. The 

role of law in Zittrainnet’s rule of law is extremely limited. Laws concerning 

jurisdiction, privacy, free speech, copyrights, and trademarks often transmogrify 

into cyberlaw when they are invoked in an Internet context, but they exist and 

evolve offline too, which prevents their total capture by cyberlaw scholars. 

Zittrain’s redefinition of cyberlaw compresses debates that engage complicated, 

intersecting bodies of law into a much narrower conversation about the value of 

generativity, and how best to secure the appropriate level of it. In general 

Zittrain seems quite pessimistic about whether cyberlaw can achieve anything 

positive beyond somehow—he never tells us how—fostering generativity. At 

one point in the book he even describes the enforcement of laws online as 

something that could result in net social losses, and therefore a mechanism of 

Internet governance that is inferior to “retention of generative technologies.” 70 

Zittrain seems to have a lot more confidence in technologists than in attorneys. 

He waxes rhapsodic about the wisdom and forethought of the “framers” of the 

Internet throughout the tome. 71 One of “the primary” ways he proposes to 

address tetheredness and its associative ills is “a series of conversations, 

arguments, and experiments whose participants span the spectrum between 

network engineers and PC software designers, between expert users with time 

It is true. The field before this book was us cheerleaders trying to convince a 

skeptical (academic) world about the importance and value of certain central 

features of the network. Zittrain gives these features a name—generativity— 

and then shows us an aspect of this generative net that we cheerleaders 

would rather you not think much about: the extraordinary explosion of 

malware and the like that the generative net has also generated. 

70 See ZITTRAIN, THE FUTURE OF THE INTERNET, supra note 3, at 113-114. Zittrain states: 

Technologies that lend themselves to an easy and tightly coupled expression 

of governmental power simply will be portable from one society to the next. 

It will make irrelevant the question about how firms like Google and Skype 

should operate outside their home countries. 

This conclusion suggests that although some social gain may result from 

better enforcement of existing laws in free societies, the gain might be more 

than offset by better enforcement in societies that are less free—under 

repressive governments today, or anywhere in the future. If the gains and 

losses remain coupled, it might make sense to favor retention of generative 

technologies to put what law professor James Boyle has called the 

“Libertarian gotcha” to authoritarian regimes: if one wants technological 

progress and the associated economic benefits, one must be prepared to 

accept some measure of social liberalization made possible with that 

technology. 

71 See ZITTRAIN, THE FUTURE OF THE INTERNET, supra note 3, at 7, 27, 31, 33, 34, 69, 99.


to spend tinkering and those who simply want the system to work—but who 

appreciate the dangers of lockdown” (p. 173). On the Zittrainnet, with the 

exception of a select few cyberlaw professors, academics in disciplines other 

than law, particularly computer science, are going to be the true benevolent 

dictators of cyberlaw, mediating disputes with technological innovations and 

enforcing their judgments through code. 

The Private Sector 

Zittrain quite understandably doubts that for-profit entities will selflessly 

prioritize the well-being of the Internet over their own commercial gain. So, 

they are unlikely to consistently adhere to pro-generative business plans unless 

they can be convinced that doing so will benefit them. One of Zittrain’s 

objectives in writing the book was to educate the reader about the ways that 

extensive generativity can serve commercial goals. However, while corporate 

actors may find Zittrain’s book of interest, I suspect actual experiences in the 

marketplace will be what drives their decisions about tethers and generativity. 

Zittrain opens his book with what is framed as an apocryphal tale: Apple II 

computers were revolutionary because they facilitated the development of new 

and original uses by outsiders; but thirty years later the same company launched 

an anti-generativity counterrevolution of sorts by releasing its innovative iPhone 

in a locked format intended to discourage the use of applications that were not 

developed or approved by Apple. 72 

But how would Zittrain change this? Surely when the company made this 

decision, it knew even more than Zittrain about the role that generativity played 

in the success of the Apple II, but still chose a different strategy for the iPhone. 

Affirmative curtailment of its generativity initially lowered the risk that iPhones 

would be plagued by viruses or malware, and allowed Apple to control the ways 

that most consumers use them. Would Zittrain have forced generativity into 

the mechanics of the iPhone by law? Or, would he strip Apple of its ability to 

use the law to interfere when others hack the iPhone and make it more 

customizable? Or, would he instead simply wait for the market to show Apple 

the error of its degenerative ways? He never specifies. What he says at the end 

of his iPhone discussion is: 

A lockdown on PCs and a corresponding rise of tethered 

appliances will eliminate what today we take for granted: a 

world where mainstream technology can be influenced, even 

revolutionized, out of left field. Stopping this future depends 

72 See generally ZITTRAIN, THE FUTURE OF THE INTERNET, supra note 3, at 86–87 (summarizing 

work by Eric von Hippel on the subject).


on some wisely developed and implemented locks, along with 

new technologies and a community ethos that secures the keys 

to those locks among groups with shared norms and a sense of 

public purpose, rather than in the hands of a single gatekeeping 

entity, whether public or private. 73 

It sounds like Zittrain wants to prevent Apple from interfering when consumers 

modify their iPhones. But how he proposes to achieve this is addressed only 

generally, much later in the book when he suggests vague, persuasion-based 

solutions. My inner pragmatist thinks strong consumer protection laws might 

be a viable option to this and many other problems he articulates in the book, 

but Zittrain mentions that possibility only glancingly, in the context of 

maintaining data portability. 74 

In July of 2008, Apple began allowing software developers to sell software for 

the iPhone, and tens of thousands of applications have subsequently been 

independently developed for the iPhone, 75 suggesting either successful 

deployment of a strategic multistep product rollout Apple had planned all along, 

or a midcourse marketing correction. In either event, after the App Store the 

iPhone cannot accurately be described as non-generative, at least as I 

understand the concept, 76 and what Zittrain characterized as a problem seems 

to have been largely solved without the intervention of cyberlaw. The iPhone is 

still tethered, of course, possibly giving consumers just enough rope to hang 

themselves if Apple decides to interfere with the contents or operation of any 

given phone. But tethering also facilitates positive interactions, such as updates 

and repairs. It is now, to use a phrase Zittrain uses in a different context, “[a] 

technology that splits the difference between lockdown and openness.” 77 

It is true that Apple could alter the iPhone’s balance between generativity and 

tetheredness without notice or reason. But there is every reason to expect that 

Apple will try to keep its customers happy, especially given increased 

competition by devices running Google’s Android operating system—with its 

73 ZITTRAIN, THE FUTURE OF THE INTERNET, supra note 3, at 5. 

74 Id. at 177. 

75 See, e.g., Jon Fortt, iPhone apps: For fun and profit?, FORTUNE TECH DAILY, July 6, 2009, 

http://money.cnn.com/2009/07/06/technology/apple_iphone_apps.fortune/index. 

htm 

76 See, e.g., Adam Thierer iPhone 2.0 cracked in hours … what was that Zittrain thesis again?, THE 

TECHNOLOGY LIBERATION FRONT, July 10, 2008, 

http://techliberation.com/2008/07/10/iphone-20-cracked-in-hours-what-was-thatzittrain-thesis-again/. 



even more open apps marketplace. 78 A recent short review of the book in The 

Observer noted: 

The problem facing books about the internet is that by the 

time they have hit the shelves, they are already dated. This is 

clear on the second page of The Future of the Internet, where 

Jonathan Zittrain writes that the iPhone is purposefully 

resistant to “applications” (programmes allowing the phone to 

do clever things apart from make calls). 79 

The problem facing this book is deeper than datedness. Zittrain is wrong in his 

assumptions about rigidity and fixedness. 80 In the abstract generativity and 

tetheredness may be opposites, but in reality they can exist within a single 

appliance. He actually makes this point when he describes computers with dual 

applications designated “red” and “green,” one generative and the other 

secure. 81 But he does not acknowledge that many technological devices already 

78 Yi-Wyn Yen & Michal Lev-Ram, Google’s $199 Phone to Compete with the iPhone, TECHLAND, 

Sept. 17, 2008, http://techland.blogs.fortune.cnn.com/2008/09/17/googles-199phone-to- 

compete-with-the-iphone/. 

79 Helen Zaltzman, The Future of the Internet by Jonathan Zittrain, OBSERVER (London), June 14, 

2009, at 26, available at http://www.guardian.co.uk/books/2009/jun/14/futureinternet-zittrain-review. 

80 See Adam Thierer, Review of Zittrain’s “Future of the Internet”, THE TECHNOLOGY LIBERATION 

FRONT, Mar. 23, 2008, http://techliberation.com/2008/03/23/review-of-zittrainsfuture-of-the-internet/. 

Thierer writes: 

My primary objection to Jonathan’s thesis is that (1) he seems to be overstating 

things quite a bit; and in doing so, (2) he creates a false choice of 

possible futures from which we must choose. What I mean by false choice is 

that Jonathan doesn’t seem to believe a hybrid future is possible or desirable. 

I see no reason why we can’t have the best of both worlds—a world full of 

plenty of tethered appliances, but also plenty of generativity and openness. 

See also Timothy B. Lee, Sizing Up “Code” With 20/20 Hindsight, FREEDOM TO TINKER, May 

14, 2009, http://www.freedom-to-tinker.com/blog/tblee/sizing-code-2020-hindsight. 

Lee writes: 

I think Jonathan Zittrain’s The Future of the Internet and How to Stop It 

makes the same kind of mistake Lessig made a decade ago: overestimating 

regulators’ ability to shape the evolution of new technologies and 

underestimating the robustness of open platforms. The evolution of 

technology is mostly shaped by engineering and economic constraints. 

Government policies can sometimes force new technologies underground, 

but regulators rarely have the kind of fine-grained control they would need 

to promote “generative” technologies over sterile ones, any more than they 

could have stopped the emergence of cookies or DPI if they’d made 

different policy choices a decade ago. 

81 ZITTRAIN, THE FUTURE OF THE INTERNET, supra note 3, at 154-57.


shift between tethered and generative functions, driven by the demands of their 

users. 

Making assumptions about consumer preferences can be hazardous, especially 

for folks who tend to associate mostly with people who share common 

interests, common backgrounds, a common race, a common gender. The 

Zittrainnet’s netizens, being human, are likely to engage in all manner of 

typecasting and generalizing when they redesign their Internet sectors of 

interest. If the leading netizens echo the demographic pattern of the cyberlaw 

scholars, white men with elite educations will be making most of the calls. 82 

And Internet governance will be exceedingly top-down. 

At present companies can dramatically alter the levels of tetheredness and 

generativity in their products and services for any reason or no reason at all, and 

Zittrain never explains what sort of regulations or market interventions he 

thinks are necessary to achieve or preserve the Zittrainnet. He is critical of 

companies that assist totalitarian governments with surveillance or censorship 

initiatives, 83 but fails to acknowledge the reason that many technologies that can 

be readily employed to spy on people are developed: Companies want to be able 

to shadow and scrutinize their customers themselves. Consumers usually agree 

to this scrutiny in nonnegotiable EULA terms and conditions. For companies, 

closely following the acts and omissions of their customers or client base is 

generative behavior, even though it relies on tethers. Information about 

consumers can lead to innovations in goods and services as well as in marketing 

them. 

Governments 

Zittrain expresses grave concerns about government intervention on the 

Internet. He does not seem to believe that government actors can competently 

safeguard users, or effectively regulate technology. And he fears governments 

will further harness the Internet to advance surveillance and censorship agendas 

that are anathema to freedom. Zittrain writes with deep foreboding: 

The rise of tethered appliances significantly reduces the 

number and variety of people and institutions required to apply 

the state’s power on a mass scale. It removes a practical check 

on the use of that power. It diminishes a rule’s ability to attain 

82 See Anupam Chander, Whose Republic?, 69 U. CHI. L. REV. 1479, 1484–85 (2002) (reviewing 

CASS SUNSTEIN, REPUBLIC.COM (2001)). 

83 ZITTRAIN, THE FUTURE OF THE INTERNET, supra note 3, at 112–13.


legitimacy as people choose to participate in its enforcement, 

or at least not stand in its way. 84 

So it seems strange to learn that his solution to too much tethering is “a latterday 

Manhattan Project.” 85 The Manhattan Project was, of course, the code 

name for the U.S. government’s secret project to develop a nuclear bomb. It 

may have been staffed by scientists, many of whom were academics, but it was 

organized, funded, and strictly controlled by the government. 86 An analogous 

initiative to formulate the Zittrainnet would hardly be open and accessible to 

the online public. Moreover, governments generally take some kind of 

proprietary interest in the outcomes of projects they fund. Even under the 

Bayh-Dole Act, 87 which allows universities in the United States to patent 

inventions developed with federal funding, the U.S. government retains marchin 

rights. 88 Zittrain seems to want the resources that governments can provide 

without any of the restrictions or obligations governments will, as experience 

suggests, inevitably impose. It’s possible that a well-crafted Zittrainet Project 

could receive the unconditional support of government actors, but I don’t think 

this is terribly likely to happen. 

Surprisingly, one of the success stories for generativity that Zittrain references is 

the Digital Millennium Copyright Act of 1998. 89 Not only did this require 

government intervention in the form of traditional law, but it also relied on 

tethering. Web sites could not take down potentially infringing material without 

retaining a level of control that enables this. 

In addition to generativity, one of the defining principles of the Zittrainnet will 

be adherence to First Amendment principles. Zittrain’s descriptions of online 

freedom and autonomy suggest a strong belief that all the countries of the world 

84 Id. at 118. 

85 Id. at 173. 

86 U.S. DEP’T OF ENERGY, OFFICE OF HISTORY & HERITAGE RES., Early Government Support, in 

THE MANHATTAN PROJECT: AN INTERACTIVE HISTORY, 

http://www.cfo.doe.gov/me70/manhattan/1939-1942.htm (last visited July 30, 2009); 

The Manhattan Project (and Before), in THE NUCLEAR WEAPON ARCHIVE, 

http://nuclearweaponarchive.org/Usa/Med/Med.html (last visited Oct. 4, 2009); U.S. 

DEP’T OF ENERGY, OFFICE OF HISTORY & HERITAGE RES., A Tentative Decision to Build the 

Bomb, in THE MANHATTAN PROJECT: AN INTERACTIVE HISTORY, 

http://www.cfo.doe.gov/me70/manhattan/tentative_decision_build.htm (last visited 

July 30, 2009). 

87 35 U.S.C. §§ 200–212 (2006). 

88 Id. § 203. 

89 See Pub. L. No. 105–304, 112 Stat. 2860 (1998). See also ZITTRAIN, THE FUTURE OF THE 

INTERNET, supra note 3, at 119–20 (stating Zittrain’s discussion of the DMCA).


should honor and implement the free-speech values of the First Amendment, 

whether they want to or not. 90 This raises complicated issues of state 

sovereignty and international law that Zittrain does not address. 

Conclusion 

I’ve been very hard on The Future of the Internet in this review, but I truly did 

enjoy reading it. The book is very informative, if you can sift through the 

portions contrived to illustrate an unconvincing macro theory of the Internet. I 

wish Zittrain had written a book that set out only to describe the history and 

state of the Internet, rather than one that was formulated to support 

questionable generalizations and grandiose prescriptions. He could have told 

many of the same extremely interesting stories, but with more balance and less 

of a blatant “big think” agenda. 

The book is woefully lacking in specifics, in terms of advancing the reforms 

Zittrain asserts are necessary. Even if I were willing to buy into Zittrain’s claim 

that preserving and enhancing generativity should be the organizing principle of 

the Internet governance interventions, the mechanics of how this could be 

pursued holistically are never revealed. And the technicalities by which good 

generativity could be fostered while bad generativity was simultaneously 

repressed are similarly unstated. The only extensively developed account of a 

generative system Zittrain unabashedly admires is Wikipedia, which he admits is 

undemocratic. 91 It is also a system that facilitates repression of unpopular 

viewpoints, and this is likely to affect outsider groups most dramatically. 

Who will step forward to somehow cultivate the Zittrainnet is a mystery. The 

future of the Internet, Zittrain asserts, would be much safer in the hands of 

those who can competently safeguard it. He describes these people in very 

general terms as being skilled and of good faith. These hands do not belong to 

people who are affiliated with dot-coms, because they use tethering to constrain 

generativity when doing so is profitable. Nor do they belong to dot-gov 

bureaucrats, who are at best uninformed and at worst eager to use the Internet 

to enforce regimes of totalitarian rule. Readers of the book learn a lot more 

about who Zittrain thinks should not be in control of the Internet than who 

should be. But there are a number of hints and suggestions scattered 

throughout its pages that he believes he and his colleagues are capable of 

directing the Internet’s future wisely and beneficently. If they are going to 

attempt to do this by writing books, perhaps Zittrain’s offering makes sense as a 

90 Contra Joel R. Reidenberg, Yahoo and Democracy on the Internet, 42 JURIMETRICS 261 (2002). 

91 See ZITTRAIN, THE FUTURE OF THE INTERNET, supra note 3, at 141 (“And Wikipedia is 

decidedly not a democracy: consensus is favored over voting and its head counts.”).


declaration of first principles. Maybe his next book will describe the steps along 

the path to the Zittrainnet more concretely.

138 CHAPTER 2: IS THE GENERATIVE INTERNET AT RISK?


The Case for Internet Optimism, 

Part 2: Saving the Net from Its 

Supporters 

By Adam Thierer * 

In an earlier essay, I argued that two distinct strands of “Internet pessimism” increasingly 

dominate Internet policy discussions. The pessimism of “Net skeptics” is rooted in a general 

skepticism of the supposed benefits of cyberspace, digital technologies, and information 

abundance. Here, I respond to a very different strand of Internet pessimism—one expressed by 

fans of the Internet and cyberspace who nonetheless fear that dark days lie ahead unless steps 

are taken to “save the Net” from a variety of ills, especially the perceived end of “openness.” 

Introduction: Is the 

Web Really Dying? 

“The Death of the Internet” is a hot meme 

in Internet policy these days. Much as a 

famous Time magazine cover asked “Is 

God Dead?” in 1966, 1 Wired magazine, the 

magazine for the modern digerati, 

proclaimed in a recent cover story that 

“The Web is Dead.” 2 A few weeks later, 

The Economist magazine ran a cover story 

fretting about “The Web’s New Walls,” 

wondering “how the threats to the 

Internet’s openness can be averted.” 3 The 

primary concern expressed in both essays: 

* Adam Thierer is a senior research fellow at the Mercatus Center at George Mason 

University where he works with the Technology Policy Program. 

1 “Is God Dead?” TIME, April 8, 1966, 

www.time.com/time/covers/0,16641,19660408,00.html 

2 Chris Anderson & Michael Wolff, The Web Is Dead. Long Live the Internet, WIRED, Aug. 17, 

2010, www.wired.com/magazine/2010/08/ff_webrip/all/1. Incidentally, there’s a long 

history of pundits declaring just about everything “dead” at some point, from email, RSS, 

and blogging to eReaders, browser, and even Facebook and Twitter. See Harry McCracken, 

The Tragic Death of Practically Everything, TECHNOLOGIZER, Aug. 18, 2010, 

http://technologizer.com/2010/08/18/the-tragic-death-of-practically-everything 

3 The Web’s New Walls, THE ECONOMIST, Sept. 2, 2010, 

www.economist.com/research/articlesBySubject/displayStory.cfm?story_id=169435 

79&subjectID=348963&fsrc=nwl


The wide-open Internet experience of the 

past decade is giving way to a new regime 

of corporate control, closed platforms, and 

walled gardens. 

This fear is given fuller elucidation in 

recent books by two of the intellectual 

godfathers of modern cyberlaw: Jonathan 

Zittrain’s The Future of the Internet—And 

How to Stop It, 4 and Tim Wu’s The Master 

Switch: The Rise and Fall of Information 

Empires. 5 These books are best understood 

as the second and third installments in a 

trilogy that began with the publication of 

Lawrence Lessig’s seminal 1999 book, Code 

and Other Laws of Cyberspace. 6 

Lessig’s book framed much of how we study and discuss cyberlaw and Internet 

policy. More importantly, Code spawned a bona fide philosophical movement 

within those circles as a polemic against both cyber-libertarianism and Internet 

exceptionalism (closely related movements), as well as a sort of call to arms for 

a new Net activist movement. The book gave this movement its central 

operating principle: Code and cyberspace can be bent to the will of some 

amorphous collective or public will, and it often must be if we are to avoid any 

number of impending disasters brought on by nefarious-minded (or just plain 

incompetent) folks in corporate America scheming to achieve “perfect control” 

over users. 

It’s difficult to know what to label this school of thinking about Internet policy, 

and Prof. Lessig has taken offense at me calling it “cyber-collectivism.” 7 But 

the collectivism of which I speak is a more generic type, not the hard-edged 

Marxist brand of collectivism of modern times. Instead, it’s the belief that 

markets, property rights, and private decision-making about the future course of 

the Net must yield to supposedly more enlightened actors and mechanisms. As 

Declan McCullagh has remarked, Lessig and his students 

4 JONATHAN ZITTRAIN, THE FUTURE OF THE INTERNET—AND HOW TO STOP IT (2008). 

5 TIM WU, THE MASTER SWITCH: THE RISE AND FALL OF INFORMATION EMPIRES (2010). 

6 LAWRENCE LESSIG, CODE AND OTHER LAWS OF CYBERSPACE (1999). 

7 Adam Thierer, Our Conflict of Cyber-Visions, CATO UNBOUND, May 14, 2009, www.catounbound.org/2009/05/14/adam-thierer/our-conflict-of-cyber-visions/


prefer … what probably could be called technocratic 

philosopher kings, of the breed that Plato’s The Republic said 

would be “best able to guard the laws and institutions of our 

State—let them be our guardians.” These technocrats would be 

entrusted with making wise decisions on our behalf, because, 

according to Lessig, “politics is that process by which we 

collectively decide how we should live.” 8 

What is it, exactly, that these cyber-collectivists seek to protect or accomplish? 

To the extent it can be boiled down to a single term, their rallying cry is: 

Openness! “Openness” is almost always The Good; anything “closed” 

(restricted or proprietary) in nature is The Bad. Thus, since they recoil at the 

“cyber-collectivist” label, we might think of adherents to this philosophy as 

“Openness Evangelicals,” since they evangelize in favor of “openness” and 

seemingly make all else subservient to it. 

For example, in Future of the Internet, Zittrain argues that, for a variety of reasons, 

we run the risk of seeing the glorious days of “generative” devices and the 

“open” Internet give way to more “tethered appliances” and closed networks. 

He says: 

Today, the same qualities that led to [the success of the 

Internet and general-purpose PCs] are causing [them] to falter. 

As ubiquitous as Internet technologies are today, the pieces are 

in place for a wholesale shift away from the original chaotic 

design that has given rise to the modern information 

revolution. This counterrevolution would push mainstream 

users away from the generative Internet that fosters innovation 

and disruption, to an appliancized network that incorporates 

some of the most powerful features of today’s Internet while 

greatly limiting its innovative capacity—and, for better or 

worse, heightening its regulability. A seductive and more 

powerful generation of proprietary networks and information 

appliances is waiting for round two. If the problems associated 

with the Internet and PC are not addressed, a set of blunt 

solutions will likely be applied to solve the problems at the 

expense of much of what we love about today’s information 

ecosystem. 9 

8 Declan McCullagh, What Larry Didn’t Get, CATO UNBOUND, May 4, 2009, www.catounbound.org/2009/05/04/declan-mccullagh/what-larry-didnt-get 

9 Zittrain, supra note 4 at 8.


In other words, Zittrain fears most will flock to tethered appliances in a search 

for stability or security. That’s troubling, he says, because those tethered 

appliances are less “open” and more likely to be “regulable,” either by large 

corporate intermediaries or government officials. Thus, the “future of the 

Internet” Zittrain is hoping to “stop” is a world dominated by tethered digital 

appliances and closed walled gardens because they are too easily controlled by 

other actors. 

My primary beef with these “Openness Evangelicals” is not that openness and 

generativity aren’t fine generic principles but that: 

1. They tend to significantly overstate the severity of this problem (the 

supposed decline of openness or generativity, that is); 

2. I’m more willing to allow evolutionary dynamism to run its course within 

digital markets, even if that means some “closed” devices and platforms 

remain (or even thrive); and, 

3. It’s significantly more likely that the “openness” advocated by Openness 

Evangelicals will devolve into expanded government control of cyberspace 

and digital systems than that unregulated systems will become subject to 

“perfect control” by the private sector, as they fear. 

More generally, my problem with this movement—and Zittrain’s book, in 

particular—comes down to the dour, depressing “the-Net-is-about-to-die” fear 

that seems to fuel this worldview. The message seems to be: “Enjoy the good 

old days of the open Internet while you can, because any minute now it will be 

crushed and closed-off by corporate marauders!” Lessig started this nervous 

hand-wringing in Code when he ominously predicted that “Left to itself, 

cyberspace will become a perfect tool of control.” 10 Today, his many disciples 

in academia (including Zittrain and Wu) and a wide variety of regulatory 

advocacy groups continue to preach this gloomy gospel of impending digital 

doom and “perfect control” despite plenty of evidence that supports the case 

for optimism. 

For example, Wu warns there are “forces threatening the Internet as we know 

it” 11 while Zittrain worries about “a handful of gated cloud communities whose 

proprietors control the availability of new code.” 12 At times, this paranoia of 

10 LAWRENCE LESSIG, CODE AND OTHER LAWS OF CYBERSPACE (1999) at 5-6. 

11 WU, supra note 5 at 7. 

12 Jonathan Zittrain, Lost in the Cloud, NEW YORK TIMES, July 19, 2009, 

www.nytimes.com/2009/07/20/opinion/20zittrain.html.


some in the Openness Evangelical clan borders on outright hysteria. In August 

2008, a Public Knowledge analyst likened Apple’s management of applications 

in its iPhone App Store to the tyranny of Orwell’s 1984! 13 In other words, the 

Big Brother they want us to fear is Corporate Big Brother. Someday very soon, 

we are repeatedly told, the corporate big boys will toss the proverbial “master 

switch,” suffocating Internet innovation and digital freedom, and making us all 

cyber-slaves within their commercialized walled gardens. The possibility of 

consumers escaping from these walled gardens or avoiding them altogether is 

treated as remote—if the notion is entertained at all. 

We might think of this fear as “The Great Closing,” or the notion that, unless 

radical interventions are pursued—often through regulation—a Digital Dark 

Age of Closed Systems will soon unfold, complete with myriad America Onlinelike 

walled gardens, “sterile and tethered devices,” corporate censorship, and 

gouging of consumers. Finally, the implicit message in the work of all these 

hyper-pessimistic critics is that markets must be steered in a more sensible 

direction by those technocratic philosopher kings (although the details of their 

blueprint for digital salvation are often scarce). 

Problems with “The Great Closing” Thesis 

There are serious problems with the “Great Closing” thesis as set forth in the 

high-tech threnody of Lessig, Zittrain, Wu, and other Openness Evangelicals, or 

“”as The New York Times has called them, digital “doomsayers.” 14 

No Clear Definitions of Openness or Closedness; 

Both Are Matters of Degree 

“Open” vs. closed isn’t as black and white as some Openness Evangelicals 

make it out to be. For example, Zittrain praises the supposedly more open 

nature of PCs and the openness to innovation made possible by Microsoft’s 

Windows operating system. How ironic, since so many have blasted Windows 

as the Great Satan of closed code! Meanwhile, while most others think of 

Apple as “everyone’s favorite example of innovation,” 15 Zittrain makes the 

13 Alex Curtis, Benefits of iPhone App Store Tainted by 1984-like Control, Public Knowledge Blog, 

Aug. 11, 2008, www.publicknowledge.org/node/1703 The tech gadget website 

Gizmodo recently ran a similar Apple-as-Big-Brother essay: Matt Buchanan, Big Brother Apple 

and the Death of the Program, GIZMODO, Oct. 22, 2010, http://gizmodo.com/5670812/bigbrother-apple-and-the-death-of-the-program. 

14 Eric Pfanner, Proclaimed Dead, Web is Showing Signs of New Life, NEW YORK TIMES, Oct. 31, 

2010, www.nytimes.com/2010/11/01/technology/01webwalls.html 

15 Amar Bhide, Don’t Expect Much From the R&D Tax Credit, WALL STREET JOURNAL, Sept. 11, 

2010, 

http://online.wsj.com/article/SB10001424052748704644404575481534193344088.html


iPhone and iPad out to be “sterile, tethered” appliances. But the company’s 

App Store has offered millions of innovators the opportunity to produce almost 

every conceivable type of mobile application the human mind could imagine for 

those devices. 16 Moreover, those Apple devices don’t block completely “open” 

communications applications or interfaces, such as Web browsers, email and 

SMS clients, or Twitter. “In the abstract,” notes University of South Carolina 

School of Law professor Ann Bartow, “generativity and tetheredness may be 

opposites, but in reality they can exist within a single appliance.” 17 

While the Apple devices seem to prove that, in reality, almost all modern digital 

devices and networks feature some generative and “non-generative” attributes. 

“No one has ever created, and no one will ever create, a system that allows any 

user to create anything he or she wants. Instead, every system designer makes 

innumerable tradeoffs and imposes countless constraints,” note James 

Grimmelmann and Paul Ohm. 18 “Every generative technology faces … 

tradeoffs. Good system designers always restrict generativity of some kinds in 

order to encourage generativity of other kinds. The trick is in striking the 

balance,” they argue. 19 Yet, “Zittrain never fully analyzes split-generativity 

systems, those with generative layers built upon non-generative layers, or viceversa.” 

20 

The zero-sum fear that the ascendancy of mobile apps means less “generativity” 

or the “death of the Web” is another myth. Nick Bilton of The New York Times 

notes: 

Most of these apps and Web sites are so intertwined that it’s 

difficult to know the difference. With the exception of 

downloadable games, most Web apps for news and services 

require pieces of the Web and Internet to function properly. So 

as more devices become connected to the Internet, even if 

they’re built to access beautiful walled gardens, like mobile 

16 Apple, Apple’s App Store Downloads Top Three Billion, Jan. 5, 2010, 

www.apple.com/pr/library/2010/01/05appstore.html 

17 Ann Bartow, A Portrait of the Internet as a Young Man, 108 MICHIGAN LAW REVIEW 6, at 1102- 

03, www.michiganlawreview.org/assets/pdfs/108/6/bartow.pdf 

18 James Grimmelmann & Paul Ohm, Dr. Generative or: How I Learned to Stop Worrying and Love 

the iPhone, MARYLAND LAW REVIEW (2010) at 940-41. 

19 Id. at 941. 

20 Id. at 944. (emphasis in original).


apps or TV-specific interfaces, they will continue to access the 

Web too, enabling each platform to grow concurrently. 21 

Ironically, it was Chris Anderson, editor of Wired and author of the apocalyptic 

“Web is Dead” cover story, who best explained why fears of “The Great 

Closing” are largely overblown: 

Ecommerce continues to thrive on the Web, and no company 

is going to shut its Web site as an information resource. More 

important, the great virtue of today’s Web is that so much of it 

is noncommercial. The wide-open Web of peer production, the 

so-called generative Web where everyone is free to create what 

they want, continues to thrive, driven by the nonmonetary 

incentives of expression, attention, reputation, and the like. 22 

And Jeff Bertolucci of PC World makes it clear generative computing is alive and 

well: 

The next big computing platform won’t be a version of 

Apple’s Mac OS, Google’s Android, or Microsoft’s Windows. 

It’s already here—and it’s the Web. And the drive to offer the 

most compelling window to the Web possible, via the browser, 

is intense. The browser is spreading beyond the PC and 

smartphone to new types of gadgetry, including TV set-top 

boxes and printers. This is a trend that will accelerate in the 

coming years. 23 

The Evils of Closed Systems or Digital 

“Appliances” Are Greatly Over-Stated 

Openness Evangelicals often fail to appreciate how there obviously must have 

been a need / demand for some “closed” or “sterile” devices or else the market 

wouldn’t have supplied them. Why shouldn’t people who want a simpler or more 

secure digital experience be offered such options? Wu worries that devices like 

the iPad “are computers that have been reduced to a strictly limited set of 

functions that they are designed to perform extremely well.” 24 Needless to say, 

21 Nick Bilton, Is the Web Dying? It Doesn’t Look That Way, NEW YORK TIMES BITS BLOG, Aug. 

17, 2010, http://bits.blogs.nytimes.com/2010/08/17/the-growth-of-the-dying-web 

22 Anderson & Wolff, supra note 2. 

23 Jeff Bertolucci, Your Browser in Five Years, PC WORLD, June 16, 2010, 

www.pcworld.com/article/199071/your_browser_in_five_years.html 

24 Wu, supra note 5 at 292.


it will be hard for many consumers to sympathize with Wu’s complaint that 

products work too well! 

However, as noted throughout this essay, it’s also not quite true that those 

devices are as closed or crippled as their critics suggest. As Grimmelmann and 

Ohm aptly note, “restricting generativity in one place (for example, by building 

computers with fixed circuit boards rather than a tangle of reconfigurable wires) 

can massively enhance generativity overall (by making computers cheap and 

usable enough that everyone can tinker with their software).” 25 For example, in 

November 2010, Damon Albarn, lead singer of the popular band “Gorillaz,” 

announced that the group’s next album would be recorded entirely on an iPad. 26 

Regardless, just how far would these critics go to keep devices or platform 

perfectly “generative” or “open” (assuming we can even agree on how to define 

these concepts)? Do the Openness Evangelicals really think consumers would 

be better served if they were forced to fend for themselves with devices that 

arrived totally unconfigured? Should the iPhone or iPad, for example, be 

shipped to market with no apps loaded on the main screen, forcing everyone to 

go find them on their own? Should TiVos have no interactive menus out-ofthe-box, 

forcing consumers to go online and find some “homebrew” code that 

someone whipped up to give users an open source programming guide? 

Some of us are able to do so, of course, and those of us who are tech geeks 

sometimes find it easy to look down our noses at those who want their hand 

held through cyberspace, or who favor more simplistic devices. But there’s 

nothing wrong with those individuals who seek simplicity, stability, or security 

in their digital devices and online experiences—even if they find those solutions 

in the form of “tethered appliances” or “walled gardens.” Not everyone wants 

to tinker or to experience cyberspace as geeks do. Not everyone wants to 

program their mobile phones, hack their consoles, or write their own code. 

Most people live perfectly happy lives without ever doing any of these things! 

Nonetheless, many of those “mere mortals” will want to use many of the same 

toys that the tech geeks use, or they may just want to take more cautious steps 

into the occasionally cold pool called cyberspace—one tippy toe at a time. Why 

shouldn’t those users be accommodated with “lesser” devices or a “curated” 

Web experience? Kevin Kelly argues that there’s another way of looking at 

these trends. Digital tools are becoming more specialized, he argues, and “with 

the advent of rapid fabrication … specialization will leap ahead so that any tool 

can be customized to an individual’s personal needs or desires.” 27 Viewed in 

25 Grimmelmann & Ohm, supra note 18, at 923. 

26 Damon Albarn Records New Gorillaz Album on an iPad, NME NEWS, November 12, 2010, 

http://www.nme.com/news/gorillaz/53816 

27 Kevin Kelly, What Technology Wants (2010) at 295-6.


this light, the Openness Evangelicals would hold back greater technological 

specialization in the name of preserving market norms or structures they prefer. 

The best argument against digital appliancization is that the desire for more 

stable and secure systems will lead to a more “regulable” world—i.e., one that 

can be more easily controlled by both corporations and government. As 

Zittrain puts it: 

Whether software developer or user, volunteering control over 

one’s digital environment to a Manager means that the 

manager can change one’s experience at any time—or worse, 

be compelled to by outside pressures. … The famously 

ungovernable Internet suddenly becomes much more 

governable, an outcome most libertarian types would be 

concerned about. 28 

No doubt, concerns about privacy, child safety, defamation, cybersecurity, 

identity theft and so on, will continue to lead to calls for more intervention. At 

the corporate level, however, some of that potential intervention makes a great 

deal of sense. For example, if ISPs are in a position to help do something to 

help alleviate some of these problems—especially spam and viruses—what’s 

wrong with that? Again, there’s a happy balance here that critics like Zittrain 

and Wu fail to appreciate. Bruce Owen, an economist and the author of The 

Internet Challenge to Television, discussed it in his response to Zittrain’s recent 

book: 

Why does Zittrain think that overreaction is likely, and that its 

costs will be unusually large? Neither prediction is self-evident. 

Faced with the risk of infection or mishap, many users already 

restrain their own taste for PC-mediated adventure, or install 

protective software with similar effect. For the most risk-averse 

PC users, it may be reasonable to welcome “tethered” PCs 

whose suppliers compete to offer the most popular 

combinations of freedom and safety. Such risk-averse users are 

reacting, in part, to negative externalities from the poor 

hygiene of other users, but such users in turn create positive 

externalities by limiting the population of PCs vulnerable to 

contagion or hijacking. As far as one can tell, this can as easily 

produce balance or under-reaction as overreaction—it is an 

empirical question. But, as long as flexibility has value to users, 

28 Jonathan Zittrain, Has the Future of the Internet Happened? Sept. 7, 2010, CONCURRING 

OPINIONS blog, www.concurringopinions.com/archives/2010/09/has-the-future-ofthe-internet-come-about.html


suppliers of hardware and interconnection services will have 

incentives to offer it, in measured ways, or as options. 29 

Indeed, we can find happy middle-ground solutions that balance openness and 

stability—and platform operators must be free to discover where that happy 

medium is through an ongoing process of trial and error, for only through such 

discovery can the right balance be struck in a constantly changing landscape. A 

world full of hybrid solutions would offer more consumers more choices that 

better fit their specific needs. 

Finally, to the extent something more must be done to counter the supposed 

regulability of cyberspace, the solution should not be new limitations on 

innovation. Instead of imposing restrictions on code or coders to limit 

regulability, we should instead place more constraints on our government(s). 

Consider privacy and data collection concerns. While, as a general principle, it 

is probably wise for companies to minimize the amount of data they collect 

about consumers to avoid privacy concerns about data breaches, there are also 

benefits to the collection of that data. So rather than legislating the “right” data 

retention rules, we should hold companies to the promises they make about 

data security and breaches, and tightly limit the powers of government to access 

private information through intermediaries in the first place. 

Most obviously, we could begin by tightening up the Electronic 

Communications Privacy Act (ECPA) and other laws that limit government 

data access. 30 More subtly, we must continue to defend Section 230 of the 

Communications Decency Act, which shields intermediaries from liability for 

information posted or published by users of their systems, because (among 

many things) such liability would make online intermediaries more susceptible 

to the kind of back-room coercion that concerns Zittrain, Lessig and others. If 

we’re going to be legislating the Internet, we need more laws like that, not those 

of the “middleman deputization” model or those that would regulate code to 

achieve this goal. 

Companies Have Strong Incentives to Strike 

the Right Openness/Closedness Balance 

Various social and economic influences help ensure the scales won’t be tipped 

completely in the closed or non-generative direction. The Web is built on 

29 Bruce Owen, As Long as Flexibility Has Value to Users, Suppliers Will Have Incentives to Offer It, 

BOSTON REVIEW, March/April 2008, www.bostonreview.net/BR33.2/owen.php 

30 A broad coalition has proposed such reforms. See www.digitaldueprocess.org.


powerful feedback mechanisms and possesses an extraordinary level of 

transparency in terms of its operations. 

Moreover, the breaking news cycle for tech developments can be measured not 

in days, but in minutes or even seconds. Every boneheaded move meets 

immediate and intense scrutiny by bloggers, tech press, pundits, gadget sites, etc. 

Never has the white-hot spotlight of public attention been so intense in helping 

to shine a light on corporate missteps and forcing their correction. We saw this 

dynamic at work with the Facebook Beacon incident, 31 Google’ Buzz debacle, 32 

Amazon 1984 incident, 33 Apple’s Flash restrictions, 34 the Sony rootkit episode, 35 

and other examples. 

Things Are Getting More Open 

All the Time Anyway 

Most corporate attempts to bottle up information or close off their platforms 

end badly. The walled gardens of the past failed miserably. In critiquing 

Zittrain’s book, Ann Bartow has noted that “if Zittrain is correct that 

CompuServe and America Online (AOL) exemplify the evils of tethering, it’s 

pretty clear the market punished those entities pretty harshly without Internet 

governance-style interventions.” 36 Indeed, let’s not forget that AOL was the 

big, bad corporate boogeyman of Lessig’s Code and yet, just a decade later, it has 

been relegated to an also-ran in the Internet ecosystem. 

31 See Nancy Gohring, Facebook Faces Class-Action Suit Over Beacon, NETWORKWORLD.COM, Aug. 

13, 2008, http://www.networkworld.com/news/2008/081308-facebook-faces-classaction-suit-over.html. 

32 See Ryan Paul, EPIC Fail: Google Faces FTC Complaint Over Buzz Privacy, ARS TECHNICA, Feb. 

17, 2010, http://arstechnica.com/security/news/2010/02/epic-fail-google-facescomplaint-over-buzz-privacy-issues.ars. 

33 See John Timmer, Amazon Settles 1984 Suit, Sets Limits on Kindle Deletions, ARS TECHNICA, Oct. 

2, 2009, http://arstechnica.com/web/news/2009/10/amazon-stipulates-terms-ofbook-deletion-via-1984-settlement.ars. 

34 See Rob Pegoraro, Apple Ipad’s Rejection of Adobe Flash Could Signal the Player’s Death Knell, THE 

WASHINGTON POST, Feb. 7, 2010, http://www.washingtonpost.com/wpdyn/content/article/2010/02/05/AR2010020501089.html. 

35 See Wikipedia, Sony BMG CD Copy Protection Scandal, 

http://en.wikipedia.org/wiki/Sony_BMG_CD_copy_protection_scandal (last 

accessed Dec. 9, 2010). 

36 Bartow, supra note 17 at 1088, 

www.michiganlawreview.org/assets/pdfs/108/6/bartow.pdf


The America Online Case Study: 

Remembering Yesterday’s Face of “Closed” Evil 

When it comes to “closed” systems, evil has a face, but it seems the face is 

always changing. When Lessig penned Code a decade ago, it was American 

Online (AOL) that was set to become the corporate enslaver of cyberspace. For 

a time, it was easy to see why Lessig and others might have been worried. 

Twenty five million subscribers were willing to pay $20 per month to get a 

guided tour of AOL’s walled garden version of the Internet. Then AOL and 

Time Warner announced a historic mega-merger that had some predicting the 

rise of “new totalitarianisms” 37 and corporate “Big Brother.” 38 

But the deal quickly went off the rails. 39 By April 2002, just two years after the 

deal was struck, AOL-Time Warner had already reported a staggering $54 

billion loss. 40 By January 2003, losses had grown to $99 billion. 41 By September 

2003, Time Warner decided to drop AOL from its name altogether and the deal 

continued to slowly unravel from there. 42 In a 2006 interview with the Wall 

Street Journal, Time Warner President Jeffrey Bewkes famously declared the 

death of “synergy” and went so far as to call synergy “bullsh*t”! 43 In early 2008, 

Time Warner decided to shed AOL’s dial-up service 44 and in 2009 spun off 

AOL entirely. 45 Further deconsolidation followed for Time Warner, which 

37 Norman Soloman, AOL Time Warner: Calling The Faithful To Their Knees, Jan. 2000, 

www.fair.org/media-beat/000113.html 

38 Robert Scheer, Confessions of an E-Columnist, Jan. 14, 2000, ONLINE JOURNALISM REVIEW, 

www.ojr.org/ojr/workplace/1017966109.php 

39 Adam Thierer, A Brief History of Media Merger Hysteria: From AOL-Time Warner to Comcast- 

NBC, Progress & Freedom Foundation, PROGRESS ON POINT 16.25, Dec. 2, 2009, 

www.pff.org/issues-pubs/pops/2009/pop16.25-comcast-NBC-merger-madness.pdf 

40 Frank Pellegrini, What AOL Time Warner’s $54 Billion Loss Means, April 25, 2002, TIME 

ONLINE, www.time.com/time/business/article/0,8599,233436,00.html 

41 Jim Hu, AOL Loses Ted Turner and $99 billion, CNET NEWS.COM, Jan. 30, 2004, 

http://news.cnet.com/AOL-loses-Ted-Turner-and-99-billion/2100-1023_3- 

982648.html 

42 Id. 

43 Matthew Karnitschnig, After Years of Pushing Synergy, Time Warner Inc. Says Enough, WALL 

STREET JOURNAL, June 2, 2006, 

http://online.wsj.com/article/SB114921801650969574.html 

44 Geraldine Fabrikant, Time Warner Plans to Split Off AOL’s Dial-Up Service, NEW YORK 

TIMES, Feb. 7, 2008, 

www.nytimes.com/2008/02/07/business/07warner.html?_r=1&adxnnl=1&oref=slog 

in&adxnnlx=1209654030-ZpEGB/n3jS5TGHX63DONHg 

45 Press Release, Time Warner, Time Warner Inc. Completes Spin-off of AOL Inc. (Dec. 10, 

2009), http://www.timewarner.com/corp/newsroom/pr/0,20812,1946835,00.html.


spun off its cable TV unit and various other properties. Looking back at the 

deal, Fortune magazine senior editor at large Allan Sloan called it the “turkey of 

the decade.” 46 

In the larger scheme of things, AOL’s story has already become an afterthought 

in our chaotic cyber-history. But we shouldn’t let those old critics forget about 

their lugubrious lamentations. To recap: the big, bad corporate villain of 

Lessig’s Code attempted to construct the largest walled garden ever, and partner 

with a titan of the media sector in doing so—and this dastardly plot failed miserably. 

The hysteria about AOL’s looming monopolization of instant messaging—and 

with it, the rest of the Web—seems particularly silly: Today, anyone can 

download a free chat client like Digsby or Adium to manage multiple IM 

services from AOL, Yahoo!, Google, Facebook and just about anyone else, all 

within a single interface, essentially making it irrelevant which chat service your 

friends use. 

From this case study one would think the Openness Evangelicals would have 

gained a newfound appreciation for the evolutionary and dynamic nature of 

digital markets and come to understand that, in markets built upon code, the 

pace and nature of change is unrelenting and utterly unpredictable. Indeed, 

contra Lessig’s lament in Code that “Left to itself, cyberspace will become a 

perfect tool of control,” cyberspace has proven far more difficult to “control” 

or regulate than any of us ever imagined. The volume and pace of technological 

innovation we have witnessed over the past decade has been nothing short of 

stunning. 

Critics like Zittrain and Wu, however, wants to keep beating the cyber-sourpuss 

drum. So, the face of corporate evil had to change. Today, Steve Jobs has 

become the supposed apotheosis of all this closed-system evil instead of AOL. 

Jobs serves as a prime villain in the books of Zittrain and Wu and in many of 

the essays they and other Openness Evangelicals pen. It’s worth noting, 

however, that their enemies list is growing longer and now reads like a “Who’s 

Who” of high-tech corporate America. According to Zittrain and Wu’s books, 

’we need to worry about just about every major player in the high-tech 

ecosystem—telcos, cable companies, wireless operators, entertainment 

providers, Facebook, and others. 

Even Google—Silicon Valley’s supposed savior of Internet openness—is not 

spared their scorn. “Google is the Internet’s switch,” Wu argues. “In fact, it’s 

46 Allan Sloan, ‘Cash for…’ and the Year’s Other Clunkers, WASHINGTON POST, Nov. 17, 2009, 

www.washingtonpost.com/wpdyn/content/article/2009/11/16/AR2009111603775.html


the world’s most popular Internet switch, and as such, it might even be 

described as the current custodian of the Master Switch.” More ominously, he 

warns, “it is the switch that transforms mere communications into 

networking—that ultimately decides who reaches what or whom.” 47 

It seems, then, that the face of “closed” evil is constantly morphing. Shouldn’t 

that tell us something about how dynamic these markets are? 

There are few reasons to believe that today’s efforts to build such walled 

gardens would end much differently. Indeed, increasingly when companies or 

coders erect walls of any sort, holes form quickly. For example, it usually 

doesn’t take long for a determined group of hackers to find ways around 

copy/security protections and “root” or “jailbreak” phones and other devices. 48 

Once hacked, users are usually then able to configure their devices or 

applications however they wish, effectively thumbing their noses at the 

developers. This process tends to unfold in a matter of just days, even hours, 

after the release of a new device or operating system. 

Number of Days Before New Devices Were “Rooted” or “Jailbroken” 49 

original iPhone 10 days 

original iPod Touch 35 days 

iPhone 3G 8 days 

iPhone 3GS 1 day 

iPhone 4 38 days 

iPad 1 day 

T-Mobile G1 (first Android phone) 13 days 

Palm Pre 8 days 

Of course, not every user will make the effort—or take the risk 50—to hack their 

devices in this fashion, even once instructions are widely available for doing so. 

47 Wu, supra note 5 at 280. 

48 “In living proof that as long as there’s a thriving geek fan culture for a device, it will never be 

long for the new version to be jailbroken: behold iOS 4.1. Most people are perfectly willing 

to let their devices do the talking for them, accept what’s given, and just run sanctioned 

software. But there are those intrepid few—who actually make up a fairly notable portion of 

the market—who want more out of their devices and find ways around the handicaps built 

into them by the manufacturers.” Kit Dotson, New iOS for Apple TV Firmware Released, 

Promptly Decrypted, SiliconAngle, Sept. 28, 2010, http://siliconangle.com/blog/2010/09/ 

28/new-ios-for-apple-tv-firmware-released-promptly-decrypted 

49 Original research conducted by author and Adam Marcus based on news reports.


Nonetheless, even if copyright law might sometimes seek to restrict it, the 

hacking option still exists for those who wish to exercise it. Moreover, because 

many manufacturers know their devices are likely to be hacked, they are 

increasingly willing to make them more “open” right out of the gates or offer 

more functionality/flexibility to make users happy. 

Innovation Continues to Unfold Rapidly 

in Both Directions along the “Open” 

vs. “Closed” Continuum 

As noted above, part of Zittrain and Wu’s lament seems to be that the devices 

that the hoi polloi choose might crowd out those favored by tinker-happy tech 

geeks (of which I count myself a proud member). But we geeks need not fear 

such foreclosure. Just because there are some “closed” systems or devices on 

the market, it doesn’t mean innovation has been foreclosed among more 

“open” systems or platforms. A hybrid future is both possible and desirable. 

Again, we can have the best of both worlds—a world full of plenty of closed 

systems or even “tethered appliances,” but also plenty of generativity and 

openness. As Web 2.0 pioneer Tim O’Reilly notes: 

I’m not terribly taken in by the rhetoric that says that because 

content silos are going up, and we’re seeing more paid content, 

the open web is over. Individuals, small companies, 

entrepreneurs, artists, all have enormous ability to share and 

distribute their work and find an audience. I don’t see that 

becoming less in today’s environment. 51 

Consider the battle between the Apple iPhone and Google Android mobile 

phone operating systems. Zittrain says Android is “a sort of canary in the coal 

mine” 52 for open platforms, but ignores the frantic pace of its growth, now 

accounting for one-quarter of mobile Web traffic just three years after its 

inception 53 and stealing away Apple’s marketshare in the process. 54 Beyond 

50 Rooting or jailbreaking a smartphone creates the risk of “bricking” the device—rendering it 

completely inoperable (and thus no more useful than a brick). Additionally, hacking devices 

in this fashion typically voids any manufacturer warranty. 

51 The Web is Dead? A Debate, WIRED, Aug. 17, 2010, 

www.wired.com/magazine/2010/08/ff_webrip_debate/all/1 


OPINIONS blog, www.concurringopinions.com/archives/2010/09/has-the-future-ofthe-internet-come-about.html 

53 Sean Hollister, Android Accounts for One-Quarter of Mobile Web Traffic, Says Quantcast, 

ENGADGET, Sept. 4, 2010, www.engadget.com/2010/09/04/android-accounts-for-onequarter-of-mobile-web-traffic-says-qua; 

Android Most Popular Operating System in U.S.


downplaying Android’s success as a marketplace triumph for openness (and 

proof of the non-governmental forces that work to force a balance between 

openness and closedness), Zittrain also reverts to the “kill switch” boogeyman: 

He warns us that any day now Google could change its mind, close the Android 

platform, and “kill an app, or the entire phone” remotely. 55 But where’s the 

business sense in that? What’s the incentive for Google to pursue such a course 

of action? Would Google be able to produce all those millions of apps 

currently produced by independent developers? That seems both unlikely and 

unpopular. Meanwhile, how many times has supposedly control-minded Apple 

actually thrown the dreaded “kill switch” on apps? There are tens of millions of 

apps in Apple’s App Store and hundreds of billions of downloads. If Steve Jobs 

is supposed to be the great villain of independent innovation, he seems to be 

doing a pretty bad job at it! “The App Store is, by some estimates, now a multibillion-dollar-a-year 

business,” note Grimmelmann and Ohm. 56 “The iPhone is 

a hotbed of creative tinkering; people are doing amazing things with it.” 57 

In fact, Wu admits Apple’s App Store offers a “seemingly unlimited variety of 

functions” and that “Apple does allow outsiders to develop applications on its 

platform” since “the defeat of the Macintosh by Windows taught Jobs that a 

platform completely closed to outside developers is suicide.” 58 That should be 

the end of the story. Yet Wu’s fear of that big proverbial “kill switch” overrides 

all: Any day now, that switch will be thrown and Lessig’s pessimistic predictions 

of “perfect control” will finally come to pass, he implies. As Wu says, “all 

innovation and functionality are ultimately subject to Apple’s veto.” 59 And 

consider the lament of Tom Conlon of Popular Science: “Once we replace the 

personal computer with a closed-platform device such as the iPad, we replace 

Among Recent Smartphone Buyers, NIELSEN WIRE, Oct. 5, 2010, 

http://blog.nielsen.com/nielsenwire/online_mobile/android-most-popularoperating-system-in-u-s-among-recent-smartphone-buyers 

54 Tricia Duryee, Apple Continued To Lose U.S. Marketshare Despite Spike From iPhone 4 Sales, 

MOCONEWS.NET, Sept. 15, 2010, http://moconews.net/article/419-apple-continuedto-lose-u.s.-marketshare-despite-spike-from-iphone-4-sa; 

Miguel Helft, The iPhone 

Has a Real Fight on Its Hands, NEW YORK TIMES BITS, Oct. 5, 2010, 

http://bits.blogs.nytimes.com/2010/10/05/the-iphone-has-a-real-fight-on-itshands/ 


OPINIONS blog, www.concurringopinions.com/archives/2010/09/has-the-future-ofthe-internet-come-about.html 

56 Grimmelmann & Ohm, supra note 18 at 923. 

57 Id. 

58 Wu, supra note 5 at 292. 

59 Id.


freedom, choice, and the free market with oppression, censorship, and 

monopoly.” 60 But Apple is hardly the only game in town, and each time Apple 

creates a new product category (iPod, iPhone, iPad, etc.), other companies are 

quick to follow with their own, usually more open systems, often running 

Google’s Android operating system. 

Neither Wu nor Zittrain, however, spend much time investigating how often 

their proverbial kill switch is actually thrown—by Apple or anyone else. There 

have been a handful of examples, but those are hardly the rule. The vast 

majority of all applications are immediately accepted and offered on the 

platform. Moreover, if they were blocked, they could quickly be found on other 

platforms. Again, there are plenty of alternatives to Apple products if you don’t 

like their (somewhat) more restrictive policies regarding application 

development. 

Bottom line: Today’s supposed “walled gardens” are less “walled” than ever 

before, and “closed” systems aren’t really so closed. 

The Internet Was Never Quite 

So Open or Generative 

At times, Zittrain and others seem to have created an Internet imago; an 

idealized conception of a supposed better time when cyberspace was more open 

and vibrant. But let’s face it, the “good ol’ days” that many Openness 

Evangelicals seem to be longing for weren’t really so glorious. Were you online 

back in 1994? Did you enjoy Trumpet Winsock and noisy 14.4 baud modems? 

Did you like loading up multiple 5¼-inch floppy disks just to boot your 

machine? Needless to say, most of us don’t miss those days. 

Here’s the other forgotten factor about the Net’s early history: Until the Net 

was commercialized, it was an extremely closed system. As Geert Lovink 

reminds us: 

[In] [t]he first decades[,] the Internet was a closed world, only 

accessible to (Western) academics and the U.S. military. In 

order to access the Internet one had to be an academic 

computer scientist or a physicist. Until the early nineties it was 

not possible for ordinary citizens, artists, business[es] or 

activists, in the USA or elsewhere, to obtain an email address 

60 Tom Conlon, The iPad’s Closed System: Sometimes I Hate Being Right, POPULAR SCIENCE, Jan. 29, 

2010, www.popsci.com/gadgets/article/2010-01/ipad%E2%80%99s-closed-systemsometimes-i-hate-being-right


and make use of the rudimentary UNIX-based applications. … 

It was a network of networks—but still a closed one. 61 

Ironically, it was only because Lessig and Zittrain’s much-dreaded AOL and 

CompuServe came along that many folks were even able to experience and 

enjoy this strange new world called the Internet. “The fact that millions of 

Americans for the first time experienced the Internet through services like AOL 

(and continue to do so) is a reality that Zittrain simply overlooks,” notes 

Lovink. 62 Could it be that those glorious “good ol’ days” Zittrain longs for were 

really due to the way closed “walled gardens” like AOL and CompuServe held 

our hands to some extent and gave many new Netizens a guided tour of 

cyberspace? 

Regardless, we need not revisit or reconsider that history. That’s ancient history 

now because the walls around those gardens came crumbling down. 

Summary 

When you peel away all the techno-talk and hand-wringing, what Zittrain and 

other Openness Evangelicals object to is the fact that some people are making 

choices that they don’t approve of. To be generous, perhaps it’s because they 

believe that the “mere mortals” don’t fully understand the supposed dangers of 

the choices they are making. But my contention here has been that things just 

aren’t as bad as they make them out to be. More pointedly, who are these critics 

to say those choices are irrational? 

Again, so what if some mere mortals choose more “closed” devices or 

platforms because they require less tinkering and “just work?” It isn’t the end of 

the world. Those devices or platforms aren’t really as closed as they suggests— 

in fact, they are far more open in some ways that the earlier technologies and 

platforms Zittrain, et.al. glorify. And it simply doesn’t follow that just because 

some consumers choose to use “appliances” that it’s the end of the generative 

devices that others so cherish. “General-purpose computers are so useful that 

we’re not likely to abandon them,” notes Princeton University computer science 

professor Ed Felten. 63 For example, a October 2010 NPD Group survey 

61 Geert Lovink, Zittrain’s Foundational Myth of the Open Internet, NET CRITIQUE BY GEERT 

LOVINK, Oct. 12, 2008, 

http://networkcultures.org/wpmu/geert/2008/10/12/zittrains-foundational-mythof-the-open-internet/ 

62 Id. 

63 Ed Felten, iPad to Test Zittrain’s “Future of the Internet” Thesis, FREEDOM TO TINKER blog, Feb. 

4, 2010, www.freedom-to-tinker.com/blog/felten/ipad-test-zittrains-future-internetthesis


revealed that “contrary to popular belief, the iPad isn’t causing cannibalization 

in the PC market because iPad owners don’t exhibit the same buying and 

ownership patterns as the typical consumer electronics customer.” 64 According 

to NPD, only 13% of iPad owners surveyed bought an iPad instead of a PC, 

while 24% replaced a planned e-reader purchase with an iPad. Thus, to the 

extent the iPad was replacing anything, it would be other “non-generative” 

devices like e-readers. 

In a similar vein, James Watters, Senior Manager of Cloud Solutions 

Development at VMware, argues: 

Innovation will be alive and well because the fundamental 

technologies at the core of cloud computing are designed for 

massive, vibrant, explosive, awesome, and amazing application 

innovation. There will always be a big place in the market for 

companies who achieve design simplicity by limiting what can 

be done on their platforms—Apple and Facebook may march 

to massive market share by this principle—but as long as the 

technologies underpinning the network are open, 

programmable, extensible, modular, and dynamic as they are 

and will be, innovation is in good hands. 65 

Thus, we can have the best of both worlds—a world full of plenty of “tethered” 

appliances, but also plenty of generativity and openness. We need not make a 

choice between the two, and we certainly shouldn’t be demanding someone else 

make it for us. 

Against the Stasis Mentality 

& Static Snapshots 

There are some important practical questions that the Openness Evangelicals 

often fail to acknowledge in their work. Beyond the thorny question of how to 

define “openness” and “generativity,” what metric should be used when existing 

yardsticks become obsolete so regularly? 

This points to two major failings in the work of all the cyber-collectivists— 

Lessig in Code, Zittrain in Future of the Internet, and Wu in The Master Switch: 

64 Nearly 90 Percent of Initial iPad Sales are Incremental and not Cannibalizing the PC Market, According 

to NPD, NPD Group PRESS RELEASE, October 1, 2010, 

www.npd.com/press/releases/press_101001.html 

65 James Watters, NYT Kicks Off Cloud Paranoia Series, SILICONANGLE blog, July 21, 2009, 

http://siliconangle.com/blog/2009/07/21/nyt-kicks-off-cloud-paranoia-editorialseries


1. They have a tendency to adopt a static, snapshot view of markets and 

innovation; and, 

2. They often express an overly nostalgic view of the past (without 

making it clear when the “good ‘old days” began and ended) while 

adopting an excessively pessimist view of the present and the chances 

for progress in the future. 

This is what Virginia Postrel was referring to in The Future and Its Enemies when 

she criticized the stasis mentality because “It overvalues the tastes of an 

articulate elite, compares the real world of trade-offs to fantasies of utopia, 

omits important details and connections, and confuses temporary growing pains 

with permanent catastrophes.” 66 And it is what economist Israel Kirzner was 

speaking of when warned of “the shortsightedness of those who, not 

recognizing the open-ended character of entrepreneurial discovery, repeatedly 

fall into the trap of forecasting the future against the background of today’s 

expectations rather than against the unknowable background of tomorrow’s 

discoveries.” 67 

Indeed, there seems to be a complete lack of appreciation among the Openness 

Evangelicals for just how rapid and unpredictable the pace of change in the 

digital realm has been and will likely continue to be. The relentlessness and 

intensity of technological disruption in the digital economy is truly 

unprecedented but often under-appreciated. We’ve had multiple mini-industrial 

revolutions within the digital ecosystem over the past 15 years. Again, this is 

“evolutionary dynamism” at work. (Actually, it’s more like revolutionary 

dynamism!) Nothing—absolutely nothing—that was sitting on our desks in 1995 

is still there today (in terms of digital hardware and software). It’s unlikely that 

much of what was on our desk in 2005 is still there either—with the possible 

exception of some crusty desktop computers running Windows XP. Thus, at a 

minimum, analysts of innovation in this space “should … extend the time 

horizon for our assessment of the generative ecosystem” 68 to ensure they are 

not guilty of the static snapshot problem. 

Speaking of Windows, it perfectly illustrates the complexity of defining 

generative systems. Compare the half-life of Windows PC operating systems— 

which Zittrain indirectly glorifies in his book as generativity nirvana—to the 

half-life of Android operating systems. Both Apple and Android-based devices 

66 VIRGINIA POSTREL, THE FUTURE AND ITS ENEMIES (1998), at xvii-xviii. 

67 ISRAEL KIRZNER, DISCOVERY AND THE CAPITALIST PROCESS (University of Chicago Press, 

1985), at xi. 

68 Grimmelmann & Ohm, supra note 18 at 947.


have seen multiple OS upgrades since release. Some application developers 

actually complain about this frantic pace of mobile OS “revolutions,” especially 

with the Android OS, since they must deal with multiple devices and OS 

versions instead of just one Apple iPhone. They’d rather see more OS 

consistency among the Android devices for which they’re developing to 

facilitate quicker and more stable rollouts. They also have to consider whether 

and how to develop the same app for several other competing platforms. 

Meanwhile, Windows has offered a more “stable” developing platform for 

developers because Microsoft rolls out OS upgrades at a much slower pace. 

Should we should consider an OS with a slower upgrade trajectory more 

“generative” than an OS that experiences constant upgrades if, in practice, the 

former allows for more “open” (and potentially rapid) independent innovation 

by third parties? Of course, there other factors that play into the “generativity” 

equation, 69 but it would be no small irony to place the Windows PC model on 

the higher pedestal of generativity than the more rapidly-evolving mobile OS 

ecosystem. 

Conclusion: Toward Evolutionary 

Dynamism & Technological Agnosticism 

Whether we are debating where various devices sit on a generativity continuum 

(of “open” versus “closed” systems), or what fits where on a “code failure” 

continuum (of “perfect code” versus “market failure”), the key point is that the 

continuum itself is constantly evolving and that this evolution is taking place at a much 

faster clip in this arena than it does in other markets. Coders don’t sit still. 

People innovate around “failure.” Indeed, “market failure” is really just the 

glass-is-half-empty view of a golden opportunity for innovation. Markets 

evolve. New ideas, innovations, and companies are born. Things generally 

change for the better—and do so rapidly. 

69 “[G]enerativity is essential but can never be absolute. No technological system is perfectly 

generative at all levels, for all users, forever. Tradeoffs are inevitable.” Grimmelmann and 

Ohm, supra note 18 at 923.


In light of the radical revolutions constantly unfolding in this space and 

upending existing models, it’s vitally important we avoid “defining down” 

market failure. This is not based on a blind faith in free markets, but rather a 

profound appreciation for the fact that in markets built upon code, the pace and nature 

of change is unrelenting and utterly unpredictable. Contra Lessig’s lament in Code that 

“Left to itself, cyberspace will become a perfect tool of control”—cyberspace 

has proven far more difficult to “control” or regulate than any of us ever 

imagined. Again, the volume and pace of technological innovation we have 

witnessed over the past decade has been nothing short of stunning. 

We need to give evolutionary dynamism a chance. Sometimes it’s during what 

appears to be a given sector’s darkest hour that the most exciting things are 

happening within it—as the AOL case study illustrates. It’s easy to forget all the 

anxiety surrounding AOL and its “market power” circa 1999-2002, when 

scholars like Lessig predicted that the company’s walled garden approach would 

eventually spread and become the norm for cyberspace. As made clear in the 

breakout above, however, the exact opposite proved to be the case. The critics 

said the sky would fall, but it most certainly did not. 

Similarly, in the late 1990s, many critics—including governments both here and 

in the EU—claimed that Microsoft dominated the browser market. Dour 

predictions of perpetual Internet Explorer lock-in followed. For a short time, 

there was some truth to this. But innovators weren’t just sitting still; exciting 

things were happening. In particular, the seeds were being planted for the rise 

of Firefox and Chrome as robust challengers to IE’s dominance—not to 

mention mobile browsers. Of course, it’s true that roughly half of all websurfers


still use a version of IE today. But IE’s share of the market is falling rapidly 70 as 

viable, impressive alternatives now exist and innovation among these 

competitors is more vibrant than ever. 71 That’s all that counts. The world 

changed, and for the better, despite all the doomsday predictions we heard less 

than a decade ago about Microsoft’s potential dominance of cyberspace. 

Moreover, all the innovation taking place at the browser layer today certainly 

undercuts the gloomy “death of the Net” thesis set forth by Zittrain and others. 

Thus, as O’Reilly argues, this case study again shows us the power of open 

systems and evolutionary dynamism: 

Just as Microsoft appeared to have everything locked down in 

the PC industry, the open Internet restarted the game, away 

from what everyone thought was the main action. I guarantee 

that if anyone gets a lock on the mobile Internet, the same 

thing will happen. We’ll be surprised by the innovation that 

starts happening somewhere else, out on the free edges. And 

that free edge will eventually become the new center, because 

open is where innovation happens. […] it’s far too early to call 

the open web dead, just because some big media companies 

are excited about the app ecosystem. I predict that those same 

big media companies are going to get their clocks cleaned by 

small innovators, just as they did on the web. 72 

In sum, history counsels patience and humility in the face of radical uncertainty 

and unprecedented change. More generally, it counsels what we might call 

“technological agnosticism.” We should avoid declaring “openness” a 

sacrosanct principle and making everything else subservient to it without regard 

to cost or consumer desires. As Anderson notes, “there are many Web 

triumphalists who still believe that there is only One True Way, and will fight to 

the death to preserve the open, searchable common platform that the Web 

represented for most of its first two decades (before Apple and Facebook, to 

name two, decided that there were Other Ways).” 73 The better position is one 

based on a general agnosticism regarding the nature of technological platforms 

and change. In this view, the spontaneous evolution of markets has value in its 

70 Tim Stevens, Internet Explorer Falls Below 50 Percent Global Marketshare, Chrome Usage Triples, 

ENGADGET, Oct. 5, 2010, www.engadget.com/2010/10/05/internet-explorer-fallsbelow-50-percent-global-marketshare-chr 

71 Nick Wingfield & Don Clark, Browsers Get a Face-Lift, WALL STREET JOURNAL, Sept. 15, 2010, 

http://online.wsj.com/article/SB10001424052748704285104575492102514582856.html 

72 The Web is Dead? A Debate, WIRED, Aug. 17, 2010, 

www.wired.com/magazine/2010/08/ff_webrip_debate/all/1 

73 Id.


own right, and continued experimentation with new models—be they “open” 

or “closed,” “generative” or “tethered”—should be permitted. 

Importantly, one need not believe that the markets in code are “perfectly 

competitive” to accept that they are “competitive enough” compared to the 

alternatives—especially those re-shaped by regulation. “Code failures” are 

ultimately better addressed by voluntary, spontaneous, bottom-up, marketplace 

responses than by coerced, top-down, governmental solutions. Moreover, the 

decisive advantage of the market-driven, evolutionary approach to correcting 

code failure comes down to the rapidity and nimbleness of those responses. 

Let’s give those other forces—alternative platforms, new innovators, social 

norms, public pressure, etc.—a chance to work some magic. Evolution happens, 

if you let it.

CHAPTER 3 

IS INTERNET EXCEPTIONALISM DEAD? 

The Third Wave of Internet Exceptionalism 165 

Eric Goldman 

A Declaration of the Dependence of Cyberspace 169 

Alex Kozinski and Josh Goldfoot 

Is Internet Exceptionalism Dead? 179 

Tim Wu 

Section 230 of the CDA: Internet Exceptionalism 

as a Statutory Construct 189 


Internet Exceptionalism Revisited 209 


163

164 CHAPTER 3: IS INTERNET EXCEPTIONALISM DEAD?


The Third Wave of Internet 

Exceptionalism 

By Eric Goldman * 

From the beginning, the Internet has been viewed as something special and 

“unique.” For example, in 1996, a judge called the Internet “a unique and 

wholly new medium of worldwide human communication.” 1 

The Internet’s perceived novelty has prompted regulators to engage in “Internet 

exceptionalism”: crafting Internet-specific laws that diverge from regulatory 

precedents in other media. Internet exceptionalism has come in three distinct 

waves: 

The First Wave: Internet Utopianism 

In the mid-1990s, some people fantasized about an Internet “utopia” that 

would overcome the problems inherent in other media. Some regulators, 

fearing disruption of this possible utopia, sought to treat the Internet more 

favorably than other media. 

47 U.S.C. § 230 (“Section 230”—a law still on the books) is a flagship example 

of mid-1990s efforts to preserve Internet utopianism. The statute categorically 

immunizes online providers from liability for publishing most types of third 

party content. It was enacted (in part) “to preserve the vibrant and competitive 

free market that presently exists for the Internet and other interactive computer 

services, unfettered by Federal or State regulation.” 2 The statute is clearly 

exceptionalist because it treats online providers more favorably than offline 

publishers—even when they publish identical content. 

The Second Wave: Internet Paranoia 

Later in the 1990s, the regulatory pendulum swung in the other direction. 

Regulators still embraced Internet exceptionalism, but instead of favoring the 

Internet, regulators treated the Internet more harshly than analogous offline 

activity. 

For example, in 2005, a Texas website called Live-shot.com announced that it 

would offer “Internet hunting.” The website allowed paying customers to 

* Associate Professor and Director, High Tech Law Institute, Santa Clara University School of 

Law. Email: egoldman@gmail.com. Website: http://www.ericgoldman.org. 

1 American Civil Liberties Union v. Reno, 929 F. Supp. 824 (E.D. Pa. 1996). 

2 47 U.S.C. § 230(b)(2).

166 CHAPTER 3: IS INTERNET EXCEPTIONALISM DEAD? 

control, via the Internet, a gun on its game farm. An employee manually 

monitored the gun and could override the customer’s instructions. The website 

wanted to give people who could not otherwise hunt, such as paraplegics, the 

opportunity to enjoy the hunting experience. 3 

The regulatory reaction to Internet hunting was swift and severe. Over threedozen 

states banned Internet hunting. 4 California also banned Internet fishing 

for good measure. 5 However, regulators never explained how Internet hunting 

is more objectionable than physical space hunting. 

For example, California Sen. Debra Bowen criticized Internet hunting because it 

“isn’t hunting; it’s an inhumane, over the top, pay-per-view video game using 

live animals for target practice … . Shooting live animals over the Internet 

takes absolutely zero hunting skills, and it ought to be offensive to every 

legitimate hunter.” 6 

Sen. Bowen’s remarks reflect numerous unexpressed assumptions about the 

nature of “hunting” and what constitutes fair play. In the end, however, 

hunting may just be “hunting,” in which case the response to Internet hunting 

may just be a typical example of adverse Internet exceptionalism. 7 

The Third Wave: 

Exceptionalism Proliferation 

The past few years have brought a new regulatory trend. Regulators are still 

engaged in Internet exceptionalism, but each new advance in Internet 

technology has prompted exceptionalist regulations towards that technology. 

For example, the emergence of blogs and virtual worlds has helped initiate a 

push towards blog-specific and virtual world-specific regulation. In effect, 

Internet exceptionalism has splintered into pockets of smaller exceptionalist 

efforts. 

3 Sylvia Moreno, Mouse Click Brings Home Thrill of the Hunt, WASH. POST, May 8, 2005. 

4 Internet Hunting Bans, The Humane Society of the United States, 

http://www.hsus.org/web-files/PDF/internethunting_map.pdf (last visited Aug. 23, 

2010). 

5 Zachary M. Seward, Internet Hunting Has Got to Stop – If It Ever Starts, WALL ST. J., Aug. 10, 

2007. 

6 Michael Gardner, Web ‘Hunts’ in Cross Hairs of Lawmakers, S.D. UNION-TRIBUNE, Apr. 6, 

2005. 

7 Eric Goldman, A Web Site for Hunting Poses Questions About Killing, S.J. MERCURY NEWS, July 

25, 2005.


Regulatory responses to social networking sites like Facebook and MySpace are 

a prime example of Internet exceptionalism splintering. Rather than regulating 

these sites like other websites, regulators have sought social networking sitespecific 

laws, such as requirements to verify users’ age, 8 combat sexual 

predators 9 and suppress content that promotes violence. 10 The result is that the 

regulation of social networking sites differs not only from offline enterprises but 

from other websites as well. 

Implications 

Internet exceptionalism—either favoring or disfavoring the Internet—is not 

inherently bad. In some cases, the Internet truly is unique, special or different 

and should be regulated accordingly. Unfortunately, more typically, anti- 

Internet exceptionalism cannot be analytically justified and instead reflects 

regulatory panic. 

In these cases, anti-Internet regulatory exceptionalism can be harmful, especially 

to Internet entrepreneurs and their investors. It can distort the marketplace 

between Web enterprises and their offline competition by hindering the Web 

business’ ability to compete. In extreme cases, such as Internet hunting, 

unjustified regulatory intervention may put companies out of business. 

Accordingly, before enacting any exceptionalist Internet regulation (and 

especially any anti-Internet regulation), regulators should articulate how the 

Internet is unique, special or different and explain why these differences justify 

exceptionalism. Unfortunately, emotional overreactions to perceived Internet 

threats or harms typically trump such a rational regulatory process. Knowing 

this tendency, perhaps we can better resist that temptation. 

8 Nick Alexander, Attorneys General Announce Agreement With MySpace Regarding Social Networking 

Safety, NAA GAZETTE, Jan. 18, 2008, 

http://www.naag.org/attorneys_general_announce_agreement_with_myspace_rega 

rding_social_networking_safety.php; Brad Stone, Facebook Settles with New York, N.Y. 

TIMES BITS BLOG, Oct. 16 2007, http://bits.blogs.nytimes.com/2007/10/16/facebooksettles-with-new-york/. 

9 KIDS Act of 2007 (H.R. 719/S. 431) (requiring sexual predators to register their email 

addresses and other screen names and enabling social networking sites to access those 

electronic identifiers so that the sexual predators can be blocked from registering with the 

social networking sites). 

10 H. Res. 224 (2007) (resolution requesting that social networking sites proactively remove 

“enemy propaganda from their sites,” such as videos made by terrorists).



A Declaration of the 

Dependence of Cyberspace 

By Hon. Alex Kozinski * & Josh Goldfoot ** 

Governments of the Industrial World, you weary giants of flesh and steel, 

I come from Cyberspace, the new home of Mind. On behalf of the future, 

I ask you of the past to leave us alone. You are not welcome among us. 

You have no sovereignty where we gather. 1 

That was the opening of “A Declaration of the Independence of Cyberspace.” 

The would-be Cyber-Jefferson who wrote it was John Perry Barlow, a cofounder 

of the Electronic Frontier Foundation, a noted libertarian and a 

Grateful Dead lyricist. He delivered the Declaration on February 8, 1996, the 

same day that President Clinton signed into law the Communications Decency 

Act. That Act was chiefly an early effort to regulate Internet pornography. 

Many had concerns about that law, and, indeed, the Supreme Court would 

eventually declare most of it unconstitutional. 2 

Barlow’s argument invoked what he believed was a more decisive criticism than 

anything the Supreme Court could come up with. Barlow saw the Internet as 

literally untouchable by our laws. Extolling the power of anonymity, he taunted 

that “our identities have no bodies, so, unlike you, we cannot obtain order by 

physical coercion.” Unlike the Declaration of Independence, this was not a 

declaration that cyberspace was newly independent; it was an observation that 

cyberspace had always been independent, and will always remain independent, 

because its denizens were beyond the law’s reach. 

Needless to say, the weary giants of flesh and steel did not take kindly to the 

Declaration. They fought back hard and won numerous battles: witness the fall 

of Napster, Grokster, Aimster and innumerable other file-sharing and childpornography-trading 

sites and services. Ironically, the Department of 

* Chief Judge, United States Court of Appeals for the Ninth Circuit. 

** B.A., Yale University; J.D., University of Virginia School of Law; Trial Attorney, Department 

of Justice, Criminal Division, Computer Crime & Intellectual Property Section. This essay 

was originally published in 32 COLUMBIA JOURNAL OF LAW & THE ARTS, no. 4, 2009 at 365. 

The views expressed in this essay are the views of the authors and do not necessarily 

represent the views of the U.S. Department of Justice or the United States. 

1 John Perry Barlow, A Declaration of the Independence of Cyberspace (Feb. 8, 1996), 

http://homes.eff.org/~barlow/Declaration-Final.html. 

2 Reno v. American Civil Liberties Union, 521 U.S. 844 (1997).


Homeland Security now has a “National Strategy to Secure Cyberspace.” 3 Even 

the cyber-libertarians have shifted their focus: The Electronic Frontier 

Foundation, which Barlow co-founded, now accepts that there may be a place 

for so-called “network neutrality” regulation, even though it regulates how 

subscribers access the Internet and how content reaches them. 4 

In other ways, the Declaration has proved prescient. As far back as 1996, 

Barlow had identified that the Internet poses a significant problem for 

governments. Then, as now, people used the Internet to break the law. The 

Internet gives those people two powerful tools that help them escape the law’s 

efforts to find and punish them. First, the Internet makes anonymity easy. 

Today any 11-year-old can obtain a free e-mail account, free website and free 

video hosting. The companies that provide these things ask for your name, but 

they make no effort to verify your answer; as a result, only Boy Scouts tell them 

the truth. You can be tracked through your Internet protocol (IP) address, but 

it is not too tough to use proxies or some neighbor’s open Wi-Fi connection to 

get around that problem. Thus, if your online conduct ever hurts someone, it 

will be difficult for the victim to ever find out who you are and sue you. 

Second, the Internet makes long-distance international communication cheap. 

This allows the world’s miscreants, con-artists and thieves easy access to our 

gullible citizens. When people find out they’ve been had, they often find that 

they have no practical recourse because of the extraordinary difficulties involved 

in pursuing someone overseas. The Internet’s global nature makes it easy for 

people to hide from our courts. 

These two advantages of Internet law-breakers pose a serious and recurring 

problem. That problem has been particularly painful for intellectual property 

rights holders. It is common knowledge that instead of buying music or 

movies, you can use the Internet to download perfect copies for free from 

individuals known only by their IP addresses. In some cases, wrongdoers have 

become so bold that they demand payment in exchange for the opportunity to 

download infringing material. 

The situation seemed unsolvable to Barlow and others in 1996. Armed with 

anonymity and invulnerability, Internet actors could ignore efforts to apply law 

to the Internet. Barlow concluded that the Internet’s nature posed an 

insurmountable barrier to any effort at legal enforcement. Some scholars even 

3 The National Strategy to Secure Cyberspace, Feb. 2003, 

http://www.dhs.gov/xlibrary/assets/National_Cyberspace_Strategy.pdf. 

4 See https://www.eff.org/files/filenode/nn/EFFNNcomments.pdf.


began work on theorizing how the diverse denizens of cyberspace might join 

together and go about creating their own indigenous legal system. 5 

But over time, a solution to Barlow’s problem appeared. Let us entertain, for a 

moment, the conceit that there is a “cyberspace,” populated by people who 

communicate online. The denizens of cyberspace exist simultaneously in 

cyberspace and in the real flesh-and-steel world. Their cyberspace selves can be 

completely anonymous; their real-life selves are easier to identify. Their 

cyberspace selves have no physical presence; their real-life selves both exist and 

have base material desires for PlayStations, Porsche Boxsters and Battlestar 

Galactica memorabilia. Their physical selves can be found in the real world and 

made to pay in real dollars or serve real time behind real bars for the damage 

their cyber-selves cause. 

The dilemma that online law-breakers face is that their cyberspace crimes have 

real-life motives and fulfill real-life needs. Therefore, they need some way to 

translate their online misdeeds into offline benefits. The teenager downloads a 

MP3 so that he can listen to it. The con-artist asks for money to be wired to 

him so that he can withdraw it and buy things with it. The fringe activist who emails 

a death threat to a judge does so in the hopes that the judge will change 

his behavior in the real world. 

These Internet actors usually rely on real-world institutions to get what they 

want. They use Internet Service Providers (ISPs) and hosting companies to 

communicate, and they use banks and credit card companies to turn online 

gains into cash. Without these institutions, they either could not accomplish 

their online harms, or they would not be able to benefit from them in the real 

world. Unlike anonymous cyberspace miscreants, however, these institutions 

have street addresses and real, physical assets that can satisfy judgments in the 

United States. By placing pressure on those institutions to cut off service to 

customers who break the law, we can indirectly place pressure on Internet 

wrong-doers. Through this pressure, we have a powerful tool to promote 

online compliance with the law. 

In some cases, for some offenses, we have the legal tools to do this already. For 

intellectual property cases, the tool for holding those institutions liable is 

secondary liability: contributory and vicarious infringement. The Ninth Circuit 

has led the way in developing the law in this area. In Perfect 10 v. Google, the 

court noted the cases that had applied contributory infringement to Internet 

actors, and summarized their holdings as saying that “a computer system 

operator can be held contributorily liable if it has actual knowledge that specific 

infringing material is available using its system … and can take simple measures 

5 See, e.g., David G. Post, Anarchy, State, and the Internet: An Essay on Law-Making in Cyberspace, 

JOURNAL OF ONLINE LAW, Article 3, (1995), available at http://ssrn.com/abstract=943456.


to prevent further damage to copyrighted works … yet continues to provide 

access to infringing works.” 6 In other words, if people are using your stuff to 

infringe copyrights, and you know about it, and you can easily stop them, but 

you do not, then you are on the hook. 

The motive behind secondary liability is simple. Everyone agrees that the direct 

infringers ideally should be the ones to pay. But there might be too many of 

them to sue; or, they might be anonymous; or, they might be in Nigeria. This 

can make them apparently invulnerable to lawsuits. That invulnerability has a 

cause: someone is providing the tools to infringe and looking the other way. 

The doctrine of secondary liability says that such behavior is unacceptable. 

Those who provide powerful tools that can be used for good or evil have some 

responsibility to make sure that those tools are used responsibly. 

Put more directly: with some changes to the law, the institutions that enable the 

anonymity and invulnerability of cyberspace denizens can be held accountable 

for what their anonymous and invulnerable customers do. The anonymity of 

cyberspace is as much a creation of men as it is a creation of computers. It is 

the result of policy choices. We have accepted, without serious examination, 

that it is perfectly fine for a business to grant free Web space and e-mail to any 

schmuck who comes off the street with an IP address, and then either keep no 

record of that grant or discard the record quickly. Businesses that do this are 

lending their powerful and potentially harmful capabilities and demanding little 

accountability in return. That arrangement has obvious benefits but also 

obvious costs. The victims of online torts and crimes bear these costs, and 

those victims are, overwhelmingly, third parties. They include big movie 

studios, middle-aged Internet newbies and, unfortunately in some cases, young 

children. 

If the legal rules change, and companies are held liable more often for what 

their users do, then the cost of anonymity would shift away from victims and 

toward the providers. In this world, providers will be more careful about 

identifying users. Perhaps online assertions of identity will be backed up with 

offline proof; providers will be more careful about providing potential scam 

artists in distant jurisdictions with the tools to practice their craft. All this 

would be expensive for service providers, but not as expensive as it is for 

injured parties today. 

Secondary liability should not reach every company that plays any hand in 

assisting the online wrong-doer, of course. Before secondary liability attaches, 

the plaintiff must show that the defendant provided a crucial service, knew of 

the illegal activity, and had a right and a cost-justified ability to control the 

6 Perfect 10, Inc. v. Amazon.com, Inc., 508 F.3d 1146, 1172 (9th Cir. 2007) (quotations, 

citations and italics omitted).


infringer’s actions. This rule will in almost every case exclude electrical utilities, 

landlords, and others whose contributions to illegal activity are minuscule. 

While we have come a long way from Barlow’s Declaration of the 

Independence of Cyberspace, the central idea behind it—that the Internet is a 

special place, separate somehow from the brick and mortar world, and thus 

subject to special rules and regulations, or no rules and regulations—lingers. 

The name itself has a powerful influence: we don’t speak of “telephone-space” 

or “radio-space” or “TV-space”—though we do have Television City in 

Hollywood. Prior technological advances that aided in connecting people were 

generally recognized as tools to aid life in the real world; no one claimed that 

they made up a separate dimension that is somehow different and separate from 

the real word. Every time we use the term “cyberspace” or the now-outmoded 

“Information Superhighway,” we buy into the idea that the world-wide network 

of computers that people use for electronic commerce and communication is a 

separate, organic entity that is entitled to special treatment. 

This idea of cyberspace as a separate place subject to a different set of rules— 

one where courts ought to tread lightly lest they disturb the natural order of 

things and thereby cause great harm—still arises in many court cases. 7 

The first of these is Perfect 10 v. Visa—a case where one of the authors of this 

piece was in the dissent. 8 The facts are simple: plaintiff produces and owns 

pictures of scantily-clad young women, which it sells online. It alleged that 

unknown parties had copied the pictures and were selling them online, at a 

lower price, using servers in remote locations where the legal system was not 

hospitable to copyright and trademark lawsuits, and, moreover, they could fold 

up their tents and open up business elsewhere if anyone really tried to pursue 

them. So the plaintiff didn’t try to sue the primary infringers; instead, it went 

after the credit card companies that were processing the payments for what they 

claimed were pirated photographs. 

7 Some disclaimers: One of the authors of this piece (Chief Judge Kozinski) sat on the panel 

that decided some of the cases given as examples here. He wants to make it clear that he 

won’t re-argue the cases here. Both involved split decisions, and his views as to how those 

cases should have come out is set out in his opinions in those cases. His colleagues on the 

other side are not present to argue their positions and, in any event, it’s unseemly to continue 

a judicial debate after the case is over. Furthermore, despite his disagreement with his 

colleagues, he respects and appreciates their views. The judges that came out the other way 

are some of the dearest of his colleagues, and some of the finest judges anywhere. The 

disagreement is troubling, because they bring a wealth of intelligence, diligence, talent, 

experience and objectivity to the problem, and he can’t quite figure out why they see things 

so differently. 

8 Perfect 10, Inc. v. Visa Int’l Serv. Ass’n, 494 F.3d 788 (9th Cir. 2007).


This was by far not the first case that applied the doctrine of secondary 

infringement to electronic commerce. The cases go back at least to the 1995 

case of Religious Technology Center v. Netcom, 9 a case involving the liability of an ISP 

for damage caused when it posted copyrighted Scientology documents to 

USENET, at the direction of one of its users. And, of course, the Napster, 

Aimster and Grokster cases all dealt with the secondary liability of those who 

assist others in infringement. 10 Perfect 10, though, presented a novel question: 

how do you apply the doctrine of secondary infringement to people who help 

the transaction along, but never have any physical contact with the protected 

work? 

Two excellent and conscientious Ninth Circuit jurists, Judges Milan Smith and 

Stephen Reinhardt, said there was no liability, whereas the dissenting judge 

concluded that there was. Visa, the dissent argued, was no different from any 

other company that provided a service to infringers, knew what it was doing, 

and had the ability to withdraw its service and stop the infringement, but did 

nothing. 

This debate fits within a larger context. In the majority’s rejection of 

contributory liability, it cited a public policy decision that found that the 

Internet’s development should be promoted by keeping it free of legal 

regulation. Relatedly, the majority distinguished some precedent by saying that 

its “tests were developed for a brick-and-mortar world” and hence “do not lend 

themselves well to application in an electronic commerce context.” 11 

This argument channels Barlow’s declaration that users of the Internet are 

entitled to special treatment (or, as he would have it, entitled to no treatment). 

The chief justification for this argument is that the Internet is so new, exotic 

and complicated that the imposition of legal rules will chill, stifle, discourage or 

otherwise squelch the budding geniuses who might otherwise create the next 

Google, Pets.com, or HamsterDance.com. For example, the Electronic 

Frontier Foundation argued to the Supreme Court during the Grokster case that 

if the Ninth Circuit’s opinion were reversed, the effect would “threaten 

innovation by subjecting product design to expensive and indeterminate judicial 

9 Religious Tech. Ctr. v. Netcom, 907 F.Supp. 1361 (N.D. Cal. 1995). 

10 See A&M Records, Inc. v. Napster, Inc., 284 F.3d 1091 (9th Cir. 2002); In re Aimster 

Copyright Litig., 334 F.3d 643 (7th Cir. 2003); Metro-Goldwyn-Mayer Studios Inc. v. 

Grokster, Ltd., 545 U.S. 913 (2005). 

11 Perfect 10, 494 F.3d at 798, n.9.


second-guessing.” 12 The Ninth Circuit was reversed, and if that decision slowed 

the pace of product design, no one seems to have noticed. 

This argument became particularly central in a second case, Fair Housing Council 

of San Fernando Valley v. Roommates.com. 13 The case involved a claim that the 

commercial website Roommates.com violated state and federal fair housing 

laws by helping to pair up roommates according to their personal preferences, 

the exercise of which is allegedly prohibited by law. Again, one of the authors 

of this piece was a judge on that case, and was in the majority at both the panel 

and the en banc level—despite the efforts of some conscientious and brilliant 

dissenting judges, of whose intellectual rigor and commitment to the rule of law 

no one can doubt. 

The majority mostly held that Roommates.com could be held liable, if the 

plaintiff’s allegations were proven true. The court held essentially that an online 

business had to be held to the same substantive law as businesses in the brickand-mortar 

world. The dissenters saw things quite differently; to them, the 

majority placed in jeopardy the survival of the Internet. Here is a taste of the 

dissent: 

On a daily basis, we rely on the tools of cyberspace to help us 

make, maintain, and rekindle friendships; find places to live, 

work, eat, and travel; exchange views on topics ranging from 

terrorism to patriotism; and enlighten ourselves on subjects 

from “aardvarks to Zoroastrianism.” … The majority’s 

unprecedented expansion of liability for Internet service 

providers threatens to chill the robust development of the 

Internet that Congress envisioned … . We should be looking 

at the housing issue through the lens of the Internet, not from 

the perspective of traditional publisher liability. 14 

And finally, the unkindest cut of all: “The majority’s decision, which sets us 

apart from five circuits, … violates the spirit and serendipity of the Internet.” 15 

The argument that a legal holding will bring the Internet to a standstill makes 

most judges listen closely. Just think of the panic that was created when the 

12 See Brief for Respondents, Metro-Goldwyn-Mayer Studios Inc. v. Grokster, Ltd., No. 04-480 

(9th Cir. Mar. 1, 2005), available at 2005 WL 508120 and at 

http://w2.eff.org/IP/P2P/MGM_v_Grokster/20050301_respondents_brief.pdf. 

13 Fair Housing Council of San Fernando Valley v. Roommates.Com, LLC, 521 F.3d 1157 (9th 

Cir. 2008). 

14 Id. at 1176-77 (footnote omitted). 

15 Id. at 1177.


Blackberry server went down for a few hours. No one in a black robe wants to 

be responsible for anything like that, and when intelligent, hard-working, 

thoughtful colleagues argue that this will be the effect of one of your rulings, 

you have to think long and hard about whether you want to go that way. It 

tests the courage of your convictions. 

Closely related is the argument that, even if you don’t bring down the existing 

structure, the threat of liability will stifle innovation, so that the progress we 

have seen in recent years—and the gains in productivity and personal 

satisfaction—will stop because the legal structure has made innovation too risky 

or expensive. The innovation argument is partly right but mostly wrong. 

Certainly, some innovators will shy away from legally murky areas. It’s hard to 

think of a worse recipe for creativity than having a lawyer attend every 

engineering meeting. But promoting innovation alone cannot be a sufficient 

justification for exempting innovators from the law. An unfortunate result of 

our complex legal system is that almost everyone is confused about what the 

law means, and everyone engaged in a business of any complexity at some point 

has to consult a lawyer. If the need to obey the law stifles innovation, that 

stifling is just another cost of having a society ruled by law. In this sense, the 

Internet is no different than the pharmaceutical industry or the auto industry: 

They face formidable legal regulation, yet they continue to innovate. 

There is an even more fundamental reason why it would be unwise to exempt 

the innovators who create the technology that will shape the course of our lives: 

Granting them that exemption will yield a generation of technology that 

facilitates the behavior that our society has decided to prohibit. If the Internet 

is still being developed, then we should do what we can to guide its 

development in a direction that promotes compliance with the law. 

For example, what use is “innovation” in creating a job hunting site if the 

innovators produce a site that invites employers to automatically reject any 

applicant from a particular race? Perhaps the job site is a bold new innovation 

that makes hiring far easier and more efficient than it has ever been. But if this 

site is used widely, it will facilitate racial discrimination in hiring—conduct that 

society has already decided it must prohibit. Similarly, is a file-sharing service 

such as Grokster worth the harm it causes by offering no built-in tools for 

identifying participants or establishing they have the right to “share” the files 

they copy? Far from exempting this growing industry from the law, we should 

vigorously enforce the law as the industry grows, so that when it is mature, its 

services won’t guide behavior toward conduct that society has decided to 

discourage. As difficult as it might be for innovators today, it is easier than the 

alternatives: forcing them to rebuild everything ten years down the road, or 

grudgingly accepting that we have surrendered key aspects of our ability to 

govern our society through law.


It is Barlow who is generally credited with taking the word “cyberspace” from 

the science fiction of William Gibson and applying it to the Internet. 16 In doing 

so, he launched the conceit that such a “space” exists at all. This was wholly 

unjustified. It is a mistake to fall into Barlow’s trap of believing that the set of 

human interactions that is conducted online can be neatly grouped together into 

a discrete “cyberspace” that operates under its own rules. Technological 

innovations give us new capabilities, but they don’t change the fundamental 

ways that humans deal with each other. The introduction of telephones and 

cars did create new legal questions. Those questions all revolved around what 

the acceptable uses of the new technologies were. How closely can you follow 

the car in front of you on the highway? Can you repeatedly dial someone’s 

phone to annoy them? Can you tap into a phone conversation or put a tape 

recorder in a phone booth? Over time, courts and legislatures answered these 

questions with new legal rules. They had to; the essence of the controversy 

arose from the new technological abilities. But no one thought that telephones 

and cars changed the legal rules surrounding what was said on a telephone or 

where a car traveled. Can an oral contract be formed with a telephone call? Of 

course; it is still two people speaking. Is it trespassing to drive across my 

neighbor’s front yard? Of course; you are on his land. 

Like cars and telephones, the Internet prompts new questions about the 

acceptable uses of the new technology. Is port-scanning a form of hacking? 

When does title to a domain name legally transfer? While analogies to settled 

legal rules are helpful in answering these questions, they are not conclusive. 

Answers to these questions will look like new legal rules. 

But when the Internet is involved in a controversy only because the parties 

happened to use it to communicate, new legal rules will rarely be necessary. 

When the substance of the offense is that something was communicated, then 

the harm occurs regardless of the tools used to communicate. If an attorney 

betrays a client’s confidence, the duty to the client is breached regardless of 

whether the attorney used a telephone, a newspaper, a radio station, or the 

Internet. The choice of communication medium might affect the magnitude of 

the harm, but if it is illegal for A to communicate X to B without C’s 

permission, there is no reason to fashion new rules of liability that depend on 

the mode of communication used. 

There are some ways that the Internet might require courts to re-think legal 

rules. The Internet makes long-distance communication cheaper than it was 

before. To the extent that existing legal rules were premised on the assumption 

that communications were expensive, the Internet might require a reappraisal. 

Courts are already reevaluating, for example, what it means to do business 

16 See John Perry Barlow, Crime and Puzzlement: In Advance of the Law on the Electronic Frontier, 

WHOLE EARTH REV., Sept. 22, 1990, at 44, 45.


within a state, for purposes of the long-arm statute, when the defendant’s 

“business establishment” is a server located in Uzbekistan. 

Yet the vast majority of Internet cases that have reached the courts have not 

required new legal rules to solve them. It has been fifteen years since America 

Online unleashed its hordes of home computing modem-owners on e-mail and 

the Internet and fifteen years since the release of the Mosaic Web browser. 

After all that time, we have today relatively few legal rules that apply only to the 

Internet. Using the Internet, people buy stocks, advertise used goods and apply 

for jobs. All those transactions are governed by the exact same laws as would 

govern them if they were done offline. 

Those who claim the Internet requires special rules to deal with these ordinary 

controversies have trouble explaining this history. Despite this dearth of 

Internet-specific law, the Internet is doing wonderfully. It has survived 

speculative booms and busts, made millionaires out of many and, unfortunately, 

rude bloggers out of more than a few. The lack of a special Internet civil code 

has not hurt its development. 

The Internet, it turns out, was never so independent or sovereign as early 

idealists believed. It was an astounding social and technological achievement, 

and it continues to change our lives. But it has not proven to be invulnerable to 

legal regulation—at least, not unless we choose to make it invulnerable. As 

intriguing as Barlow’s Declaration of Independence was, the original 1776 

Declaration is more profound in its understanding of the purpose and abilities 

of government: men have rights of “Life, Liberty and the pursuit of 

Happiness,” and “to secure these rights, Governments are instituted among 

Men.” The government that we have instituted retains its purpose of securing 

those rights, and it accomplishes that purpose through the law. We have seen 

that our government has many tools at its disposal through which it can bring 

law to the Internet’s far reaches. The Internet might pose obstacles toward that 

job, but those obstacles can be overcome. The question is whether we will do it.


Is Internet Exceptionalism Dead? 

By Tim Wu * 

In 1831, Alexis de Tocqueville released Democracy in America, the founding text 

of “American exceptionalism.” After long study in the field, America, he had 

concluded, was just different than other nations. In an often-quoted passage, de 

Tocqueville wrote: 

The position of the Americans is therefore quite exceptional, 

and it may be believed that no democratic people will ever be 

placed in a similar one. Their strictly Puritanical origin—their 

exclusively commercial habits—even the country they inhabit, 

which seems to divert their minds from the pursuit of science, 

literature, and the arts—the proximity of Europe, which allows 

them to neglect these pursuits without relapsing into 

barbarism—a thousand special causes, of which I have only 

been able to point out the most important—have singularly 

concurred to fix the mind of the American upon purely 

practical objects. His passions, his wants, his education, and 

everything about him seem to unite in drawing the native of 

the United States earthward; his religion alone bids him turn, 

from time to time, a transient and distracted glance to heaven. 1 

Is there such a thing as Internet exceptionalism? If so, just what is the Internet 

an exception to? It may appear technical, but this is actually one of the big 

questions of our generation, for the Internet has shaped the United States and 

the world over the last twenty years in ways people still struggle to understand. 

From its beginnings the Internet has always been different from the networks 

that preceded it—the telephone, radio and television, and cable. But is it 

different in a lasting way? 

The question is not merely academic. The greatest Internet firms can be 

succinctly defined as those that have best understood what makes the Internet 

different. Those that have failed to understand the “Network of Networks”— 

say, AOL, perished, while those that have, like Google and Amazon, have 

flourished. Hence the question of Internet exceptionalism is often a multibillion 

dollar question. The state of the Internet has an obvious effect on 

national and international culture. It is also of considerable political relevance, 

* Professor, Columbia Law School; Fellow, New America Foundation 

1 ALEXIS DE TOCQUEVILLE, DEMOCRACY IN AMERICA 519 (Henry Reeve trans., D. Appleton 

and Company 1904) (1831).


both for enforcement of the laws, and the rise of candidates and social 

movements. 

What makes the question so interesting is that the Internet is both obviously 

exceptional and unexceptional at the same time. It depends on what you might 

think it is an exception to. It is clear that the Internet was a dramatic revolution 

and an exception to the ordinary ways of designing communications systems. 

But whether it enjoys a special immunity to the longer and deeper forces that 

shape human history is, shall we say, yet to be seen. 

* * * 

In the early 2000s, Jack Goldsmith and I wrote Who Controls the Internet? 2 The 

book is an explicitly anti-exceptionalist work. It addressed one particular way 

that the Internet might be an exception, namely, the susceptibility, as it were, of 

the Internet to regulation by the laws of nations. From the mid-1990s onward it 

was widely thought that the Internet would prove impossible to control or 

regulate. Some legal scholars, in interesting and provocative work, argued that 

in some ways the Network might be considered to have its own sovereignty, 

like a nation-state. 3 That was the boldest claim, but the general idea that the 

Internet was difficult or impossible to regulate was, at the time, a political, 

journalistic and academic commonplace, taken for granted. For example, 

reflecting his times, in 1998 President Clinton gave a speech about China’s 

efforts to control the Internet. “Now, there’s no question China has been 

trying to crack down on the Internet—good luck” he said. “That’s sort of like 

trying to nail Jello to the wall.” 4 

That was the conventional wisdom. In our book we suggested that despite the 

wonders of the Network it did not present an existential challenge to national 

legal systems, reliant, as they are, on threats of physical force. 5 We predicted 

that nations would, and to some degree already had, reassert their power over 

the Network, at least, for matters they cared about. They would assert their 

power not over the Network in an abstract sense, but the actual, physical 

humans and machinery who lie underneath it. Many of the book’s chapters 

ended with people in jail; unsurprisingly, China provided the strongest example 

of what a State will do to try to control information within its borders. 

2 TIM WU & JACK GOLDSMITH, WHO CONTROLS THE INTERNET (2006). 

3 David Post & David Johnson, Law and Borders–The Rise of Law in Cyberspace, 48 STAN. L. REV. 

1367 (1996). 

4 R. MICHAEL ALVAREZ & THAD E. HALL, POINT, CLICK AND VOTE: THE FUTURE OF 

INTERNET VOTING 3 (2004). 

5 JOHN AUSTIN, THE PROVINCE OF JURISPRUDENCE DETERMINED (Wilfrid E. Rumble, ed., 

Cambridge Univ. Press 1995) (1832).


Drama aside, in a deeper way, we were interested in what you might call the 

persistence of physicality. Despite its virtual qualities, behind the concept of a 

global network were living human beings, blood and flesh. The human body’s 

susceptibility to pain and imprisonment is a large part of what the nation-state 

bases its rule on, and that had not changed. We predicted that the nation’s 

threat of physical force, otherwise known as laws, would therefore shape the 

Network as much as its founding ambitions. 

Here is how we put the point in the introduction to our book, written in about 

2005 or so: 

Our age is obsessed with the search for the newest “new 

thing.” Our story, by contrast, is about old things—ancient 

principles of law and politics within nations, cooperation and 

clashes between nations, and the enduring relevance of 

territory, and physical coercion. It is a story where Thomas 

Hobbes is as important as Bill Gates. Like it or not, these old 

things are as important to the Net’s development, if not more 

so, than any technological or intellectual breakthrough. 

In these pages we present a strong resistance to Internet 

exceptionalism, or any arguments that new technologies can 

only be understood using novel intellectual frameworks. Like 

other revolutionary communication technologies, the Internet 

has changed the way we live, and fostering undreamt of new 

forms of social organization and interaction. But also like 

other revolutionary communication technologies, the Internet 

has not changed the fundamental roles played by territorial 

government. 

We are optimists who love the internet and believe that it can 

and has made the world a better place. But we are realistic 

about the role of government and power in that future, and 

realists about the prospects for the future. 

I regret to say that it has been the Chinese government that has done the most 

to prove our basic thesis correct. The Jello was, somehow, nailed to the wall. 

Despite nearly a decade of Westerners (most particularly Western newspaper 

columnists) assuming or hoping that the Net would bring down the Chinese 

state, it didn’t happen; indeed it never even came close. And so, five years later 

the basic ideas in our book seem hard to contest. Consequently, this one 

particular species of Internet exceptionalism—the idea that the network has its 

own sovereignty in a sense, or is an exception to law—has weakened and may 

be dead.


In the summer of 2010, in fact, as if to hammer to point home, the Chinese 

government released a new White Paper on “Internet Policy.” It made its 

centerpiece the phrase coined by the Internet exceptionalists of the 1990s: 

“Internet Sovereignty.” However, that phrase did not mean what it did in the 

1990s. Rather as the People’s Daily, the state newspaper, explained, “Internet 

Sovereignty” means that “all foreign IT companies operating in China must 

abide by China’s laws and [be] subject to Beijing’s oversight.” 6 

* * * 

Leaving law aside, however, the larger questions of Internet Exceptionalism 

remain unanswered. It is surely one thing for the Internet to be a living 

exception to the legal system, a sovereign unto itself in some way. But is the 

Network an exception as an information network, as a means for a nation or world 

to communicate? Here, surely, the exceptionalist is on far stronger ground. 

Whatever you might say about efforts to use the Internet to avoid law, we 

cannot doubt that the “Networks of Networks” has changed the way we 

communicate in dramatic fashion. Technologically, and in its effects on 

business, culture and politics, the Internet seems, by almost any account, an 

exception, different from the way other systems of mass communications have 

operated, whether the telephone, radio, or the television. 

This point seems so obvious as to be commonplace to anyone who’s lived 

through the 1990s. Unlike television, radio and newspapers, which all are 

speech outlets for a privileged few, the Internet allows anyone to be a publisher. 

Unlike the private cable networks, the Internet is public and, in its totality, 

owned by no one. Unlike the telephone system, it carries video, graphics, the 

Web, and supports any idea anyone can come up with. It has played host to 

generations of new inventions, from email and the World Wide Web to the 

search engine, from shops like eBay and Amazon to social networking and 

blogging. It has challenged and changed industries, from entertainment to 

banking and travel industries. These features and others are what have made 

the Network so interesting for so many years. 

The question is whether, however, the Internet is different in a lasting way. 

What do I mean, “a lasting way?” I rely on the sense that certain ideas, once 

spread, seem to lodge permanently, or for centuries at least—e.g., the idea of 

property, civil rights, or vaccination. Each is an idea that, once received, has a 

way of embedding itself so deeply as to be nearly impossible to dislodge. In 

contrast are ideas that, while doubtlessly important, tend, in retrospect, to form 

a rather interesting blip in history, a revolution that came and went. Will we 

6 Information Office of the State Council of the People’s Republic of China, The Internet in 

China, 2010, http://www.china.org.cn/government/whitepaper/node_7093508.htm; 

White paper explains ‘Internet Sovereignty’, PEOPLE’S DAILY ONLINE, June 9, 2010, 

http://english.peopledaily.com.cn/90001/90776/90785/7018630.html.


think of the open age of the Internet the way we think of communism, or the 

hula-hoop? 7 

If the Internet is exceptional in a lasting way, it must be for its ideology as 

expressed in its technology. And in this sense its exceptionalism is similar to 

American exceptionalism. Both the Nation and the Network were founded on 

unusual and distinct ideologies, following a revolution (one actual, another 

technological). In a typical account, writer Seymour Martin Lipset writes in 

American Exceptionalism: A Double-Edged Sword: “The United States is exceptional 

in starting from a revolutionary event … it has defined its raison d’être 

ideologically.” 8 Or, as one-time Columbia professor Richard Hofstadter wrote 

in the 20th century, “it has been our fate as a nation to not to have ideologies, 

but to be one.” 9 De Tocqueville put American exceptionalism down to 

particular features of the United States—the religiosity of its founding, its 

proximity to yet freedom from Europe, and, as he wrote, “a thousand special 

causes.” 10 

Looking at the Internet, its founding and its development, we can find the same 

pattern of a revolution, an ideology, and many “special causes.” While much of 

it was purely technical, there were deeply revolutionary ideas, even by 

technological standards, at the heart of the Internet, even if sometimes they 

were arrived at in accidental fashion or for pragmatic reasons. 

Of course, fully describing all that makes the Internet different would take 

another Democracy in America, and we have the benefit of many writers who’ve 

tried to do just that, whether in Katie Hafner and Matthew Lyon’s Where 

Wizards Stay up Late, the oral accounts of its creators, classic works like J.H. 

Saltzer et al., End-to-End Arguments in System Design, or Jonathan Zittrain’s The 

Future of the Internet. 11 

7 I’ve spent some time thinking about these questions, and I want to suggest that it isn’t really 

possible to answer the question in full without understanding the story of the networks that 

preceded the Internet. My fullest answer to the question I’ve posed, then, is in THE 

MASTER SWITCH (Knopf 2010), an effort to try and find the patterns, over time, that 

surround revolutionary technologies. This time, unlike in WHO CONTROLS THE INTERNET, 

when it comes to the broader question of the Internet as a way of moving information, I 

tend to side with the exceptionalists, though it is a close call. 

8 SEYMOUR MARTIN LIPSET, AMERICAN EXCEPTIONALISM 18 (1996). 

9 JAMES M. JASPER, RESTLESS NATION 38 (2000). 

10 ALEXIS DE TOCQUEVILLE, DEMOCRACY IN AMERICA 519 (Henry Reeve trans., D. Appleton 

and Company 1904) (1831). 

11 KATIE HAFNER & MATTHEW LYON, WHERE WIZARDS STAY UP LATE: THE ORIGINS OF THE 

INTERNET (1996);J. H. Saltzer, D. P. Reed & D. D. Clark, End-To-End Arguments in System 

Design, 2 ACM TRANSACTIONS ON COMPUTER SYSTEMS (TOCS) 277-288 (1984); JONATHAN 

ZITTRAIN, THE FUTURE OF THE INTERNET—AND HOW TO STOP IT (2009).


To understand what makes the Internet different, the origins of the Internet 

bear careful examination. First, the Network’s predecessors (the telephone, 

cable, etc.) were all commercial enterprises first and foremost, invented and 

deployed (in the U.S.) by private firms. The Internet, in contrast, was founded 

as a research network, explicitly non-commercial and public for the first decade 

of its existence. Private companies were involved, yes, but it was not a 

commercial operation in the same sense that, say, the cable networks always 

were. 

Perhaps, thanks to its origins, the Internet was founded with an ideology that 

was far more explicit than most—a kind of pragmatic libertarianism whose 

influence remains. The early Internet scientists had various principles that they 

were proud of. One example is David Clark’s memorable adage. “We reject: 

kings, presidents, and voting. We believe in: rough consensus and running 

code.” Another is found in a famous Request For Comments written by 

Internet founder Jon Postel, setting forth the following as a principle for 

network operators: “Be conservative in what you do. Be liberal in what you 

accept from others.” 12 

The Network constituted not just a technological advance, though it was that as 

well, but also a rejection of dominant theories of system design and, in a deeper 

sense, a revolution in information governance. The early Internet researchers 

were designing a radically decentralized network in an age—the mid-1960s— 

when highly centralized systems ran nearly every aspect of American and world 

life. In communications this was represented by AT&T, the great monopolist, 

with its mighty and near-perfect telephone network. But it could also be found 

in other aspects of society, from the enlarged Defense Department that ran the 

Cold War, the new, giant government agencies that ran social programs, and 

enormous corporations like General Motors, IBM, and General Electric. 

So when Vint Cerf and his colleagues put the Internet on the TCP/IP protocol 

in 1982 (its effective “launch”), most information networks—and I don’t mean 

this is a pejorative sense—could be described as top-down dictatorships. One 

entity—usually a firm or a part of the State (or both), like AT&T or the BBC, 

decided what the network would be. The Internet, in contrast, has long been 

governed more like a federation of networks, and in some respects, like a 

Republic of Users. That is implicit in the ability of anyone to own an IP 

address, set up a website, and publish information—something never true, and 

still not true, on any other network. 

12 Paulina Borsook, How Anarchy Works, WIRED (Oct. 1995), 

http://www.wired.com/wired/archive/3.10/ietf.html; Jon Postel, Information Sciences 

Institute of the University of Southern California, DOD Standard Transmission Control 

Protocol 13 (1980), available at http://tools.ietf.org/html/rfc761#section-2.10.


Throughout its history, the universal Network has, true to the governance 

structure, seen a pattern of innovation that is unlike any other. This too is the 

subject of much scholarship and popular recognition—the mode of 

“decentralized innovation” that had led every several years or so to the next 

wonder, starting with email, through the Web, search engines, online retail, Web 

video, social networking, and onward. These innovations arrived in a highly 

disorganized fashion often led by amateurs and outsiders. The spread of 

computer-networking itself began with amateur geeks glorified in 1980s films 

like War Games. 13 It is hard to think of a truly important Internet invention that 

came from a firm that predated the Internet. Society-changers like Craigslist, 

eBay, Wikipedia and blogs are obviously the products of geeks. 

* * * 

Can it last? Can the Internet remain, in this sense, exceptional? Whatever the 

Internet’s original ideas, it is easy to argue that all this, too, shall pass. The 

argument from transience suggests that all that seems revolutionary about the 

Internet is actually just a phase common to speech inventions. In other words, 

the Internet is following a path already blazed by other revolutionary inventions 

in their time, from the telephone to radio. Such disruptive innovations usually 

do arrive as an outsider of some kind, and will pass through what you might call 

a “utopian” or “open” phase—which is where we are now. But that’s just a 

phase. As time passes, even yesterday’s radical new invention becomes the 

foundation and sole possession of one or more great firms, monopolists, or 

sometimes, the state, particularly in totalitarian regimes like the Soviet Union or 

the Third Reich. The openness ends, replaced with better production value and 

tighter controls. It is, in other words, back to normal, or at least what passed 

for normal for most of human history. 

We might learn from the fate of the broadcast radio, the darling new technology 

of the 1920s. 14 In the 1920s, opening a radio station was relatively easy, not 

quite as easy as a website, but within the reach of amateurs. American radio was 

once radically decentralized, open and rather utopian in its aspirations. But by 

the 1930s, broadcast in the United States was increasing controlled by the 

chains—most of all, the National Broadcast Company, NBC, who brought 

better programming, but also much less of the amateur, open spirit. But that’s 

nothing compared to countries like Germany and the Soviet Union, where radio 

became the domain of the state, used to control and cajole. In Germany, every 

citizen was issued a “people’s receiver” tuned only to Nazi channels, and within 

13 War Games (Metro-Goldwyn-Mayer 1983) 

14 This story of radio can be found in TIM WU, THE MASTER SWITCH, chaps 3, 5 (2010).


the space of a decade, the free radio had became what Joseph Goebbels called 

the “spiritual weapon of the totalitarian state.” 15 

Yet I find it hard to imagine such a dramatic or immediate fate for the Internet. 

It seems in so many ways too established, its values too enmeshed in society, to 

disappear in an instant. 

Perhaps it would be more accurate to suggest that there are aspects of the 

Internet ideology that are more and less likely to fade, to become yesterday’s 

ideas. At one extreme, the Internet’s core technological ideas, protocol layering 

& packet-switching, seem unlikely to go anywhere. The reason is that these 

techniques have become the basis of almost all information technology, not just 

the Internet itself. The telephone networks are today layered and packetswitched, 

even if they don’t rely on the Internet Protocol. 

More vulnerable, however, are the Internet’s early ideas of openness and 

decentralized operation—putting the intelligence in the edges, as opposed to the 

center of the network. Originally described by engineers as the E2E principle, 

and popularly contained in the catch-phrase “Net Neutrality,” these principles 

have survived the arrival of broadband networks. Yet by its nature, Net 

Neutrality seems easier to upset, for discrimination in information systems has 

long been the rule, not the exception. There are, importantly, certain 

commercial advantages to discriminatory networking that are impossible to 

deny, temptations that even the Internet’s most open firms find difficult to 

resist. So while I may personally think open networking is important for 

reasons related to innovation and free speech, it seems obvious to me that open 

networking principles can be dislodged from their current perch. 

Another open question is whether some of the means of production and 

cultural creativity that are associated with the Internet are destined for lasting 

importance. We have recently lived through an era when it was not unusual for 

an amateur video or blog to gain a greater viewership than films made for tens 

of millions. But is that, Lessig’s “remix culture,” 16 a novelty of our times? We 

also live in era where free software is often better than that which you pay for. 

They are the products of open production systems, the subject of Yochai 

Benkler’s The Wealth of Networks, the engines behind Linux and Wikipedia and 

other mass projects—as discussed in Benkler’s essay in this collection. 17 Of 

course such systems have always existed, but will they retreat to secondary 

15 Quoted in Garth S. Jowett, GARTH JOWETT & VICTORIA O’DONNELL, READINGS IN 

PROPAGANDA AND PERSUASION 132 (2005). 

16 LAWRENCE LESSIG, REMIX: MAKING ART AND COMMERCE THRIVE IN THE HYBRID 

ECONOMY (2008). 

17 YOCHAI BENKLER, THE WEALTH OF NETWORKS (2007).


roles? Or will they perhaps become of primary importance for many areas of 

national life? 

The only honest answer is that it is too early to tell. And yet, at the same time, 

the transience of all systems suggests that at least some of what we take for 

granted right now as intrinsic to our information life and to the nature of the 

Internet will fade. 

The reasons are many. It might simply be that the underlying ideas just 

discussed turn out to have their limits. Or that they are subject to an almost 

natural cycle—excessive decentralization begins to make centralization more 

attractive, and vice versa. More sinisterly, it might be because forces 

disadvantaged by these ideas seem to undermine their power—whether 

concentrated forces, like a powerful state, or more subtle forces, like the human 

desire for security, simplicity and ease that has long powered firms from the 

National Broadcasting Corporation to Apple, Inc. 

Whatever the reasons, and while I do think the Internet is exceptional (like the 

United States itself), I also think it will, come to resemble more “normal” 

information networks—indeed, it has already begun to do so in many ways. 

Exceptionalism, in short, cannot be assumed, but must be defended. 

* * * 

I began this essay with a comparison between Internet and American 

exceptionalism. Yet I want to close by suggesting we can learn from the 

comparison in a slightly different sense. I’ve suggested that there is a natural 

tendency for any exceptional system to fade and transition back to observed 

patterns. But even if that’s true, what is natural is not always normatively good, 

not always what we want. For example, it may very well be “natural” for a 

democracy, after a few decades or less, to ripen into a dictatorship of some 

kind, given the frustrations and inefficiencies of democratic governance. 

Cromwell and Napoleon are the bearers of that particular tradition, and it has 

certainly been the pattern over much of history. 

But the idea of American Exceptionalism has included a commitment to trying 

to avoid that fate, even if it may be natural. Despite a few close calls, the 

United States remains an exception to the old rule that Republics inevitably 

collapse back into dictatorship under the sway of a great leader. The Internet, 

so far, is an exception to the rule that open networks inevitably close and 

become dominated by the State or a small number of mighty monopolists. 

Twenty-five years after .COM, we might say we still have a republic of 

information—if we can keep it.


Bibliography 

In lieu of extensive footnotes, I thought I’d provide here the books and articles 

that have, implicitly or explicitly, taken on the question of Internet 

Exceptionalism. Notice that, for those familiar in the field, this may lead to 

some unusual groupings—but the fundamental question is whether the project 

in question tries to argue the Internet is magically different or a repeat of ageold 

problems. 

Exceptionalist Works 

� PETER HUBER, LAW AND DISORDER IN CYBERSPACE (1997). 

� J. H. Saltzer, D. P. Reed & D. D. Clark, End-To-End Arguments in 

System Design, 2 ACM TRANSACTIONS ON COMPUTER SYSTEMS (TOCS) 

277-288 (1984). 

� NICHOLAS NEGROPONTE, BEING DIGITAL (1996). 

� LAWRENCE LESSIG, THE FUTURE OF IDEAS (2002). 

� David R. Johnson & David Post, Law & Borders: The Rise of Law in 

Cyberspace, 48 STAN. L. REV. 1367 (1996). 

� Mark A. Lemley & Lawrence Lessig, The End of End-to-End: Preserving the 

Architecture of the Internet in the Broadband Era, 48 UCLA L. REV. 925 

(2001). 

� Susan P. Crawford, The Internet and the Project of Communications Law, 55 

UCLA L. REV. 359 (2007). 

� DAVID POST, IN SEARCH OF JEFFERSON’S MOOSE (2009). 

� TIM WU, THE MASTER SWITCH (2010). 

Anti-Exceptionalist Works 

� LAWRENCE LESSIG, CODE: AND OTHER LAWS OF CYBERSPACE 

(2006). 

� TIM WU & JACK GOLDSMITH, WHO CONTROLS THE INTERNET? 

(2008). 

� Tim Wu, Cyberspace Sovereignty?, 10 HARV. J.L. & TECH. 647 (1997). 

� Christopher S. Yoo, Would Mandating Broadband Network Neutrality Help 

or Hurt Competition? A Comment on the End-to-End Debate, 3 J. 

TELECOMM. & HIGH TECH. L. 23 (2004). 

� YOCHAI BENKLER, THE WEALTH OF NETWORKS (2007). 

� CORY DOCTOROW, LITTLE BROTHER (2008). 

On the Topic / Mixed 

� JONATHAN ZITTRAIN, THE FUTURE OF THE INTERNET—AND HOW 

TO STOP IT (2009).


Section 230 of the CDA: 

Internet Exceptionalism as a 

Statutory Construct 

By H. Brian Holland * 

Introduction 

Since its enactment in 1996, Section 230 of the Communications Decency Act 

has become perhaps the most significant statute in the regulation of online 

content, and one of the most intensely scrutinized. Many early commentators 

criticized both Congress, for its apparent inability to craft the more limited 

statute it intended, and the courts, for interpreting the statute broadly and failing 

to limit its reach. Later commentators focus more clearly on policy concerns, 

contending that the failure to impose liability on intermediaries fails to 

effectuate principles of efficiency and cost avoidance. More recently, 

commentators have argued that Section 230 immunity should be limited 

because it contributes to the proliferation of anonymous hate speech, 

intimidation, and threats of violence against traditionally marginalized groups. 

Acknowledging the validity of these concerns, this essay nevertheless takes the 

opposing view, defending broad Section 230 immunity as essential to the 

evolving structure of Internet governance. Specifically, Section 230 provides a 

means of working within the sovereign legal system to effectuate many of the 

goals, ideals, and realities of the Internet exceptionalist and cyber-libertarian 

movements. By mitigating the imposition of certain external legal norms in the 

online environment, Section 230 helps to create the initial conditions necessary 

for the development of a modified form of exceptionalism. With the impact of 

external norms diminished, Web 2.0 communities, such as wikis 1 and social 

network services, 2 have emerged to facilitate a limited market in norms and 

values and to provide internal enforcement mechanisms that allow new 

communal norms to emerge. Section 230 plays a vital role in this process of 

* Associate Professor, Texas Wesleyan School of Law. A modified version of this essay 

originally appeared in the University of Kansas Law Review. In Defense of Online Intermediary 

Immunity: Facilitating Communities of Modified Exceptionalism, 56 U. Kan. L. Rev. 369 (2008), 


1 A wiki is a website designed to allow visitors to easily create and edit any page on the site. 

For more information, see Wikipedia, Wiki, http://en.wikipedia.org/wiki/Wiki (last 


2 Social network services are online services designed for users to share messages, links, and 

media (photos and video) with friends or others with similar interests. Some popular social 

network services are Facebook, MySpace, and Twitter.


building heterogeneous communities that encourage collaborative production 

and communication. Efforts to substantially reform or restrict Section 230 

immunity are therefore largely unnecessary and unwise. 

The essay begins with a brief introduction to Section 230. As interpreted and 

applied by the judiciary, this statute is now conceived as a broad grant of 

immunity from tort liability—broad not only in terms of those who can claim 

its protection but also in terms of predicate acts and causes of action to which 

such immunity extends. 

Working from this foundation, I then seek to position the courts’ expansion of 

Section 230 immunity within the larger debate over Internet governance, 

suggesting that proponents of expanded immunity are successfully creating what 

might be characterized as a modified, less demanding form of cyber-libertarian 

exceptionalism than what Eric Goldman calls, in his essay in this book, the 

“First Wave of Internet Exceptionalism” (one of “Internet Utopianism”), as 

articulated in the mid-1990s. The dramatic expansion of Section 230 immunity 

has in a limited sense effectuated a vision of a community in which norms of 

relationship, thought and expression are yet to be formed. The tort liability 

from which Section 230 provides immunity is, together with contract, a primary 

means by which society defines civil wrongs actionable at law. In the near 

absence of these external norms of conduct regulating relationships among 

individuals, the online community is free to create its own norms, its own rules 

of conduct, or none at all. It is a glimpse of an emergent community existing 

within, rather than without, the sovereign legal system. 

Finally, I make the case for preserving broad Section 230 immunity. As an 

initial matter, many of the reforms offered by commentators are both 

unnecessary and unwise because the costs of imposing indirect liability on 

intermediaries are unreasonable in relationship to the harm deterred or 

remedied by doing so. Moreover, the imposition of liability would undermine 

the development of Web 2.0 communities as a form of modified exceptionalism 

that encourages the development of communal norms, efficient centers of 

collaborative production, and open forums for communication. 

The Expansion of Section 230 Immunity 

In May of 1995, a New York trial court rocked the emerging online industry 

with its decision in Stratton Oakmont, Inc. v. Prodigy Services Co., 3 holding the 

Prodigy computer network liable for defamatory comments posted on one of its 

bulletin boards by a third-party. The key factor in this result was Prodigy’s 

attempt to create a more family-friendly environment through the exercise of 

editorial control over the bulletin boards and moderating for offensive content. 

3 No. 31063/94, 1995 WL 323710 (N.Y. Sup. Ct. May 24, 1995).


Prodigy was therefore treated as a publisher of the information, rather than a 

mere distributor, and held strictly liable for actionable third-party content. 

Representatives of the online industry argued that the Prodigy decision placed 

service providers in an untenable position by creating a “Hobson’s choice” 4 

between monitoring content and doing nothing, thereby insulating the service 

from liability. Congress responded to the decision by amending the draft 

Communications Decency Act (CDA) to include a tailored immunity provision 

addressing the online industry’s concerns. As one element of what came to be 

known as the Good Samaritan provisions of the CDA, Section 230 was 

generally intended to provide online service providers and bulletin board hosts 

with immunity from tort liability for the defamatory acts of their users. This 

was accomplished by addressing those specific elements of common law 

defamation at issue in the Prodigy decision—editorial control and the distinct 

treatment of publishers and distributors under the law. To that end, Section 

230 provided that no interactive computer service should be treated as the 

publisher or speaker of third-party content, and that efforts to moderate 

content should not create such liability. 

In the years following the enactment of Section 230, courts consistently 

extended its application. This trend began in 1997 with the watershed decision 

in Zeran v. America Online, Inc., 5 in which the Fourth Circuit applied Section 230 

to claims that America Online (AOL) should be held liable for the defamatory 

content posted by one of its users. The plaintiffs claimed liability arose in part 

because AOL had allegedly failed to remove third-party defamatory messages 

from its bulletin board system within a reasonable time, refused to post 

retractions to defamatory messages, and failed to screen for similar defamatory 

messages thereafter. The court found the plaintiff’s tort claims were preempted 

by Section 230, which rendered AOL immune. In reaching this result, the court 

rejected a strict reading of Section 230 as being limited to its terms. Although 

the statute failed to make any explicit reference to distributor liability, which the 

Prodigy decision appeared to leave intact, the court read distributor immunity 

into the statute, finding distributor liability to be an included subset of the 

publisher liability foreclosed by the statute. By collapsing the publisherdistributor 

distinction, the Fourth Circuit adopted the most expansive reading 

possible of both defamation law and Section 230. Thus, even though AOL 

knew the statements were false, defamatory, and causing great injury, AOL 

could simply refuse to take proper remedial and preventative action without fear 

of liability. 

4 SAMUEL FISHER, THE RUSTICK’S ALARM TO THE RABBIES (1660), as cited in Hobson’s choice, 

Wikipedia, http://en.wikipedia.org/wiki/Hobson%27s_choice (last accessed Dec. 1, 

2010). 

5 129 F.3d 327 (4th Cir. 1997).


Following Zeran, and building on that court’s reading of both the statute and the 

policies sought to be effected, courts have extended the reach of Section 230 

immunity along three lines: (1) by expanding the class who may claim its 

protections; (2) by limiting the class statutorily excluded from its protections; 

and (3) by expanding the causes of action from which immunity is provided. 6 

As to the first, courts have interpreted the provision of immunity to interactive 

computer services to include such entities as Web hosting services, email service 

providers, commercial websites like eBay and Amazon, individual and company 

websites, Internet dating services, privately-created chat rooms, and Internet 

access points in copy centers and libraries. The additional provision of 

immunity to users of those services promises similar results. Already, one 

decision has held that a newsgroup user cannot be held liable for re-posting 

libelous comments by a third party, 7 while another court found a website 

message board to be both a provider and a user of an interactive computer 

service. 8 

The second line of extension results from a narrow reading of the term 

“information content provider,” which defines the class for whom there is no 

immunity. Specifically, courts have held that minor alterations to third-party 

content does not constitute the provision of content itself, so long as the 

provider does not induce the unlawful content through the provision of 

offending raw materials of authorship and where the basic form and message of 

the original is retained. 9 The third point of expansion has been to extend 

Section 230 immunity beyond causes of action for defamation and related 

claims to provide immunity from such claims as negligent assistance in the 

sale/distribution of child pornography, 10 negligent distribution of pornography 

of and to adults, 11 negligent posting of incorrect stock information, 12 sale of 

fraudulently autographed sports memorabilia, 13 invasion of privacy, 14 and 

misappropriation of the right of publicity. 15 

6 But see Fair Housing. Council of San Fernando Valley v. Roommates.com, LLC, 521 F.3d 

1157 (9th Cir. 2008) (declining to extend Section 230 immunity to Roommates.com for 

certain categories of content solicited by the site for users in violation of federal fair housing 

laws). 

7 Barrett v. Rosenthal, 146 P.3d 510, 527 (Cal. 2006). 

8 DiMeo v. Max, 433 F. Supp. 2d 523, 531 (E.D. Pa. 2006). 

9 Batzel v. Smith, 333 F.3d 1018, 1031 (9th Cir. 2003). See also Donato v. Moldow, 865 A.2d 

711, 724 (N.J. Super. Ct. App. Div. 2005) (quoting Batzel v. Smith). 

10 Doe v. Am. Online, Inc., 783 So. 2d 1010, 1017 (Fla. 2001). 

11 Does v. Franco Prods., No. 99 C 7885, 2000 WL 816779, at *5 (N.D. Ill. June 22, 2000), aff’d 

sub nom. Doe v. GTE Corp., 347 F.3d 655 (7th Cir. 2003). 

12 Ben Ezra, Weinstein & Co. v. Am. Online, Inc., 206 F.3d 980, 986 (10th Cir. 2000). 

13 Gentry v. eBay, Inc., 121 Cal. Rptr. 2d 703, 715 (Cal. Ct. App. 2002).


Section 230, Internet 

Governance & Exceptionalism 

Situated within the larger debate over Internet governance, the concept of 

Internet exceptionalism presumes that cyberspace cannot be confined by 

physical borders or controlled by traditional sovereign governments, and thus 

that cyber-libertarian communities will emerge in which norms of relationship, 

thought and expression are yet to be formed. Although these ideas have been 

subjected to intense criticism and somewhat obscured by recent developments 

in the governance debates, they remain a touchstone for the cyber-libertarian 

ideal. This part of the essay seeks to clear space in the governance debates for 

this vision of exceptionalism, and argues that Section 230 is in some limited way 

facilitating the emergence of cyber-libertarian communities in a modified, less 

demanding form. 

Foundational Arguments of 

Internet Governance 

The debate over Internet governance evolved in two surprisingly distinct, albeit 

convergent stages. The first stage of the governance debate focused on law and 

social norms, and whether these traditional models of regulating human 

relations could be validly applied to the online environment. In this context, 

exceptionalism was conceptualized as a state of being to which the Internet had 

naturally evolved, apart from terrestrial space. The second stage of the debate 

introduced network architecture as an important and potentially dominant 

means of regulating the online environment. In this context, exceptionalism 

became an objective to be pursued and protected as a matter of choice, rather 

than a natural state. At a more exacting level, these debates implicated 

fundamental questions of legitimacy, preference, politics, democracy, collective 

decision-making, and libertarian ideals. 

In the early 1990s, as the Internet began to reach the masses with the advent of 

the World Wide Web, a particular vision of the online environment emerged to 

advocate and defend Internet exceptionalism. Described as digital 

libertarianism or cyber-libertarianism, the vision was one of freedom, liberty, 

and self-regulation. Cyber-libertarians believed the Internet could and would 

develop its own effective legal institutions through which rules would emerge. 

These norms would emerge from collective discourse around behavior, 

14 Carafano v. Metrosplash.com, Inc., 339 F.3d 1119, 1124 (9th Cir. 2003). 

15 See id. at 1122, 1125 (extending § 230 immunity to defendant in claim “alleging invasion of 

privacy, misapproriation of the right of publicity, defamation and negligence”). See also 

Perfect 10, Inc. v. CCBill LLC, 488 F.3d 1102, 1118–19 (9th Cir. 2007) (finding that § 230 

immunity extends to state-law intellectual property claims, including unfair competition, false 

advertising, and right of publicity).


relationship, and content, rather than from the control and regulation of 

network architecture. Control of architecture was seen almost exclusively as an 

instrument by which to enforce emerging social norms, and not as a means of 

determining the norms themselves. By the mid-1990s this process of selfregulation 

was well underway. 

At the same time, however, sovereign nations and their constituents increasingly 

sought to impose existing offline legal regimes on this emerging, resource-rich 

environment. Many in the online community resisted, perceiving this regulation 

as a threat to the exceptional nature of the Internet. Advocates of selfregulation 

envisioned cyberspace as a distinct sphere, apart from physical space. 

These cyber-libertarian exceptionalists saw the imposition of existing offline 

legal systems grounded in territorially-based sovereignty as inappropriate. They 

believed that the online environment should instead be permitted to develop its 

own discrete system of legal rules and regulatory processes. Self-regulation was 

preferable in its own right because it had proven so effective in creating the 

environment sought to be preserved, and also because the alternative seemed 

devastating. The imposition of external, territorially-based legal regimes would 

be, the exceptionalists argued, infeasible, ineffective, and fundamentally 

damaging to the online environment. 

Faced with the attempted imposition of offline legal regimes, cyber-libertarians 

responded by attacking the validity of exercising sovereign authority and 

external control over cyberspace. According to Professors David Johnson and 

David Post, two leading proponents of self-governance, external regulation of 

the online environment would be invalid because Internet exceptionalism—the 

state of being to which the Internet naturally evolved—destroys the link 

between territorially-based sovereigns and their validating principles of power, 

legitimacy, effect, and notice. 16 Most importantly, the Internet’s decentralized 

architecture deprives territorially-based sovereigns of the power, or ability, to 

regulate online activity. Likewise, extraterritorial application of sovereign law 

fails to represent the consent of the governed, or to effectuate exclusivity of 

authority based on a relative comparison of local effects. The loss of these 

limiting principles results in overlapping and inconsistent regulation of the same 

activity with significant spillover effect. Deprived of these validating principles, 

it would be illegitimate to apply sovereign authority and external control in 

cyberspace. 

16 David R. Johnson & David Post, Law and Borders—The Rise of Law in Cyberspace, 48 Stan. L. 

Rev. 1367 (1996).


A primary challenge to these cyber-libertarian arguments came from Professor 

Goldsmith, who engaged both their descriptive and normative aspects. 17 In 

terms of the legitimacy of sovereign regulation, Goldsmith criticized Johnson 

and Post’s limited view of sovereignty and over-reliance on the relationship 

between physical proximity and territorial effects. Moreover, he argued that 

they had overstated the impossibility of regulation, mistaking ability for cost; 

failed to recognize the deterrent effect on extraterritorial actors of local 

enforcement against end users and network components located within the 

territory; and mistakenly equated valid regulation with some measure of nearperfect 

enforcement. Finally, where true conflicts between sovereigns existed, 

Goldsmith argued that these could be resolved with the same tools used in the 

offline world—rules of jurisdiction, conflict of laws, enforcement, etc. 

Throughout, Goldsmith struck at Johnson and Post’s exceptionalist view of the 

Internet, implicitly rejecting the ultimate significance of both the technical and 

communal aspects of that ideal. This critique proved devastating to these early 

cyber-libertarian arguments. 

The governance debate entered its second phase in 1999 with the publication of 

Professor Lessig’s book, Code and Other Laws of Cyberspace. 18 Prior to Lessig’s 

book, the governance debate had focused primarily on behavioral and property 

norms, with the assumption that either existing sovereign law or the law 

emerging from Internet self-governance would prevail. Network architecture 

merely provided the means to enforce these norms, particularly those emerging 

from self-governance. Lessig reconceived Internet exceptionalism as a two-part 

phenomenon, one regulatory and the other cultural. The former recognizes that 

many of those features that make the Internet exceptional (in the cyberlibertarian 

sense) are merely coding choices, and not the innate nature of 

cyberspace. Within the network, architecture and code are the most basic forms 

of regulation. Code can be easily changed. Thus, Lessig argued, to protect the 

cultural aspects of exceptionalism, we must first recognize the exceptional 

regulatory power of architecture and code within cyberspace, and its pivotal role 

in preserving or destroying that culture. 

Lessig first pointed out that law and social norms are but two means of 

regulating human behavior. In cyberspace, unlike real space, it is possible for 

architecture to dominate regulatory structures. Architecture acts as a regulator 

in the offline world as well—in the form of time, nature, physics, etc.—but our 

laws and social norms are generally conceived with these regulators assumed. 

Alteration of that architecture is unusually difficult if not practically impossible. 

In cyberspace, by comparison, architecture in the form of code is remarkably 

17 Jack L. Goldsmith, Against Cyberanarchy, 65 U. Chi. L. Rev. 1199 (1998); Jack L. Goldsmith, 

The Internet and the Abiding Significance of Territorial Sovereignty, 5 Ind. J. Global Legal Stud. 475 

(1998). 

18 LAWRENCE LESSIG, CODE AND OTHER LAWS OF CYBERSPACE (1999).


fluid. Code effectuates a series of choices, from data collection, to anonymity, 

to access. And code can be changed. Not only is code fluid, but within 

cyberspace it is a uniquely powerful form of regulation. Rather than regulating 

behavior and relationships through punishment, deterrence and post-violation 

corrective action, code provides the means to exercise perfect control and thus 

perfect regulation—regulation not just of effects, but of the very universe of 

choices from which an individual actor is able to select. 

With this shift in focus, the debate itself evolved. Lessig cautioned that the 

greatest threat to the exceptional culture of cyberspace comes from the union of 

perfect control and market forces of commerce. The architectural components 

that provide the means of perfect control are held almost exclusively by private 

entities with commercial and political interests distinct from the collective. The 

invisible hand, Lessig argued, cannot resist the promise of perfect control, and 

has little or no motivation to protect the fundamental values promoted by 

cyber-libertarian exceptionalism. According to the cyber-libertarian narrative, 

barriers that are present in the real world do not exist or are de minimus in the 

online environment. In the context of Internet architecture, exceptionalism can 

be found in original principles of network design that rely on open protocols 

and non-discriminatory data transfer—a network that is decentralized, 

borderless, and with the potential for nearly unlimited data capacity. Indeed, 

the digital data flowing through this system is itself exceptional, because it is 

easy to create and manipulate, easy to copy with no degradation in quality, and 

easy to access and distribute. In the context of online relationships, 

exceptionalism resides (at the very least) in the interactivity, immediacy, and 

potential scope of interaction, as well as the opportunity for anonymity. 

However, the very promise of perfect control is to eliminate many of these 

choices and the fundamental values they reflect as subservient to commercial 

goals. In cyberspace, control over coded architecture supplies the means for 

making this election. Building on this assertion, Lessig argued that in order to 

protect fundamental values, decisions regarding architecture should emerge 

from the body politic and collective decision-making, rather than being 

concentrated in private actors. 

For many cyber-libertarians, Lessig’s message presented great problems. 

Although many had already abandoned the argument that the exercise of 

sovereign authority in cyberspace was normatively invalid, they had not given 

up (as a matter of preference) the vision of an emergent, self-governed, digital 

libertarian space. Sovereign legal regimes were still seen as the greatest threat to 

that vision. Territorial governments should, the cyber-libertarians argued, 

simply leave cyberspace alone to flourish. From this perspective, Lessig’s 

arguments about the unique regulatory power of architecture and code in 

cyberspace were largely convincing. But his description of the corrupting 

influence of perfect control and concentrated private power, and particularly his


call for government regulation to counteract those influences and preserve 

fundamental values, were difficult to square with most libertarian views. 

The debate on net neutrality provides a glimpse of this division. Many 

commentators, including Lessig, are concerned that the private owners that 

control the physical/infrastructure layer of the network will, in pursuit of crosslayer 

vertical integration and increased revenues, privilege certain content or 

applications. They therefore endorse regulatorily-mandated neutrality as a 

means of preserving one aspect of Internet exceptionalism. Not surprisingly, 

many libertarians reject this approach, endorsing instead market-based solutions 

for effectuating individual choice. 

The irony of this debate is fairly apparent. Many who might otherwise have 

characterized themselves as cyber-libertarian, or at least sympathetic to that 

vision, are now conflicted. Net neutrality would necessarily be imposed by 

external sovereign legal systems and subordinated to the control of commercial 

entities, rather than emerging as a common norm. In the extremes, the issue 

seems to present a choice between entrenched political power and unregulated 

market forces, with neither providing adequate protection for individuals. Thus, 

many of the Internet exceptionalists who sought to segregate the Internet from 

territorial boundaries, who assumed existing sovereign governments and legal 

regimes were the greatest threat to the online community, who believed that the 

computer scientist would remain in control of the network (and thus in control 

of enforcement), found themselves asking Congress to protect the Internet 

from private actors and market forces. 

What’s Left of Exceptionalism? 

What then is left of Internet exceptionalism? In his revolutionary essay A 

Declaration of the Independence of Cyberspace, John Perry Barlow described 

cyberspace as consisting not of computers, wires, or code, but of “transactions, 

relationships, and thought itself.” 19 It was this vision, this perception of an 

evolving social space, that guided Barlow’s ideal of the culture he sought to 

preserve—a distinct vision of potential worthy of protection. Indeed, to many 

early inhabitants of cyberspace, communal control and regulation of network 

architecture appeared a given, if for no other reason than that perfect external 

control seemed almost impossible. Freedom of choice in individual expression, 

human behavior, and relationships were the heart of the online cultural and 

social ideal that stirred Barlow and other cyber-libertarians. 

As it evolved, the governance debate fractured this largely unified vision, 

distinguishing validity from preference, law and social norms from architecture 

19 John Perry Barlow, A Declaration of the Independence of Cyberspace (Feb. 8, 1996), 

http://homes.eff.org/~barlow/Declaration-Final.html.


and code, technical exceptionalism from cultural exceptionalism, government 

power from private commercial power, and even libertarian from libertarian. 

Lessig argued persuasively that the greatest threat to digital libertarianism arose 

from private actors, unbounded by fundamental values (including constitutional 

values) and with the ability to exercise perfect control over choice. Lessig’s 

analysis, generally speaking, was focused on the treatment of data as data, based 

primarily on the identity of its owner and the commercial interests represented. 

Choice in action was to be controlled by the regulation of owned data, 

discriminatory treatment of data to the benefit of certain owners, restriction of 

network access, and similar means. These technical controls would then be 

bolstered by traditional sovereign law validating those measures. 

What seems somewhat obscured in Lessig’s architecture-and-code approach 

(which clearly remains the central concern of the governance debate) is Barlow’s 

original vision of relational libertarianism, with its focus on expression of 

individual choice and the development of new communal social norms within a 

system of self-governance. This is the part of Internet exceptionalism that was, 

in a sense, overwhelmed by the debate over architecture and code. Yet there 

are some choices, primarily relational, that remain largely unaffected by that 

debate. In this sphere, the question is not access to choice, the ability to 

choose, or the available universe of choices, but rather what norms apply to the 

choices being made outside those controls. 

Post argues that fundamental normative values could “best be protected by 

allowing the widest possible scope for uncoordinated and uncoerced individual 

choice among different values and among different embodiments of those 

values.” 20 He believes that the imposition of sovereign legal regimes in 

cyberspace, rather than promoting fundamental values as Lessig argued, would 

instead deny the digital libertarian culture the opportunity to develop apart from 

the offline world, with its own set of fundamental values. He argues it is better 

to serve the private interest (even if powerful and commercially motivated) than 

the interest of terrestrial sovereigns. Indeed, he sees exceptionalism as requiring 

self-governance, to the exclusion of external legal norms imposed by sovereign 

powers, as a precondition to the emergence of a new system of norms. 

Section 230 as a Form of 

Cyber-Libertarian Exceptionalism 

Most would say that Barlow and Post lost the battle. However, this particular 

strain of Internet exceptionalism, envisioned as self-governance and emerging 

social norms applicable to relationships between individuals (as opposed to data 

as data), has been preserved in a modified, less demanding form. Ironically, it is 

because of sovereign law, not in spite of it, that this occurred. The dramatic 

20 David Post, Against “Against Cyberanarchy,” 17 Berkeley Tech. L.J. 1365 (2002).


expansion of Section 230 immunity has effectuated many of the ideals 

promoted by Post, Barlow, and others, albeit on a limited scale. This expansion 

has created an environment in which many of the norms and regulatory 

mechanisms present in the offline world are effectively inapplicable. This is so 

not because the very nature of cyberspace makes such application impossible, or 

because sovereign law is necessarily ineffective or invalid, but rather because 

sovereign law has affirmatively created that condition. 

The torts for which Section 230 provides immunity are, together with contract 

law, the primary means by which society defines civil wrongs actionable at law. 

These norms of conduct regulate relationships among individuals: articulating 

wrongs against the physical and psychic well-being of the person (e.g., assault, 

battery, emotional distress), wrongs against property (e.g., trespass to land, 

trespass to chattels, conversion), wrongs against economic interests (e.g., fraud, 

tortious interference), and wrongs against reputation and privacy (e.g., 

defamation, misappropriation of publicity, invasion of privacy). Section 230 has 

been interpreted and applied to provide expansive immunity from tort liability 

for actions taken on or in conjunction with computer networks, including the 

Internet. Statutory language defining who may claim the protections of Section 

230 immunity, including providers of interactive computer services and the 

users of such services, has been broadly extended. In contrast, the primary 

limitation on the range of claimants to Section 230 immunity, which is 

statutorily unavailable to the allegedly tortious information content provider, 

has been construed fairly narrowly. Moreover, the immunity provided to this 

expansive cross-section of online participants now reaches well beyond 

defamation to include a wide range of other tortious conduct and claims. As 

such, many of the norms of conduct regulating relationships among individuals 

in the offline world—those civil wrongs actionable at (tort) law—simply do not 

apply to many in the online world. 

Even where the online entity is alleged to be aware of the illegal acts of their 

users, and to be either actively facilitating those illegal acts or refusing to stop 

them, the intermediary retains Section 230 immunity. This is true even where 

the intermediary has the knowledge, technical ability, and contractual right to 

take remedial action. In the offline world, such active and knowing facilitation 

would likely violate social norms established in tort law. In the online world, 

however, the defendants are immune from liability. Established norms, as 

expressed through the mechanisms of tort law, are neutralized by Section 230 

and its judicial interpretations. 

In the near absence of these external legal norms, at least within the range of 

choices being made outside the data-as-data architectural controls, the online 

community is free to create its own norms, its own rules of conduct, or none at 

all. The inhabitants may not have a blank slate—criminal law, intellectual 

property law, and contract law still apply—but much of what Barlow embraced


as central tenets (mind, identity, expression) remain undefined. Section 230 

offers a modified version of cyber-libertarian exceptionalism, less demanding of 

the sovereign and existing offline social norms, and therefore less satisfying. 

But it is nonetheless a glimpse of that society, maintained by the sovereign legal 

regime rather than against it. The law now applies to nearly every tort that can 

be committed in cyberspace. It is nibbling at the edges of intellectual property 

rights. It protects against the civil liability components of criminal acts. It 

generally extends to all but the first speaker, who may well get lost in the 

network to escape liability even without immunity. 

A Case for Preserving 

Section 230 Immunity 

As interpreted by the courts, the immunity provisions of Section 230 have been 

heavily criticized. Many commentators have argued that by failing to impose 

indirect liability on intermediaries, significant harms will go undeterred or 

unremedied, and that Section 230 should be reformed to serve the interests of 

efficiency and cost allocation. This part of the essay addresses these criticisms 

directly, concluding that substantially reforming the statute is both unnecessary 

and unwise because the cost of such liability is unreasonable in relation to the 

harm deterred or remedied. Indeed, given Section 230’s role in facilitating the 

development of Web 2.0 communities, reforming the statute to narrow the 

grant of immunity would significantly damage the online environment—both as 

it exists today and as it could become. 

Evaluating Calls for Reform 

Early critics of Section 230 tended to focus on the issues of congressional intent 

and broad interpretation by the courts. More recent commentators have moved 

beyond these issues to engage the larger implications of providing such 

sweeping immunity to online intermediaries, suggesting amendments to Section 

230 intended to effectuate policies of efficiency and cost allocation. This 

critique begins with the premise that in the online environment, individual bad 

actors are often beyond the reach of domestic legal authorities. This creates a 

situation in which significant individual harms cannot be legally deterred or 

remedied, and the fear that the Internet’s potential as a marketplace will not be 

realized. Given these negative conditions, where a third party maintains a 

certain level of control, the imposition of indirect liability is desirable. The 

failure to do so may create inefficiencies by failing to detect and deter harmful 

behavior where the cost of doing so is reasonable. Commentators have argued 

that, in the online environment, intermediaries are in the best position to deter 

negative behavior, to track down primary wrongdoers, and to mitigate damages. 

This is particularly true in regard to information-based torts, the damages of 

which might be mitigated in many circumstances simply by taking down, 

prohibiting, or blocking the objectionable content.


At the heart of this attack on Section 230 immunity is the idea that, in the 

absence of indirect intermediary liability, significant harms will go undeterred or 

unremedied. These fears are either misplaced or overstated. As an initial 

matter, it is not clear that a significant number of bad actors are beyond the 

reach of the law. Advances in technology are making it increasingly possible to 

locate and identify bad actors online, such that online anonymity is difficult to 

maintain. Likewise, where the bad actor is identified but is found outside the 

jurisdiction, sovereign governments have developed methods for resolving 

disputes to permit the direct extraterritorial application of domestic law, such as 

rules of jurisdiction, conflict of laws, and recognition of judgments. Indeed, 

anti-exceptionalists have strenuously argued that the application of sovereign 

authority to online activity originating outside the jurisdiction is legitimate and 

valid in large part because of these rules. 

Moreover, although the immunity provided by Section 230 arguably mitigates 

the legal incentives for online intermediaries to deter and remedy certain 

negative behavior, it does not eliminate those legal incentives. Section 230 

expressly states that it has no effect on criminal law, intellectual property law, or 

communications privacy law. These external norms remain applicable to and 

enforceable against both content providers and intermediaries in the online 

environment. Perhaps even more significantly, although Section 230 removes 

legal incentives to enforce the norms expressed in tort law, law is certainly not 

the only incentive for an intermediary to act. Communal, commercial and other 

incentives also play a role. Indeed, Section 230 immunity allows intermediaries 

the freedom to intervene in a multitude of ways. Thus, individual harms and 

marketplace security can be addressed through alternate legal regimes and 

internal incentives. 

Furthermore, proponents of indirect intermediary liability concede that even 

where harms do exist, intermediaries may only rightly be held liable for failing to 

detect and deter harmful behavior where the cost of doing so is reasonable. It 

is unclear, however, that the costs of intermedial regulation are reasonable. In 

terms of remedies and reforms, critics generally suggest some form of the 

detect-deter-mitigate model, imposing a duty upon the intermediary with the 

potential for liability in cases of breach. The two most common models are 

traditional liability (damages) regimes and notice-and-takedown schemes. 

Proponents of traditional liability schemes generally find theoretical fault with 

the exceptionalist view of the Internet, and analytical fault with broad judicial 

interpretations of the statute that collapse distributor-with-knowledge liability 

into immunity from publisher liability. Proponents of a notice-and-takedown 

scheme likewise work from a distributor-with-knowledge model that imposes a 

limited duty of care on intermediaries, but generally acknowledge some degree 

of exceptionalism that requires a distinct scheme. Most suggest some variation


utilizing elements of the Digital Millennium Copyright Act (DMCA) 21 and the 

European Union’s E-Commerce Directive, 22 wherein intermediary liability is 

triggered by actual notice of the objectionable content or a standard of 

reasonable care, and requiring remedial action (e.g., taking down the content at 

issue). 

The costs of these indirect intermediary liability schemes could be great. Under 

traditional liability rules, intermediaries may be forced to adopt a least-commondenominator 

approach, resulting in overly-broad restrictions on expression and 

behavior. A modified distributor-with-knowledge approach, usually in the form 

of a takedown scheme similar to that employed by the DMCA, may produce the 

same type of chilling effect. This is potentially exacerbated by the use of a 

should-have-known standard that can trigger the need to patrol for harmful 

content, raising costs and leading to even greater overbreadth in application. 

Moreover, indirect liability reduces incentives to develop self-help technology, 

such as location or identity tracking software and end-user filters, the 

development of which was one of Section 230’s primary policy goals. Thus, if 

the scale of undeterred or unremedied harms is minimal, and the negative 

impact of a detect-deter-mitigate model is significant, then the cost associated 

with the imposition of indirect intermediary liability is not reasonable. 

Resisting the Urge Toward Homogeny 

The case for preserving Section 230 immunity begins by recasting intermediary 

immunity in terms of exceptionalism, self-governance and norms, because it is 

precisely the gap between the offline social norms expressed in tort law and the 

broad immunity provided to online participants that has led to the rather strong 

criticism of Section 230. As a conceptual matter, communal enforcement 

presents the greatest challenge to effectuating some modified version of the 

exceptionalist ideal. When external legal norms are excluded, internal 

enforcement mechanisms facilitate the emergence of new communal norms to 

take their place. Much of the criticism of Section 230 stems from the lack of 

legal enforcement that accompanies immunity, and the resulting inability to 

form new social norms to replace those of the sovereign. It is important to 

recognize, however, that Web 2.0 communities, such as wikis and social 

networks, represent a real and significant manifestation of the exceptionalist 

vision, because they both facilitate a market in norms and values, and provide 

the internal enforcement mechanisms necessary for internal norms to emerge. 

Section 230 plays a vital role in the development of these communities by 

21 Digital Millenium Copyright Act, Pub. L. 105-304, 112 Stat. 2860 (1998). 

22 Directive 2000/31/EC of the European Parliament and of the Council of 8 June 2000 on 

certain legal aspects of information society services, in particular electronic commerce, in 

the Internal Market (‘Directive on electronic commerce’), http://eurlex.europa.eu/LexUriServ/LexUriServ.do?uri=CELEX:32000L0031:EN:NOT.


substantially and continually mitigating the primacy of external legal norms 

within the confines of the community. This permits choice, empowers the 

intermediary to create a market in social norms, and allows alternate forms and 

gradations of enforcement. The architecture of the community gives these 

choices form and substance, backed by an enforcement model, such that 

communal norms have the opportunity to develop. In this sense, Section 230 

and the Web 2.0 model effectuate the emergence of a modified form of 

exceptionalism. The reforms proposed by most commentators would have a 

negative impact on these communities, with little benefit beyond those 

communal norms that are likely to emerge, and should be rejected. 

Exceptionalism, Self-Governance & Social Norms 

Exceptionalism does not argue for the absence of social norms. Instead, 

exceptionalism embraces the idea of cyberspace as an environment in which the 

authority of external legal regimes is minimal, and where an open market in 

norms and values works in concert with self-governance to permit the online 

community to establish its own substantive social norms. Section 230 helps to 

effectuate a modified form of exceptionalism by moderating the imposition of 

external legal norms so as to permit a limited range of choices—bounded, at 

least, by criminal law, intellectual property law and contract law—in which the 

online community is free to create its own norms and rules of conduct. 

However, the development of social norms within this environment requires 

not only the ability to exercise broad individual choice among different values 

and embodiments of those values, but also some mechanism of communal 

enforcement through which to effectuate some form of self-governance. 

Early proponents of exceptionalism were able to focus on relational libertarian 

ideals, viewing the Internet as a unique social space in which norms governing 

thought, expression, identity, and relationship should be permitted to evolve. 

This focus developed precisely because the mechanisms of enforcement 

required for self-governance and the evolving definition of emergent social 

norms were taken for granted. The architecture of enforcement was primarily 

controlled by a community involved in the process as adherents to the 

exceptionalist ideal, who could be trusted both to ensure broad individual 

choice and to utilize the means of enforcement as a tool of self-governance as 

norms emerged. 

As a means of effectuating exceptionalism, the primary weakness of Section 230 

is the lack of an enforcement component. Although the modified 

exceptionalism enabled by Section 230 permits a range of choices, it does 

nothing to provide enforcement mechanisms to solidify emerging communal 

norms. Where immunity exists, legal enforcement mechanisms are never 

triggered. Likewise, the architecture of enforcement relied upon by early 

exceptionalists is no longer communal or likely committed to the vision of a 

distinct cyber-libertarian space, but is instead concentrated in private


commercial entities. As a consequence, Section 230 immunity creates a gap: 

Certain external legal norms are excluded, but internal communal norms are 

often unable to coalesce to take their place. It is this gap, resulting from the 

lack of architectural enforcement controls, which fuels criticism of the 

immunity provision. In application, however, an enforcement model has 

emerged that mediates the tension between the broad availability of individual 

value choices and the ability to effectively self-govern so as to permit the 

development of communal norms. 

Communities of Modified Exceptionalism 

Web 2.0 communities are structured as a limited commons and are built on an 

architecture of participation that operates as a platform for user-created content 

and collaboration. At the core are principles of open communication, 

decentralized authority, the freedom to share and re-use, and an idea of the 

Internet as a social forum or market for exchanging opinions and ideas in search 

of norms to create a culture based on sharing. Section 230 plays a vital role in 

the development and maintenance of these architectures by providing 

intermediaries with limited immunity from liability for the tortious content 

provided by users. Indeed, in this sense, Section 230 seems to favor the 

development of Web 2.0 services and the provision of user-based content over 

the traditional model of providing first-party institutional content. 

The parallels between Web 2.0 and Barlow’s vision of a communal social space 

are evident, albeit in modified form. Barlow embraced the potential of an 

environment premised upon freedom of choice in individual expression, human 

behavior and relationships. To achieve that potential, he and others believed 

that regulation by existing sovereign powers must be rejected in favor of selfgovernance, 

so that new communal social norms might have the opportunity to 

emerge. At the heart of this ideal was an affirmation that values participation in 

the market of expression, ideas and action without the constraint of 

preconceived value judgments. Web 2.0 promises a somewhat limited version 

of this environment—existing within sovereign authority, narrowed by certain 

enduring norms, and confined to segmented communities administered by 

private entities—by facilitating the market by which norms are tested. 

Two of the most common models of these Web 2.0 services, wikis and social 

networks, are indicative of how Section 230 can effectuate the modified form of 

cyber-libertarian exceptionalism described above. Partly as a result of the 

immunity from liability provided by Section 230, these services facilitate the 

market in social norms by creating enclaves in which users may exercise broad 

(although not unbounded) individual choice among competing values. At the 

same time, the intermediary retains control over the architecture and thus the 

means of enforcement. As the market defines social good through the 

evolution of communal norms, that architecture may be employed as a 

mechanism of governance. In the absence of legal incentives, the enforcement


of communal norms is driven by internal incentives, such as the need for 

financial support from community donations, a communal desire for 

information integrity, or the need to build an audience for advertising. In some 

communities, participants may be incentivized by credibility and stature in the 

form of temporal seniority, post count, rank within the community’s governing 

body, etc. 

The online encyclopedia Wikipedia is a specific example of a Web 2.0 

community of collective action. Each entry in the Wikipedia database is created 

and edited by volunteers who are guided by three primary principles: the 

Neutral Point of View policy, the No Original Research policy, and the 

Verifiability policy. Registered users can originate new articles, and any user, 

whether registered or anonymous, can edit an existing article. In the period 

between Wikipedia’s inception in 2001 and 2010, this experiment in voluntary 

collaborative action produced more than ten million articles. 

These activities are overseen by two levels of administrators, administrators and 

bureaucrats. Administrators (historically called sysops, short for system 

operators) have the power to edit pages, delete or undelete articles and article 

histories, protect pages, and block or unblock user accounts or IP addresses. 

Bureaucrats have the further power to create additional sysops with the 

approval of the community. In February 2006, in response to a series of 

significant and persistent acts of vandalism, the co-founder of Wikipedia created 

an additional layer of protection: Administrators can protect any article so that 

all future changes must be approved by an administrator. 23 Administrators help 

facilitate dispute resolution and enforcement. Low-level disputes are resolved 

in talk pages. Here, moderators guide members to resolution with reference to 

policies and guidelines developed over the life of the community. Thus, 

principle values and norms can lead to more specific rules. This approach 

works in most cases. More serious violations, such as malicious editing of an 

article (or vandalism), are addressed through fast-repair mechanisms executed 

by community members. Wikipedia administrators are also able to block user 

accounts or IP addresses. 

As described, the Wikipedia community reflects a modified form of the 

exceptionalist model, initially allowing for individual choice among a range of 

values, facilitating a market in social norms, and providing a means of 

enforcement to effectuate norms as they develop. Indeed, recent studies reflect 

not only that norms have emerged from this market, but that those norms have 

solidified and expanded. Through this process, the Wikipedia community is 

23 See Wikipedia, Wikipedia: Protection Policy, 

http://en.wikipedia.org/wiki/Wikipedia:Protection_policy (last accessed Dec. 1, 

2010).


moving from an immediate focus on particular articles to more generalized 

concerns for quality of content and community. 

Not unexpectedly, open source projects such as Wikipedia are not immune to 

abuse. In terms of community health, and to protect against these abuses, 

Wikipedia has adopted a code of conduct and principles of etiquette that stress 

civility and discourage personal attacks. As discussed above, these norms are 

enforced through an architecture that is designed to reinforce those norms with 

an eye towards the health of the community. At the most basic level, this 

occurs through routine editing by participants. Over time, more complex 

mechanisms for dispute resolution and enforcement have developed, such that 

in the past few years administrative and coordination activities have gained 

importance. 

The relationship between architecture and social norms is fascinatingly apparent 

both in the Wikipedia’s architectural choice to track and correlate the IP address 

of any anonymous user who edits the encyclopedia, as well as the development 

of a monitoring system that tracks those changes for analysis. This system 

serves as a mechanism for enforcing social norms, particularly the norm of 

neutrality in more controversial areas. In terms of more formal enforcement, 

some edits that might previously have been overlooked are now being 

reexamined in light of the organization from which they originated. Less 

formally, but perhaps even more effectively, organizations which are perceived 

to have breached the norms of the community have faced, and will face, 

recriminations. Moreover, the entire community is now aware that enforcement 

of those norms is now more effective, presumably creating a deterrence effect. 

The Wikipedia example illuminates a constant process, as choices are narrowed 

by communal norms that develop and are given life through enforcement 

mechanisms, such that principle norms generate a breadth of more particular 

rules. Section 230 immunity plays an important role in this process, permitting 

the community to evolve and structure itself in the most efficient manner. To a 

limited extent, Section 230 immunity permits uncoordinated and uncoerced 

individual choice among different values and among different embodiments of 

those values. It further allows the intermediary to play an active role in 

facilitating the market in social norms and in creating enforcement mechanisms 

as a tool of self-governance. Those enforcement mechanisms can then 

themselves adapt. This allows not only for the development of distinct 

community values, but also for a means of tapping into incentives, adapting to 

evolving norms and conditions, and reducing costs associated with disputes. 

Within this framework, greater variations in community norms are possible. As 

communities grow, niche communities are formed at low cost. It is not the 

global vision of early exceptionalism, but rather a more limited and localized 

form of modified exceptionalism that functions as a laboratory for testing social 

norms and values.


Conclusion 

Critics of Section 230 have both overstated the harms arising from immunity 

and understated the costs of alternate schemes for imposing indirect liability on 

online intermediaries. At the same time, they have ignored the important role 

Section 230 plays in the development of online communities. The immunity 

provided by Section 230 helps to create the initial conditions necessary for the 

development of a modified form of exceptionalism by mitigating the effect of 

external legal norms in the online environment. Web 2.0 communities are then 

able to facilitate a market in norms and provide the architectural enforcement 

mechanisms that give emerging norms substance. Given Section 230’s crucial 

role in this process, and the growing importance of Web 2.0 communities in 

which collaborative production is yielding remarkable results, reforming the 

statute to substantially narrow the grant of immunity is both unnecessary and 

unwise.



Internet Exceptionalism Revisited 

By Mark MacCarthy * 

Introduction 

In the mid-1990s, commentators began debating the best way for governments 

to react to the development of the Internet as a global communications 

medium. Internet exceptionalists argued that the borderless nature of this new 

medium meant that the application of local law to online activities would create 

insoluble conflicts of law. The exceptionalists believed that as the Internet 

grew, reliance on local governments to set rules for the new online world would 

not scale well. Their alternative was the notion of cyberspace as a separate place 

that should be ruled by norms developed by self-governing communities of 

users. 1 

Critics of the exceptionalist view responded with a vision of a bordered Internet 

where local governments could apply local law. 2 In this view, cyberspace is not 

a separate place. It is simply a communications network that links real people in 

real communities with other people in different jurisdictions. Governments can 

regulate activity on this new communications network in many different ways, 

including by relying on the local operations of global intermediaries. Global 

intermediaries are the Internet service providers (ISPs), payment systems, search 

engines, auction sites, and other platform and application providers that provide 

the infrastructure necessary for Internet activity. Although they are often global 

in character, they also have local operations subject to local government control. 

According to critics of the exceptionalist view, governments have the right and 

the obligation to use this regulatory power over intermediaries to protect their 

citizens from harm. 3 Conflicts that might arise from this regulatory activity can 

* Mark MacCarthy is Adjunct Professor in the Communications Culture and Technology 

Program at Georgetown University. Formerly, he was Senior Vice President for Public 

Policy at Visa Inc. Substantial portions of this essay were originally published as Mark 

MacCarthy, What Payment Intermediaries are Doing About Online Liability and Why It Matters, 25 

BERKELEY TECH. L. J. 1037 (2010), available at http://btlj.org/data/articles/25_2/1037- 

1120%20MacCarthy%20WEB.pdf. 

1 See, e.g., David R. Johnson & David Post, Law and Borders — The Rise of Law in Cyberspace, 48 

STAN. L. REV. 1367, 1387-92 (1996). 

2 E.g., Jack L. Goldsmith, Against Cyberanarchy, 65 U. CHI. L. REV. 1199 (1998). 

3 See id. at 1238-39.


be resolved through the normal mechanisms governments use to resolve 

conflict of law questions. 4 

Governments generally followed the advice of the proponents of regulation, not 

the regulatory skeptics. 5 And despite some set-backs in First Amendment 

cases, 6 regulators have continued a steady march toward controlling the Internet 

by regulating intermediaries. 7 Some legal scholars argue that government 

reliance on intermediaries to control unlawful behavior on the Internet is 

justified because putting the enforcement burden on intermediaries is the least 

expensive way for governments to effectively assert jurisdiction. 8 The key 

rationale is that governments cannot easily find wrong-doers on the Internet, 

but intermediaries can. They are best positioned to monitor their own systems. 

As Mann and Belzley put it, they are the “least-cost avoider.” 9 

The defenders of local government jurisdiction over the Internet often rely on 

historical analogies to buttress their case that local control is inevitable and 

desirable. Debra Spar developed the thesis that society’s reaction to new 

technologies follows a predictable sequence of innovation, commercial 

exploitation, creative anarchy, and then government rules. 10 In the innovative 

stage a new technology is developed, in the second stage it is used in 

commercial ventures, in the third stage there is a tension between the anarchist 

impulse and the need for commercial order and stability, and in the final stage 

society reaches out to regulate the now mature technology to create and 

4 Id. at 1200-01 (arguing that “regulation of cyberspace is feasible and legitimate from the 

perspective of jurisdiction and choice of law”). 

5 The U.S. exception is § 230 of the Telecommunications Act of 1996 which immunizes many 

Internet actors from liability in many contexts for the illegal activity of their users. 47 U.S.C. 

§ 230(c) (2006). 

6 See, e.g,, Reno v. ACLU, 521 U.S. 844, 885 (1997) (“The interest in encouraging freedom of 

expression in a democratic society outweighs any theoretical but unproven benefit of 

censorship.”); Ctr. for Democracy & Tech. v. Pappert, 337 F. Supp. 2d 606, 665 (E.D. Pa. 

2004) (finding that a statute requiring ISPs to block access to websites displaying child 

pornography violated the First Amendment). 

7 See generally JACK GOLDSMITH & TIM WU, WHO CONTROLS THE INTERNET?: ILLUSIONS OF A 

BORDERLESS WORLD (2006) (citing many examples of this trend). This Article documents 

further examples in which payment systems were induced by laws, regulations, pressure, and 

notions of corporate responsibility to take actions to control the illegal online behavior of 

people using their systems. 

8 See, e.g., Ronald J. Mann & Seth R. Belzley, The Promise of Internet Intermediary Liability, 47 WM. 

& MARY L. REV. 239, 249-50 (2005). 

9 Id. at 249. 

10 DEBORA L. SPAR, RULING THE WAVES: CYCLES OF DISCOVERY, CHAOS, AND WEALTH FROM 

THE COMPASS TO THE INTERNET 11-22 (2001).


maintain the needed stability. 11 The development of radio is the standard 

example of this pattern. Radio’s initial pioneers thought its ability to wirelessly 

broadcast information from one point to many made government control 

difficult and unnecessary. 12 But later commercial enterprises actively sought out 

government regulation in order to end the chaos on the airwaves that prevented 

broadcasters from reaching their intended audience. 13 Applying Spar’s analysis 

here, the Internet is somewhere between stage three and stage four, where we 

can expect further regulation of Internet activity under the watchful eye of 

government. The historical example demonstrates that although every new 

technology is thought to be outside the jurisdiction of government, this belief 

usually gives way in time to the realities of government control. 

In the case of the Internet, the advent of government control prompted many 

observers to think the Internet exceptionalists had been routed. 14 However, 

Internet exceptionalism is still a widely held belief, 15 and the notion that 

government control of cyberspace is both impossible and illegitimate still 

motivates much discussion of Internet policy. 16 Moreover, the initial legislative 

expression of Internet exceptionalism—Section 230 of the 1996 

Telecommunications Act—is still on the books. This section provides a safe 

harbor from indirect liability for what might be called pure Internet 

intermediaries—those entities providing Internet access service or online 

11 Id.; see also Mann & Belzley, supra note 9, at 243-44; GOLDSMITH & WU, supra note 7, at 124 

(relying on Spar’s work). 

12 See generally SPAR, supra note 10, at 124-90 (describing the history of radio technology 

development). 

13 Id. at 171-72. 

14 See GOLDSMITH & WU, supra note 7, at 14 (asserting that “notions of a self-governing 

cyberspace are largely discredited”). 

15 See generally DAVID G. POST, IN SEARCH OF JEFFERSON’S MOOSE (David Kairys ed., 2009) 

[hereinafter Post, IN SEARCH OF JEFFERSON’S MOOSE ] (demonstrating an elegant take on 

Internet exceptionalism). The heart of the response to Goldsmith is that scale matters and 

that while it is physically possible and permissible under current “settled” law of crossborder 

jurisprudence, it is not “workable” to subject all websites to perhaps hundreds of 

different and possibly conflicting jurisdictions. See David G. Post, Against “Against 

Cyberanarchy”, 17 BERKELEY TECH. L.J. 1365, 1384 (2002) [hereinafter Post, Against “Against 

Cyberanarchy”]. 

16 See H. Brian Holland, supra (adapted from H. Brian Holland,In Defense of Online Intermediary 

Immunity: Facilitating Communities of Modified Exceptionalism, 56 U. KAN. L. REV. 369, 397 

(2007)). Holland’s version of modified exceptionalism is closely connected with the legal 

principle that online intermediaries are not liable for third party conduct. He asserts that the 

immunity from liability created by § 230 of the Communications Decency Act “helps to 

effectuate a modified form of exceptionalism by moderating the imposition of external legal 

norms so as to permit a limited range of choices—bounded, at least, by criminal law, 

intellectual property law and contract law—in which the online community is free to create 

its own norms and rules of conduct.” Id. at 397.


services. 17 Despite a growing call to revisit this immunity, 18 it has been 

extended several times. The Internet gambling law, which creates liability for 

traditional intermediaries such as payment systems, contains a limitation on 

liability for pure Internet intermediaries. 19 Similarly, the recently passed online 

pharmacy law exempts pure Internet intermediaries from a general duty to avoid 

aiding or abetting unauthorized Internet sales of controlled substances. 20 The 

adoption of these provisions in recent laws might be merely § 230 on automatic 

pilot, but more likely, some version of Internet exceptionalism is at work in 

these legislative distinctions. 

A recent speech by the Obama Administration’s senior communications 

policymaker, Lawrence Strickling, provides further evidence of the continuing 

relevance of the Internet exceptionalist perspective. 21 In defending Section 

230’s limitation on liability, Assistant Secretary Strickling argued: 

This limitation on liability has enabled the creation of 

innovative services such as eBay and YouTube, which host 

content provided by others, without requiring that those 

services monitor every single piece of content available on their 

sites. Absent this protection against liability, it is hard to 

imagine that these services would have been as successful as 

they turned out to be. 22 

Internet exceptionalism is the view that the normal rules that apply to realworld 

providers of goods and services should not apply to online entities. 

Secretary Strickling argues for this view on policy grounds. Without it, he 

asserts, the innovative character of the Internet would come to a halt. The next 

17 47 U.S.C. § 230(c)(1) (2006) (“No provider or user of an interactive computer service shall 

be treated as the publisher or speaker of any information provided by another information 

content provider.”). The interpretation of this provision is quite broad. See, e.g., Zeran v. Am. 

Online, Inc., 129 F.3d 327, 330-31 (4th Cir. 1997) (finding that plaintiff ’s tort claims of 

defamation were preempted by § 230). The immunity does not extend to criminal law, 

contract law, or intellectual property law. 47 U.S.C. § 230(e)(1)-(4) (2006). 

18 See, e.g., Doug Lichtman & Eric Posner, Holding Internet Service Providers Accountable, 14 U. CHI. 

SUP. CT. ECON. REV. 221 (2006), John Palfrey and Urs Gasser, BORN DIGITAL 106 (2008), 

and Daniel Solove, THE FUTURE OF REPUTATION 125-160 (2007). 

19 31 U.S.C. § 5365(c) (2006). 

20 Ryan Haight Online Pharmacy Consumer Protection Act of 2008, Pub. L. No. 110-425, § 

(h)(3)(A)(iii), 122 Stat. 4829-30. 

21 Remarks by Lawrence Strickling, Assistant Secretary of Commerce for Communications and 

Information, to Internet Society’s INET Series: Internet 2020: The Next Billion Users April 29, 

2010 available at 

http://www.ntia.doc.gov/presentations/2010/InternetSociety_04292010.html 

22 Id.


YouTube or Google could never emerge because the legal liabilities associated 

with running such a new business would strangle it. 

If the Internet exceptionalists rested their case on the literal impossibility of 

extending local law to cyberspace then there is not much left to their argument. 

A “bordered Internet” where intermediaries try to control behavior prohibited 

by local law is becoming a reality. Most Internet intermediaries have explicit 

policies that prohibit illegal activities. 23 These general policies are supplemented 

with specific policies and procedures designed to prevent the use of these 

systems for specific illegal activities. 

Moreover, it is not just voluntary efforts by Internet intermediaries that show 

how Internet activity can be controlled. Governments have been effectively 

extending their control over Internet activity through imposing obligations on 

intermediaries. It has been estimated that at least 26 countries impose some 

kind of filtering obligations on Internet entities. 24 Recent government actions in 

France and the United Kingdom impose “graduated response” obligations on 

ISPs, requiring them to cut off Internet access for alleged repeat copyright 

violators. 25 It is possible to challenge these extensions of government power 

23 Participants in Google’s advertising programs “shall not, and shall not authorize any party to 

… advertise anything illegal or engage in any illegal or fraudulent business practice.” Google 

Inc. Advertising Program Terms 4 (Aug. 22, 2006), available at 

https://adwords.google.com/select/tsandcsfinder. MasterCard has rules for both 

merchants and their acquiring banks: “A Merchant must not submit for payment into 

interchange … and an Acquirer must not accept from a Merchant for submission into 

interchange, any Transaction that is illegal.” MASTERCARD, MASTERCARD RULES 5.9.7 (2008), 

available at http://www.merchantcouncil.org/merchantaccount/downloads/mastercard/MasterCard_Rules_5_08.pdf. 

MasterCard prohibits 

its issuing banks from engaging in illegal transactions. Id. at 3.8.4. Visa has similar rules, for 

example: “A Merchant Agreement must specify that a Merchant must not knowingly submit, 

and an Acquirer must not knowingly accept from a Merchant, for submission into the Visa 

payment system, any Transaction that is illegal or that the Merchant should have known was 

illegal.” VISA, VISA INTERNATIONAL OPERATING REGULATIONS § 4.1.B.1.c (2008), available at 

http://usa.visa.com/download/merchants/visa-international-operatingregulations.pdf. 

Visa’s regulations also specify acquirer penalties for merchants engaging in 

illegal cross-border transactions. Id. § 1.6.D.16. 

24 RONALD DEIBERT, JOHN PALFREY, RAFAL ROHOZINSKI, JONATHAN ZITTRAIN, ACCESS 

DENIED: THE PRACTICE AND POLICY OF GLOBAL INTERNET FILTERING 1 

(2008). 

25 Eric Pfanner, U.K. Approves Crackdown on Internet Pirates, NEW YORK TIMES, April 8, 2010 at 

http://www.nytimes.com/2010/04/09/technology/09piracy.html?scp=1&sq=digital 

%20economy%20bill%20uk&st=cse. Eric Pfanner, France Approves Wide Crackdown on Net 

Piracy, NEW YORK TIMES, October 22, 2009, 

http://www.nytimes.com/2009/10/23/technology/23net.html?_r=1. Sometimes the 

ISPs cooperate in a graduated response policy to settle legal claims. For a review of 

government and private sector efforts to control online copyright violations, see Christina 

Angelopoulos, Filtering the Internet for Copyrighted Content in Europe, IRIS PLUS, March 2009 

available at http://www.obs.coe.int/oea_publ/iris/iris_plus/iplus4_2009.pdf.en


over Internet activity as unwise, or as a violation of a human right to Internet 

access or as too costly. But it is no longer plausible to maintain that they are 

simply impossible. 

This conclusion is discussed at length in another essay in this collection that 

focuses on the traditional payment intermediaries, payment card companies 

such as Visa, MasterCard, and American Express, as an instructive category of 

intermediary platforms. 26 Developments over the last several years conclusively 

demonstrate that these payment intermediaries can control specific illegal 

activities on the Internet and governments can extend their control to these 

payment intermediaries. 

Thus, the debate over Internet exceptionalism has shifted from the “nature” of 

the Internet as something intrinsically beyond the control of governments to a 

problem of choice. 27 Intermediaries can control illegal behavior on the Internet 

and governments can control intermediaries, but should they? And if government 

should exert control over intermediaries in order to control Internet activities, 

how should the global legal order be restructured to accommodate their role? 

This essay explores the extent to which the experience of payment systems in 

controlling the illegal online behavior of their users illuminates the debate 

among the Internet exceptionalists, defenders of the bordered Internet, and the 

internationalists. It concludes that exceptionalism, in either its original or 

modified forms, is not the right framework for Internet governance because 

intermediaries should not defer to the judgments of self-governing communities 

of Internet users when the judgments conflict with local law. The 

exceptionalists are correct that a “bordered Internet” will not scale up, but the 

experience of traditional payment systems points towards international 

harmonization. If governments are going to use intermediaries to regulate the 

Internet, they need to coordinate their own laws to make that role possible. 

The essay addresses each of the three main approaches to Internet governance: 

exceptionalism, the bordered Internet, and internationalism. The first section, 

on exceptionalism, begins with a discussion of the original Internet 

exceptionalist perspective, which viewed government regulation of the Internet 

as infeasible and normatively less desirable than government deference to the 

rules developed by self-governing Internet communities. This is followed by a 

discussion of Brian Holland’s revised version of exceptionalism. Under this 

approach, the various immunities from intermediary liability established by local 

jurisdictions enable the development of autonomous Internet norms. Both 

versions are shown to have significant limitations when viewed in light of 

26 See MacCarthy, Online Liability for Payment Systems, infra at 230. 

27 See Holland, supra note 16, at 376-77 (“In this context, exceptionalism became an objective to 

be pursued and protected as a matter of choice, rather than a natural state.”).


payment system experiences. The next section explores the “bordered 

Internet,” the idea that in certain cases local governments may properly and 

unilaterally extend their jurisdiction over Internet activities through 

intermediaries. Payment intermediaries use standard measures to resolve 

conflicts of law and follow a practical rule that treats a transaction as illegal if it 

is illegal in the jurisdiction of either the merchant or the cardholder. This 

section then discusses limitations on this method of resolving cross-border 

jurisdictional conflicts. The final section concludes with a discussion and 

endorsement of the internationalist perspective, according to which local 

governments should only exercise control over specific Internet activities in a 

coordinated fashion. 

Internet Exceptionalism: 

The Original Version 

In February 1996, John Perry Barlow identified Internet exceptionalism when 

he declared cyberspace to be independent of national governments, roughly on 

the grounds that cyberspace “does not lie within your borders” and that it “is a 

world that is both everywhere and nowhere, but it is not where bodies live.” 28 

Conflicts in cyberspace would be resolved not with the territorially-based “legal 

concepts of property, expression, identity, movement, and context,” which “do 

not apply,” to cyberspace because they “are all based on matter, and there is no 

matter here.” 29 Rather, in cyberspace “governance will arise according to the 

conditions of our world, not yours.” 30 Cyberspace “is different.” 31 

Almost concurrently, legal scholars David Johnson and David Post made a 

similar case for Internet exceptionalism. 32 In their view, the Internet destroys 

“the link between geographical location” and “the power of local governments to 

assert control over online behavior; [and] … the legitimacy of a local sovereign’s 

efforts to regulate global phenomena ….” 33 The Internet destroys the power of 

local governments because they cannot control the flow of electrons across 

their physical boundaries. If they attempted to do so, determined users would 

just route around the barriers. Moreover, if one jurisdiction could assert control 

over Internet transactions, all jurisdictions could, resulting in the impossibility 

28 Declaration of John P. Barlow, Cognitive Dissident, Co-Founder, Elec. Frontier Found., A 

Declaration of the Independence of Cyberspace (Feb. 8, 1996), available at 

http://w2.eff.org/Censorship/Internet_censorship_bills/barlow_0296.declaration. 

29 Id. 

30 Id. 

31 Id. 

32 See generally Johnson & Post, supra note 1. 

33 Id. at 1370 (emphasis added).


that all “Web-based activity, in this view, must be subject simultaneously to the 

laws of all territorial sovereigns.” 34 The Internet destroys the legitimacy of local 

jurisdiction because legitimacy depends on the consent of the governed and 

“[t]here is no geographically localized set of constituents with a stronger and 

more legitimate claim to regulate it than any other local group. The strongest 

claim to control comes from the participants themselves, and they could be 

anywhere.” 35 Since “events on the Net occur everywhere but nowhere in 

particular … no physical jurisdiction has a more compelling claim than any 

other to subject these events exclusively to its laws.” 36 

Behind these arguments seemed to be an appealing political vision. The ideal 

envisaged self-organizing groups of people making the rules that applied to their 

conduct. These rules would not be imposed from the outside, but would be 

freely chosen by the active participation of the community members. The key 

was deliberation by free, rational agents in their communities, not imposition of 

rules by an arbitrary act of will by a distant sovereign. This ideal of participatory 

democracy was intended, in part, to offset the alienating effects of large-scale 

modern democracies, which in practice had long failed to provide their 

members with the sense of community participation that alone seemed to justify 

the imposition of collective rules. 

The way this vision would be implemented on the Internet would be through 

the development of autonomous communities of Internet users. These Internet 

communities were largely isolated from “real world” communities. Since it took 

special care and effort to reach out to participate in them, only those people 

who really wanted to participate would, and the effects of activities in those 

communities would be limited to those who chose to participate. Given the 

structure of the Internet as a communications network, which moved almost all 

major decisions on content to the edges of the network, a diversity of law could 

arise in cyberspace as each community developed its own norms for regulating 

the conduct of its members. People would be free to participate in the 

communities they wanted, but could easily avoid those they did not like. 

Enforcement of the community rules would be accomplished through peer 

pressure, reputational systems, informal dispute resolution mechanisms, and 

ultimately, banishment. The system as a whole would evolve through a process 

analogous to biological evolution, where diverse and potentially competing rule 

sets as embodied in different communities would vie for acceptance in a free 

marketplace of rules. 

34 Id. at 1374. 

35 Id. at 1375. 

36 Id. at 1376.


Internet exceptionalism is thus the view that activity on the Internet should be 

regulated by Internet community norms, not laws of territorial jurisdictions or 

globally harmonized laws. 37 It is hard to avoid the sense that the political vision 

pre-dated the Internet—that the feasibility argument masked the underlying 

vision and the arrival of the Internet simply created the possibility of 

implementing the vision in a way that the “real” world did not. To see this, 

imagine the reaction of Internet exceptionalists to the idea of a world 

government that would establish uniform global laws. This would eliminate the 

conflict of law problem. But exceptionalists are even more appalled at the idea 

of world government control over the Internet than with the idea of nationstate 

control over it. This suggests that the issue is not feasibility of control, but 

the value of participative community decision making and diversity. 

This early cyber libertarian vision was immediately attacked by those who 

defended the feasibility and legitimacy of extending local laws to cover Internet 

activity. 38 As they note, “[t]he mistake here is the belief that governments 

regulate only through direct sanctioning of individuals…. Governments can … 

impose liability on intermediaries like Internet service providers or credit card 

companies.” 39 Government action against these intermediaries “makes it harder 

for local users to obtain content from, or transact with, the law-evading content 

providers abroad. In this way, governments affect Internet flows within their 

borders even though they originate abroad and cannot easily be stopped at the 

border.” 40 And these efforts to bring order to the Internet through pressure on 

intermediaries are often legitimate because they provide “something invisible 

but essential: public goods like criminal law, property rights, and contract 

enforcement … that can usually be provided only by governments.” 41 

The debate took an interesting twist through the work of Larry Lessig. A key 

element of the early exceptionalist framework was the idea that the Internet had 

37 Mann and Belzley describe their view as “consciously exceptionalist” because “specific 

characteristics of the Internet make intermediary liability relatively more attractive than it has 

been in traditional offline contexts because of the ease of identifying intermediaries, the 

relative ease of intermediary monitoring of end users, and the relative difficulty of directly 

regulating the conduct of end users.” Mann & Belzley, supra note 9, at 250-51. But this is an 

odd way of framing the issue. Internet exceptionalism is not simply the view that the 

Internet should be treated differently from the offline world. The claim is more specifically 

that the Internet should be free of local jurisdictions. Mann and Belzley’s view, which implies 

that the Internet should be brought under local jurisdictions through the mechanism of 

intermediary liability, is thus the very opposite of exceptionalism. It is one version of 

Internet non-exceptionalism. 

38 See generally Goldsmith, supra note 4 (challenging the regulation skeptics). 

39 Goldsmith, supra note 4, at 1238. 

40 GOLDSMITH & WU, supra note 7, at 68. 

41 Id. at 140.


a fundamental nature, which governments did not control, could not alter, and 

which effectively prevented them from imposing local rules. In his influential 

book, Code and Other Laws of Cyberspace,Lessig took aim at this idea. 42 He 

pointed out that computer systems, software applications, and communications 

networks were human creations and that the choices of the architects of these 

systems were embodied in the code that made it possible for these systems to 

run. Far from being a natural object, these systems were subject to the 

decisions of the parties (usually non-governmental entities) that had the right 

and the ability to create, maintain and alter them. 

The initial openness and transparency of the Internet was therefore something 

that could not be assumed as a fact of nature, but something that needed to be 

maintained against possible opponents. But unlike the early cyber libertarians, 

Lessig did not focus on the dangers that local governments might try to control 

choices by controlling code. He thought the openness of the Internet had to be 

maintained against the interests of non-governmental parties seeking to advance 

their own strategic interests. Lessig’s initial private sector targets were the 

network carriers who were seeking to alter the “end-to-end” design of the 

network in order to pursue their own strategic interests at the expense of 

application providers, service providers and end users who relied on the 

neutrality of the Internet to conduct their ordinary activities. In this way, the 

Internet exceptionalist debate merged with the net neutrality debate and the 

original defenders of exceptionalism seemed to be faced with the (to them) 

unattractive dilemma of using local governments to promote Internet values of 

openness or allowing their Internet choices to be dictated by unaccountable 

private entities that controlled the fundamental architecture of the Internet. 43 

This attack was so effective that many believe that these notions of a “selfgoverning 

cyberspace are largely discredited.” 44 But modified versions accept 

the basic premise that the Internet should be free of local regulation and 

governed instead by its users. One version of the revived exceptionalism, 

defended by Brian Holland, focuses on Web 2.0 communities. 45 This view 

42 LAWRENCE LESSIG, CODE AND OTHER LAWS OF CYBERSPACE (1999). 

43 See Holland, supra note 16 at 108-119 for a summary of this way of connecting the Internet 

exceptionalist debate with the net neutrality debate. 

44 Id. at 14. 

45 Holland writes: 

By mitigating the imposition of certain external legal norms in the online 

environment, § 230 helps to create the initial conditions necessary for the 

development of a modified form of exceptionalism. With the impact of 

external norms diminished, Web 2.0 communities, such as wikis and social 

networks, have emerged to facilitate a limited market in norms and values 

and to provide internal enforcement mechanisms that allow new communal 

norms to emerge.


argues that together with the immunity provisions of Section 230 of 

Communications Decency Act, these communities have the potential to allow 

internal community norms to take the place of external territorially based laws. 46 

Critique of Internet Exceptionalism 

The experience of global payment intermediaries described in a companion 

article in this volume confirms the view that intermediaries can effectively 

control illegal activity in cyberspace. This still leaves the question of whether 

intermediaries should resist governmental pressure to control the behavior of 

their users. As a general matter, they should not defer to the judgments of selfgoverning 

communities of Internet users when these judgments conflict with 

local law. As corporate citizens, they have an obligation to obey the laws of the 

jurisdictions in which they operate, and they simply have no basis to excuse 

themselves from that duty in order to let online communities determine their 

own fate. But even when local law does not require them to take action against 

illegal behavior, their responsibility to keep their systems free of illegal activity 

means that they often should take specific steps to stop these activities. 

The fundamental objection, even to Holland’s modified exceptionalism, is that 

the “law” of Internet communities is not really the law of that community. It is 

a commercial contract enforceable under the rules of some local jurisdiction, 

and the terms of the contract are subject to the same kinds of legal and 

regulatory oversight that bind contracts between people in local jurisdictions. 

Deferring to these contracts does not usually mean democratic community selfgovernment. 

Local regulations are needed to fully protect the members of these 

communities. 47 Moreover, in some cases, the legal discretion granted to 

intermediaries to control the conduct of their members may be too broad and 

should be limited by replacing intermediary judgment with public authority 

decisions. The remainder of this section develops these points. 

Even if Internet communities could substantially exclude a significant portion 

of external legal norms, it still does not follow that internal norms will 

necessarily emerge from the process of debate and deliberation that Holland 

envisages. As Holland notes, “external legal norms are excluded, but internal 

communal norms are often unable to coalesce to take their place” because 

enforcement is “concentrated in private commercial entities.” 48 The hope of his 

modified Internet exceptionalism is that the intermediaries who control the new 

Web 2.0 platforms will be driven by internal incentives to accommodate the 

46 Id. 

Holland, supra note 16, at 369. 

47 This Section focuses on competition policy, privacy, and consumer protection as examples. 

48 Holland, supra note 16, at 398.


wishes of the online communities they create, allowing users to establish norms 

for their own communities. 49 

But it is not clear that Web 2.0 platforms are likely to grant this kind of 

democratic self-governance. For example, intermediaries can be subject to 

pressure. Craig Newmark, the operator of Craigslist, has insisted that he made 

his decision to remove ads for erotic services as a result of consultation with his 

online community. 50 But it is also true that Craigslist was under criminal 

investigation by a number of state attorneys general for violation of state laws 

against prostitution. 51 One could argue immunity in this case, but Craigslist did 

not. 52 It complied with a law enforcement request to remove certain postings 

and the decision to remove these ads will be subject to ongoing oversight by 

these law enforcement agencies. 53 However, the question remained whether or 

not Craigslist would take the legal risk if the community voted to keep these ads 

in place. 

These communities are not typically governed by democratic voting procedures 

that guarantee the consent of the governed. They are governed by contractual 

terms of service. Often prospective members of these communities have a 

simple take-it-or-leave-it choice when they decide to join. 54 

49 These internal incentives include “the need for financial support from community donations, 

a communal desire for information integrity, or the need to build an audience for 

advertising.” Id. at 400; see also Matthew Schruers, Note: The History and Economics of ISP 

Liability for Third Party Content, 88 VA. L. REV. 205, 261 (“ISPs respond to content-based 

complaints as a matter of good business practice for the purpose of maintaining customer 

goodwill and satisfaction.”). 

50 Craigslist Founder Seeks Larger DC Role, NAT’L J., June 2, 2009, available at 

http://techdailydose.nationaljournal.com/2009/06/craigslist-founder-seekslarge.php 

(reporting Craig Newmark’s comments to the Computers Freedom and Privacy 

Conference). 

51 See Brad Stone, Craigslist to Remove ‘Erotic’ Ads, N.Y. TIMES, May 14, 2009, at B1. Craigslist’s 

attorneys asserted immunity under § 230, but chose voluntarily to remove the ads to which 

various state attorneys general had objected. Id. State Attorneys General felt confident that 

they could bring a case under state criminal law despite the immunity granted by § 230. Id. 

The case was given national attention when a medical student was accused of killing a 

masseuse whom he met through Craigslist. Id. 

52 Id. 

53 Id. 

54 See Johnson & Post, supra note 1, at 1380 (describing AOL and Compuserve terms of service 

as examples of law in cyberspace). Johnson & Post view the rules for an Internet community 

to be “a matter for principled discussion, not an act of will by whoever has control of the 

power switch.” Id. But it is hard to see how terms of service for a typical Internet service or 

application is anything other than an act of will by the person who controls the service or 

application. It might satisfy certain legal standards for informed consent, but it is not the 

product of principled discussion. And this might be the way consumers want it. Online


If consumers do not like the terms of service, then protest can be effective, as 

in the recent case of users objecting to the change in terms of service 

unilaterally offered by Facebook. By threatening the privacy rights of the 

community, the platform stirred up substantial community unrest, and 

ultimately the new terms of service were withdrawn. 55 But this exit right is not 

the same as democratic self-governance, and it is not always effective. What if 

Facebook had not responded to community objections? Would people actually 

have left, and where would they have gone? Lock-in is a real restriction in social 

networks. 

The exemption from liability based on Section 230 does not mean that online 

entities are exempt from local law. Often, local law is needed to protect 

consumers from the actions of Internet intermediaries. Regulation of online 

communities by governments seems especially timely and urgent in three areas: 

competition policy, privacy, and consumer protection. 

With respect to competition, concentration in particular sectors of the online 

world should be examined because it can so significantly reduce consumer 

choice. The Department of Justice has indicated, for example, that it is going to 

take a more active approach in this area. 56 Along with the Federal Trade 

communities might not offer to determine their online laws through a political process 

because the members of the community cannot be bothered. People visit many different 

websites and use many different web services. It is hard to believe that they want full 

democratic participation rights to set up the rules for each of these services. And it is 

implausible that they would actually spend the time, if they were offered the opportunity. 

The example of privacy policies makes the point. A recent study concluded that if all U.S. 

consumers read all the privacy policies for all the web sites they visited just once a year, the 

total amount of time spent on just reading the policies would be 53.8 billion hours per year 

and the cost to the economy of the time spent doing this would be $781 billion per year. 

Aleecia M. McDonald & Lorrie F. Cranor, The Cost of Reading Privacy Policies, 4 I/S: J.L. & 

POL’Y FOR INFO. SOC’Y 543, 565 (2008). 

55 N.Y. Times, Facebook, Inc., 

http://topics.nytimes.com/top/news/business/companies/facebook_inc/index.ht 

ml?8qa&scp=1-spot&sq=facebook&st=nyt (last updated May 27, 2009). In 2007, the 

company had created a community backlash when it introduced an advertising service that 

allowed a user’s online activities to be distributed to other community members. Epic.org: 

Electronic Privacy Information Center, Social Networking Privacy, 

http://epic.org/privacy/socialnet/default.html (last visited Feb. 3, 2009). In the face 

of this protest, it provided a simple way for users to decline to participate. Id. In February 

2009, it proposed new privacy rules according to which users will own and control their own 

information, and in April it allowed a vote of its users on these new principles. Over 75% 

of those voting endorse them, and on July 1, 2009 it adopted them. Id. 

56 Press Release, U.S. Dep’t of Justice, Justice Department Withdraws Report on Antitrust 

Monopoly Law: Antitrust Division to Apply More Rigorous Standard with Focus on the 

Impact of Exclusionary Conduct on Consumers (May 11, 2009), available at 

http://www.justice.gov/atr/public/press_releases/2009/245710.pdf.


Commission (FTC), they have initiated inquiries focused on the search engine 

market. 57 

Privacy and security rules need to be defined as well. The FTC has taken major 

action in this area, and is stepping up its enforcement. 58 They are also focusing 

on the development of a new privacy framework to analyze the basis for the 

harms associated with privacy violations. 59 Furthermore, the FTC has focused 

on developing rules for online behavioral advertising. 60 In addition, rules 

governing privacy for online cloud computing services need to be clarified, 

perhaps by additional legislation. 61 

Consumer protection rules should be updated to apply more effectively to new 

developments in electronic commerce including the growth of mobile 

commerce and user-generated content, the greater availability of digital goods 

online, and increased numbers of consumers acting as online sellers, and new 

developments in accountability and payment protection. A timely development 

might be the harmonization of consumer redress and liability rights across 

various payment mechanisms. 62 

Finally, the discretion given to Internet intermediaries over which transactions 

to allow must be subject to public scrutiny. Today, intermediaries exercise 

57 See, e.g., Miguel Helft, U.S. Inquiry Is Confirmed into Google Books Deal, N.Y. TIMES, July 3, 2009, 

at B3; Miguel Helft & Brad Stone, Board Ties at Apple and Google Scrutinized, N.Y. TIMES, May 

5, 2009, at B1; Peter Whoriskey, Google Ad Deal Is Under Scrutiny: Yahoo Agreement Subject of 

Antitrust Probe, Sources Say, WASH. POST, July 2, 2008, at D1. 

58 See Press Release, Fed. Trade Comm’n, Sears Settles FTC Charges Regarding Tracking 

Software (June 4, 2009), available at http://www.ftc.gov/opa/2009/06/sears.shtm 

(reporting that in the Sears case the FTC obtained a settlement from Sears after charging 

that their consent practices in regard to installing an online tracking program on customers’ 

computers constituted an unfair or deceptive practice). 

59 See Stephanie Clifford, Fresh Views at Agency Overseeing Online Ads, N.Y. TIMES, Aug. 5, 2009, at 

B1 (stating that David Vladeck, the new head of the FTC’s consumer protection division, is 

rethinking privacy). Vladeck said that “[t]he frameworks that we’ve been using historically for 

privacy are no longer sufficient.” Id. In his view the FTC will begin to consider not just 

whether companies caused monetary harm, but whether they violated consumers’ dignity 

because, for example, “[t]here’s a huge dignity interest wrapped up in having somebody 

looking at your financial records when they have no business doing that.” Id. 

60 See Press Release, Fed. Trade Comm’n, FTC Staff Revises Online Behavioral Advertising 

Principles (Feb. 13, 2009), available at http://www.ftc.gov/opa/2009/02/behavad.shtm. 

61 See generally ROBERT GELLMAN, WORLD PRIVACY FORUM, PRIVACY IN THE CLOUDS: RISKS TO 

PRIVACY AND CONFIDENTIALITY FROM CLOUD COMPUTING (2009) (discussing these cloud 

computing issues). 

62 Legal payment protections now differ depending on the type of payment product used 

(debit or credit) and the nature of the payment provider—traditional payment providers like 

Visa face legal requirements while new payment providers such as cell phone companies do 

not.


judgment over which transactions are subject to such legal risk that they cannot 

be allowed. These decisions are made in the context of the business interests 

and technological capabilities of the intermediaries themselves, but they have 

important effects on the rights and interests of other parties. Some examples, 

explained in a companion essay in this volume, include: 

� Payment systems effectively decide which Internet gambling 

transactions are illegal. By choosing to block all coded gambling 

transactions, the system disadvantages horseracing, state lottery, and 

Indian gaming transactions that are arguably legal. 

� Payment systems take complaints from third parties, make an 

independent legal assessment of the merits of the case, and withdraw 

service based on these assessments. In effect, they adjudicate these 

copyright cases. 

These decisions are sound and sensible ways to balance complex and competing 

interests. However, they are private sector judgments, inevitably subjective and 

influenced by the particular interests of the parties involved. 

Other intermediaries also have enforcement abilities that they can use at their 

own discretion. For instance, in June 2009, it was reported that a British ISP 

had agreed to disconnect subscribers who were accused of three instances of 

infringement by a copyright owner. 63 Allegations of violations would be made 

by a contractor working for the content owner and transmitted to the ISP. 64 At 

this point, these decisions are largely up to the payment intermediaries and the 

ISPs themselves, although in some jurisdictions they are dictated by government 

requirements, 65 yet their decisions will have profound effects on the shape and 

63 See, e.g., Danny O’Brien, Irish ISP Agrees to Three Strikes Against Its Customers, DEEPLINKS 

BLOG, http://www.eff.org/deeplinks/2009/01/irish-isp-agrees-three-strikes-againstits-users 

(Jan. 28, 2009). 

64 Under the agreement the music labels, instead of going to court to get an order to have the 

ISP shut off a subscriber’s connection, provide evidence of infringement to the ISP directly. 

Id. As O’Brien noted, 

Id. 

The difference is that an ISP is not a court; and its customers will never have 

a chance to defend themselves against the recording industry’s accusations 

and “proof.” To whom, without judicial oversight, has the ISP obligated 

itself to provide meaningful due process and to ensure that the standard of 

proof has been met? 

65 The movement toward graduated response would replace this discretion with government 

processes. Under the recently passed Haute Autorité pour la Diffusion des Œuvres et la 

Protection des Droits sur Internet” (High Authority of Diffusion of the Art Works and 

Protection of the (Copy)Rights on Internet) (“HADOPI”) law, French ISPs would be 

required to suspend Internet access for subscribers who have been subject to three 

allegations of copyright violations. Catherine Saez, French HADOPI Law, Now Complete, Can 

Brandish Its Weapons, INTELL. PROP. WATCH, Oct. 23, 2009, http://www.ip-


direction of electronic commerce. Deferring to the norms of the Internet 

community in this context means deferring to these private judgments by 

intermediaries. 

There is a role for Internet community decision-making. The best 

circumstances for deference to law constructed for and by particular Internet 

communities is when an Internet community’s norms do not “fundamentally 

impinge upon the vital interests of others who never visit this new space.” 66 To 

the extent that an Internet community is self-contained or its activities affect 

others only on a voluntary basis, then there is a case for deferring. 67 

Payment Systems & 

the Bordered Internet 

Goldsmith and Wu attack Internet exceptionalism, but they also construct a 

positive vision of a “bordered Internet.” 68 This world would work pretty much 

as the world worked before the Internet. New regulations would be crafted to 

deal with the new dangers specifically created by the Internet, but there would 

be no fundamental need to adjust the basic domestic or international 

framework. 69 

watch.org/weblog/2009/10/23/french-hadopi-law-now-complete-can-brandish-itsweapons/. 

A court review would be required before suspension. Id. A similar graduated 

response program was adopted in Britain in April 2010. See Eric Pfanner, U.K. Approves 

Crackdown on Internet Pirates, NEW YORK TIMES, April 8, 2010, available at 

http://www.nytimes.com/2010/04/09/technology/09piracy.html. Whether these 

graduated response programs are needed is a point of controversy, but they replace ISP 

discretion with a system of public accountability. 

66 Johnson & Post, supra note 1, at 1389. 

67 See POST, IN SEARCH OF JEFFERSON’S MOOSE, supra note 15, at 178-86 (describing “massively 

multi-player online games” or MMOGS as good candidates for this effort at online rule 

creation). This might be. However, Linden Labs, the creator of Second Life, one of the 

most famous MMOGs, found it necessary to rely on external banking regulators when it 

decided to ban the offering of interest or any return on investment in-world without proof 

of an applicable government registration statement or financial institution charter. Kend 

Linden, New Policy Regarding In-World “Banks”, SECOND LIFE BLOGS, Jan. 8, 2008 06:43:56 

PM, https://blogs.secondlife.com/community/features/blog/2008/01/08/newpolicy-regarding-in-world-banks. 

Linden Labs properly concluded that it “isn’t, and can’t 

start acting as, a banking regulator.” Id. New rule-making institutions will emerge only if 

people think that they are real. For this reason, a policy to defer in certain cases should be 

public and stable in order to provide the opportunity for the development of alternative 

rules. 

68 GOLDSMITH & WU, supra note 7, at viii. 

69 Id. at 149.


Jurisdictional disputes would be one significant problem with the bordered 

Internet. The initial Internet exceptionalist argument was that Internet activity 

is simultaneously present in multiple overlapping and inconsistent jurisdictions, 

and that no one jurisdiction has a better claim to regulate the activity than any 

other jurisdiction. It would be better to think of the activity as taking place in a 

separate jurisdiction altogether and have the territorial governments of the 

world defer to the community norms created there. Goldsmith and Wu’s 

response was that Internet activity was real world activity, taking place in 

particular jurisdictions, and that local governments could exert control over this 

activity by attaching obligations to the local operations of global Internet 

intermediaries. 70 This indirect liability for intermediaries would make it easier to 

extend local law to the bad actor. 71 Conflict of laws would be handled by the 

normal mechanisms for resolving these disputes, and ultimately enforced by 

actions taken against local operations of global intermediaries. 72 

Jurisdiction in cyberspace is a complex topic with many different approaches to 

assigning both the applicable law and the court of jurisdiction. 73 Questions 

include determining the location of the transaction, the jurisdiction, and the 

interests of the parties. 74 An early attempt to deal with these issues in the 

Internet context was the FTC’s approach to consumer protection in the global 

marketplace. 75 The simplest cross-border electronic transaction implicates 

transnational concerns. Choice of law debates inevitably follow. The FTC 

considered arguments for the “country of origin” approach and the “country of 

destination” approach. 76 Under the country of origin approach, the law of the 

70 Id. at 68-72. 

71 Mann & Belzley, supra note 9, at 259 (“[On the Internet it is] easier for even solvent 

malfeasors engaged in high-volume conduct to avoid responsibility either through anonymity 

or through relocation to a jurisdiction outside the influence of concerned policymakers.”). 

Mann and Belzley also argue that indirect liability makes sense in “cases in which the retailer 

is located in a jurisdiction outside the United States that will not cooperate with the relevant 

state regulators.” Id. at 277. 

72 GOLDSMITH & WU, supra note 7, at 158-61. 

73 See, e.g., Paul S. Berman, Towards a Cosmopolitan Vision of Conflict of Laws: Redefining 

Governmental Interests in a Global Era, 153 U. PA. L. REV. 1819, 1822 (2005) (arguing that judges 

should adopt a cosmopolitan approach in Internet cases involving choice of law and foreign 

judgment issues, grounded in the “idea that governments have an interest not only in helping 

in-state litigants win the particular litigation at issue, but a more important long-term interest 

in being cooperative members of an international system and sharing in its reciprocal 

benefits and burdens”). 

74 See generally Goldsmith, supra note 4 (discussing many of these theories); see also Berman, supra 

note 73, at 1839-40 (discussing various choice-of-law theories that address these questions). 

75 FED. TRADE COMM’N, CONSUMER PROTECTION IN THE GLOBAL ELECTRONIC 

MARKETPLACE: LOOKING AHEAD (2000). The FTC’s discussion of applicable law and 

jurisdiction is especially relevant. Id. at 4-11. 

76 Id.


merchant would apply and the courts of the merchant’s country would 

adjudicate any disputes. 77 Under the country of destination approach, the law 

of the consumer would apply and the courts of the consumer’s country would 

adjudicate disputes. 78 

The defense of the country of origin approach relied on the difficulty of 

applying any other legal framework to the electronic marketplace. 79 Only this 

country of origin framework seems to allow for the growth of global ecommerce. 

The framework considers problems encountered by small 

businesses selling in many countries of creating and applying a standard for 

some variety of “purposeful” targeting. Creating a default rule of the country of 

origin was deemed to better provide needed uniformity and predictability for 

online businesses. 

This approach has defects. First, it forces consumers to rely on unfamiliar 

consumer protections. If merchants cannot be expected to know the laws of 

180 countries, neither can consumers. Second, it creates a “race to the bottom,” 

whereby unscrupulous merchants can simply locate in a country with weak 

consumer protections. Third, consumers cannot reasonably be expected to 

travel to the country of origin to obtain redress. Fourth, consumers could not 

rely on their own consumer protection agencies for redress either, since these 

agencies would also be unable to enforce the consumer’s home jurisdiction 

protections. 

So neither default rule seemed to suffice. As a practical matter, consumer 

education, self-regulatory efforts, and the development of codes of conduct by 

multinational organizations were the means chosen to address the cross–border 

consumer protection issue. 80 For other issues that could not be addressed 

77 Id. at 2. 

78 Id. The European Union appeared to take the side of the country of origin in its E- 

Commerce Directive. European Commission, E-Commerce Directive, 

http://ec.europa.eu/internal_market/e-commerce/directive_en.htm (last visited Feb. 

15, 2010). The Directive contains an Internal Market clause “which means that information 

society services are, in principle, subject to the law of the Member State in which the service 

provider is established.” Id. 

79 FED. TRADE COMM’N, supra note 75, at 4 (discussing the “two fundamental challenges” to a 

country-of-destination framework, including “the use of physical borders to determine 

rights in a borderless medium” and compliance costs). 

80 In 1999, the OECD issued its Guidelines for Consumer Protection in the Context of 

Electronic Commerce, which address principles that could be used by electronic commerce 

merchants in the absence of global consumer protection rules. ORG. FOR ECON. CO- 

OPERATION & DEV., GUIDELINES FOR CONSUMER PROTECTION IN THE CONTEXT OF 

ELECTRONIC COMMERCE (1999) [hereinafter OECD GUIDELINES], available at 

http://www.oecd.org/document/51/0,3343,en_2649_34267_1824435_1_1_1_1,00.html. 

The FTC and the OECD held a 10th year anniversary of the release of these guidelines in


through these means, the traditional tools of international conflict of law 

resolution would have to suffice. 81 

Some commentators such as Paul Berman attempted to reach beyond the 

traditional dispute resolution mechanisms for resolving conflict of law cases 

with principles that take into account the realities of multiple community 

affiliations. 82 His “cosmopolitan pluralism” was “cosmopolitan” because it 

went beyond the laws of any one particular jurisdiction and recognized the 

legitimacy of norms created by private parties and communities. 83 It was plural 

because it did not dissolve the multiplicity of community affiliations and their 

associated norms into a single world-wide standard. Diversity and conflict 

would endure and would need to be resolved according to a series of principles 

that recognized the need to balance competing national norms. 84 

These approaches to resolving jurisdictional disputes in cyberspace have various 

advantages and disadvantages. However, payment system intermediaries 

needed a mechanism to address the jurisdictional question that was easy to 

apply, effective in resolving the dispute, and minimized legal risk to the system 

or its members. It could not wait for unpredictable, after-the-fact judgments by 

courts. The idea they developed, discussed in chapter 6 of this book, was that a 

December 2009. OECD, OECD Conference on Empowering E-Consumers, 

http://www.oecd.org/ict/econsumerconference (last visited Sep. 1, 2010). 

81 In an interesting twist, some commentators used the presence of these dispute resolution 

mechanisms to argue against indirect liability for intermediaries. Why deputize intermediaries 

to stop illegal activities on the Internet when governments can reach the bad actors and 

resolve any disputes in the normal way? Responding to the argument that indirect liability is 

needed because the bad actor is unreachable by law enforcement or aggrieved parties, 

Holland says: 

As an initial matter, it is not clear that a significant number of bad actors are 

beyond the reach of the law. Advances in technology are making it 

increasingly possible to locate and identify bad actors online, such that online 

anonymity is difficult to maintain. Likewise, where the bad actor is identified 

but is found outside the jurisdiction, sovereign governments have developed 

methods for resolving disputes to permit the direct extraterritorial 

application of domestic law, such as rules of jurisdiction, conflicts of laws, 

and recognition of judgments. 

Holland, supra note 16, at 393. 

82 Berman, supra note 73, at 1862. 

83 Id. 

84 Id. Berman’s work has affinities with that of political philosophers working in the area of 

national sovereignty in a global world. See, e.g., Thomas W. Pogge, WORLD POVERTY 

AND HUMAN RIGHTS 168-95 (2002).


transaction is unacceptable in the payment system if it is illegal in the 

jurisdiction of either the buyer or the seller. 85 

The payment card approach provides a simple default rule for intermediaries to 

apply when determining whether to allow transactions in their systems. It 

eliminates the heavily fact-based balancing assessments needed to determine, on 

a case-by-case basis, whose law applies. The default rule also does not simply 

adopt a country of origin or country of destination perspective, each of which is 

limited. Nor does it leave the transaction in a legal limbo where no law 

applies. 86 

The payment system experience leads to several observations. First, direct 

conflicts of law are not as frequent as some anticipated. Technology and 

payment system practices effectively reduce these conflicts to the rare instance 

85 Visa’s policy is stated in International Piracy: The Challenges of Protecting Intellectual Property in the 

21st Century: Hearing Before the Subcomm. on Courts, the Internet, and Intellectual Property of the H. 

Comm. on the Judiciary, 110th Cong. 73–82 (2007) at 71 (statement of Mark MacCarthy, Senior 

Vice President for Global Public Policy, Visa Inc.). Other payment intermediaries have 

similar procedures, such as eBay’s restriction about selling and shipping illegal goods to the 

country where they are illegal. eBay, Offensive Material Policy, 

http://pages.ebay.com/help/policies/offensive.html (last visited Feb. 4, 2010) 

(“[B]ecause eBay is a worldwide community, many of our users live in countries where the 

possession or sale of items associated with hate organizations is a criminal offense. We can’t 

allow the sale or shipping of these items there.”). 

86 The internal application of this rule involves system efficiency and the balance of interests 

among the stakeholders in the system. If the merchant is in violation of its own country’s 

law, then enforcement is conceptually easy. Merchants discovered in violation of local law 

either have to stop the transactions or be removed from the system. If the merchant is in 

violation of the law in a different jurisdiction, things are more complicated. Should the bank 

of the merchant or the bank of the customer be burdened with the enforcement 

responsibility? If the merchant has this responsibility, then he must not introduce the illegal 

transaction into the system and the merchant’s bank must not try to process it, then steps 

must be taken at the merchant’s end to stop the transaction. These steps could include: a 

system decision requiring the merchant to stop these transactions entirely; coding and 

programming modifications by the merchant, the merchant’s processor, or the system 

operator that would block transactions at the merchant end from entering the system if the 

customer was from a jurisdiction where the transaction would be illegal; or restricting the 

transaction to the merchant’s own jurisdiction. Alternatively, the enforcement measures 

could be put on the cardholder side. Merchants could introduce properly-coded transactions 

into the system and rely on action on the cardholder’s side to stop the transaction. This 

seems to fit the case of Internet gambling, where U.S. law makes Internet gambling illegal 

for U.S. citizens, and the payment networks responded to the Unlawful Internet Gambling 

Enforcement Act of 2006 (UIGEA) with a coding and blocking system that allowed 

merchants to continue their services in countries where Internet gambling was illegal, as 

discussed earlier in this Article. For instance, should merchants be responsible for knowing 

the laws of all the countries of all the customers they deal with? Perhaps not, but if 90% of 

their sales are from an offshore jurisdiction, they should be responsible for knowing that 

sales of their product are legal in that jurisdiction. Violations of the policy would largely be 

dealt with on a complaint basis.


where the law of one country demands what the law of another country forbids. 

Directly contradicting laws are more common in “political” areas, where 

governments are seeking information from intermediaries to enforce local laws 

against their own citizens. 87 

Second, regulating the Internet by focusing on the local affiliates of global 

payment operations does not require the use of either the traditional or the new 

“cosmopolitan” conflict resolution methods. By relying on global payment 

intermediaries, local jurisdictions reach out to the local affiliates that are totally 

within their jurisdiction. They do not put burdens on entities in foreign 

jurisdictions at all. There is literally no conflict and thus nothing to which 

normal mechanisms of conflict resolution may attach. 88 

Some commentators have correctly pointed out that when the laws of different 

jurisdictions apply to a single transaction, the ability of any particular jurisdiction 

to unilaterally regulate the Internet is limited. 89 But intermediaries can reduce 

these conflicts. Global payment systems can simplify transactions to events in 

which only a buyer in one jurisdiction and a seller in another are implicated. By 

concentrating enforcement on intermediaries instead of individuals or 

merchants, local jurisdictions can take advantage of the economies that these 

institutions make possible. 

The experience of payment intermediaries reveals that, within limits, the 

differences among conflicting jurisdictions can be managed. The bordered 

87 See, e.g., Press Release, Privacy Int’l, Europe’s Privacy Commissioners Rule Against SWIFT 

(Nov. 23, 2006), available at 

http://www.privacyinternational.org/article.shtml?cmd[347]=x-347-546365 

(describing the SWIFT case, where SWIFT was required to comply with U.S. demands for 

access to financial information about European customers in virtue of its operations on US 

soil, while such compliance put them in violation of the European data protection directive). 

In addition, passage of the Global Online Freedom Act (GOFA) could put Internet 

intermediaries in a conflict of law situation with China and other countries. See Global 

Online Freedom Act of 2007, H.R. 275, 110th Cong. (2007). H.R. 275 was introduced by 

Representative Chris Smith on January 5, 2007 and would require U.S. intermediaries to resist 

certain orders from countries in which they are doing business. Id. 

88 Antigua brought a complaint against the U.S. for the enforcement of its gambling laws, but 

its success was based only on (1) the U.S.’s failure to exclude Internet gambling from the list 

of services that required open treatment and (2) the idiosyncrasies of U.S. gambling law 

which appear to allow domestic horse racing to engage in Internet gambling while denying 

similar opportunities to offshore Internet gambling merchants. But these are technical 

obstacles created by the interaction of complex U.S. law and international WTO law and are 

not real conflict of law problems. See Appellate Body Report, United States—Measures 

Affecting the Cross-Border Supply of Gambling and Betting Services, 358-64, WT/DS285/AB/R 

(Apr. 7, 2005). Op cit. supra note 130. 

89 See, e.g., H. Brian Holland, The Failure of the Rule of Law in Cyberspace?: Reorienting the Normative 

Debate on Borders and Territorial Sovereignty, 24 J. MARSHALL J. COMPUTER & INFO. L. 1, 26 

(2005).


Internet works on a small scale. The scale is currently small for two reasons: 

First, he number of cases of governments reaching across borders to inflict their 

laws on Internet merchants in other jurisdictions is still relatively small. 

Moreover, in contrast to the rhetoric about the Internet creating a global 

marketplace, the scope of cross-border commerce itself is still limited. The 

reality is that the volume of cross-border transactions is not large enough to 

create a truly substantial cross-border jurisdictional crisis. Currently, only four 

percent of the sales for electronic commerce merchants in the U.S. come from 

abroad. 90 And data from Europe show that cross border online transactions are 

not increasing as fast as overall e-commerce transactions, staying relatively 

stable from 2006 to 2008 at six to seven percent. 91 

As David Post has warned, the problem the Internet creates for local 

jurisdictions is one of scale. 92 The bordered Internet simply does not scale up. 

Global payment systems cannot accommodate an enforcement burden in which 

each jurisdiction uses payment system mechanisms to enforce each of its local 

laws on the Internet. 

It is not hard to see how we can get into a kind of tragedy of the commons in 

this area. Each individual extension of local jurisdiction into cyberspace seems 

small and costless, but collectively the burden becomes unbearable. 

Governments might feel free to exploit this enforcement mechanism, in the 

same way that grazers use the commons—under the impression that it is an 

unlimited resource. However, one of two outcomes will occur as the crossborder 

rules pile up: Either cross-border transactions will remain small and the 

potential for the Internet to be a global channel of commerce will not be 

realized, or the political costs of each government attempting to regulate the e- 

90 This is based on transaction data from the Visa system. See International Piracy Hearing, supra 

note 85, at 75 (statement of Mark MacCarthy, Senior Vice President for Global Public 

Policy, Visa Inc.). 

91 Comm’n of the European Cmtys., Commission Staff Working Document: Report on Cross-Border 

E-commerce in the EU 3, SEC (2009) 283 final (Mar. 5, 2009), available at 

http://ec.europa.eu/consumers/strategy/docs/com_staff_wp2009_en.pdf (“From 

2006 to 2008, the share of all EU consumers that have bought at least one item over the 

Internet increased from 27% to 33% while cross-border e-commerce remained stable (6% to 

7%).”). 

92 See Post, Against “Against Cyberanarchy”, supra note 15, at 1377 (stating that “scale matters”); see 

also Holland, supra note 89, at 29. Holland states: 

Id. 

The online actor cannot know, as a practical matter, the many laws applicable 

to a particular act, nor when one or more sovereigns may decide to attempt 

regulatory action. This is particularly true in those areas of regulation in 

which morality, religion and culture are at their most influential, such as 

speech, race, sex, and even intellectual property. Moreover, it is not simply 

one actor or a few legal systems. It is an exponential multitude.


commerce activities of other countries will mount. Either development reveals 

the limitations of the bordered Internet as a long-term framework for Internet 

governance. 

Goldsmith and Wu suggest that enforcement of Internet regulations through 

intermediaries is necessarily limited in size. 93 They suggest that maybe the 

system will not be able to scale up, but it won’t have to. 94 Small countries such 

as Antigua cannot enforce Internet rules because global intermediaries can 

simply pull up stakes and leave if the rules are too strict. 95 However, there are a 

sufficiently large number of countries that global intermediaries will not feel 

capable of abandoning. If all of them use the intermediary enforcement 

mechanism, the system will be overwhelmed. 

Internationalism 

The fundamentally correct insight of the Internet exceptionalists is that the 

unilateral imposition of one nation’s law onto all Internet activities that cross 

borders won’t scale. 96 

Internationalism might be the way out. It is the idea that the Internet will 

eventually be governed, at least for some services, by global institutions and 

arrangements, and that this is the right public policy for local governments to 

follow in their dealings with illegal cross border Internet transactions. 97 This 

policy could be implemented through a uniform global standard, or any of a 

variety of techniques such as World Trade Organization rules that bring local 

laws into harmony. The basic justification for this policy is similar to the 

justification for establishing a single uniform national policy that prevents the 

clash of inconsistent rules at the state level: When activities have widespread 

and significant effects on those outside the local jurisdiction, then uniform 

principles or some other coordinating mechanism should be adopted at the 

higher level. 98 This universalism could promise better laws, whereby the 

93 GOLDSMITH & WU, supra note 7, at 81-82. 

94 Id. at 81. 

95 See id. at 160 (suggesting that acting as the Internet police is just a normal cost of doing 

business for global companies, which they can avoid in a particular case by leaving a country 

that tried to impose costs that exceeded the benefits of continued presence in the country 

and thus creating another objection to the bordered Internet to effectively give larger 

countries a greater role in Internet governance than smaller ones). 

96 See Johnson & Post, supra note 1, at 1390 (“One nation’s legal institutions should not 

monopolize rule-making for the entire Net.”). 

97 GOLDSMITH & WU, supra note 7, at 26. 

98 Id. (“If the nations of the world agree to a single global law for questions like libel, 

pornography, copyright, consumer protection, and the like, the lives of Internet users


“[i]nternational standards could reflect a kind of collection of best practices 

from around the world — the opposite of the tyranny of the unreasonable.” 99 

Goldsmith and Wu make several criticisms of internationalism. First, a system 

of universal laws would be unattractive; it would leave the world divided and 

discontent because the universal law would be unpopular in large segments of 

the world population. Second, the system of local national laws would better 

reflect differences among people. Diversity is a good thing and cannot be taken 

into account by a universal code that overrides local differences. Third, it is not 

needed. The conflicts of laws, extraterritoriality, and other considerations are 

perfectly manageable within the current international framework. For example, 

since most Internet users do not have assets in other countries, they are 

effectively subject only to the laws of the country where they live. Only large 

multinational companies with assets all over the world face the 

multijurisdictional problem, and they already have to live with that because they 

are already global. Compliance with a plurality of international laws is simply a 

cost of doing business for global companies. There’s nothing new here that 

would justify a move to a more harmonized global order. There are extra costs 

to be sure, but nothing so onerous or burdensome that it would require a move 

to global law. 100 

The responses to these criticisms are straightforward. An unpopular global law 

is not the goal. Neither is suppression of diversity the goal. The idea is to 

integrate local laws in some fashion when the regular conflicts among them 

prove to be intolerable. When diversity does not create this difficulty, there is 

no need for integration. If, for example, local governments value diversity 

enough to refrain from using intermediaries to enforce local laws against actors 

in other jurisdictions, then there is no need for harmonization of these 

enforcement efforts. But to the extent that governments want to take global 

enforcement steps, they also need to take steps to integrate the laws they want 

become much simpler: no conflicting laws, no worries about complying with 175 different 

legal systems, no race to the bottom.”). 

99 Id. at 27. Reidenberg also argues that as jurisdictions increasingly conflict there will need to 

be an overarching harmonization of international rules: 

[O]nline enforcement with electronic blockades and electronic sanctions will 

cause serious international political conflicts. These conflicts arise because of 

the impact on territorial integrity. Such conflicts are likely to force 

negotiations toward international agreements that establish the legal criteria 

for a state to use technological enforcement mechanisms. This progression 

leads appropriately to political decisions that will define international legal 

rules. 

Joel R. Reidenberg, States and Internet Enforcement, 1 U. OTTAWA L. & TECH. J. 213, 

230 (2003-2004). 

100 See GOLDSMITH & WU, supra note 7, at 152-60.


to enforce. The reason for this is that global intermediaries’ costs to mediate 

the conflicts associated with unilateral attempts at local regulation of the 

Internet will be so onerous and burdensome that they will cause an unwarranted 

and unnecessary decline in global interaction. 101 

Berman also describes how the internationalist hope for global standards avoids 

the conflict of law problem: “if we constructed one universal ‘world 

community’ with one set of governing rules, there would never need to be a 

‘choice of law’ in the sense that conflict-of-laws scholars use the term.” 102 

However, he is critical of this universal world community for two reasons. 

First, he is critical of this community because of its potential to dissolve 

community affiliations that provide important emotional connections and 

opportunities for normative discussion of those connections. Second, he views 

this universal community as fundamentally unrealistic given the dominance of 

current notions of nation-state sovereignty. 103 

These objections can be met at the level of generality at which they are cast. We 

do not need to think of ourselves as primarily world citizens in order to endorse 

specific global approaches. We can still have deep attachments to local 

communities and can still debate the relative importance of the overlapping 

communities we participate in. The global approach endorses the view that 

self-government “requires a politics that plays itself out in a multiplicity of 

settings, from neighborhoods to nations to the world as a whole” and “citizens 

who can abide the ambiguity associated with divided sovereignty, who can think 

and act as multiply situated selves.” 104 But participation in global community 

and the wisdom to know when the global perspective should take precedence 

over more local concerns is essential to this vision of self-government in a 

global world. 

The internationalist proposal is to provide global coordination only when 

necessary. It is to move to global standards when, as a practical matter, the 

burdens of allowing diverse local rules are too high. The model of national 

uniform standards is appropriate: not everything has to be done at the national 

level, but some things should be done there in order to have an efficient and fair 

national system. Similarly, there is no need to move from the current system to 

101 Interestingly, the earlier Jack Goldsmith seemed more inclined to accept these practical 

considerations as a rationale for international harmonization: “When in particular contexts 

the arbitrariness and spillovers become too severe, a uniform international solution remains 

possible.” Goldsmith, supra note 4, at 1235. 

102 Berman, supra note 73, at 1860. 

103 Id. at 1860-61. 

104 MICHAEL J. SANDEL, PUBLIC PHILOSOPHY: ESSAYS ON MORALITY IN POLITICS 34 (2005).


a world government. But if there are practical ways to improve Internet 

governance through global harmonization, they should be taken. 

If governments are going to use payment intermediaries as enforcers of local 

law, there are a number of steps that could be taken to coordinate their efforts, 

including: 

� In the Internet gambling context, a move to an internationallyinteroperable 

licensing system that would require each jurisdiction that 

allows Internet gambling to defer to the licensing decisions of other 

jurisdictions 

� In the copyright context, the continued evolution of uniform copyright 

rules. 

International agreements are one mechanism to create coordinated action. 

Although controversial because of the secrecy involved in its development, and 

the sense that affected parties were excluded from participation, the Anti- 

Counterfeiting Trade Agreement (ACTA) is a reasonable, though flawed, model 

for action in this area. 105 There are many mechanisms for international 

coordination. Decisions regarding which mechanisms to use depend on the 

issue and the fora available for resolution. 

Internationalism has its dangers. Why should each jurisdiction have the same 

regulations on hate speech and the same regulations on alcohol consumption? 

The answer is that there will be no harmonization where there are such 

fundamental differences. Intermediaries will be called upon to resolve the issue 

themselves or they will be caught between warring governments and forced to 

choose sides. But efforts should be made to minimize such differences when 

these differences have global consequences, especially when they are superficial 

differences that reflect no fundamental divisions. For the same reason that we 

want uniform global technical standards for information and communications 

technologies, if possible, we want similar legal frameworks if governments are 

going to enforce laws on the Internet. 

These efforts to ease the friction involved in extending government authority to 

the Internet through a global framework are in line with other efforts to create 

global frameworks that promote the growth of the Internet. For example, the 

thirty-first International Conference of Data Protection and Privacy 

Commissioners, held in Madrid in November 2009, adopted a set of global 

105 See Media Statement, Participants in ACTA Negotiations, Anti-Counterfeiting Trade Agreement 

(ACTA), June 12, 2009, available at 

http://www.med.govt.nz/templates/Page____40974.aspx. For a summary of the 

ACTA process and the content of the agreement, see THE ANTI-COUNTERFEITING TRADE 

AGREEMENT – SUMMARY OF KEY ELEMENTS UNDER DISCUSSION (2009), available at 

http://www.med.govt.nz/templates/MultipageDocumentTOC____40563.aspx.


privacy standards. 106 There is also likely to be a renewed push for global 

consumer protection on the occasion of the tenth anniversary of the 

Organisation for Economic Co-operation and Development’s Guidelines for 

Consumer Protection in the Context of Electronic Commerce. 107 

Both these efforts relate to the growth of the Internet as a vibrant international 

marketplace. They do this by building online trust. Global information security 

standards reassure people that their information is safe no matter what the 

physical location of the websites they visit. Establishing global privacy 

standards means that the collection and use of online information will be 

governed by common principles regardless of a website’s jurisdiction and will 

make it easier for global business to transfer information from one jurisdiction 

to another in a seamless manner. Finally, effective global consumer protection 

rules will mean that people will have the information and redress rights they 

need to shop confidently online no matter where the website is located. 

Conclusion 

The initial demand from Internet exceptionalists that the online world be left 

alone by governments has morphed into the idea that governments should 

create a global framework to protect and spur the growth of the Internet. The 

intervening steps in this development are not hard to trace: Internet 

exceptionalists confused their ideal of self-governing Internet communities with 

the idea that the Internet was ungovernable because it was a global 

communications network that crossed borders. This idea of an intrinsically 

ungovernable Internet was undermined by the recognition that the coding that 

underlies Internet applications and services is a matter of choice, not 

106 Artemi R. Lombarte, Dir., Agencia Española de Protección de Datos, Slide Presentation: 

International Standards on Data Protection & Privacy (2009), available at 

https://www.agpd.es/portalweb/canaldocumentacion/comparecencias/common/I 

APP_Privacy_Summit_09.pdf. He describes one of the main criteria of the global privacy 

standards project as “To elaborate a set of principles and rights aimed to achieve the 

maximum degree of international acceptance, ensuring at once a high level of protection.” Id. 

(emphasis in original). For the standards adopted, see THE MADRID PRIVACY DECLARATION 

(Nov. 3, 2009), http://thepublicvoice.org/TheMadridPrivacyDeclaration.pdf. 

107 OECD GUIDELINES, supra note 80; see also Org. for Econ. Co-operation & Dev., Conference 

on Empowering E-Consumers: Strengthening Consumer Protection in the Internet 

Economy, Programme (2009), available at 

http://www.oecd.org/dataoecd/33/22/44045376.pdf (describing the conference). The 

OECD endorsed steps toward global enforcement of some consumer protection rules in a 

2003 report on cross-border fraud and a 2007 report on consumer dispute resolution and 

redress. See Comm. on Consumer Policy, Org. for Econ. Co-operation & Dev., OECD 

Guidelines for Protecting Consumers from Fraudulent and Deceptive Commercial Practices 

Across Borders (2003), available at http://www.oecd.org/dataoecd/24/33/2956464.pdf; 

Comm. on Consumer Policy, Org. for Econ. Co-operation & Dev., OECD Recommendation 

on Consumer Dispute Resolution and Redress (2007), available at 

http://www.oecd.org/dataoecd/43/50/38960101.pdf.


unchangeable nature. If something about this system created difficulties for 

government control, this could be changed. Further, the idea that governments 

cannot control the Internet was undermined by the need for the local 

operations of global intermediaries to provide essential Internet services and the 

practical ability of governments to control these intermediaries. 

Internet intermediaries can control the content of the activities on their online 

communities, and government can compel or pressure intermediaries to take 

these steps. Intermediaries have a general obligation to follow the law, and 

except in extreme cases, they have no right to resist these lawfully established 

burdens. The establishment of these laws needs to follow all the rules of good 

policymaking, including imposing an obligation only when the social benefits 

exceed the social costs. However, a bordered Internet in which each country 

attempts to use global intermediaries to enforce its local laws will not scale. 

This is the fundamentally correct insight of the Internet exceptionalists. If 

governments are going to use intermediaries to enforce local laws, they are 

going to have to harmonize the local laws they want intermediaries to enforce.

CHAPTER 4 

HAS THE INTERNET FUNDAMENTALLY 

CHANGED ECONOMICS? 

Computer-Mediated Transactions 239 

Hal R. Varian 

237 

Decentralization, Freedom to Operate 

& Human Sociality 257 



From Dismal Science to Strange Tales 273 

Larry Downes 

The Regulation of Reputational Information 293 

Eric Goldman

238 CHAPTER 4: HAS THE INTERNET FUNDAMENTALLY CHANGED ECONOMICS?


Computer-Mediated Transactions 

By Hal R. Varian * 

Every now and then a set of technologies becomes available that sets off a 

period of “combinatorial innovation.” Think of standardized mechanical parts 

in the 1800s, the gasoline engine in the early 1900s, electronics in the 1920s, 

integrated circuits in the 1970s, and the Internet in the last few decades. 

The component parts of these technologies can be combined and recombined 

by innovators to create new devices and applications. Since these innovators 

are working in parallel with similar components, it is common to see 

simultaneous invention. There are many well-known examples such as the 

electric light, the airplane, the automobile, and the telephone. Many scholars 

have described such periods of innovation using terms such as “recombinant 

growth,” “general purpose technologies,” “cumulative synthesis” and “clusters 

of innovation.” 1 

The Internet and the Web are wonderful examples of combinatorial innovation. 

In the last 15 years we have seen a huge proliferation of Web applications, all 

built from a basic set of component technologies. 

The Internet itself was a rather unlikely innovation; I like to describe it as a “lab 

experiment that got loose.” Since the Internet arose from the research 

community rather than the private sector, it had no obvious business model. 

Other public computer networks, such as AOL, CompuServe, and Minitel, 

generally used subscription models, but were centrally controlled and offered 

little scope for innovation at the user level. The Internet won out over these 

alternatives, precisely because it offered a flexible set of component 

technologies that encouraged combinatorial innovation. 

The earlier waves of combinatorial innovation required decades, or more, to 

play out. For example, David Hounshell argues that the utopian vision of 

* Univ. of Cal., Berkeley and Google. hal@ischool.berkeley.edu. 

1 See, e.g., Martin Weitzman, Recombinant Growth, 113 Q. J. OF ECON. 331-360 (1998); Timothy 

Bresnahan & M. Trajtenberg, General Purpose Technologies: Engines of Growth?, 65 J. OF 

ECONOMETRICS 83-108 (1995), available at 

http://ideas.repec.org/a/eee/econom/v65y1995i1p83-108.html; Timothy Bresnahan, 

General Purpose Technologies, in HANDBOOK OF THE ECONOMICS OF INNOVATION (Bronwyn 

Hall & Nathan Rosenberg, eds., 2010); Nathan Rosenberg, Technological Change in the Machine 

Tool Industry, in PERSPECTIVES IN TECHNOLOGY 9-31 (1976); ABBOTT PAYSON USHER, A 

HISTORY OF MECHANICAL INVENTION (revised ed., Dover Publ’ns 1998); and Joseph A. 

Schumpeter, The Analysis of Economic Change, in ESSAYS ON ENTREPRENEURS, INNOVATIONS, 

BUSINESS CYCLES AND THE EVOLUTION OF CAPITALISM 134-149 (Richard V. Clemence, ed., 

2000) (originally published in Review of Economic Statistics, May 1935).

240 CHAPTER 4: HAS THE INTERNET FUNDAMENTALLY CHANGED ECONOMICS? 

interchangeable parts took more than a century to be realized. 2 The Web was 

invented in the early 1990s, but it did not become widely used until the mid- 

1990s. Since then, we have seen a huge number of novel applications—from 

Web browsers, to search engines, to social networks—to mention a few 

examples. As with the Internet, the Web initially had no real business model, 

but offered a fertile ground for combinatorial innovation. 

Innovation was so rapid on the Internet because the component parts were all 

bits. They were programming languages, protocols, standards, software 

libraries, productivity tools and the like. There was no time to manufacture, no 

inventory management, and no shipping delay. You never run out of HTML, 

just like you never run out of email. New tools could be sent around the world 

in seconds and innovators could combine and recombine these bits to create 

new Web applications. 

This parallel invention has led to a burst of global innovation in Web 

applications. While the Internet was an American innovation, the Web was 

invented by an Englishman living in Switzerland. Linux, the most used 

operating system on the Web, came from Finland 3, as did MySQL, a widely 

used database for Web applications. 4 Skype, which uses the Internet for voice 

communication, came from Estonia. 5 

Of course, there were many other technologies with worldwide innovation, 

such as automobiles, airplanes, photography, and incandescent lighting. 

However, applications for the Internet, which is inherently a communications 

technology, could be developed everywhere in the world in parallel, leading to 

the rapid innovation we have observed. 

Computer-Mediated Transactions 

My interest in this essay is in the economic aspects of these technological 

developments. I start with a point so mundane and obvious, it barely seems 

worth mentioning: Nowadays, most economic transactions involve a computer. 

2 David A. Hounshell, From the American System to Mass Production, 1800-1932: The 

Development of Manufacturing Technology in the United States (1985). 

3 See Linus Torvalds & David Diamond, Just for Fun: The Story of an Accidental Revolution 

(2002). 

4 See Oracle Corporation, From Visions to Reality: An Interview with David Axmark, Co-Founder of 

MySQL AB, July 2007, http://dev.mysql.com/tech-resources/interviews/davidaxmark.html. 

5 See Andreas Thomann, Skype: A Baltic Success Story, CREDIT SUISSE GROUP, June 9, 2006, 

http://emagazine.creditsuisse.com/app/article/index.cfm?fuseaction=OpenArticle&aoid=163167&coid=78 

05&lang=EN.


Sometimes this computer takes the form of a smart cash register, part of a 

sophisticated point of sale system, or a website. In each of these cases, the 

computer creates a record of the transaction. 

This record-keeping role was the original motivation for having the computer as 

part of the transaction. Creating a record of transactions is the first step in 

building an accounting system, thereby enabling a firm to understand its 

financial status. 

Now that computers are in place, they can, however, be used for many other 

purposes. In this essay, I explore some of the ways that computer mediation 

can affect economic transactions. These computer mediated transactions, I 

argue, have enabled significant improvements in the way transactions are carried 

out and will continue to impact the economy in the foreseeable future. 

I classify the impact of computer mediated transactions into four main 

categories according to the innovation they facilitate: 

� New forms of contract; 

� Data extraction and analysis; 

� Controlled experimentation; 

� Personalization and customization. 

Enable New Forms of Contract 

Contracts are fundamental to commerce. The simplest commercial contract 

says, “I will do X if you do Y,” as in “I will give you $1 if you give me a cup of 

coffee.” Of course, this requires that the actions to be taken are verifiable. Just 

asking for coffee does not mean that I will get it. As Abraham Lincoln 

supposedly remarked, “If this is coffee, please bring me some tea; but if this is 

tea, please bring me some coffee.” 6 

A computer used in a transaction can observe and verify many aspects of that 

transaction. The record produced by the computer allows the contracting 

parties to condition the contract on terms that were previously unobservable, 

thereby allowing for more efficient transactions. 

I am not claiming that increased observation will necessarily lead to more 

efficient contracts. There are counterexamples to the assertion that “more 

6 Susan L. Rattiner, Food and Drink: A Book of Quotations (2002).


information is better” such as the Hirshleifer example. 7 I am merely claiming 

that additional information allows for more efficient contracts. 

Of course, the study of contracts is a highly developed field in economics. As 

such, it is hardly novel to suggest that contractual form depends on what is 

observable. What is interesting, however, is the way that progress in 

information technology enables new contractual forms. 

Consider, for example, a rental-car agency that buys insurance based on 

accident rates, and that accident rates, in turn, depend on the speed of a vehicle. 

All renters would prefer to drive within the speed limit if they are compensated 

with a lower rental fee. However, if there is no way to monitor the speed of a 

rental car, such a contractual provision is unenforceable. Putting a computer 

transmitter in the trunk of the car that records the vehicle’s speed makes the 

contract enforceable and potentially makes everyone better off.. 8 

The transportation sector has capitalized on the availability of computerized 

transmitters to create more efficient contracts in many areas. 

� Car dealers are selling cars with “starter interrupt” devices that inhibit 

operations if car payments are missed. 9 

� Similar interrupt devices attached to breath analyzers are mandated for 

drunk driving offenders in many states. 

� Parents can buy a device known as “MyKey” which allows them to 

limit auto speed, cap the volume on the radio, require seat belt use and 

encourage other safe-driving habits for teenage drivers. 10 

� In the relevant economics literature, Hubbard and Baker examine a 

variety of ways that vehicular monitoring systems have impacted the 

trucking industry. 11 

7 Jack Hirshleifer, The Private and Social Value of Information and the Reward to Inventive Activity, 61 

THE AM. ECON. REV. 561-74 (Sept. 1971), available at 

http://faculty.fuqua.duke.edu/~qc2/BA532/1971%20AER%20Hirshleifer.pdf. 

8 This is a particularly simple case. If drivers have heterogeneous preferences, those who 

prefer to speed may be made worse off by the availability of such a device. 

9 Associated Press, For Some High-risk Auto Buyers, Repo Man is a High-tech Gadget, L.A. TIMES, 

June 13, 2006, http://articles.latimes.com/2006/jun/13/business/fi-late13. 

10 Nick Bunkley & Bill Vlasic, Ensuring Junior Goes for a Mild Ride, N.Y. TIMES, Oct. 6, 2008, 

http://www.nytimes.com/2008/10/07/automobiles/07auto.html. 

11 Thomas N. Hubbard, The Demand of Monitoring Technologies: The Case for Trucking, 115 Q. J. OF 

ECON. 533-560 (2000), 

http://www.mitpressjournals.org/doi/abs/10.1162/003355300554845; George Baker & 

Thomas N. Hubbard. Contractibility and Asset Ownership: On-board Computers and Governance in 

US Trucking, 119 Q. J. OF ECON. 1443-1479 (2004), 

http://www.mitpressjournals.org/doi/abs/10.1162/0033553042476152.


There are many other examples of computer-mediated contracts. The work of 

Dana & Spier and Mortimer, provides examples that describe the efficiency 

gains resulting from revenue sharing in the video tape rental industry. 12 

Video tapes were originally purchased by retail stores from distributors for 

about $65 per tape. Since the videos were so expensive, stores only bought a 

few. As a result, the popular videos quickly disappeared from the shelves, 

making everyone unhappy. 

In 1998, retailers and distributors adopted a new business model: a revenue 

sharing arrangement in which stores paid a small upfront fee of $3 to $8, but 

split the revenue when the video was rented, with 40% to 60% going to the 

retailer. Stores no longer had an incentive to economize on purchases, and all 

parties to the transaction—retailers, distributors, and customers—were made 

better off. 

Sharing revenue at point of sale requires that both parties be able to monitor the 

transaction. The technological innovations of bar code scanning, the 

computerized cash register, and computer networks enabled revenue-sharing 

arrangements. 

Of course, when a transaction takes place online, revenue-sharing is much 

easier. Online advertising is a case in point where revenue from an advertiser 

for an ad impression or click may be split among publishers, ad exchanges, ad 

networks, affiliates and other parties based on contractual arrangements. 

Although the benefits from computers offering more information to contracting 

parties have only been discussed thus far, there are also cases in which 

computers can be used to improve contractual performance by hiding 

information using cryptographic methods. A picturesque example is the 

“cocaine auction protocol” which describes an auction mechanism designed to 

hide as much information as possible. 13 

Finally, “algorithmic game theory” is an exciting hybrid of computer science 

and economic theory that deserves mention. This subject brings computational 

considerations to game theory (how a particular solution can be computed) and 

12 James D. Dana & Kathryn E. Spier. Revenue Sharing and Vertical Control in the Video Rental 

Industry, XLIX Q. J. OF ECON. 223-245 (2001), available at 

http://www3.interscience.wiley.com/journal/118972449/abstract; Julie H. Mortimer, 

Vertical Contracts in the Video Rental Industry, 75 REV. OF ECON. STUDIES 165-199 (2008), 

http://www3.interscience.wiley.com/journal/119395822/abstract. 

13 Frank Stajano & Ross Anderson, The Cocaine Auction Protocol: On the Power of Anonymous 

Broadcast, in PROCEEDINGS OF INFORMATION HIDING WORKSHOP, LECTURE NOTES IN 

COMPUTER SCIENCE, 1999, http://www.cl.cam.ac.uk/ rja14/Papers/cocaine.pdf.


strategic considerations to algorithm design (whether a particular algorithm is 

actually incentive-compatible). 14 

Some History of Monitoring Technologies 

Though I have emphasized computer mediated transactions, a computer can be 

defined quite broadly. The earliest example of an accounting technology I 

know of that enabled new forms of contract involves Mediterranean shipping 

circa 3300 B.C. 

The challenge was how to write a “bill of lading” for long distance trade in 

societies that were pre-literate and pre-numerate. The brilliant solution was to 

introduce small clay tokens, known as “bullae,” which were small 

representations of the material being transported. As each barrel of olive oil 

was loaded onto a ship, a barrel-shaped token was placed in a clay envelope. 

After the loading was completed, the envelope was baked in a kiln and given to 

the ship’s captain. At the other end of the voyage, the envelope was then 

broken open and the tokens were compared to the barrels of oil on the ship as 

they were unloaded. If the numbers matched, the contract was verified. Later, 

marks were scratched on the outside of the envelope to indicate the number of 

tokens inside. Some authors believe that this innovation led to the invention of 

writing between 3400 and 3300 B.C. 15 

A somewhat more recent example is the invention of the cash register in 1883 

by James Ritty. 16 Ritty, a saloon owner, discovered that his employees were 

stealing money. In response, he developed a device to record each transaction 

on paper tape, an invention that he patented under the name of “the 

incorruptible cashier.” 17 Ritty’s machine became the basis of the National Cash 

Register (NCR) Company founded in 1884. The NCR device added a cash 

drawer and a bell that sounded “ka-ching” whenever the drawer was opened, to 

alert the owner of the transaction, thereby discouraging pilfering. This 

improved monitoring technology made retailers willing to hire employees 

14 See NOAM NISAN, TIM ROUGHGARDEN, EVA TARDOS, AND VIJAY V. VAZIRANI, EDS., 

ALGORITHMIC GAME THEORY (2007) for a comprehensive collection of articles and Hal R. 

Varian, Economic Mechanism Design for Computerized Agents, in USENIX WORKSHOP ON 

ELECTRONIC COMMERCE 13-21 (1995), 

http://www.sims.berkeley.edu/hal/Papers/mechanism-design.pdf for an early 

contribution to this theory. 

15 Jean-Jacques Glassner, Zainab Bahrani, & Marc Van de Miero, The Invention of Cuneiform: 

Writing in Sumer (2005). 

16 MIT School of Engineering, Inventor of the Week Archive: James Ritty, Cash Register, April 2002, 

http://web.mit.edu/invent/iow/ritty.html. 

17 Cash Register and Indicator, U.S. Patent 271,363 (filed Feb 15, 1882), available at 

http://www.google.com/patents?hl=en&lr=&vid=USPAT271363.


outside their immediate families, leading to larger and more efficient 

establishments. 18 

Enabling Online Advertising 

Online advertising serves as a poster child for algorithmic mechanism design. A 

Pasadena company called GoTo began ranking search results using an auction. 19 

Users did not like this particular form of search, so GoTo switched to using an 

auction to rank advertisements. In the original auction, ads were ranked by “bid 

per click” and advertisers paid the amount they bid. After consultation with 

auction theorists, GoTo moved to a second-price auction: An advertiser paid a 

price per click determined by the bid of the advertiser in the next lower 

position. 20 

There is a fundamental divergence of incentives in advertising. The publisher 

(i.e. the content provider) has space on its Web page for an ad and wants to sell 

ad impressions to the highest bidders. The advertiser does not care directly 

about ad impressions, but does care about visitors to its website, and ultimately, 

the sale of its products. Hence, the publisher wants to sell impressions, but the 

advertiser wants to buy clicks. 

This is similar to an international trade transaction where the buyer wants to pay 

in euros and the seller wants to receive dollars. The solution in both cases is the 

same: an exchange rate. In the context of online advertising, the exchange rate 

is the predicted click-through rate, an estimate of how many clicks a particular 

ad impression will receive. This allows one to convert the advertiser’s offered 

bid per click to an equivalent bid per impression. The publisher can thus sell 

each impression to the highest bidder. 

This mechanism aligns the interests of the buyer and the seller, but creates 

unintended consequences. If the advertiser only pays for clicks, he has no 

direct incentive to economize on impressions. Excessive impressions, however, 

18 See JoAnne Yates, Business Use of Information and Technology from 1880-1950, in A NATION 

TRANSFORMED BY INFORMATION: HOW INFORMATION HAS SHAPED THE UNITED STATES 

FROM COLONIAL TIMES TO THE PRESENT 107-135. (Alfred D. Chandler and James Cortada, 

eds., 2000) (detailing the role of office machinery in the development of commercial 

enterprises). 

19 GoTo.com Posts Strong Relevancy Ranking in NPD Survey of Search Engines, BUSINESS WIRE, April 

11, 2000, available at http://www.highbeam.com/doc/1G1-61423181.html. 

20 For accounts of the development of these auctions, see John Battelle, THE SEARCH: HOW 

GOOGLE AND ITS RIVALS REWROTE THE RULES OF BUSINESS AND TRANSFORMED OUR 

CULTURE (2005); Steve Levy, Secret of Googlenomics: Data-fueled Recipe Brews Profitability, WIRED, 

2009, http://www.wired.com/culture/culturereviews/magazine/17- 

06/nep_googlenomics?currentPage=all.


impose an attention cost on users, so further attention to ad quality is important 

to ensure that ad impressions remain relevant to users. 

Nowadays, the major providers of search engine advertising all estimate clickthrough 

rates along with other measures of ad quality and use auctions to sell 

these ads. Economists have applied game theory and mechanism design to 

analyze the properties of these auctions. 21 

Enabling Data Extraction & Analysis 

The data from computer-mediated transactions can be analyzed and used to 

improve the performance of future transactions. 

The Sabre air passenger reservation system offered by American Airlines is an 

example of this. The original conception, in 1953, was to automate the creation 

of an airline reservation. However, by the time the system was released in 1960, 

it was discovered that such a system could also be used to study patterns in the 

airline reservation process: The acronym Sabre stands for Semi-Automatic 

Business Research Environment. 22 

The existence of airline reservation systems enabled sophisticated differential 

pricing (also known as “yield management”) in the transportation industry. 23 

Many firms have built data warehouses based on transaction-level data which 

can then be used as input for analytic models of customer behavior. A 

prominent example is supermarket scanner data which has been widely used in 

economic analyses. 24 Scanner data has also been useful in constructing price 

21 See, e.g., Susan Athey & Glenn Ellison, Position Auctions with Consumer Search, 2007. 

http://kuznets.fas.harvard.edu/~athey/position.pdf; Benjamin Edelman, Michael 

Ostrovsky, & Michael Schwartz, Internet Advertising and the Generalized Second Price Auction, 97 

AM. ECON. REV. 242-259 (March 2007); Hal R. Varian, Online Ad Auctions, 99 AM. ECON. 

REV. 430-434 (2009). 

22 Sabre, History, available at http://web.archive.org/web/20080225161359/ 

http://www.sabreairlinesolutions.com/about/history.htm. 

23 Barry C. Smith, John F. Leimkuhler, & Ross M. Darrow, Yield Management at American Airlines, 

22 INTERFACES 8-31 (1992), available at http://www.jstor.org/pss/25061571 (on the 

history of yield management in the airline industry). Kalyan T. Talluri & Garrett J. van 

Ryzin, THE THEORY AND PRACTICE OF REVENUE MANAGEMENT (Kluwer Academic 

Publishers 2004), http://books.google.com/books?id=hogoH5LXmyIC (a textbook 

explanation of yield management). 

24 Aviv Nevo & Catherin Wolfram, Why Do Manufacturers Issue Coupons? An Empirical Analysis of 

Breakfast Cereals, 22 THE RAND J. OF ECON. 319-339 (2002), 

http://research.chicagobooth.edu/marketing/databases/dominicks/docs/2002_W 

hy_Do_Manufacturers.pdf; Igal Hendel & Aviv Nevo, Measuring the Implications of Sales and 

Consumer Inventory Behavior, 74 ECONOMETRICA 1637-1673 (2006), 

http://faculty.wcas.northwestern.edu/~ieh758/measuring.pdf.


indexes, 25 since it allows for much more direct and timely access to prices. The 

fact that the data is timely is worth emphasizing, since it allows for real time 

analysis and intervention for businesses and at the policy level. 

Hyunyoung Choi and I have used real-time publicly-available search engine data 

to predict the current level of economic activity for automobile, real estate, retail 

trade, travel, and unemployment indicators. 26 There are many other sources of 

real-time data such as credit card, package delivery, and financial data. This has 

been referred to as “nowcasting” to describe the use of real-time data in 

estimating the current state of the economy. 27 A variety of econometric 

techniques are used to deal with the problems of variable selection, gaps, lags, 

structural changes and so on. Much of the real-time data is also available at 

state and city levels, allowing for regional macroeconomic analysis. 

In the last 20 years, the field of machine learning has made tremendous strides 

in “data mining.” This term was once pejorative, at least among 

econometricians, but now enjoys a somewhat better reputation due to the 

exciting applications developed by computer scientists and statisticians. 28 One 

of the main problems with data mining is over-fitting, but various sorts of 

cross-validation techniques have been developed to mitigate this problem. 

Econometricians have only begun to utilize these techniques. 29 

25 ROBERT C. FEENSTRA & MATTHEW SHAPIRO, EDS., SCANNER DATA AND PRICE INDEXES 

(2003); Farm Foundation, Food CPI, Prices, and Expenditures: A Workshop on the Use of Scanner 

Data in Policy Analysis, June 2003, http://www.ers.usda.gov/ 

briefing/CPIFoodAndExpenditures/ScannerConference.htm. 

26 Hyunyoung Choi & Hal R. Varian, Predicting the Present with Google Trends, GOOGLE RESEARCH 

BLOG, April 2, 2009, http://googleresearch.blogspot.com/2009/ 

04/predicting-present-with-google-trends.html; Hyunyoung Choi & Hal R. Varian, 

Predicting Initial Claims for Unemployment Benefits, GOOGLE RESEARCH BLOG, July 22, 2009, 

http://googleresearch.blogspot.com/2009/07/posted-by-hal-varian-chiefeconomist.html. 

27 See M. P. CLEMENTS & DAVID F. HENDRY, GREAT BRITAIN STATISTICS COMMISSION, 

FORECASTING IN THE NATIONAL ACCOUNTS AT THE OFFICE FOR NATIONAL STATISTICS 

(2003); Jennifer L. Castle & David Hendry, Nowcasting from Disaggregates in the Face of Location 

Shifts, June 18, 2009, http://www.economics.ox.ac.uk/members/ 

jennifer.castle/Nowcast09JoF.pdf [hereinafter Castle & Hendry, Nowcasting]. 

28 For a technical overview, see TREVOR HASTIE, JEROME FRIEDMAN, & ROBERT TIBSHIRANI, 

THE ELEMENTS OF STATISTICAL LEARNING: DATA MINING, INFERENCE, AND PREDICTION 

(2d ed. 2009). 

29 See Castle & Hendry, Nowcasting, supra note 27.


Enabling Experimentation 

As Ronald Coase has said, “If you torture the data long enough it will 

confess.” 30 It is difficult to establish causality from retrospective data analysis. 

It is thus noteworthy that computer mediation allows one to measure economic 

activity and also conduct controlled experiments. 

In particular, it is relatively easy to implement experiments on Web-based 

systems. Such experiments can be conducted at the query level, user level, or 

geographic level. 

In 2008, Google ran 6,000 experiments involving Web search which resulted in 

450-500 changes in the system. 31 Some of these experiments were with the user 

interface and some were basic changes to the algorithm. 32 The ad team at 

Google ran a similar number of experiments, tweaking everything from the 

background color of the ads, to the spacing between the ads and search results, 

to the underlying ranking algorithm. 

In the 1980s, Japanese manufacturers touted their “kaizen” system that allowed 

for “continuous improvement” of the production process. 33 In a well-designed 

Web-based business, there can be continuous improvement of the product 

itself—the website. 

Google and other search engines also offer various experimental platforms to 

advertisers and publishers such as “Ad Rotation,” which rotates ad creatives 

(i.e., the wording of the ad) among various alternatives to choose the one that 

performs best and “Website Optimizer,” a system that allows websites to try 

different designs or layouts and determine which performs best. 

Building a system that allows for experimentation is critical for future 

improvement, but it is too often left out of initial implementation. This is 

unfortunate, since it is the early versions of a system that are often most in need 

of improvement. 

30 Gordon Tullock, A Comment on Daniel Klein’s ‘A Plea to Economists Who Favor Liberty’, 27 

EASTERN ECONOMIC JOURNAL 205 (No. 2, Spring 2001), available at 

http://college.holycross.edu/RePEc/eej/Archive/Volume27/V27N2P203_207.pdf. 

31 Rob Hoff, Google Search Guru Singhal: We Will Try Outlandish Ideas, BUS. WEEK, Oct. 2009, 

http://www.businessweek.com/the_thread/ 

techbeat/archives/2009/10/google_search_g.html. 

32 Id. 

33 For more information on the Japanese kaizen philosophy, see MASAAKI IMAI, KAIZEN: THE 

KEY TO JAPAN’S COMPETITIVE SUCCESS (1986).


Cloud computing, which I will discuss later in the essay, offers a model for 

“software as service,” which typically means software is hosted in a remote data 

center and accessed via a Web interface. There are numerous advantages to this 

architecture. It allows for controlled experiments which can, in turn, lead to 

continuous improvement of the system. Alternatives such as packaged software 

make experimentation much more difficult. 

Ideally, experiments lead to understanding of causal relations that can then be 

modeled. In case of Web applications there are typically two “economic 

agents”: the users and the applications. The applications are already modeled 

via the source code that is used to implement them, so all that is necessary is to 

model the user behavior. The resulting model will often take the form of a 

computer simulation that can be used to understand how the system works. 

Some examples of this are the Bid Simulator and Bid Forecasting tools offered 

by Google and Yahoo!. 34 These tools give an estimate of the cost and clicks 

associated with possible bids. The cost per click is determined by the rules of 

the auction and can be calculated directly; the clicks are part of user behavior 

and must be estimated with economic forecasting. Putting them together 

creates a model of the auction outcomes. 

How Experiments Change Business 

Because computer mediation drastically reduces the cost of experimentation, 

there have been changes for the role of management. As Kohavi et al. have 

emphasized, decisions should be based on carefully controlled experiments 

rather than “the Highest Paid Person’s Opinion (HiPPO).” 35 

If experiments are costly, utilizing expert opinions by management is a plausible 

way to make decisions. When experiments are inexpensive, however, they are 

likely to provide more reliable answers than opinion, even the opinions of 

highly paid experts. Furthermore, even when experienced managers have 

better-than-average opinions, it is likely that there are more productive uses of 

their time than to sit around a table debating which background colors will 

appeal to Web users. The right response from managers to such questions 

should be to “run an experiment.” 

34 For more information on Bid Simulator and Bid Forecasting, see Louise Rijk, Bid Simulator 

Adds More Transparency to Google AdWords Bidding, INTERNET MKTG. & BUS. REV., Aug. 10, 

2009, http://www.advmediaproductions.com/newsletter/ 

NL_google-adwords-bid-simulator.html. 

35 Ron Kohavi, Roger Longbotham, Dan Sommerfield, & Randal M. Henne, Controlled 

Experiments on the Web: Survey and Practical Guide, 19 DATA MINING & KNOWLEDGE 

DISCOVERY 140-181 (2008) http://www.springerlink.com/content/r28m75k77u145115.


Businesses have always engaged in experimentation in one form or another. 

The availability of computer mediated transactions has, however, made these 

experiments much more inexpensive and flexible than in the past. 

Enabling Customization & Personalization 

Finally, computer mediated transactions allow for customization and 

personalization of interactions by basing current transactions on earlier 

transactions or other relevant information. 

Instead of a “one size fits all” model, the Web offers a “market of one.” 

Amazon.com, for example, makes individual suggestions of items to purchase 

based on an individual’s previous purchases, or on purchases of consumers like 

that individual. These suggestions can be based on “recommender systems” of 

various sorts. 36 

In addition to content, prices may also be personalized, leading to various forms 

of differential pricing. There are certainly welfare effects of such personalized 

pricing. Acquisiti and Varian examine a model in which firms can condition 

prices based on past history. 37 The ability of firms to extract surplus, they 

discover, is quite limited when consumers are sophisticated. In fact, firms have 

to offer “enhanced services” to justify higher prices. 

I have previously suggested that there is a “third welfare theorem” that applies 

to (admittedly extreme) cases with perfect price discrimination and free entry: 

Perfect price discrimination results in the optimal amount of output sold while 

free entry pushes profits to zero, conferring all benefits to consumers. 38 

The same type of personalization can occur in advertising. Search engine 

advertising is inherently customized since ads are shown based on a user’s 

query. Google and Yahoo! offer services that allow users to specify their areas 

of interest and then view ads related to those interests. It is also relatively 

common for advertisers to use various forms of “re-targeting” that allow them 

to show ads based on users’ previous responses to related ads. 

36 Paul Resnick & Hal R. Varian, Recommender Systems, 3 COMM’CNS OF THE ASSOC. FOR 

COMPUTER MACH. 56-58 (March 1997), http://cacm.acm.org/magazines/1997/3/8435recommender-systems/pdf. 

37 Alessandro Acquisiti & Hal R. Varian, Conditioning Prices on Purchase History. 24 MKTG. SCI. 

367-381 (2005), http://www.sims.berkeley.edu/hal/Papers/privacy.pdf. 

38 Hal R. Varian, Competition and Market Power, in JOSEPH FARRELL, CARL SHAPIRO, & HAL R. 

VARIAN, EDS., THE ECONOMICS OF INFORMATION TECHNOLOGY: AN INTRODUCTION, 1-46 

(Cambridge Univ. Press 2005). For a theoretical analysis of first-degree price discrimination, 

see David Ulph & Nir Vulkan, Electronic Commerce, Price Discrimination, and Mass Customisation, 

Nov. 2007, http://vulkan.worc.ox.ac.uk/wp-content/images/combined-paper.pdf.


Transactions Among Workers 

Thus far, there has been an emphasis on transactions among buyers, sellers and 

advertisers. But computers can also mediate transactions among workers. The 

resulting improvements in communication and coordination can lead to 

productivity gains, as documented in the literature on the impact of computers 

on productivity. 

In a series of works, Paul David has drawn an extended analogy between the 

productivity impact of electricity at the end of the nineteenth century and the 

productivity impact of computing at the end of the twentieth century. 39 

Originally, factories were powered by waterwheels which drove a shaft and all 

of the machines in the factory had to connect to this central shaft. The 

manufacturing process involved moving the piece being assembled from station 

to station during assembly. 

The power source evolved from waterwheels to steam engines to electric 

motors. Eventually electric motors were attached to each machine, which 

allowed more flexibility in how the machines were arranged within the factory. 

However, factories still stuck to the time-honored arrangements, grouping the 

same sort of machines in the same location-all the lathes in one place, saws in 

another, and drills in yet another. 

In the first decade of the twentieth century, Henry Ford invented the assembly 

line. Then, the flexibility offered by electric motors became well appreciated. 40 

As David demonstrates, the productivity impact of the assembly line was 

significant, and over the last century, manufacturing has become far more 

efficient. 41 

39 See Paul David, The Dynamo and the Computer: An Historical Perspective on the Modern Productivity 

Paradox, 80 AM. ECON. REV. 355-61 (May 1990) 

http://ideas.repec.org/a/aea/aecrev/v80y1990i2p355-61.html [hereinafter David, 

Productivity Paradox]; Paul David, General Purpose Engines, Investment, and Productivity Growth: 

From the Dynamo Revolution to the Computer Revolution., in E. Deiaco, E. Hornel, & G. Vickery, 

eds., TECHNOLOGY AND INVESTMENT: CRUCIAL ISSUES FOR THE 90S (1991); Paul David, 

Computer and the Dynamo: The Modern Productivity Paradox in the Not-too-distant Mirror, in 

TECHNOLOGY AND PRODUCTIVITY: THE CHALLENGE FOR ECONOMIC POLICY 315-348 

(1991). 

40 Ford suggests that the inspiration for the assembly line came from observing the 

meatpacking plants in Chicago, where animal carcasses were hung on hooks and moved 

down a line where workers carved off different pieces. If you could use this process to disassemble 

a cow, Ford figured you could use it to assemble a car. See HENRY FORD, MY LIFE 

AND WORK (Doubleday, Page & Co. 1923). 

41 I do not mean to imply that the only benefit from electric motors came from improved 

factory layout. Motors were also more efficient than drive belts and the building 

construction was simpler. See David, Productivity Paradox, supra note 39.


I want to extend David’s assembly line analogy to examine “knowledge worker 

productivity.” 42 Prior to the widespread use of the personal computer, 

producing office documents was a laborious process. A memo was dictated to a 

stenographer who later typed the document, making carbon copies. The typed 

manuscript was corrected by the author and circulated for comments. As with 

pre-assembly line production, the partially-produced product was carried around 

to different stations for modification. When the comments all came back, the 

document was re-typed, re-produced and re-circulated. 

In the latter half of the twentieth century, there were some productivity 

enhancements for this basic process, such as White-Out, Post-it Notes, and 

photocopy machines. Nonetheless, the basic production process remained the 

same for a century. 

When the personal computer became widespread, editing became much easier, 

and the process of collaborative document production involved floppy disks. 

The advent of email allowed one to eliminate the floppy disk and simply mail 

attachments to individuals. 

All of these effects contributed to improving the quantity and quality of 

collaborative document production. However, they all mimicked the same 

physical process: circulating a document to individuals for comments. Editing, 

version control, tracking changes, circulation of the documents and other tasks 

remained difficult. 

Nowadays, there is a new model for document production enabled by “cloud 

computing.” 43 In this model, documents live “in the cloud,” meaning in some 

data center on the Internet. The documents can be accessed at any time, from 

anywhere, on any device, and by any authorized user. 

Cloud computing dramatically changes the production process for knowledge 

work. There is now a single master copy that can be viewed and edited by all 

relevant parties, with version control, check points and document restore built 

in. All sorts of collaboration, including collaboration across time and space, 

have become far easier. 

42 See Peter F. Drucker, Knowledge-worker Productivity: The Biggest Challenge, 41 CAL. MGMT. REV. 

79-94 (1999). 

43 Michael Armbrust, Armando Fox, Rean Griffith, Anthony D. Joseph, Randy H. Katz, 

Andrew Konwinski, Gunho Lee, David A. Patterson, Ariel Rabkin, Ion Stoica, & Matei 

Zaharia, Above the Clouds: A Berkeley View of Cloud Computing, Feb. 10, 2009, 

http://www.eecs.berkeley.edu/Pubs/TechRpts/2009/EECS-2009-28.html 

[hereinafter Above the Clouds]. See also, Wikipedia, Cloud Computing, 

http://en.wikipedia.org/wiki/Cloud_computing.


Instead of passing the document amongst collaborators, a single master copy of 

the document can be edited by all interested parties (simultaneously if desired). 

By allowing workflow to be re-organized, cloud computing changes knowledge 

worker productivity the same way that electricity changed the productivity of 

physical labor. 

Enabling Deployment of Applications 

As previously mentioned, cloud computing offers what is referred to as 

“software as a service.” This architecture reduces support costs and makes it 

easier to update and improve applications. 

Cloud computing, however, does not only offer “software as a service.” It also 

offers “platform as a service,” which means that software developers can deploy 

new applications using the cloud infrastructure. 

Nowadays, it is possible for a small company to purchase data storage, hosting 

services, an application development environment, and Internet connectivity 

“off the shelf” from vendors such as Amazon.com, Google, IBM, Microsoft, 

and Sun. 

The “platform as a service” model turns a fixed cost for small Web applications 

into a variable cost, dramatically reducing entry costs. Computer engineers can 

both explore the combinatorial possibilities of generic components to create 

new inventions and can actually purchase standardized services in the market in 

order to deploy those innovations. 

This development is analogous to the recent history of the book publishing 

industry. At one time, publishers owned facilities for printing and binding 

books. Today, due to the strong economies of scale inherent in this process, 

most publishers have outsourced the actual production process to a few 

specialized book production facilities. 

Similarly, in the future, it is likely that there will be a number of cloud 

computing vendors that will offer computing on a utility-based model. This 

production model dramatically reduces the entry costs of offering online 

services, and will likely lead to a significant increase in businesses that provide 

such specialized services. 44 

The hallmarks of modern manufacturing are routinization, modularization, 

standardization, continuous production, and miniaturization. These practices 

have had a dramatic impact on manufacturing productivity in the twentieth 

44 Above the Clouds, supra note 43.


century. The same practices can be applied to knowledge work in the twentyfirst 

century. 

Computers, for example, can automate routine tasks such as spell-checking and 

data retrieval. Communications technology allows tasks to be modularized and 

routed to the workers best able to perform those tasks. Similar to how the 

miniaturization of the electric motor allowed physical production to be 

rearranged in 1910, the miniaturization of the computer—from the mainframe, 

to the workstation, to the PC, to the laptop, and to the mobile phone—allows 

knowledge production to be rearranged on a local and global scale. 

Enabling Micro-Multinationals 

An interesting implication of computer mediated transactions among 

knowledge workers is that interactions are no longer constrained by time or 

distance. 

Email and other tools allow for asynchronous communication over any 

distance, which allows for optimization of tasks on a global basis. Knowledge 

work can be subdivided into tasks, much like physical work in Adam Smith’s 

hypothetical pin factory. 45 But even more, those tasks can be exported around 

the world to where they can most effectively be performed. 

For example, consultants at McKinsey routinely send their PowerPoint slides to 

Bangalore for beautification. There are many other cognitive tasks of this sort 

that can be outsourced, including translation, proofreading, document research, 

etc. Amazon.com’s Mechanical Turk is an intriguing example of how computers 

can aid in matching up workers and tasks. 46 As of March 2007, there were 

reportedly more than 100,000 workers from 100 countries who were providing 

services via the Mechanical Turk. 47 

The dramatic drop in communications costs in the last decade has led to the 

emergence of what I have termed “micro-multinationals.” 48 Nowadays, a 10- or 

12-person company can have communications capabilities that only the largest 

45 ADAM SMITH, AN INQUIRY INTO THE NATURE AND CAUSES OF THE WEALTH OF NATIONS 18- 

21 (Edwin Cannan, ed., Methuen & Co., Ltd. 1904) (1776), 

http://www.econlib.org/library/Smith/smWN.html. 

46 Wikipedia. Amazon Mechanical Turk, 

http://en.wikipedia.org/wiki/Amazon_Mechanical_Turk. 

47 Jason Pontin, Artificial Intelligence, with Help From the Humans, N.Y. TIMES, March 25 2007, 

www.nytimes.com/2007/03/25/business/yourmoney/25Stream.html 

?ex=1332475200en=cd1ce5d0bee647d5ei=5088partner=rssnytemc=rss. 

48 Hal Varian, Technology Levels the Business Playing Field, N.Y. TIMES, Aug. 25 2005, 

http://www.nytimes.com/2005/08/25/business/25scene.html.


multinationals could afford 15 years ago. Using tools like email, websites, wikis, 

voice over IP, and video conferencing, tiny companies can coordinate workflow 

on a global basis. By sending work from one time zone to the next, these 

companies effectively work around the clock, giving them a potential 

competitive advantage over firms that are restricted to one time zone. 

Many micro-multinationals share a common history: A student comes to the 

United States for graduate school. They use the Internet and the collaborative 

tools available in scientific workgroups. Some get bitten by the start-up bug. 

They draw on their friends and colleagues back home, who have other contacts 

living abroad. The collaborative technologies previously mentioned allow such 

loose groups to collaborate on producing computer code, which may end up as 

a working product. 

As Saxenian has pointed out, “emigration” means something quite different 

now than it did 30 years ago. 49 As she puts it, a “brain drain” has been replaced 

by a “brain circulation.” We now have a host of collaborative technologies that 

allow an immigrant to maintain ties to his social and professional networks in 

his home country. 

Conclusion 

I began this essay with a discussion of combinatorial innovation and pointed 

out that innovation has been so rapid in the last decade because innovators 

around the world can work in parallel, exploring novel combinations of 

software components. When the innovations are sufficiently mature to be 

deployed, they can be hosted using cloud computing technology and managed 

by global teams, and even by tiny companies. Ideally, these new services can 

serve as building blocks for new sorts of combinatorial innovation in business 

processes that will offer a huge boost to knowledge worker productivity in the 

future. 

49 ANNALEE SAXENIAN, THE NEW ARGONAUTS: REGIONAL ADVANTAGE IN A GLOBAL 

ECONOMY (2006).



Decentralization, Freedom to 

Operate & Human Sociality 

By Yochai Benkler * 

Three Stories of Innovation in the 

Networked Information Economy 

In 1994, two groups of software engineers were working on the next generation 

of critical software: a Web server; the software that a website runs to respond to 

requests from users. One group was within Microsoft, understanding that the 

next generation of critical infrastructure would be the Web, and trying to extend 

its market from the operating system to the Web server. The other was a group 

of developers led by Brian Behlendorf, formerly from the group of academic 

computing engineers from the University of Illinois in Urbana-Champagne, 

who were patching up the server developed in tandem with the development of 

Mosaic, the first graphical interface to access the Web, at Urbana-Champagne. 

They called it a patchy server, which became the name of the resulting open 

source project: Apache server. Anyone who would have predicted that the 

system, built by a scrappy set of developers, who adopted a licensing approach 

that asserted no exclusive rights over their output, and were working in an area 

considered strategically critical by the largest company in the field and that was 

developing a product in direct competition, would win would have been 

laughed out of the room. And yet, it moves (as Galileo famously said, 

defending his theories of the Earth’s orbit around the Sun). Over 15 years, 

through two boom and bust cycles, Apache has held 50-60% of the market 

share in Web servers (in the summer of 2010, it was about 55%), while 

Microsoft’s Web server market share has hovered between 25% and 35% 

(about 25% in summer of 2010). 

In 1999, two of the most insightful economists looking at the new rules for the 

information economy opened their book with an analysis of how Microsoft’s 

move into the market in encyclopedias embodied the new challenges created by 

the digital economy. In February of 2001, the developer of one of several 

ongoing efforts to develop an online encyclopedia half gave up and dumped 

about 900 stubs onto an open source platform, under a license that let anyone 

edit it and gave no one power to veto. This made participation easy, but control 

relatively hard. And no one was paid to write or edit the encyclopedia. It was 

probably the ugliest technical system for encyclopedia development being 

experimented with at the time. Five years later, this ugly duckling would be 

* Yochai Benkler is the Berkman Professor of Entrepreneurial Legal Studies at Harvard, and 

faculty co-director of the Berkman Center for Internet and Society.


identified by a study done by the staff at Nature as having roughly similar error 

rates as Britannica for science articles. By 2009, Microsoft’s Encarta 

encyclopedia product was discontinued. Wikipedia has come to embody the 

fundamental changes we have to deal with when trying to understand the 

networked information economy. 

In 2001 a Swedish and Danish entrepreneur invested in software developed by 

three Estonian programmers and released a brilliant new solution for peer-topeer 

file sharing: Kazaa. Thanks to the fact that the firm was based in the 

Netherlands, where Dutch law provided it greater immunity from suit by record 

labels, Kazaa quickly became a major platform after the demise of Napster. 1 By 

2003, the same group of entrepreneurs and programmers had launched a peerto-peer 

voice telephony application built on the same basic architecture as 

Kazaa: Skype. Theoretically, Skype should not have worked. For close to two 

decades, the Internet Protocol’s “first-come, first-served,” treat-all-packets-ona-best-efforts-basis 

approach was thought to prevent serious voice over Internet 

applications from working well. And yet, here was this small company 

providing better quality, encrypted, end-to-end communications, using the 

users’ own computers and connections as its basic infrastructure. They did not 

need to control the flow of packets in the network to provide Quality-of-Service 

assurances. They just provided service of a quality that was good enough for 

the price: free for calls from one Skype user to another, soon followed by very 

low rates for calls to regular phones. In 2005, eBay bought Skype for over $2.5 

billion. 

Radical Decentralization of 

Physical, Human & Social Capital 

The three stories above outline the basic transformative elements of the 

networked information economy. We have seen a radical decentralization of 

the most important forms of capital in the most advanced sectors of the 

economy: physical, human, and social capital. For the first time since the 

Industrial Revolution, the most important inputs into the core economic 

activities of the most advanced economies are widely distributed in the 

population. Technologically, the change begins with physical capital: 

Processing, storage, communications, and sensing hardware have come to be 

developed in packages of sufficiently low cost to be put in service by individuals 

for their own personal use. These advances are capable of mixing consumer use 

with production activities. The rapid increase in physical capabilities 

emphasizes continuous rapid innovation as a core dimension of growth and 

welfare which, in turn, emphasizes human capital. 

1 Napster itself was a college dorm room experiment, one of many that flourished at that 

time, which dramatically and permanently changed the landscape of the music industry.


Human capital too is, by nature, widely distributed in the population, and is 

extremely sticky and hard to aggregate or transfer effectively from one 

individual to another. While we measure education when we try to quantify 

human capital, that is far from all of what human capital really entails. Certainly 

it does partly entails acquired, codified knowledge of the kind we get in 

education; but that is only one part of it. Creativity, insight, experience—all 

these go into answering the critical question: Will this individual come up with 

an idea, and even more importantly, will this interaction and conversation 

among a given set of individuals result in an interesting set of ideas emerging? 

Organizationally, the increased emphasis on interactions among human beings 

responding to surprising new opportunities has increased the importance of 

loosely-coupled interactions beyond slower-moving group boundaries like 

firms. These new organizational frameworks and the cooperative dynamics they 

require depend on lightweight, flexible mechanisms that we all carry for 

interacting with other people. That is, they depend on human sociality. This 

basic set of protocols for non-destructive human interaction is also 

fundamentally and widely shared in the population. They are not locked up in 

the cabinets of smart corporate lawyers’ incorporation forms or major deal 

documents. They are the core social and psychological features of human 

beings that have co-evolved, physically and culturally, to allow us to be the 

kinds of social creatures we in fact are—warts and all. 

This distributed network of human beings, possessing the physical, human, and 

social capital that they do, are now connected in a global network of 

communications and exchange that allows much greater flow and conversation, 

so that many new connections are possible on scales never before seen. 

Together, these mean that conversations and new ideas—but more importantly, 

pilots, experiments, and toy implementations of these new ideas—are cheap and 

widespread, and innovation happens everywhere, all the time, at low cost. The 

vast majority of ideas go nowhere, just as the vast majority of experiments fail. 

But the sheer scale of experimentation has meant that the network has reliably 

provided the flow of innovation that we have come to expect and depend on, 

and it has largely come from unpredictable corners rather than from yesterday’s 

innovators or the previous decades’ large firms. 

As a result of these basic dynamics, in the networked information economy, 

experimentation, continuous learning and improvement, low-cost prototyping, 

deployment, iteration, and adoption are more important than well-behaved 

innovation investments. Social behavior plays a much larger productive 

economic role than it could when physical capital requirements meant that, 

however good an idea someone had, transitioning it to a platform that could 

actually be adopted by consumers/users was simply too expensive to do except 

through a system of contracts and investment—through a more-or-less formal 

corporate model. In the networked information economy, freedom to operate is


more important than power to appropriate, and voluntarism and sociality are more important 

than formal contract, and play an important role alongside corporate organization. 

Freedom to Operate is More 

Important than Power to Appropriate 

The story of open source software is the core story about the importance of 

freedom to operate and loosely-coupled association that transcends contract 

and corporate structure. What makes a software development project “open 

source” is that the output of the development activity, the code, is released 

under a copyright license that allows anyone to look at, modify, and redistribute 

the code and modifications to it. This means that anyone, anywhere, can come 

to the state of the art in the code, adopt it, adapt it, and release it, building on 

the innovative contributions of others with complete freedom to operate with 

and on it. In a networked environment where human capital resides in many 

places, and where it is impossible for anyone firm to hire all the smartest people 

(or more to the point, to hire all the people who are likely to have the most 

relevant and powerful insights for any new challenge), a system that depends on 

open access to the universe of available resources, projects, and collaborators 

on them will outperform a system that only allows people who have already 

been identified, recruited, and contracted with based on past projections of 

what would be important for working on a new problem. 

The licensing aspect of open source software raises another important aspect of 

change. Historically, assuring the owner of financial capital of the soundness of 

an entrepreneur was the critical factor. To do so, it was necessary to possess 

property in core inputs, and a network of contracts for flows of what could not 

reliably or efficiently be owned, like supply relations. Today, assuring a steady 

and reliable flow of complementary contributions from other developers is as 

important to maintain a high rate of innovation, experimentation, and 

adaptation as securing the complementary financial inputs. At the early stages, 

complementary contributions from other developers are more important than 

financial inputs. 

With the rise of peer production, radically distributed collaborative production 

on the open source model, adoption of licensing terms like those of free and 

open source software, or Creative Commons, becomes an important avenue to 

secure those complimentary human investments in the project. Where it is 

impossible to assure that you will always employ the right people, open source 

licensing has become an increasingly common strategy for entrepreneurs and 

large firms alike to improve the probability that they will be able to attract the 

complimentary rapid development contributions they need, in the time frame 

they need it, on currently-unpredicted challenges to assure high-velocity 

innovation. Formalized freedom to operate, in the form of open source 

licensing, is coupled with a strong pre-commitment by the firms that undertake


the limitations it imposes on their power to appropriate, so as to assure 

potential collaborators against defection with regard to the fruits of the 

common, firm-boundary-crossing enterprise. Increasingly this is also becoming 

a way for firms that compete in some domains, such as software services, to 

engage in pre-competitive cooperation on the development of core necessary 

tools, like the Linux kernel and operating system, the Apache Web server, etc. 

Voluntarism & Sociality Become 

More Important than Formal Contract 

Another important characteristic of the networked information economy is the 

critical role of knowledge and creativity. These require uniquely human inputs, 

and are persistently uncontractible. That is, you can neither define for explicit 

codification nor characterize for monitoring over time, what it means to be 

creative, or insightful, or usably knowledgeable in context where an innovation 

challenge occurs. As a result, tacit knowledge and insight are necessarily and 

always imperfectly defined for, or monitored through, contract. To assure the 

right motivations and orientation towards finding new solutions to challenges, it 

is necessary for an economy at large, as it is for any given organization, to 

harness the non-contractible motivations of individuals to the knowledge and 

innovation task at hand. This is not new, in the sense that literature on highcommitment, 

high-performance organizations has been around for decades, and 

management theory keeps flowing back and forth between periods that 

emphasize explicit material rewards and monitoring to control employees 

shirking their responsibilities, and periods where the limitations of those 

approaches become clearer, and the benefits of models that depend on a more 

holistic, human view of what is required to create a motivated workforce 

prevail. 

In the networked information economy, where so much of what needs to be 

done is uncontractible and so many of those who need to be engaged are not 

even in a position to have a contractual relationship, the role of sociality and 

cooperative human systems designs that aim to engage, and depend on social, 

moral, and emotional motivations as well as, and often instead of, material 

motivations, has become much larger. Wikipedia, in my three stories, stands as 

the ultimate example of a system that critically depends on these non-material 

motivational vectors, mediated through a technical-social platform that is 

optimized to engage these motivations and allow people to cooperate over a 

system that provides great freedom to operate, no power to appropriate, and 

tremendous room for social organization and interaction (which all have their 

own warts and bumps). 

Rather than the traditional formal modes of organization—be they a formal 

corporation based on contracts, or formal, stable associations on the model of 

rotary clubs or unions—the new forms of social networks (not the


Facebook/MySpace-type websites, but the actual social phenomenon) permit 

people to have more loosely-coupled social associations, in which they can 

participate for some of the time, and combine their investments with many 

others who are similarly loosely-tied to each other, and may spend their time at 

different rates, and in different enterprises, during the course of their day, week, 

or year. Together, these new forms of loose association, based on social signals 

rather than price signals or a formal corporate managerial hierarchy, form what 

I have called peer production. They are not, by any stretch of the imagination, 

going to replace all production activities built on more formal, structured 

models. Anyone who claims that the argument is one of replacement 

misunderstands the claim. 

The new models of production do, however, come to play a significant 

productive role in an environment that continues and will continue to be 

occupied by more traditional forms. They create new sources of competition— 

as in the case of Wikipedia displacing Encarta—and new forms of 

complementary sources of innovation and other inputs—as in the case of open 

source software and the software services industry. They do not herald the 

death of traditional market/firm-based production. To argue otherwise would 

be silly. But it would be equally silly to simply assume away a major new 

organizational innovation. Peer production and cooperative human systems are a new way 

to harness a latent but massively productive force. They make the line between production and 

consumption fuzzy, and offer new pathways to harness the time, insight, experience, wisdom 

and creativity of hundreds of millions of people around the world to perform tasks that, until a 

decade ago, we only knew how to perform through formal models of employment and contract. 

They are an organizational innovation that anyone ignores at their peril. Just 

ask the Departments of Defense or State how they feel about WikiLeaks. 

Innovation Anywhere & 

Everywhere Over an Open Network 

The story of Skype rounds out the core changes that the networked information 

economy presents. In the mid-twentieth-century, the epitome of innovation 

was Bell Labs. With enough Nobel laureates to make the most ambitious 

academic physics departments green with envy and massive investment from 

monopoly profits, Bell Labs is where we got the transistor on which the entire 

information economy is built. It is, indeed, where we got information theory 

itself. The Bell system also epitomizes the organizational model of the midtwentieth-century. 

“One System, One Policy, Universal Service” was how the 

company’s legendary President, Theodore Vail, put it 100 years ago. 2 According 

to this model of thought, if the Internet was ever to carry that most delay- 

2 AT&T, Milestones in AT&T History, http://www.corp.att.com/history/milestones.html 

(last accessed Aug. 17, 2010).


sensitive of all services, voice, we would have to change how we manage 

packets. Best effort delivery 3 just wouldn’t do it. Someone needed to manage 

the network and decide—this packet, which carries voice, is more latency 

sensitive, while that packet, which carries email or a Web page, can wait. But, as 

it turns out, this persistent prediction was false. And the people who proved it 

false were not working for Bell Labs. Indeed, it was probably impossible for 

anyone inside one of the current incarnations of the Bell system to have done 

so. It was, instead, left to three Estonian developers and a couple of Dutch and 

Danish edgy entrepreneurs to do so. They were not the only ones to try. 

Others did too—VocalTec in Israel was among the first; but they were too 

early. 

The point is that in a global networked information environment, innovation 

can come from anywhere; insights of various forms can find each other, and 

experimentation and implementation are cheap to do from anywhere to 

anywhere else. Massive experimentation is followed by massive failures. But 

the failures are generally cheap, at least by societal standards. And the successes 

can be readily disseminated, adopted, and generalized on a major global scale in 

very short time frames. Variation, selection, adaptation and survival/replication 

through user adoption, rather than planning and high investment, have 

repeatedly offered the more robust approach in this new complex and chaotic 

environment. Rapid, low cost experimentation and adaptation on a mass scale, 

underwritten by the ease of cheap, fast implementation and prototyping, and 

cheap widespread failure punctuated by a steady flow of unpredictable successes 

have been more important to innovation and growth in the networked economy 

than models of innovation based on higher-cost, more managed innovation 

aimed at planning for predictable, well-understood returns. 

Implications for Human Systems Design 

We live our lives through systems: organizational systems, like corporations, 

states, or nonprofits; technical systems, like the interstate highway system or the 

Internet; institutional systems like law, both public and private, or social 

conventions; and cultural, as in our belief systems for how we know things to 

be true, such as religion, or science. To a great extent, these systems are too 

complex for us to construct deterministic, fully understood interventions that 

will clearly lead to desired outcomes, along whatever dimension we think is 

important: efficiency, freedom, security, or justice. But we nonetheless apply 

ourselves to the task. We try to use management science to design better 

organizational strategies; we try to use law to refine and improve our legal 

3 “Best effort delivery describes a network service in which the network does not provide any 

guarantees that data is delivered or that a user is given a guaranteed quality of service level or 

a certain priority.” Wikipedia, Best effort delivery, 

http://en.wikipedia.org/wiki/Best_effort_delivery (last accessed Aug. 17, 2010).


system; we invest enormous amounts in designing better technical systems, and 

so forth. 

The characteristics of the networked information economy require that in our 

efforts at systems design we emphasize openness and freedom to operate over 

control and power to appropriate and that we emphasize human sociality and 

diverse motivations for diverse types over optimizing for material interests and 

letting everything else sort itself out. At a practical level, technical open design 

has made the largest and most powerful steps. Anchored in the very decision to 

separate TCP from IP, and make the core Internet protocol as open as it can be, 

and continuing to the central role that open standards have played in the 

development of the Web, XML, and WiFi, to name just a few, a continuous 

emphasis on openness already has substantial support and inertia, although it is 

always under pressure from firms that think they can get an edge by owning a 

de-facto standard, or controlling a technical choke point that would allow them 

to extract rents. In management science, we are seeing, slowly and in some 

senses at the periphery, efforts to learn the lessons of open source software and 

apply them to collaboration across firm boundaries and strategic management 

of the knowledge ecology that a firm occupies. 

In law, the most important battleground in the tension between the controloriented 

approach and the freedom-to-operate approach is intellectual property. 

Only this year Amazon received a patent for social networking 4 that reads 

more-or-less like a description of Facebook, launched four years before 

Amazon had even filed its patent application. But not everything is so silly. 

This summer, the Librarian of Congress exempted iPhone jailbreaking from the 

Digital Millennium Copyright Act’s anti-circumvention provisions. 5 If there is 

any single policy domain in which it is important to apply what we have learned 

about the new networked information economy, it is in the area of intellectual 

property. It is also the area where there is the largest potential for intellectual 

and political programmatic overlap between libertarians and progressives. 

4 Stan Schroeder, Amazon Patents Social Networking System, Winks at Facebook, MASHABLE/TECH, 

June 17, 2010, http://mashable.com/2010/06/17/amazon-patents-social-networkingsystem/. 

5 Copyright Office, Rulemaking on Exemptions from Prohibition on Circumvention of 

Technological Measures that Control Access to Copyrighted Works, July, 28, 2010, 

http://www.copyright.gov/1201/.


A Common Agenda on 

Intellectual Property for the 

Networked Information Economy 

From the perspective of economic analysis, information is a public good. Once 

someone creates new information or knowledge, anyone can use it without 

reducing its availability for anyone else. Its marginal cost is therefore zero, and 

that is its efficient price. However, for information to be available at a price of 

zero, the person who produced it must find some other mechanism to extract 

value from their investment in creating the information. Otherwise, having 

information available at its marginal cost today (zero) will lead to less 

production tomorrow. 

The overwhelming majority of information, knowledge, and culture is produced 

without the need to rely on explicit, intellectual-property-based mechanisms to 

appropriate its benefits. Firms continuously innovate in their processes so as to 

lower their costs and improve their profits; but they do generally not patent 

their innovations and license them or exclude competitors from using them. 

Individuals innovate and develop experience about their workplace to improve 

their own performance; people read news and create commentary for each 

other, and appropriate the benefits of what they find socially. Governments 

invest in R&D and reap the benefits through higher growth, greater military 

might, etc. Nonprofits and academic institutions invest in information, 

knowledge, and cultural production, and so forth. All these approaches have 

their own advantages and disadvantages; but economic survey after survey for 

the past few decades has shown that even in industrial innovation, a minority of 

sectors relies on patents, and the majority relies on a range of supply-side and 

demand-side improvements in appropriability that come from developing the 

information and either using it without exchanging it or disseminating it and 

relying on first-mover advantages, network effects, marketing and reputational 

benefits, etc. 

The only industries that are still dependent on intellectual property protections 

are the pharmaceutical industry for patents and Hollywood, the recording 

industry, and much of book publishing for copyright. Even newspapers and 

magazines are not so dependent on IP. They are, rather, advertising-supported 

media. They depend on release of the information to capture demand-side 

benefits for their paying clients in a two-sided market—the advertisers. 

The reason that it is important to remember this quick recap of innovation and 

Patents & Copyrights Economics 101 is that it helps us to see that patents and 

copyrights represent a government decision to prohibit everyone from using 

ideas or information that they can practically use, in order to serve a public 

purpose—supporting a subset of business models for the creation of new 

information, knowledge, and culture. Now, it is perfectly acceptable for


government to prohibit some actions in order to serve the public good. We 

prevent companies from selling food unless it is labeled in certain ways to serve 

public health; we prohibit violence to increase public security, and so forth. But 

we try to do so only when there is indeed a good reason. 

Sometimes, we combine prohibitions with a market in permissions. Tradeable 

emissions permits are a classic example where we think it is more efficient to 

allow firms to trade in their permissions than to simply have direct regulation. 

Patents and copyrights are exactly like tradeable emissions permits. They are a 

market-based approach toward the regulatory problem of how to prevent 

people who want to use the existing universe of information and knowledge 

that they possess in ways that will undermine future knowledge production and 

innovation. We prohibit everyone from using certain classes of information and 

knowledge, and we create a market in permissions to use that information. We 

call these permissions “copyrights” or “patents.” What is important to 

remember is that these permissions markets create a drag on freedom to operate 

on current innovation and knowledge creation, and they create a drag on 

innovation in all industries that, unlike pharmaceutical or blockbuster movie 

markets, do not heavily depend on such permissions markets. 

For progressives, the best way to understand patents and copyrights is through 

the prism of free speech: These are government regulations on what and how 

we can say things; and how we can use what we know, that are implemented in 

pursuit of legitimate government ends—aiding innovation and creative 

expression by some industries—at the expense of that freedom. As with any 

limitation on speech and learning, it has to be supported by very good reasons. 

It is not at all clear whether our contemporary economic understanding of the 

functioning of copyright law in particular, and patent law to a lesser extent, 

provide sufficient support for such significant restrictions on free speech. 

For libertarians, the best way to understand patents and copyrights is as a 

regulatory system that imposes limitations on how individuals can act on 

knowledge they possess in pursuit of their own goals. It is a regulatory system 

that creates and allocates permissions to generate market-based transfer 

mechanisms; but a regulatory system in pursuit of a government program, 

which embodies the judgment that certain business models used to sustain 

innovation and expression are more effective than others, and supports those 

deemed more effective at the expense of those other approaches. 

Both approaches should lead to a significant downward revision in the level of 

acceptable intellectual property enforcement that the United States pursues. Let 

me offer one example, which provides the basic structure of the problem: How 

long should a copyright last? If we thought that copyrights were really property, 

the answer would be something like forever. The U.S. Constitution, as well as 

the laws of practically every other country, instead limit the term of copyright,


understanding that there is a big difference between the need of exclusivity in a 

thing that, if one person uses, another cannot, and exclusivity in an idea or 

expression that anyone can use without making it any less available for anyone 

else. The former is a proper object of property. The latter is a proper object of 

regulation of individual freedom for so using the thing, but only to the extent 

justified by property. 

So how long should copyright terms be? Let’s try this thought experiment: 

Imagine that you are someone with an idea for a movie. You walk in to a group 

of hypothetical investors and you tell them: “Here’s my idea, here’s the audience 

for it, and so here is my projection for how much money we will make on it.” 

The investors ask you: “What are your assumptions about timing? By when will 

we see our return?” Now, imagine that you answered: “We won’t really break 

even in the first seventy years, but just you wait until years seventy to ninetyfive: 

We’ll be making millions!” You would be laughed out of the room. “OK, 

let’s try it with not making money the first twenty years, but making a killing in 

the years twenty to thirty.” You get the point. 

If copyright is intended to assure that there is enough appropriability to attract 

investment in creating a new expression, but it is a regulatory form that restrains 

the freedom of others to operate in pursuit of that goal, then its term should be 

keyed to the term necessary to attract investors. Given today’s discount rates in 

the relevant industries—that is, how quickly investors need to turn a profit 

before they will decide to put their money in some other enterprise—that likely 

means 18 months; maybe it means three to five years. It is possible that 

different industries have different levels of patience. But fundamentally, the 

overwhelming majority of the social cost created by the 95-year term of 

copyright—let alone the repeated practice of retroactive extension of copyright 

for works already created in response to the then-existing incentives-system—is 

incurred without any benefit for investment purposes. No sane investor today 

cares about returns on an investment in these kinds of fields (as opposed to, 

say, power plants or utilities) that are ten years out. For software, maybe the 

correct period is 18 months; for novels, maybe 10 years, although even there, 

the relevant party is the publisher’s decision to publish, not the author’s 

decision to write—because copyright-based monetization runs through the 

publisher’s business decision, not the author’s. In patents, maybe the correct 

period is 20 years for pharmaceuticals. Maybe more; or maybe less. But the 

principle for all these is the same: The period of copyright or patent protection 

should be backed out of reasonable investment assumptions and discount rates, 

not pulled out of the lobbying process, which is always skewed in favor of the 

small number of firms that possess these rights and against the millions of 

potential innovators who do not yet know that this or that piece of regulated 

access to information will get in their way five years from now.


The details of what might go on a major intellectual property reform that should 

be supported by both libertarians and progressives may differ among 

commentators. The core structure of the reasons for change are the same: (a) 

Strong patents and copyrights benefit some business models over others, and in 

particular place a strong drag on the radically distributed, chaotic, innovationeverywhere-by-everyone 

model of the networked information economy in favor 

of twentieth-century models of much more stable and controlled markets like 

those of Hollywood and the recording industry; (b) There is a big difference 

between the level of exclusivity needed to attract investment at the margin, and 

the level of exclusivity that maximizes its owner’s ability to extract rents; the size 

of the difference between the minimal necessary to attract innovation and the 

rent-maximizing level of protection is equal to the amount the incumbents are 

willing to spend on lobbying to keep the line at the maximal point, as opposed 

to the minimally-necessary point; and (c) The lines in fact should be drawn 

where the marginal effect is to attract investment, not where rents can be 

maximized. The academic community has spent years trying to refine a set of 

interventions that could improve access to information, knowledge and culture, 

while having minimal impact on incentives to invest. The following represent 

some of the most promising of these ideas. 

� Copyright term: Copyright terms should be keyed to actual market 

requirements and the discount rate in the business. Copyright that is 

any longer than necessary to attract the marginal investor that makes a 

difference between the project happening or not represents pure rent 

extraction and is a drag on innovation and creativity. 

� Renewal of existing copyrights: There are mountains of existing 

materials (animal shots from documentaries from the 1960s; explosions 

and action shots from 1970s B movies; etc.) that could provide the grist 

for new models of creative mashup tools and sites, but instead sit 

unused and unusable because the rights are excessively tied up. 

Existing copyrights should be required to be renewed periodically, 

initially for a nominal price, and later on in the life of a copyright for 

escalating fees, rising to a level no greater than necessary to make a 

copyright owner think: is their any real market for this thing, rather than 

focing holders to make fine distinctions about the value of the work, on 

one hand, or simply automatically renewing everything, whether or not 

it has any market, because it’s cheaper to renew than to review 

continued viability. Those works that continue to be of even small 

commercial value will be renewed. Those that continue to be of 

emotional significance will be renewed. All others will become freely 

usable upon failure to re-register.


� Reinstate the Sony doctrine 6 by legislative reversal of the Supreme 

Court’s Grokster decision 7 – In the midst of the panic over peer-topeer 

filesharing, the Supreme Court moved away from its long-standing 

precedent that an innovator cannot be forced to foresee and prevent 

the potentially-infringing uses of its new product (in the Sony case, the 

VCR). As long as there are substantial noninfringing uses, innovators 

are immune to suit by copyright owners whose works are being 

infringed by users of the innovator’s product. In Grokster the Supreme 

Court created a more intention-based, fact-intensive inquiry that 

imposes greater litigation risk on entrepreneurs who innovate on the 

Net with anything that can possibly be used to infringe existing 

copyrights. This is an unnecessary drag on Internet innovation and 

entrepreneurship in favor of the movie and recording industries. 

� Eliminate business methods patents: Few innovations are as 

unnecessary as a law intended to give business people an incentive to 

improve their business model. The incentive to develop a new business 

model is that it makes more money for its inventor. There is no need 

for an additional government-granted monopoly on doing business in 

this way. The Federal Circuit, which created this new doctrine 12 years 

ago, tried to walk it back in the Bilski case, 8 but the Supreme Court 

recently held 9 that the particular way that the federal Circuit went about 

doing so was indefensible. Nonetheless, it appears that a majority of 

the Supreme Court would support some other, better-reasoned 

reversal. 

� Eliminate software patents: There is fairly significant evidence that 

software patents are unnecessary, and that software development is 

heavily based on service models, time to market, network effects, 

customer habits, etc. On the other hand, patents get in the way of open 

source development, and throw a monkey wrench into the model of 

rapid innovation by anyone, anywhere, distributable everywhere. They 

create unnecessary barriers to entry that reduce the freedom to operate 

and experiment, and thereby harm innovation. 

6 Sony Corp. of America v. Universal City Studios, Inc., 464 U.S. 417 (1984), available at 

http://en.wikipedia.org/wiki/Sony_Corp._of_America_v._Universal_City_Studios,_ 

Inc. 

7 MGM Studios, Inc. v. Grokster, Ltd., 545 U.S. 913 (2005), available at 

http://en.wikipedia.org/wiki/MGM_Studios,_Inc._v._Grokster,_Ltd.. 

8 In re Bilski, 545 F.3d 943, 88 U.S.P.Q.2d 1385 (Fed. Cir. 2008), available at 

http://en.wikipedia.org/wiki/In_re_Bilski. 

9 Bilski v. Kappos, 561 U.S. ___, 130 S. Ct. 3218 (2010), available at 

http://en.wikipedia.org/wiki/Bilski_v._Kappos.


� Create a “band” of exempt experimentation, both commercial 

and non-commercial, whereby use of existing copyrighted or 

patented information or knowledge does not trigger liability: 

Today, to the extent that there are exemptions they are spare and 

niggardly. Using just three notes from a prior recording and mashing 

them up into a completely new song does not, under present copyright 

law, count as “de minimis.” 10 Academic experimentation on a patented 

drug that does not result in any alternative drug that competes but 

merely begins to create the path to one does not come within patent 

law’s “research exemption.” 11 These attitudes—none forced by the 

language of the statutes—reflect a judicial temperament that seems to 

think of copyright and patents in a Blackstonian “sole and despotic 

dominion” 12 mindframe, an approach that was never true of real 

property under common law, and would, even in terms of pure theory, 

be disastrous if applied to knowledge and information. The idea would 

be to develop a relatively robust space for experimentation which, if it 

led to products and sales, would entitle the owner of the prior, enabling 

innovation or creative expression to claim some share of the profits of 

the downstream innovator or creator. The critical point of such an 

approach would be to allow millions of experiments to run without 

liability or its risk, while at the same time assuring that truly enabling 

innovations for those experiments that do succeed can share in the 

commercial upside of their contributions to downstream innovation. 

� Continue to expand the exemptions from the Digital Millennium 

Copyright Act 13 wherever that Act’s provisions place a drag on 

interoperability and innovation in systems that depend on access 

to existing platforms and systems: Federal courts have begun to 

reject claims under the DMCA that are efforts by copyright owners to 

use digital rights management to throw a monkey wrench into the 

works of a competitor. For example, Lexmark tried to make it hard for 

competitors who wanted to compete on toner for its printers by 

creating a chip and software handshake between the printer and the 

toner cartridge. When a competitor reversed engineered the handshake 

so that their microchip-enabled toner cartridge could work with 

10 Bridgeport Music, Inc. v. Dimension Films, 410 F.3d 792 (6th Cir. 2005), available at 

http://en.wikipedia.org/wiki/Bridgeport_Music,_Inc._v._Dimension_Films. 

11 Madey v. Duke University, 307 F.3d 1351 (2002). 

12 SIR WILLIAM BLACKSTONE, COMMENTARIES ON THE LAWS OF ENGLAND, Book II Ch. II , 

Clarendon Press (Oxford) 1765-1769, available at 

http://en.wikipedia.org/wiki/Commentaries_on_the_Laws_of_England. 

13 17 U.S.C. §§ 1201-1205, available at 

http://en.wikipedia.org/wiki/Digital_Millennium_Copyright_Act#Anticircumvention_exemptions.


Lexmark printers, Lexmark argued that in order to build their 

competing cartridge, the competitor had to make a copy of the 

handshake software, which in turn required them to get around the 

encryption protecting that piece of copyrighted software. In other 

words, the competitor had violated the DMCA by circumventing the 

digital rights management encryption that protected their copyrighted 

handshake software. The court rejected the argument, emphasizing 

that a program copy whose core function was interoperability did not 

violate the DMCA. 14 

In this short a piece, I neither aim for an exhaustive list nor offer a detailed 

analysis of each of the proposals identified. Instead, I offer these as an initial 

draft of a range of policies that would increase freedom to operate in the 

networked information economy, and reduce the drag of the current system of 

copyrights and patents on both commercial and social enterprises that have 

played a critical role in the explosive innovation we have experienced on the 

Internet in the past decade and a half. 

Conclusion 

The networked information environment has introduced a period of radically 

decentralized capitalization of some of the core economic sectors in the most 

advanced economies. As a result, growth is coming to depend increasingly on 

innovation from individuals and companies at the edges, operating as the few 

successful experiments out of thousands of similar experiments that go 

nowhere. Many of these experiments are commercial. Many are noncommercial. 

Many combine the two. As a system, this open, chaotic, complex 

innovation system requires freedom to operate. It needs to take advantage of its 

technical, economic, and social structure more than it needs power to control 

uses as in prior models of well-behaved appropriation. Moreover, the diversity 

of models of experimentation, and the increasingly fuzzy line between 

production and consumption, between the social and the economic, suggest 

that for purposes of economic production and growth, formal contract and 

corporate structure are playing a less important role than they did in the prior 

century relative to the increasingly important role played by loosely-structured 

voluntarism and human sociality. 

14 Lexmark International, Inc. v. Static Control, 387 F.3d 522 (6th Cir. 2004).




From Dismal Science 

to Strange Tales 

By Larry Downes * 

Heroes 

It was a fight over nothing. 

In 2008, 12,000 members of the Writers Guild of America staged a withering 

strike against the major Hollywood studios. It lasted three months, interrupted 

dozens of TV series, and delayed several big-budget films. The two sides 

reportedly lost more than $2 billion. Yet the sole issue in the dispute was when 

and how revenues from the Internet and other digital distribution of 

entertainment would be allocated. 1 

So far, no such revenues exist. 

Online distribution of movies and especially TV is a recent phenomenon, 

powered by ever-faster data transmission speeds, the continued spread of 

broadband technologies into the home, and improved protocols for file 

compression. It seems certain that profitable models for delivering Hollywood 

content to computers, personal digital assistants (PDAs), cell phones, and other 

non-TV devices will emerge. But in these early days, as with music before it, it 

isn’t clear what those models will be. Will they be supported by advertising? 

Will content be pay-per-view or based on all-you-can-eat subscriptions? Will 

consumers prefer to own or rent? 

As industry ponders these unanswerable questions, consumers are doing much 

of the innovating themselves as they did with earlier, less bandwidth-intensive 

content such as text and music. Users of YouTube, BitTorrent, and all 

variations of video streaming or file-sharing applications, in the interest of speed 

* Larry Downes is an Internet analyst and consultant, helping clients develop business 

strategies in an age of constant disruption caused by information technology. He is the 

author of UNLEASHING THE KILLER APP: DIGITAL STRATEGIES FOR MARKET 

DOMINANCE (Harvard Business School Press 1998) and, most recently, of THE LAWS OF 

DISRUPTION: HARNESSING THE NEW FORCES THAT GOVERN LIE AND BUSINESS IN THE 

DIGITAL AGE (Basic Books 2009). This essay is adapted from THE LAWS OF DISRUPTION. 

1 Michael White & Andy Fixmer, Hollywood Workers Return to Work After Ending Strike, 

BLOOMBERG, Feb. 13, 2008, http://www.bloomberg.com/apps/news? 

pid=newsarchive&sid=aKdwR9oC54WM.


and experimentation, do not bother with the niceties of obeying the law 

(Viacom’s $1 billion lawsuit against YouTube and Google is currently on 

appeal). 2 

Why did the two sides risk so much fighting over revenue that doesn’t yet exist 

from channels that haven’t been invented? The writers say they took a stand in 

large part because they did not do so in the early days of videocassette sales and 

rentals. When the profitable models finally arrived, the writers believed they got 

a much worse deal. Because media sales and rentals now represent the largest 

share of entertainment income, missing the boat has been painful for writers. 

The studios argued that until it is clear how and when money is to be made 

from digital distribution, pre-assigning residual royalties to writers would limit 

the studios’ ability to experiment with different distribution and partnership 

models. They made the same argument with videotapes. 

Ultimately, new rates for residual royalties were agreed upon for categories 

including downloaded rentals and sales, ad-supported streaming media, short 

clips, and promotional uses. Whether these prove to be favorable rates, or even 

the right categories, remains to be seen. Either way, media will continue to 

migrate to the Internet at the expense of other forms of distribution. 

The Strange Behavior 

of Non-Rivalrous Goods 

It is hard to say if anyone made the right decisions in the writers’ strike, in part 

because the tools for valuing information products and services, even for 

present uses, are terrible. You will look in vain at the balance sheets of 

companies whose sole assets are information—including much of the 

entertainment industry, as well as professional services such as doctors, lawyers, 

and consultants—to find any useful measure of the current or future value of 

the company’s real assets. While management gurus sing the praises of 

developing a company’s intellectual capital, financial reporting systems ignore it. 

Accountants refer to all the valuable information in a business—its information 

assets—as intangibles. As the name suggests, these are assets that never take a 

physical form as do factories and inventory. Unlike physical assets, information 

assets are generally not counted in calculating the total worth of an enterprise. 

For the most part, a company’s human resources, brands, and good 

relationships with customers and suppliers—let alone its copyrights, patents, 

trade secrets, and trademarks—are left off its balance sheet. The value of the 

2 See Adam Ostrow, Viacom Loses $1 Billion Again YouTube, MASHABLE, June 23, 2010, 

http://mashable.com/2010/06/23/youtube-wins-viacom-lawsuit/.


company’s information, at least as far as accounting is concerned, is basically 

nothing. 

Why? Accountants have argued for years that information and other intangibles 

are so different economically from material goods that traditional methods of 

valuation just don’t apply. As an asset, the explanation goes, information 

behaves precisely the opposite of its tangible counterparts: Capital assets lose 

value as they are used, equipment becomes obsolete, and raw materials are 

depleted. Brands and reputations, by contrast, become more valuable the more 

they are exercised, in theory generating revenue forever. You cannot determine 

the price of a logo or a customer relationship with the same tools you use to 

depreciate a tractor. 

Fair enough. But that doesn’t explain why accountants have done so little to 

develop valuation techniques that apply to information assets. A dangerous 

result of that failure is that few managers understand information or how it 

generates value. Even CEOs of large companies regularly get it wrong when 

they talk casually about “trademarking an idea” or “copyrighting a word.” (You 

cannot do either.) 

As information becomes more central to economic performance, the failure to 

account for its value has become dangerous. Executives, especially in public 

businesses, are compensated based on the health of their companies’ balance 

sheets. To the extent that information value doesn’t appear there, it’s 

understandable that many companies don’t put much, if any, effort into 

developing or managing those assets. 

That’s unfortunate because the strategic cultivation of information assets is 

beneficial in many ways. Consider Harrah’s Entertainment, which operates 

casinos worldwide in places where gambling is legal. 

When former business school professor Gary Loveman joined the company as 

chief operating officer in 1998, he decided to look for underutilized assets on 

the company’s balance sheet. He found them in Harrah’s data warehouse. Like 

most casino chains, Harrah’s had implemented a rewards program that gave 

customers special benefits for using their membership cards while playing slot 

machines. Harrah’s was collecting vast amounts of information on its “factory 

floor,” but had done very little to put that data to use. 

A detailed review of the collected information upended several long-standing 

myths about where Harrah’s made the most money. Most of the company’s 

profits came from a quarter of its customers. Those customers were not, 

however, the “cuff-linked, limousine-riding high rollers [Harrah’s] and [their]


competitors had fawned over for many years.” 3 Instead, Harrah’s discovered 

that the high-profit customers were regular visitors, many of them recent 

retirees. They made frequent trips to the casino and spent steadily, if modestly, 

at its gaming tables, restaurants, and hotels. 

Harrah’s quickly reconfigured its customer-facing activities, including check-in, 

complimentary meals, and special promotions, orienting them toward the actual, 

as opposed to the presumed, best customers. The result was a changed 

enterprise, one that consistently outperforms its competition. 

Even though the balance sheet never reported the value of the diamonds 

Loveman found when he looked in his data mine, his information assets were 

by no means worthless. In 2006, Harrah’s was sold to a private equity 

partnership at a price that valued the information at more than $1 billion, 

representing a 30% premium in the total purchase price. 4 Today, Gary 

Loveman remains CEO of the company, a position he has held since 2003. 

The writers’ strike and the Harrah’s story teach an important lesson about the 

economics of information. Just because information value is indeterminate 

doesn’t mean it’s worthless. Not by a long shot. The Hollywood writers and 

producers clearly did not think so, nor did the buyers of Harrah’s. 

Consider another example: Search giant Google has $20 billion in assets, mostly 

cash, on its balance sheet. The company, however, even on the lowest day of 

the stock market in ten years, was worth nearly $100 billion—more than five 

times its book value. Somebody has figured out, at least in part, how to value 

the company’s information assets. 

Digital life is made up of information. It comes in a wide range of types, 

including private data, speech, news and entertainment, business practices, and 

information products and services such as films, music, inventions, and 

software. But all information operates under a common set of economic 

principles. So to thrive in the next digital decade, you must understand the 

basic elements of information economics. 

In modern economic terminology, goods are categorized as either “private” or 

“public” goods. Most goods in our industrial economy are private goods. 

Purely private goods are those that can be possessed by only one person at a 

3 Gary Loveman, Diamonds in the Data Mine, HARV. BUS. REV., May 2003. See also Julie 

Schlosser, Teacher’s Bet, FORTUNE, March 8, 2004, http://money.cnn.com/magazines/ 

fortune/fortune_archive/2004/03/08/363688/index.htm. 

4 Ryan Nakashima, Harrah’s Entertainment Accepts Buyout Bid from Private Equity Group, USA 

TODAY, Dec. 19, 2006, http://www.usatoday.com/money/industries/2006-12-19harrahbuyout_x.htm.


time (“rivalrous”) and whose use can be limited to that person or with 

whomever she might share it (“excludable”). If I own a barrel of oil, then you 

don’t own it, unless I sell it to you, in which case I no longer have it. Once it’s 

used, it’s gone forever—no one has it anymore. 

Public goods, by contrast, can be used by more than one person at the same 

time (“non-rivalrous”), and limiting access to them is difficult, if not impossible 

(“non-excludable”). The classic example in economics textbooks is national 

defense. Either everyone has this good or nobody does. The military protects 

everyone, including those who do not pay taxes. Defensive missiles cannot be 

programmed to leave a single house unguarded. 

Information is an archetypically non-rivalrous good. As Thomas Jefferson 

famously wrote, “He who receives an idea from me, receives instruction himself 

without lessening mine; as he who lights his taper at mine, receives light without 

darkening me.” 5 Once a composer completes a song, there’s no physical limit 

to how many people can perform it simultaneously. There is a cost associated 

with its creation, but the composer incurs no additional cost no matter how 

many times the work is played. Regardless of how often it is performed, the 

composition still exists. In fact, it becomes more valuable the more freely it’s 

shared—it becomes more popular, maybe even a “hit.” 

So information is non-rivalrous, but is it also excludable? Until recently, the 

answer in practice was often no. That’s because many information products 

that sprang from the creativity of the human mind could not easily be 

distributed without first being copied to physical media such as books, 

newspapers, or, in the case of music, CDs and records. In that transformation 

(“demassification,” in Alvin Toffler’s terminology 6), information lost its 

excludable property, looking more like the barrel of oil than like national 

defense. It’s easy to limit access to the barrel of oil—there’s only one, after all. 

The song, once recorded and duplicated, is harder to control, but it’s still 

possible to exclude those who didn’t pay for a copy or pay for the right, as in 

radio, to broadcast it. 

Information, until recently, was a public good in theory but in practice behaved 

more like a private good. The need to reduce it to physical media masked its 

true nature, and gave rise to seemingly incongruous terminology that includes 

“stealing an idea,” “pirating content,” and, most significantly, “intellectual 

property.” After more than five hundred years of Gutenberg’s moveable type, 

we’re so conditioned to experiencing information through mass-produced 

media that we equate the cost of the media with the value of the content. As 

5 Letter from Thomas Jefferson to Isaac McPherson (Aug. 13, 1813), available at 

http://press-pubs.uchicago.edu/founders/documents/a1_8_8s12.html 

6 See generally ALVIN TOFFLER, THE THIRD WAVE (1980).


John Perry Barlow poetically put it, until the advent of electronic distribution 

through the Internet, “the bottle was protected, not the wine.” 7 

For information products, that protection—the ability to exclude—is almost 

entirely a function of law: The law of copyright makes it a crime to “copy” 

information without permission. Copyright gives the composer the exclusive 

right to make or authorize performances of a song, for example, or to record it 

and produce copies. At the same time, copyright outlaws the production of 

copies by anyone else, including someone who purchased a legal copy. 

By limiting both the performance and production of a song, copyright 

transforms non-rivalrous information into a rivalrous physical good. But the 

alchemy of copyright is starting to fail as digital technology makes it easier to 

distribute songs electronically. Given the Internet, it’s now much harder to 

limit who gets to hear a song and when. A copy no longer requires expensive 

recording and pressing equipment, or access to a costly and very visible retail 

distribution network. 

Although the composer can legally exclude those who do not buy authorized 

copies of his work, his ability to police that right is increasingly expensive, often 

costing more than it’s worth. You can’t realistically stop people from humming 

your tune, even if they do it out loud. And now you can’t really stop them from 

sharing copies of a digital recording, either. 

Other than the composer herself, however, consumers are also potentially 

harmed by the transformation of information goods back to their non-rivalrous 

state. There were and remain important reasons for legal systems that treat 

information as if it were a rivalrous good. 

Copyright, for example, is designed to maximize the value of the up-front 

investment that information producers must make. If copyright didn’t exist, 

you could simply buy a recording of the song, reproduce it, and sell your own 

copies. Because your total investment would be only the cost of a single copy, 

your version would likely be cheaper than the one marketed by the composer 

himself. In theory, the composer would find it difficult to recover his creative 

investment, making him less likely to undertake his important work in the first 

place. Ultimately, everyone would be worse off. 

But copyright’s value comes at a high price. By imposing costs on the exchange 

of information that otherwise would not exist, the law neutralizes many of the 

valuable features of non-rivalrous goods. 

7 John Perry Barlow, The Economy of Ideas, WIRED 2.03, 1994, 

http://www.wired.com/wired/archive/2.03/economy.ideas_pr.html.


Fortunately, this special power is limited. Even with copyright, some forms of 

sharing are perfectly legal. Libraries can loan out the same copy of a recording 

to as many people as want to hear or play it, one at a time. Fans who purchased 

their own copy are likewise free to loan it to their colleagues, or even to resell 

their copy to a used record store or through online services such as Amazon 

Marketplace or eBay. 

Copyrights also have an expiration date. In the United States, for example, 

copyrights lasts 70 years beyond the life of the author or 120 years for certain 

works. 8 After this period, the work is no one’s property—the public can use it 

however they want to. Anyone can perform the work, make copies of it, adapt 

it, or incorporate it into new works. It becomes forever after a purely nonrivalrous 

good. 

Copyright protection is also limited to the producer’s particular expression and 

not the underlying ideas. The ideas in a song (love conquers all, love stinks) are 

non-rivalrous from the moment the song is written. 

Consider a 1996 court case involving sports statistics. In the days before the 

Web and wireless data devices, sports fans who were not attending a game 

could get up-to-the-minute information from a dedicated paging device from 

Motorola called Sportstrax. Sportstrax employees watched sporting events on 

TV and entered key information (e.g., who had the ball or who had scored) into 

a computer system. A few minutes later, Sportstrax customers would be paged 

with short updates. 

Motorola was sued by the National Basketball Association, which claimed the 

transmission of information by pager violated its copyright in the broadcast of 

games. 9 The court disagreed. Sporting events are not “authored,” the judges 

noted, and are therefore not protected by copyright in the first place. Game 

data, including interim scores, are facts, not a particular expression of an 

information producer. Facts are non-rivalrous, outside the protection of 

copyright. 

Today, pagers have given way to cell phones that can take photographs and 

videos and share them via the Internet. Popular television programs such as 

“American Idol” have armies of fans who watch the show and write blogs 

about each performance even as they’re watching them. So long as the actual 

performances aren’t being copied, however, the commentary is perfectly legal. 

8 17 U.S.C §§ 302- 03. 

9 Nat’l Basketball Assoc. v. Motorola, Inc., 105 F.3d 841 (2d Cir. 1997), 

http://www.law.cornell.edu/copyright/cases/105_F3d_841.htm.


As these examples suggest, the challenge for copyright and other laws 

controlling information has always been to strike the right balance between 

incentives for creators and the value that the public derives from unfettered use. 

It’s a balance that is constantly being unsettled by new technologies, a problem 

that has accelerated with the advent of the digital age. 

On one hand, information technology has greatly lowered the cost of creating 

and distributing information including books, movies, and recorded music. But 

the same technologies have also made it easier to make and distribute 

unauthorized copies which are, in many cases, perfect replicas of the original. 

Should information laws be tightened or relaxed in the next digital decade? Do 

producers need more protection from copyright laws, or do consumers deserve 

greater freedom? Are new information uses made possible by software 

applications such as YouTube, Facebook and Flickr, creating more value than 

they destroy, and for whom? 

The Five Principles 

of Information Economics 

Unfortunately, many of those debating these questions—and there are many, 

including lawmakers, industry leaders, and consumer groups—don’t understand 

the economic properties of information any better than do the accountants who 

refuse to measure it. So it’s worth summarizing the five most important 

principles of information economics. It’s even better to memorize them: 

Renewability 

Information cannot be used up. It can be enhanced or challenged, it can 

become more or less valuable over time, but once it has been created, it can be 

used over and over again. In the end it exists as it began. Most new 

information, moreover, is created from other information, making it a 

renewable energy source. In electronic form, neither its production nor its use 

generates waste products that damage the environment. In that sense, 

information is the ultimate “green” energy. 

The online encyclopedia Wikipedia, for example, isn’t written by hired experts. 

It’s written by volunteers who post articles on subjects they either know or 

think they know something about. Within certain limits, anyone else can edit, 

correct, or change entries. Over time, the articles evolve into a useful and 

reliable form. No money changes hands in either the creation of Wikipedia or 

its use.


Universality 

Everyone has the ability to use the same information simultaneously. Blog 

entries, news articles, and YouTube videos can be enjoyed simultaneously by an 

unlimited number of people. The only distribution costs are the photons on a 

screen. Each consumer, moreover, may have a completely different reason for 

consuming the same information, and perhaps her own response to it. She may 

be inspired to respond with information of her own. 

Facebook, for example, is comprised almost entirely of user-generated content. 

Users are constantly commenting on status updates, photographs, and other 

information posted by their friends, and inviting each other to join interest 

groups or play information games. 

Magnetism 

Private goods operate under the law of supply and demand: The greater the 

supply, the lower the price you can charge for it, and vice versa. The value of 

information, on the other hand, increases as a function of use. Information 

value grows exponentially as new users absorb it. The more places my brand or 

logo appears, the higher the value customers attach to all my goods. Use makes 

the brand more, not less, valuable. 

This increase in value accelerates as the information spreads, creating a kind of 

magnetic pull that generates network effects. Since no one owns the Internet’s 

protocols, for example, these standards have spread easily, resulting in the 

explosive growth that began in the 1990s. The standards are now more valuable 

than when only a few people used them. 

Friction-Free 

The more easily information flows, the more quickly its value increases. In 

electronic form, information can move in any direction at the speed of light. It 

experiences no decay along the way, arriving at its destinations in the same form 

as when it departed. For many kinds of information, including languages, 

religious doctrines, and advertising, the ease of transfer helps to improve 

society—or at least the profits of those who disseminate it. The cheaper it is to 

spread the word, the more likely and quickly it will be spread. 

There is, however, an inherent paradox: The frictionless spread of information 

can undermine the incentives for its production. Content producers, including 

authors, musicians, news organizations, and movie studios, invest heavily in the 

production of new information. To recover their investment, information 

producers must charge for its use. Economically, however, even the simplest 

payment schemes (subscriptions, for example) slow the natural tendency of 

information to move freely. Since information flows along the path of least


resistance, markets look for ways to avoid fees. In that sense, new technologies 

often subvert old business paradigms, even when the inventors of those 

technologies didn’t intend for them to do so. 

Vulnerability 

Information’s value does share one property with tangible goods: It’s not 

indestructible. Value can be destroyed through misuse. If you license your 

company’s name (and thus its reputation) to an inferior product or a product 

that does not have a clear connection to your brand, you risk confusing 

consumers about what your brand stands for. Value can also be destroyed by a 

third party, perhaps a competitor offering a knockoff product that looks like 

yours but is of lesser quality. Or an identity thief can appropriate your name 

and credit history to borrow money from banks or credit card companies. 

When the thief disappears, not only is the money gone, so is your reputation. 

Information can also be a victim of its own success. It is now so easy to 

produce, distribute, and consume information that users are experiencing 

overload. Today, websites, e-mails, blogs, text messages, and even “tweets”— 

brief messages that reflect the thoughts of a user on Twitter—all compete for 

users’ limited time. As the sources of information and the volume produced 

expand rapidly, consumers find it increasingly difficult to limit their exposure to 

information of real value to them. 

* * * 

It’s easy to see these five principles in action in digital life. Consider Google. 

One might wonder how a company can be worth anything, let alone $100 

billion, when it charges absolutely nothing for its products and services. You 

can search Google’s databases and use its e-mail service day and night without 

spending a penny; you can store photos on its Picasa photo service, create 

documents with its online word processing software, view Google Maps, and 

share videos on its YouTube service for free. Indeed, the company is 

determined to have as many people as possible take complete advantage of it. 

Even though databases and other services are what consumers want and 

therefore represent the source of the company’s value, Google does not hoard 

those assets as if they were barrels of oil to be used only when necessary. It 

treats them instead as non-rivalrous goods that increase in value as more people 

use them. 

The company isn’t being generous. Google makes nearly all its money by 

renting out advertising space to companies whose products and services 

complement the things consumers do when they are using Google. If 

information “wants to be free,” then let it be as free as possible, and make all 

the profits from the collateral effects of the network. That’s the company’s


simple business strategy, one that has created remarkable new value even as it 

disrupts the assumptions of every industry the company touches. 

And for social networking sites, just giving away content isn’t good enough. 

These companies also have to find ways to get users to help them develop their 

sites in the first place. Companies like Facebook, MySpace, and their 

professional equivalent, LinkedIn, are constantly adding free tools and gadgets 

to make their products more compelling. Once a service reaches the tipping 

point (Facebook now has 500 million users!), the search for ways to make 

money begins in earnest, including premium services and targeted 

advertisements. But thanks to the weird economics of information, there 

remain powerful reasons not to charge the users for the core product—ever. 

The Problem of Transaction Costs 

There’s one additional aspect of information economics that is essential to 

understand. The frictionless transfer of information and the problem of 

information overload suggest that the economy of digital life is a kind of 

machine. Like the best engines, it can operate with remarkable efficiency— 

provided its parts are kept lubricated and free of foreign matter. Already, 

technologies perfected during the last digital decade have ruthlessly eliminated 

waste in our increasingly efficient online lives. Still, the information economy is 

not perfect. It suffers, like its physical counterpart, from a kind of inefficiency, 

what economist Ronald Coase first called “transaction costs.” 

Coase came to the United States from England as an economics graduate 

student in 1931. Only twenty years old, Coase had a revolutionary agenda. 

Struggling to reconcile the socialism of his youth with the free-market sensibility 

of his professors, Coase saw big companies as proof that centralizing activities 

could work on a grand scale. If he could learn how big companies did it, Coase 

imagined, then perhaps the lessons could be applied to big governments as well. 

Oddly enough, no one had ever asked why companies existed, and certainly no 

one had ever thought to ask the people who were running them. 

What Coase learned made him swear off socialism forever, and led to the 

publication of an article that changed economic thinking forever—an article 

cited as revolutionary sixty years later, when Coase received the Nobel Prize in 

Economics. 

In “The Nature of the Firm,” Coase argued that there is a price not only to 

what companies buy and sell, but also to the process of buying and selling it. 10 

Buyers and sellers have to find each other, negotiate deals, and then 

10 Ronald H. Coase, The Nature of the Firm, 4 ECONOMICA 368-405 (Nov., 1937), 

http://aetds.hnuc.edu.cn/uploadfile/20080316211913444.pdf.


consummate them. This activity was neither especially easy nor without costs. 

Coase therefore argued that companies were becoming bigger because markets 

were, relatively speaking, too expensive. 

Coase called the price of doing a deal its “transaction cost.” The existence of 

transaction costs, he believed, explained why companies were internalizing more 

and more activities, especially repeated functions like buying raw materials and 

marketing. For these activities, maintaining an inside function such as a 

purchasing department was cheaper than relying for each individual purchase on 

whoever might happen to be in the market. 

To understand why, let’s take a simple example. Say you work for an averagesize 

company and you’ve run out of paper clips. Almost assuredly, you will get 

your paper clips not by leaving your office to drive to the office supply store but 

by going down the hall to the supply cabinet, where your company’s purchasing 

department maintains an inventory of basic supplies. Your company will, in 

fact, keep such basic supplies on hand as a matter of course, without giving 

much thought to the cost of carrying this inventory. This holds true even if 

buying and distributing office supplies have nothing to do with what your 

business does. Your company is likely to keep paper clips on hand even if there 

is no discount for buying in bulk. 

Why? Even if you could get paper clips on your own for the same price, you 

still have to go out and get them. This means finding the stores that carry them 

and learning how much they charge. Then you have to choose between the 

closest store and the one with the best price. At the checkout stand, you need 

to make sure you are really charged what the store advertises. If the clips are 

somehow defective, you have to take them back and demand replacements or 

some other remedy. 

And that’s just for a simple transaction. Imagine instead that you’re buying raw 

materials needed to manufacture a jet airplane. There is the additional effort of 

negotiating a price, writing a contract, inspecting the goods, and, potentially, 

invoking the legal system to enforce the terms and conditions. It’s better, you 

say, to own the supplier or at least to buy in bulk and avoid all that trouble. 

That “trouble” is transaction costs. 

Working from Coase’s basic idea, economists have identified six main types of 

transaction costs: 

� Search costs: Buyers and sellers must find each other in increasingly 

diverse and distributed markets. 

� Information costs: For buyers, learning about the products and 

services of sellers and the basis for their cost, profit margins, and


quality; for sellers, learning about the legitimacy, financial condition, 

and needs of the buyer, which may lead to a higher or lower price. 

� Bargaining costs: Buyers and sellers setting the terms of a sale, or 

contract for services, which might include meetings, phone calls, letters, 

faxes, e-mails, exchanges of technical data, brochures, meals and 

entertainment, and the legal costs of contract negotiations. 

� Decision costs: For buyers, comparing the terms of one seller to 

other sellers, and processes such as purchasing approval designed to 

ensure that purchases meet the policies of the organization; for sellers, 

evaluating whether to sell to one buyer instead of another buyer or not 

at all. 

� Policing costs: Buyers and sellers taking steps to ensure that the good 

or service and the terms under which the sale was made, which may 

have been ambiguous or even unstated, are translated into the behavior 

expected by each party. This might include inspecting the goods and 

any negotiations having to do with late or inadequate delivery or 

payment. 

� Enforcement costs: Buyers and sellers agreeing on remedies for 

incomplete performance. These include everything from mutual 

agreements for a discount or other penalties to expensive litigation. 

As this list suggests, transaction costs range from the trivial (turning over a box 

of paper clips to see the price) to amounts greatly in excess of the transaction 

itself (imagine if you were seriously injured by a defective paper clip flying off 

the shelf and sticking you in the eye). In fact, economists Douglass North and 

John Wallis have estimated that up to 45%of total economic activity consists of 

transaction costs. 11 Eliminating them entirely would translate to a staggering 

$4.5 trillion in annual savings in the United States alone, eliminating much of 

the work done by accountants, lawyers, advertisers, and government agencies. 

One needn’t go that far to improve economic performance, however. Firms are 

created, Coase concluded, because the additional cost of organizing and 

maintaining them is cheaper than the transaction costs involved when 

individuals conduct business with each other using the market. Firms, while 

suffering inefficiencies of their own, are more efficient at certain types of 

activities than the market. Technologies—in 1937, Coase had in mind 

telephones, in particular—improved the performance of one or both, constantly 

resetting the balance between what was best to internalize and what was best 

left to the market. 

11 John Joseph Wallis & Douglass C. North, Measuring the Transaction Sector in the American 

Economy, in LONG-TERM FACTORS IN AMERICAN GROWTH 95-162 (Stanley L. Engerman & 

Robert E. Gallman, eds. 1986), http://www.nber.org/chapters/c9679.


So which functions should a firm perform internally? The deceptively-simple 

answer is only those activities that cannot be performed more cheaply in the 

market or by another firm. In fact, as Coase says, a firm will tend to expand 

precisely to the point where “the costs of organizing an extra transaction within 

the firm become equal to the costs of carrying out the same transaction by 

means of an exchange on the open market.” 12 

For some activities, say plumbing, the open market works relatively well, and 

the need for plumbers to form large firms to avoid transaction costs has never 

arisen. For the large-scale operations of integrated manufacturers, such as 

Boeing and General Motors, which require coordination, heavy capital 

investment, and complex distribution systems, the firm is the only economically 

viable solution. 

Coase believed economists should turn their attention to the practical problem 

of uncovering transaction costs wherever they occur and eliminating those that 

are unnecessary. Doing so, he hoped, would, among other things, help reduce 

the need for government intervention. A great deal of regulation and liability 

laws, Coase argued, were unconscious efforts to overcome transaction costs for 

certain types of activities, such as accidents and pollution. But the regulations 

themselves generate so many transaction costs that in many cases doing nothing 

at all would have produced a better result. To find out how much law and 

regulation are optimal requires a better understanding, once again, of the costs 

involved. 

Coase had hoped his elegant proof would get economists working on the real 

problem at hand. Ironically, all he did was make economics more esoteric. 

Instead of lowering themselves to the kind of empirical research that was 

common in other social sciences, economists simply dispose of Coase in an 

opening footnote. They “assume a frictionless economy” and then proceed to 

develop elaborate mathematical models of behavior in a purely theoretical 

universe. Rather than join his quest, most economists retreated to more 

abstract models of economic behavior, which Coase dismisses as little more 

than a “vast mopping-up exercise” of loose ends left by Adam Smith’s seminal 

18 th century work, The Wealth of Nations. 

Increasingly frustrated with his economist colleagues, Coase instead took up 

residence at the University of Chicago’s law school. Economists, he came to 

see, avoided information, and misused the few sources, such as government 

data, that were readily available. Economics had become a shell game. “If you 

torture the data enough,” he wrote, dismissing much of modern economic 

12 Coase, The Nature of the Firm, supra note 10.


analysis, “nature will always confess.” 13 If that was all economists could do, 

Coase decided he was no economist. Awarded the Nobel Prize in 1991, Coase 

began his acceptance speech on a note of despair. “In my long life I have 

known some great economists,” he told the committee, “but I have never 

counted myself among their number nor walked in their company.” 14 

Look at the performance of the economy over the past twenty years and it’s 

easy to sympathize with Coase’s frustration. The “rational” stock market still 

booms and busts. Cyclical industries continue to overexpand and then 

overcontract. Efforts at creating an open global economy without trade barriers 

are met with rioting mobs. National banking regulators read every tea leaf they 

can find and still go to bed wondering if they have cut rates too soon or too 

late, too much or too little, or even if their cuts have made an iota of difference. 

While most economists fiddle with formulas, the economy is burning. Without 

a better understanding of the nature of transaction costs, we’ll never be able to 

predict—let alone improve—what seem to be the most basic elements of 

economic behavior. 

That, in any case, is the real world. In the digital world, the problem is not only 

less severe, but also solvable. The free flow of information made possible by 

digital technology is decreasing the friction of transaction costs in a variety of 

interactions. From global price comparisons to searches of much of the world’s 

knowledge to auctions for anything, the cost of deal-making is plummeting. 

The Internet is driving down all six types of transaction costs. That’s what’s 

made the Internet so disruptive in the last decade, and what will continue to 

drive dramatic consumer, business, and regulatory changes in the next digital 

decade. 

Consider a few examples: 

1. Search costs: Technology connects people across geographical, time, and 

national borders. Automatic notifications for obscure collectibles on eBay, 

finding old friends through the “People You May Know” feature on 

Facebook, or letting your TiVo pick programs for you that it thinks you 

might like to watch—each of these reduces search costs, sometimes 

dramatically. Restaurant and other business reviews available directly on 

cell phones make it easier to find just the right place no matter where you 

are. There’s even an iPhone application uses GPS technology to help you 

find your car in a crowded parking lot! 

13 Ronald H. Coase, How Should Economists Choose?, G. Warren Nutter Lecture in Political 

Economy, American Enterprise Institute (1982). 

14 Ronald H. Coase, Nobel Prize Lecture, Dec. 9, 1991, 

http://nobelprize.org/nobel_prizes/economics/laureates/1991/coase-lecture.html.


2. Information costs: Technology creates standard data structures that can be 

searched and consolidated over a growing network of computers. The 

asymmetry of sellers concealing data has eroded, radically changing the way 

people buy cars, real estate, and investment securities. Free or subscription 

services including CarFax, Zillow, and Yahoo! Finance give buyers an 

abundance of valuable information that was previously inaccessible at any 

price. Online dating services such as Chemistry.com increasingly use 

sophisticated profiling technology to suggest compatible matches. 

3. Bargaining costs: The exchange of information can now take place 

digitally and is captured in databases for easy reuse in subsequent 

transactions. Instant publication of classified ads on Craigslist means many 

local transactions are completed within minutes. Business-to-business 

transactions increasingly rely on libraries of standard terms. The nonprofit 

Association for Cooperative Operations Research and Development 

(ACORD), for example, uses the XML data standard to create standard 

forms used by insurance and reinsurance agents and brokers offering life, 

property, and other lines of products. 

4. Decision costs: Visibility to expanded online markets gives both buyers 

and sellers a better picture of minute-to-minute market conditions. Several 

insurance websites, including Progressive.com, provide instant quotes and 

comparisons to the prices of their competitors. Cell phone users can 

compare prices from online merchants while shopping at retail stores, 

putting added pressure on merchants to match or beat those prices or offer 

other incentives, including delivery or after-sales support. Online gamers 

can check the reputation of potential participants to decide whether to 

allow them to join their teams. 

5. Policing costs: Transactions conducted with system-to-system data 

transfers create a more complete record of the actual performance of the 

participants, which can then be captured and queried. For goods purchased 

online, most merchants now provide direct access to detailed shipping and 

tracking information from expediters such as UPS or FedEx or even 

standard delivery from the postal service. Some merchants, including Dell 

Computers, provide information about the manufacturing process, allowing 

customers to track their products before they are even shipped. Most 

software products now collect bug and other failure information in real 

time, automatically installing updates and repairs. Players of the online 

World of Warcraft game can “speak” directly to in-game employees or 

robots whenever they have a problem. 

6. Enforcement costs: Electronic records can simplify the process of 

resolving disputes over what was agreed upon or what did or did not occur. 

Online payment services such as PayPal offer elaborate dispute resolution


functions that include mediation and arbitration when buyers and sellers 

cannot resolve their differences, along with insurance and guaranteed 

satisfaction. These are all supported by the collection of end-to-end 

transaction data documenting the actual performance of buyers and sellers. 

Bloggers can quickly whip up electronic mobs to put pressure on 

companies, politicians, or celebrities whose behavior they feel does not 

comply with agreed-upon standards. 

Conclusion: Conflicts at the Border 

Digital technology, as I argued in my 1998 book, “Unleashing the Killer App,” 15 

has created a corollary to Coase’s observation about business organizations. As 

transaction costs in the open market approach zero, so does the size of the 

firm—if transaction costs are nonexistent, then there is no reason to have large 

companies. For products constructed entirely or largely out of information, we 

now stand on the verge of what Don Tapscott and Anthony Williams call “peer 

production,” where just the right group of people come together to apply the 

right set of skills to solve complex problems, whether in business or 

otherwise. 16 I called this phenomenon “The Law of Diminishing Firms.” 

Technology is changing the dynamics of firms, making them smaller but more 

numerous. This, however, is good for the overall economy. Efficiency 

translates to savings of time, money, and decreased waste. Productivity, 

customer satisfaction, and the availability of customized products and services 

have improved dramatically. Keeping in touch across time zones and long 

distances gets easier, as does organizing diverse groups of people for social, 

political, or business reasons. The average consumer can now edit an online 

encyclopedia, post news and photos as a citizen journalist, or operate a homebased 

business that can produce and distribute just about anything. 

Now for the bad news. Our current legal system, forged in the factories of the 

Industrial Revolution, was designed to maximize the value of rivalrous goods. 

It cannot be easily modified to deal with the unique economic properties of 

information. Worse, the crushing overhead of regulations and lawsuits, which 

may no longer be cost-effective even in the physical world, adds even less value 

when applied to the lower-transaction-cost-environment of digital life. 

Increasingly, the old rules do little more than hold back innovation for the 

benefit of those who cannot or do not know how to adapt to the economics of 

digital life. In many cases, inefficient laws are propped up by failing businesses 

15 LARRY DOWNES & CHUNKA MUI, UNLEASHING THE KILLER APP: DIGITAL STRATEGIES FOR 

MARKET DOMINANCE (1998). 

16 DON TAPSCOTT & ANTHONY WILLIAMS, WIKINOMICS: HOW MASS COLLABORATION 

CHANGES EVERYTHING (2003).


that are not eager to see their advantages erased. Sometimes those fighting 

transformation are powerful business interests, including large media 

companies, real estate agents, and even some of the technology companies 

whose products fuel the digital revolution. 

Resistance may also come from the users themselves. In digital life, private 

information can be invaluable in deciding who to interact with, either for 

business or for interpersonal transactions. As advances in technology bring 

more private information online, powerful emotions have been activated. 

Citizens in much of the world believe their rights to privacy are being violated, 

not only by businesses but by their classmates and neighbors. 

Perhaps most worrisome, governments are taking advantage of lower 

transaction costs to improve the technology of surveillance, raising fears of the 

dystopic world described by George Orwell in his novel “1984.” 

Lower transaction costs have also proven useful to criminals and terrorists, who 

operate freely and anonymously in digital realms. Sometimes their crimes 

exploit the vulnerability of information, as in the case of identity theft and other 

forms of Internet fraud. More ominously, virtual gangs are able to attack the 

infrastructure of the Internet itself, releasing viruses and other harmful software 

that incapacitate servers, destroy data, or, in the case of spam, simply waste 

people’s most precious resource: time. 

Perhaps the most difficult problems of information economics, however, 

involve the plasticity of information in electronic form. Technology has made it 

possible to realize the remarkable potential of information to be shared and 

even enhanced by as many people as are interested. Inevitably, every new 

innovation that supports this creative urge runs headlong into laws protecting 

information as property—laws that treat public goods as if they were private 

goods. Although such laws may be necessary, they have proven unduly rigid in 

their current form, sparking some of the most vitriolic fights on the digital 

frontier. 

The explosion of digital technology at home, at work, and in government, 

coupled with the economics of information, has created a perfect storm. Our 

industrial-age legal system will not survive this social transformation. After the 

flood, as in previous technological revolutions, a new legal paradigm will emerge 

to guide the construction of laws better suited to digital life. 

Implementing these new laws will require a great deal of coordination and 

collaboration. Most of all, it will require considerable courage on the part of 

those who live in both the physical and digital worlds. The next digital decade, 

like the last one, will proceed in fits and starts, with surprising changes of cast 

and characters, allies becoming enemies and enemies finding common ground.


Some winners and losers will prove, in retrospect, to have been easily predicted. 

Others will come from nowhere. 

The only thing certain is the author of the script: the poorly-understood but 

increasingly critical economic properties of information.



The Regulation of 

Reputational Information 

By Eric Goldman* 

Introduction 

This essay considers the role of reputational information in our marketplace. It 

explains how well-functioning marketplaces depend on the vibrant flow of 

accurate reputational information, and how misdirected regulation of 

reputational information could harm marketplace mechanisms. It then explores 

some challenges created by the existing regulation of reputational information 

and identifies some regulatory options for the future. 

Reputational Information Defined 

Typical definitions of “reputation” focus on third-party cognitive perceptions of 

a person. 1 For example, Black’s Law Dictionary defines reputation as the “esteem 

in which a person is held by others.” 2 Bryan Garner’s A Dictionary of Modern 

Legal Usage defines reputation as “what one is thought by others to be.” 3 The 

Federal Rules of Evidence also reflect this perception-centric view of 

“reputation.” 4 

* Associate Professor and Director, High Tech Law Institute, Santa Clara University School of 

Law. Email: egoldman@gmail.com. Website: http://www.ericgoldman.org. In 

addition to a stint as General Counsel of Epinions.com, a consumer review website now 

part of the eBay enterprise, I have provided legal or consulting advice to some of the other 

companies mentioned in this essay. I prepared this essay in connection with a talk at the 

Third Annual Conference on the Law and Economics of Innovation at George Mason 

University, May 2009. 

1 As one commentator explained: 

Through one’s actions, one relates to others and makes impressions on them. 

These impressions, taken as a whole, constitute an individual’s reputation— 

that is, what other people think of you, to the extent that their thoughts arise 

from what they know about you, or think they know about you. 

Elizabeth D. De Armond, Frothy Chaos: Modern Data Warehousing and Old-Fashioned Defamation, 

41 VAL. U.L. REV. 1061, 1065 (2007). 

2 BLACK’S LAW DICTIONARY (8th ed. 2004). 

3 BRYAN A. GARNER, A DICTIONARY OF MODERN LEGAL USAGE (1990). 

4 See, e.g., FED. R. EVID. 803(19), 803(21).


Although this definition is useful so far as it goes, I am more interested in how 

information affects prospective decision-making. 5 Accordingly, I define 

“reputational information” as follows: 

information about an actor’s past performance that helps 

predict the actor’s future ability to perform or to satisfy the 

decision-maker’s preferences. 

This definition contemplates that actors create a pool of data (both subjective 

and objective) through their conduct. This pool of data—the reputational 

information—can provide insights into the actor’s likely future behavior. 

Reputation Systems 

“Reputation systems” aggregate and disseminate reputational information to 

consumers of that information. Reputation systems can be mediated or 

unmediated. 

In unmediated reputation systems, the producers and consumers of reputational 

information communicate directly. Examples of unmediated reputation systems 

include word of mouth, letters of recommendation and job references. 

In mediated reputation systems, a third-party publisher gathers, organizes and 

publishes reputational information. Examples of mediated reputation systems 

include the Better Business Bureau’s ratings, credit reports/scores, investment 

ratings (such as Morningstar mutual fund ratings and Moody bond ratings), and 

consumer review sites. 

The Internet has led to a proliferation of mediated reputation systems, and in 

particular consumer review sites. 6 Consumers can review just about anything 

online; examples include: 

� eBay’s feedback forum, 7 which allows eBay’s buyers and sellers to rate 

each other. 

� Amazon’s product reviews, which allows consumers to rate and review 

millions of marketplace products. 

� Yelp.com, which allows consumers to review local businesses. 

5 Luis M.B. Cabral, The Economics of Trust and Reputation: A Primer (June 2005 draft), 

http://pages.stern.nyu.edu/~lcabral/reputation/Reputation_June05.pdf (treating 

information about reputation as inputs into Bayesian calculations). 

6 Indeed, this has spurred the formation of an industry association, the Rating and Review 

Professional Association. http://www.rarpa.org. 

7 http://pages.ebay.com/services/forum/feedback.html.


� TripAdvisor.com, which allows consumers to review hotels and other 

travel attractions. 

� RealSelf.com, which allows consumers to review cosmetic surgery 

procedures. 

� Avvo.com, which allows consumers to rate and review attorneys. 

� Glassdoor.com, which allows employees to share salary information 

and critique the working conditions at their employers. 

� Honestly.com, 8 which allows co-workers to review each other. 

� RateMyProfessors.com, which allows students to publicly rate and 

review their professors. 

� DontDateHimGirl.com, which allows people to create and “find 

profiles of men who are alleged cheaters.” 9 

� TheEroticReview.com, which allows johns to rank prostitutes. 10 

Why Reputational Information Matters 

In theory, the marketplace works through an “invisible hand”: consumers and 

producers make individual and autonomous decisions that, without any 

centralized coordination, collectively determine the price and quantity of goods 

and services. When it works properly, the invisible hand maximizes social 

welfare by allocating goods and services to those consumers who value them 

the most. 

A properly functioning invisible hand also should reward good producers and 

punish poor ones. Consumers allocating their scarce dollars in a competitive 

market will transact with producers who provide the best cost or quality 

options. Over time, uncompetitive producers should be drummed out of the 

industry by the aggregate but uncoordinated choices of rational and informed 

consumers. 

However, given the transaction costs inherent in the real world, the invisible 

hand can be subject to distortions. In particular, to the extent information 

8 Honestly.com was previously called Unvarnished. See Evelyn Rusli, Unvarnished: A Clean, 

Well-Lighted Place For Defamation, TECHCRUNCH, Mar. 30, 2010, 

http://techcrunch.com/2010/03/30/unvarnished-a-clean-well-lighted-place-fordefamation/. 

9 PlayerBlock is a similar service, tracking undesirable dating prospects by their cellphone 

number. See Leslie Katz, Is Your Date a Player? Send a Text and Find Out, CNET News.com, 

Oct. 22, 2007, http://news.cnet.com/8301-10784_3-9802025-7.html. 

10 See Matt Richtel, Sex Trade Monitors a Key Figure’s Woes, N.Y. TIMES, June 17, 2008. PunterNet 

is another website in this category, providing reviews of British sex workers. John Omizek, 

PunterNet Thanks Harriet for Massive Upswing, THE REGISTER, Oct. 5, 2009, 

http://www.theregister.co.uk/2009/10/05/punternet_harman/.


about producers is costly to obtain or use, consumers may lack crucial 

information to make accurate decisions. To that extent, consumers may not be 

able to easily compare producers or their price/quality offerings, in which case 

good producers may not be rewarded and bad producers may not be punished. 

When information is costly, reputational information can improve the operation 

of the invisible hand by helping consumers make better decisions about 

vendors. In this sense, reputational information acts like an invisible hand 

guiding the invisible hand (an effect I call the “secondary invisible hand”), 

because reputational information can guide consumers to make marketplace 

choices that, in aggregate, effectuate the invisible hand. Thus, in an information 

economy with transaction costs, reputational information can play an essential 

role in rewarding good producers and punishing poor ones. 

Given this crucial role in marketplace mechanisms, any distortions in 

reputational information may effectively distort the marketplace itself. In effect, 

it may cause the secondary invisible hand to push the invisible hand in the 

wrong direction, allowing bad producers to escape punishment and failing to 

reward good producers. To avoid this unwanted consequence, any regulation 

of reputational information needs to be carefully considered to ensure it is 

improving, not harming, marketplace mechanisms. 

Note that the secondary invisible hand is, itself, subject to transaction costs. It 

is costly for consumers to find and assess the credibility of reputational 

information. Therefore, reputation systems themselves typically seek to 

establish their own reputation. I describe the reputation of reputation systems 

as a “tertiary” invisible hand—it is the invisible hand that guides reputational 

information (the secondary invisible hand) to guide the invisible hand of 

individual uncoordinated decisions by marketplace actors (the primary invisible 

hand). Thus, the tertiary invisible hand allows the reputation system to earn 

consumer trust as a credible source (such as the Wall Street Journal, the New 

York Times or Consumer Reports) or to be drummed out of the market for 

lack of credibility (such as the now-defunct anonymous gossip website 

JuicyCampus). 11 

Thinking About Reputation Regulation 

This part explores some ways that the regulatory system interacts with 

reputation systems and some issues caused by those interactions. 

11 Matt Ivester, A Juicy Shutdown, JUICYCAMPUS BLOG, Feb. 4, 2009, 

http://juicycampus.blogspot.com/2009/02/juicy-shutdown.html.


Regulatory Heterogeneity 

Regulators have taken divergent approaches to reputation systems. For 

example, consider the three different regulatory schemes governing job 

references, credit reporting databases and consumer review websites: 

� Job references are subject to a mix of statutory (primarily state law) and 

common law tort regulation. 

� Credit reporting databases are statutorily micromanaged through the 

voluminous and detailed Fair Credit Reporting Act. 12 

� Consumer review websites are virtually unregulated, and many potential 

regulations of consumer review websites (such as defamation) are 

statutorily preempted. 

These different regulatory structures raise some related questions. Are there 

meaningful distinctions between reputation systems that support heterogeneous 

regulation? Are there “best practices” we can observe from these 

heterogeneous regulatory approaches that can be used to improve other 

regulatory systems? These questions are important because regulatory schemes 

can significantly affect the efficacy of reputation systems. As an example, 

consider the differences between the job reference and online consumer review 

markets. 

A former employer giving a job reference can face significant liability whether 

the reference is positive or negative. 13 Giving unfavorable references of former 

employees can lead to defamation or related claims; 14 and there may be liability 

for a former employee giving an incomplete positive reference. 15 

Employers may be statutorily required to provide certain objective information 

about former employees. 16 Otherwise, given the potentially no-win liability 

regime for communicating job references, most knowledgeable employers 

12 15 U.S.C. §§ 1681-81x. 

13 See Tresa Baldas, A Rash of Problems over Job References, NAT’L L.J., Mar. 10, 2008 (“Employers 

are finding that they are being sued no matter what course they take; whether they give a bad 

reference, a good reference or stay entirely silent.”). 

14 1-2 EMPLOYMENT SCREENING § 2.05 (Matthew Bender & Co. 2008) (hereinafter 

“EMPLOYMENT SCREENING”). 

15 Randi W. v. Muroc Joint Unified Sch. Dist., 14 Cal. 4th 1066 (1997). 

16 These laws are called “service letter statutes.” See EMPLOYMENT SCREENING, supra note 14. 

Germany has a mandatory reference law requiring employers to furnish job references, but 

in response German employers have developed an elaborate system for coding the 

references. Matthew W. Finkin & Kenneth G. Dau-Schmidt, Solving the Employee Reference 

Problem, 57 AM. J. COMP. L. 387 (2009).


refuse to provide any subjective recommendations of former employees, 

positive or negative. 17 

To curb employers’ tendency towards silence, many states enacted statutory 

immunities to protect employers from lawsuits over job references. 18 However, 

the immunities have not changed employer reticence, which has led to a virtual 

collapse of the job reference market. 19 As a result, due to mis-calibrated 

regulation, the job reference market fails to provide reliable reputational 

information. 

In contrast, the online consumer review system is one of the most robust 

reputation systems ever. Millions of consumers freely share their subjective 

opinions about marketplace goods and services, and consumer review websites 

keep proliferating. 

There are several possible reasons why consumer review websites might succeed 

where offline reputation systems might fail. My hypothesis, discussed in a 

companion essay in this collection, is that the difference is partially explained by 

47 U.S.C. § 230, passed in 1996—at the height of Internet exceptionalism—to 

protect online publishers from liability for third party content. Section 230 lets 

websites collect and organize individual consumer reviews without worrying 

about crippling legal liability for those reviews. As a result, consumer review 

websites can motivate consumers to share their opinions and then publish those 

opinions widely—as determined by marketplace mechanisms (i.e., the tertiary 

invisible hand), not concerns about legal liability. 

The success of consumer review websites is especially noteworthy given that 

individual reviewers face the same legal risks that former employers face when 

providing job references, such as the risk of personal liability for publishing 

negative reputational information. Indeed, numerous individuals have been 

sued for posting negative online reviews. 20 As a result, rational actors should 

find it imprudent to submit negative reviews; yet, millions of such reviews are 

published online. A number of theories might explain this discrepancy, but one 

theory is especially intriguing: Mediating websites, privileged by their own 

liability immunity, find innovative ways to get consumers over their fears of 

legal liability. 

17 See Baldas, supra note 13. 

18 The immunizations protect employer statements made in good faith. EMPLOYMENT 

SCREENING, supra note 14. 

19 See Finkin & Dau-Schmidt, supra note 16. 

20 See, e.g., Wendy Davis, Yelp Reviews Spawn At Least Five Lawsuits, MEDIAPOST ONLINE MEDIA 

DAILY, Jan. 21, 2009, 

http://www.mediapost.com/publications/?fa=Articles.printFriendly&art_aid=9877 

8; Agard v. Hill, 2010 U.S. Dist. LEXIS 35014 (E.D. Cal. 2010).


What lessons can we draw from this comparison? One possible lesson is that 

reputation systems are too important to be left to the market. In other words, 

the tertiary invisible hand may not ensure accurate and useful information, or 

the costs of inaccurate information (such as denying a job to a qualified 

candidate) may be too excessive. If so, extensive regulatory intervention of 

reputation systems may improve the marketplace. 

An alternative conclusion—and a more convincing one to me—is that the 

tertiary invisible hand, aided by a powerful statutory immunity like Section 230, 

works better than regulatory intervention. If so, we may get better results by 

deregulating reputation systems. 

System Configurations 

Given the regulatory heterogeneity, I wonder if there is an “ideal” regulatory 

configuration for reputation systems, especially given the tertiary invisible hand 

and its salutary effect on publisher behavior. Two brief examples illustrate the 

choices available to regulators, including the option of letting the marketplace 

operate unimpeded: 

Anti-Gaming. A vendor may have financial incentives to distort the flow of 

reputational information about it. This reputational gaming can take many 

forms, including disseminating false positive reports about the vendor, 21 

disseminating false negative reports about the vendor’s competitors, or 

manipulating an intermediary’s sorting or weighting algorithm to get more credit 

for positive reports or reduce credit for negative reports. Another sort of 

gaming can occur when users intentionally flood a reputation system with 

inaccurate negative reports as a form of protest. 22 

Do regulators need to curb this gaming behavior, or will other forces be 

adequate? There are several marketplace pressures that curb gaming, including 

competitors policing each other, 23 just as they do in false advertising cases. 24 In 

21 Lifestyle Lift Holding, Inc. v. RealSelf Inc., 2:08-cv-10089-PJD-RSW (answer/counterclaims 

filed March 3, 2008), http://www.realself.com/files/Answer.pdf (alleging that Lifestyle 

Lift posted fake positive reviews about its own business to an online review website). 

22 For example, consumers protesting the digital rights management (DRM) in EA’s Spore 

game flooded Amazon’s review site with one-star reviews, even though many of them 

actually enjoyed the game. See Austin Modine, Amazon Flash Mob Mauls Spore DRM, THE 

REGISTER, Sept. 10, 2008, 

http://www.theregister.co.uk/2008/09/10/spore_drm_amazon_effect/. A similar 

protest hit Intuit’s TurboTax 2008 over its increased prices. See Steven Musil, Amazon 

Reviewers Slam TurboTax Fee Changes, CNET NEWS.COM, Dec. 7, 2008, 

http://news.cnet.com/8301-1001_3-10117323-92.html. 

23 See Cornelius v. DeLuca, 2010 WL 1709928 (D. Idaho Apr. 26, 2010) (a marketplace vendor 

sued over alleged shill online reviews posted by competitors).


addition, the tertiary invisible hand may encourage reputation systems to 

provide adequate “policing” against gaming. However, when the tertiary 

invisible hand is weak, such as with fake blog posts where search engines are the 

only mediators, 25 government intervention might be worth considering. 

Right of Reply. A vendor may wish to publicly respond to reputational 

information published about it in an immediately adjacent fashion. Many 

consumer review websites allow vendors to comment or otherwise reply to 

user-supplied reviews, but not all do. For example, Yelp initially drew 

significant criticism from business owners who could not effectively reply to 

negative Yelp reviews because of Yelp’s architecture, 26 but Yelp eventually 

relented and voluntarily changed its policy. 27 As another example, Google 

permitted quoted sources to reply to news articles appearing in Google News as 

a way to “correct the record.” 28 

Regulators could require consumer review websites and other reputation 

systems to permit an adjacent response from the vendor. 29 But such 

intervention may not be necessary; the tertiary invisible hand can prompt 

reputation systems to voluntarily provide a reply option (as Yelp and Google 

did) when they think the additional information helps consumers. 

Undersupply of Reputational Information 

There are three primary categories of reasons why reputational information may 

be undersupplied. 

24 See, e.g., Lillian R. BeVier, A Puzzle in the Law of Deception, 78 VA. L. REV. 1 (1992). 

25 See Press Release, New York Office of the Attorney General, Attorney General Cuomo 

Secures Settlement With Plastic Surgery Franchise That Flooded Internet With False Positive 

Reviews, July 14, 2009, 

http://www.ag.ny.gov/media_center/2009/july/july14b_09.html. 

26 See Claire Cain Miller, The Review Site Yelp Draws Some Outcries of Its Own, N.Y. TIMES, Mar. 3, 

2009. 

27 See Claire Cain Miller, Yelp Will Let Businesses Respond to Web Reviews, N.Y. TIMES, Apr. 10, 

2009. 

28 See Dan Meredith & Andy Golding, Perspectives About the News from People in the News, 

GOOGLE NEWS BLOG, Aug. 7, 2007, 

http://googlenewsblog.blogspot.com/2007/08/perspectives-about-news-frompeople-in.html. 

29 See Frank A. Pasquale, Rankings, Reductionism, and Responsibility, 54 CLEV. ST. L. REV. 115 

(2006); Frank A. Pasquale, Asterisk Revisited: Debating a Right of Reply on Search Results, 3 J. BUS. 

& TECH. L. 61 (2008).


Inadequate Production Incentives 

Much reputational information starts out as non-public (i.e., “private”) 

information in the form of a customer’s subjective mental impressions about 

his/her interactions with the vendor. To the extent this information remains 

non-public, it does not help other consumers make marketplace decisions. 

These collective mental impressions represent a vital but potentially 

underutilized social resource. 

The fact that non-public information remains locked in consumers’ heads could 

represent a marketplace failure. If the social benefit from public reputational 

information exceeds the private benefit from making it public, then 

presumptively there will be an undersupply of public reputational information. 

If so, the government may need to correct this failure by encouraging the 

disclosure of reputational information—such as by creating a tort immunity for 

sites that host that disclosure, as Section 230 does, or perhaps by going further. 

But there already may be market solutions to this problem, as evidenced by the 

proliferation of online review websites eliciting lots of formerly non-public 

reputational information. 

Further, relatively small amounts of publicly disclosed reputational information 

might be enough to properly steer the invisible hand. For example, the first 

consumer review of a product in a reputation system creates a lot of value for 

subsequent consumers, but the 1,000 th consumer review of the same product 

may add very little incrementally. So even if most consumer impressions remain 

non-public, perhaps mass-market products and vendors still have enough 

information produced to keep them honest. At the same time, vendors and 

products in the “long tail” 30 may have inadequate non-public impressions put 

into the public discourse, creating a valuable opportunity for comprehensive 

reputation systems to fix the omission. However, reputation systems will tackle 

these obscure marketplace options only when they can keep their costs low 

(given that consumer interest and traffic will, by definition, be low), and 

reputation system deregulation helps reduce both the costs of litigation as well 

as responding to takedown demands. 

30 Chris Anderson, The Long Tail, WIRED, Oct. 2004, 

http://www.wired.com/wired/archive/12.10/tail.html.


Vendor Suppression of Reputational Information 

Vendors are not shy about trying to suppress unwanted consumer reviews ex 

post, 31 but vendors might try to suppress such reviews ex ante. For example, 

one café owner grew so tired of negative Yelp reviews that he put a “No 

Yelpers” sign in his café’s windows. 32 

That sign probably had no legal effect, but Medical Justice offers an ex ante 

system to help doctors use preemptive contracts to suppress reviews by their 

patients. Medical Justice provides doctors with a form agreement that has 

patients waive their rights to post online reviews of the doctor. 33 Further, to 

bypass 47 U.S.C. § 230’s protective immunity for online reputation systems that 

might republish such patient reviews, the Medical Justice form prospectively 

takes copyright ownership of any patient-authored reviews. 34 (Section 230 does 

not immunize against copyright infringement). This approach effectively allows 

doctors—or Medical Justice as their designee—to get reputation systems to 

remove any unwanted patient reviews simply by sending a DMCA takedown 

notice. 35 

Ex ante customer gag orders may be illegal. In the early 2000s, the New York 

Attorney General challenged software manufacturer Network Associates’ end 

user license agreement, which said the “customer will not publish reviews of 

this product without prior consent from Network Associates, Inc.” In 

response, the New York Supreme Court enjoined Network Associates from 

restricting user reviews in its end user license agreement. 36 Medical Justice’s 

scheme may be equally legally problematic. 

From a policy standpoint, ex ante customer gag orders pose serious threats to 

the invisible hand. If they work as intended, they starve reputation systems of 

the public information necessary to facilitate the marketplace. Therefore, 

31 See Eric Goldman, Online Word of Mouth and Its Implications for Trademark Law, in TRADEMARK 

LAW AND THEORY: A HANDBOOK OF CONTEMPORARY RESEARCH 404 (Graeme B. 

Dinwoodie and Mark D. Janis eds.) (2008) (discussing lopsided databases where all negative 

reviews are removed, leaving only positive reviews). 

32 Stefanie Olsen, No Dogs, Yelpers Allowed, CNET NEWS.COM, Aug. 14, 2007, 


33 Lindsey Tanner, Doctors Seek Gag Orders to Stop Patients’ Online Reviews, ASSOCIATED PRESS, 

Mar. 3, 2009, http://www.usatoday.com/news/health/2009-03-05-doctorreviews_N.htm. 

34 Michael E. Carbine, Physicians Use Copyright Infringement Threat to Block Patient Ratings on the Web, 

AIS’S HEALTH BUSINESS DAILY, Mar. 30, 2009, 

http://www.aishealth.com/Bnow/hbd033009.html. 

35 17 U.S.C. § 512(c)(3). 

36 People v. Network Associates, Inc., 758 N.Y.S.2d 466 (N.Y. Sup. Ct. 2003).


regulatory efforts might be required to prevent ex ante customer gag orders 

from wreaking havoc on marketplace mechanisms. 

Distorted Decision-Making 

from Reputational Information 

Reputational information generally improves decision-making, but not always. 

Most obviously, reputational information relies on the accuracy of past 

information in predicting future behavior, but this predictive power is not 

perfect. 

First, marketplace actors are constantly changing and evolving, so past behavior 

may not predict future performance. For example, a person with historically 

bad credit may obtain a well-paying job that puts him or her on good financial 

footing. Or, in the corporate world, a business may be sold to a new owner 

with different management practices. In these situations, the predictive 

accuracy of past information is reduced. 37 

Second, some past behavior may be so distracting that information consumers 

might overlook other information that has more accurate predictive power. For 

example, a past crime or bankruptcy can overwhelm the predictive information 

in an otherwise-unblemished track record of good performance. 

Ultimately, a consumer of information must make smart choices about what 

information to consult and how much predictive weight to assign to that 

information. Perhaps regulation can improve the marketplace’s operation by 

shaping the information that consumers consider. For example, if some 

information is so highly prejudicial that it is likely to distort consumer decisionmaking, 

the marketplace might work better if we suppress that information 

from the decision-maker. 38 

At the same time, taking useful information out of the marketplace could create 

its own adverse distortions of the invisible hand. Therefore, we should tread 

cautiously in suppressing certain categories of information. 

37 Cf. Note, Badwill, 116 HARV. L. REV. 1845 (2003) (describing how companies can mask a 

track record of bad performance through corporate renaming). 

38 Cf. FED. R. EVID. 403 (“Although relevant, evidence may be excluded if its probative value is 

substantially outweighed by the danger of unfair prejudice, confusion of the issues, or 

misleading the jury…”). This fear underlies a French proposal to enact a “right to forget” 

statute. See David Reid, France Ponders Right-to-Forget Law, BBC CLICK, Jan. 8, 2010, 

http://news.bbc.co.uk/2/hi/programmes/click_online/8447742.stm.


Conclusion 

Although “reputation” has been extensively studied in a variety of social science 

disciplines, there has been comparatively little attention paid to how regulation 

affects the flow of reputational information in our economy. Understanding 

these dynamics would be especially valuable in light of the proliferation of 

Internet-mediated reputation systems and the irresistible temptation to regulate 

novel and innovative reputation systems based on emotion, not necessarily 

sound policy considerations.

CHAPTER 5 

WHO WILL GOVERN THE NET IN 2020? 

Imagining the Future of Global Internet Governance 307 


305 

Democracy in Cyberspace: Self-Governing Netizens & 

a New, Global Form of Civic Virtue, Online 315 


Who’s Who in Internet Politics: A Taxonomy of 

Information Technology Policy & Politics 327 

Robert D. Atkinson

306 CHAPTER 5: WHO WILL GOVERN THE NET IN 2020?


Imagining the Future of 

Global Internet Governance 

By Milton Mueller * 

When discussing who (or what) will govern the Internet in 2020, people tend to 

want predictions. They want authoritative statements from experts. They want 

you to tell them what will happen. But an honest scholar of Internet governance 

would never attempt to meet that demand. The problem is not just that the 

future of Internet governance is uncertain, subject to the influence of many 

complex variables. The problem is that its future is, literally, indeterminate. 

While it is correct that there is an ongoing struggle over the governance of the 

Internet, we cannot know how it will come out. 

Forget about predictions and forecasts. It’s better to have a clear conception of 

how we want the Internet to be governed. This means that we need to be able to 

imagine feasible futures and to create strategies to realize them. 

Let’s step back. Why is Internet governance an interesting problem in the first 

place? Why does contemplating the Internet’s future require imagination and 

creativity? Because there is a tension, even a contradiction, between the existing 

institutions for regulating communications and information, and the technical 

capabilities and processes of open internetworking. Existing institutions are 

organized around territorial, hierarchical nation-states; the process of 

internetworking, on the other hand, provides globalized and distributed 

interoperation amongst all the elements of an increasingly powerful and 

ubiquitous system of digital devices and networks. 

This technical capability puts pressure on the nation-state in five distinct ways. 

1. It globalizes the scope of communication. Its distance-insensitive cost 

structure and non-territorial addressing and routing architecture make 

borderless communication the default; any attempt to impose a 

jurisdictional overlay on its use requires additional (costly) 

interventions. 

2. It facilitates a quantum jump in the scale of communication. It massively 

enlarges our capacity for message generation, duplication, and storage. 

As a programmable environment, it industrializes information services, 

information collection, and information retrieval. The sheer volume of 

* Milton Mueller teaches and does research on the political economy of communication and 

information. His new book NETWORKS AND STATES: THE GLOBAL POLITICS OF 

INTERNET GOVERNANCE (MIT Press, 2010) provides a comprehensive overview of the 

political and economic drivers of a new global politics.

308 CHAPTER 5: WHO WILL GOVERN THE NET IN 2020? 

transactions and content on the Internet often overwhelms the capacity 

of traditional governmental processes to respond—but that same 

scalability can transform governmental processes as well. 

3. It distributes control. Combined with liberalization of the 

telecommunications sector, the Internet protocols have decentralized 

and distributed participation in and authority over networking and 

ensured that the decision-making units over network operations are not 

necessarily closely aligned with political units, as they were in the days 

of post, telephone and telegraph monopolies. 

4. It grows new institutions. Decision-making authority over standards and 

critical Internet resources rests in the hands of a transnational network 

of actors that emerged organically alongside the Internet, outside of the 

nation-state system. These relatively young but maturing institutions, 

such as the Internet Engineering Task Force (IETF), the Internet 

Corporation for Assigned Names and Numbers (ICANN), and 

Regional Internet Address Registries (RIRs) provide a new locus of 

authority for key decisions about standards and critical resources. 

5. It changes the polity. By converging different media forms and facilitating 

fully interactive communication, the Internet dramatically alters the cost 

and capabilities of group action. As a result, radically new forms of 

collaboration, discourse, and organization are emerging. This makes it 

possible to mobilize new transnational policy networks and enables 

new forms of governance as a solution to some of the problems of 

Internet governance itself. 

Transnational scope, boundless scale, distributed control, new institutions and 

radical changes in collective action capabilities—these factors are transforming 

national control and sovereignty over communication and information policy, 

setting in motion new institutional forms and new kinds of geopolitical 

competition. The governance of global Internetworking is thus a relatively new 

problem created by socio-technical change. The future of Internet governance 

will be driven by the clash between its raw technical potential and the desire of 

various incumbent interests—most notably nation-states—to assert control 

over that potential. 

While the Internet poses novel governance problems, how we solve them 

cannot be predicted. It depends vitally on our ability to accurately diagnose the 

economic, technical and political forces at work and on our ability to imagine 

strategies, mechanisms and techniques that can harness those forces to do what 

we want to do. Thus, to repeat, it is better to invest our mental resources in 

conceptualizing and enacting feasible visions of how we want the Internet to be 

governed than it is to invest in making deterministic predictions.


The Pace of Change 

2020 is not very far away. Ten years is the blink of an eye when it comes to 

institutional development at the global level. Proof of this can be found by 

glancing back ten years from our current vantage point of 2010. Despite the 

Internet’s reputation for rapid change, the basic issues and problems of Internet 

governance have not changed much since 2000. Yes, there has been turbulence 

in the market economy, with specific firms rising and falling. But the cast of 

institutional characters that regulate or govern Internetworking was already well 

in place by 2000. The organically developed Internet institutions such as the 

IETF, the Internet Society, and the Internet address registries 1 already existed. 

On the other hand, the U.S. government and its rival nation-states were 

entering the scene. ICANN, with its unilateral control by the U.S. government, 

had already emerged as the uncomfortable compromise between the a-national 

“Internet community” and the community of states. The seeds of the tensions 

among the U.S., the E.U. and the BRIC 2 nations caused by U.S. pre-eminence 

in that regime were already sown. There was already theoretical talk of cyberwar 

(though this has picked up dramatically in the last few years). There were already 

efforts to block and filter Internet content, though these have become 

increasingly refined. Peer-to-peer file sharing was already beginning to drive 

copyright holders mad (Napster was started in 2000). Whatever change has 

taken place since 2000 has been evolutionary rather than revolutionary. 

Disruption or Continuity? 

It is possible to identify some aspects of the current Internet governance regime 

that could disrupt the existing evolutionary trajectory. I divide them into two 

distinct categories: the geopolitical and the techno-economic. 

Geopolitical Factors 

The Root: One of the most important geopolitical factors is still U.S. control of 

the root of the name and numbering hierarchies. This control is bound up with 

the issue of the singularity of those roots and the universal interoperability of 

the Internet. Here the U.S. is pre-eminent, and along with that pre-eminence 

come forms of responsibility and danger. A policy misstep or mistake can 

disrupt the status quo. Will the U.S. finally fully privatize ICANN, or will it 

“internationalize” it? Will the Internet Assigned Numbers Authority (IANA) 

contract 3 be competitively bid, or routinely reassigned to ICANN, keeping it 

1 Users are assigned IP addresses by Internet service providers (ISPs), who usually obtain 

larger blocks of IP addresses from a Regional Internet Registry (RIR). 

2 Brazil, Russia, India, and China 

3 The IANA contract is a contract between the U.S. Commerce Department that authorizes 

ICANN to perform what is commonly referred to as the IANA function, a bundle of 

technical operations that includes the registration and allocation of IP addresses,


subordinate to the United States government? The governance issues related to 

the root and control/coordination hierarchies have intensified as the technical 

community (usually funded by U.S. government contracts) has moved to 

“secure” the Internet by making access to and use of critical identifier resources 

reliant upon cryptographic key hierarchies. The harder the U.S. government 

tries to rigidify its existing forms of institutional control over Internet resources, 

the more likely it becomes that the Internet will fragment. 

Cyber-Warfare: Military conflict is always a potent source of institutional 

disruption. Geopolitically, the U.S. is pursuing a dangerously contradictory 

agenda. On the one hand it insists on retaining pre-eminent control of Internet 

standards, protocols and virtual resources and maximizing the dependence of 

the rest of the world on them. On the other hand it wants to treat cyberspace as 

a “national asset” and develop an overwhelming cyber-warfare and cyberweapons 

capability based on those very same standards, protocols and 

resources. But insofar as cyberspace is militarized, its status as a globalized 

platform for information and communication among the business and civil 

society is undermined. Contradictions abound here, and as they play out, the 

chances of a structural change increase. 

Free Trade in Information Services: The U.S. approach to Internet freedom 

is driven as much by economics as by ideology and ethics. Due to its liberal 

policies, the U.S. leads the world in the supply of Internet-based information 

services. Of course the rest of the world will gradually catch up, but the natural 

state of Internet-based information services is to be transnational, accessible 

anywhere in the world, and so suppliers who would challenge the Googles and 

Facebooks must be transnational as well. The contradiction between the open 

Internet and various forms of trade protectionism in the content industries— 

including the cultural protectionism that is often disguised as support for 

“diversity”—could be a key driver of Internet governance. Advocates of civil 

liberties and communication rights need to forge common ground with 

advocates of free trade and market liberalism for anything important to happen 

here. 

Techno-Economic Factors 

Unlicensed Wireless Broadband: A great deal of the consolidation of control 

over the Internet is contingent upon the access bottleneck. The fewer market 

players in the Internet service provider space, the easier it becomes for states 

and state-favored monopolies to blunt and channel the potential of information 

and communication technology. Thus, new access technologies like unlicensed 

wireless broadband become critical factors shaping the future. If they can take 

management of the Internet root server system, and maintenance of the authoritative root 

zone file for the domain name system.


root and disrupt current market structures around the supply of Internet access, 

the arrangements for governance and control will need to be reconsidered. With 

greater choice of access arrangements, the less feasible it becomes for 

governments to impose onerous regulatory arrangement upon consumers 

through intermediaries. 

DPI and Net Neutrality: A key characteristic of the Internet so far has been 

the “end-to-end” principle, which put the processing intelligence for 

applications and services at the end points and made the network a relatively 

simple packet-forwarding system. Deep packet inspection (DPI) is a new 

technological capability that could lead to a wholesale departure from that 

principle. Developed in response to legitimate concerns about efficient 

bandwidth management and the detection and interception of malware, it 

increases the awareness and control of the network intermediary over the traffic 

coursing through its system. This is a fateful shift of control. Needless to say, 

there are demands to extend its capabilities to less technical forms of 

intervention, such as censorship, copyright protection or national securityoriented 

surveillance of communications. At the same time, concerns about 

privacy, network neutrality, and competition policy have put legal and regulatory 

checks upon the usage of DPI. This is an arena that goes to the heart of 

Internet governance in the future. 

Two Visions 

Two visions of possible futures should help to illustrate how these themes might 

play out, but more importantly, how I think they ought and ought not play out. 

The Dark Vision 

Picture a world in a long-term global recession, one that lasts the better part of 

the decade we are discussing. There is growing conservatism—by which I mean 

increasingly nationalist and ethnocentric attitudes, a growing impatience with, 

and rejection of, the rigors of market liberalism, and a greater willingness to 

trade freedom and innovation for security and stability. In such a scenario of 

recession-driven reaction, trade barriers rise. Hostility to immigration and 

“offshoring” grows. Internet-based communications become increasingly 

confined to national spaces. There is blocking and filtering of content at the 

national level; the full linkage of online identity to national identity; the licensing 

of content, application and service providers at the national level; the 

subordination of information flows to the surveillance needs of national 

governments. Infrastructure providers stop expanding and rely on national 

broadband subsidy plans. As this happens, the major U.S. Internet/media 

corporations succeed in minimizing competition and maximize rent-seeking in 

an increasingly mature, stable market. With the number of players winnowed 

down, these corporations will make disastrous concessions to governments 

seeking to extend their authority over cyberspace in areas such as online identity


and identification, online surveillance, security practices, protectionist standards 

and content regulation. Some variant of a Google-Verizon merger spawns the 

AT&T of the 21 st century, a dominant private sector entity with its own 

commercial interests, but one whose markets and fortunes follow the flag of 

U.S. policy worldwide. Reacting to its quasi-sponsorship by the State 

Department, other countries erect barriers. In this context, with national 

security and a cyber- version of the military-industrial complex becoming the 

main driver of international policy, the U.S. government eventually participates 

in the strangling of its own progeny. 

As the U.S. develops an overwhelming cyber-warfare and cyber-weapons 

capability, the rest of the world revolts. The U.S. provokes a cyber-cold war, or 

perhaps even a short “hot war” with Russia and China, and uses it to rationalize 

and extend many of the controls. The European Commission—but not 

European civil society—will side with the U.S., effectively paralyzing and 

subordinating Europe’s ability to contribute anything constructive, much less 

innovative, to the Internet governance debates. The Internet world fragments 

on linguistic grounds, with the English-speaking or English-dominant world 

drifting away from the Chinese, Korean, Russian and Japanese societies. 

The Bright Vision 

It’s easy enough to describe that scenario because it seems to be the road we are 

already on. It is much harder to imagine a better future, one that is both feasible 

and consistent with the interests and capabilities of current actors. But let’s give 

it a try. In another work, I’ve tried to describe the basic nature of what I call a 

denationalized liberalism as the guide to the future of Internet governance. 4 

At its core, a denationalized liberalism favors a universal right to receive and 

impart information regardless of frontiers, and sees freedom to communicate 

and exchange information as fundamental, primary elements of human choice 

and political and social activity. Political institutions should seek to build upon, 

not undermine or reverse, the limitless possibilities for forming new social 

aggregations around digital communications. In line with its commitment to 

freedom, this ideology holds a presumption in favor of networked, associative 

relations over hierarchical relations as a mode of transnational governance. 

Governance should emerge primarily as a byproduct of many unilateral and 

bilateral decisions by its members to exchange or negotiate with other members 

(or to refuse to do so). This networked liberalism thus moves decisively away 

from the dangerous, conflict-prone tendency to build political institutions 

around linguistic, religious, and ethnic communities. Instead of rigid, bounded 

communities that conceal domination with the pretense of homogeneity and a 

4 MILTON MUELLER, NETWORKS AND STATES: THE GLOBAL POLITICS OF INTERNET 

GOVERNANCE (MIT Press: 2010).


“collective will,” this liberalism offers governance of communication and 

information through more flexible and shifting social aggregations. 

Although committed to globalism in the communicative sector, networked 

liberalism recognizes that, for the time being, people are deeply situated within 

national laws and institutions regarding such basic matters as contracts, 

property, crime, education, and welfare. It is characterized not by absolute 

hostility to national and subnational governments as such, but rather by an 

attempt to contain them to those domains of law and policy suited to localized or 

territorialized authority. It seeks to detach the transnational operations of 

Internet infrastructure and the governance of services and content from those 

limited jurisdictions as much as possible, and to prevent states from ensnaring 

global communications in interstate rivalries and politico-military games. This 

requires a complete detachment of Internet governance institutions from 

nation-state institutions, and the creation of new, direct accountability 

relationships for Internet institutions. 

Such an ideology needs to answer tough questions about when hierarchical 

exercises of power are justified and through which instruments they are 

exercised. A realistic denationalized liberalism recognizes that emergent forms 

of control will arise from globally networked communities. It recognizes that 

authoritative interventions will be needed to secure basic rights against coercive 

attacks, and that network externalities or bottlenecks over essential facilities may 

create a concentrated power with coercive effect. It should also recognize the 

exceptional cases where the governance of shared resources requires binding 

collective action. Insofar as collective governance is necessary and unavoidable, 

a denationalized liberalism strives to make Internet users and suppliers an 

autonomous, global polity, with what might be called neodemocratic rights to 

representation and participation in these new global governance institutions. 

The concept of democracy is qualified by the realization that the specific form 

of democratic governance associated with the territorial nation-state cannot and 

should not be directly translated into the global level. However, it does maintain 

the basic objectives of traditional democracy: to give all individuals the same 

formal rights and representational status within the institutions that govern 

them so that they can preserve and protect their rights as individuals. Such a 

liberalism is not interested, however, in using global governance institutions to 

redistribute wealth. That would require an overarching hierarchical power that 

would be almost impossible to control democratically; its mere existence would 

trigger organized political competition for its levers, which would, in the current 

historical context, devolve into competition among preexisting political and 

ethnic collectivities—the very opposite of networked liberalism. 

In short, we need to find ways to translate classical liberal rights and freedoms 

into a governance framework suitable for the global Internet. There can be no


cyber-liberty without a political movement to define, defend, and institutionalize 

individual rights and freedoms on a transnational scale.


Democracy in Cyberspace: Self- 

Governing Netizens & a New, Global 

Form of Civic Virtue, Online 

By David R. Johnson * 

The Internet can be viewed as a set of wires, wireless “pipes” and servers, a set 

of protocols, or as a vast array of content and applications to which these lower 

layers of the stack provide access. None of those tangible and intangible things 

can be “governed.” They may or may not be owned or manipulated. But it is 

the actions of the people involved in creating and using the Internet that are the 

proper subject of a question regarding “governance.” Viewed with respect to 

the social and legal relationships among the people who are creating and using 

it, and who would be “governed,” the Internet is a complex system—so making 

accurate predictions about its future state is impossible. 

But it is possible to answer the question: “Who could and who should govern the 

Internet in 2020?” My answer is, in a word: netizens—the global polity of those 

who collaborate online, seek to use the new affordances of the Internet to 

improve the world, and care about protecting an Internet architecture that 

facilitates new forms of civic virtue. 

The Internet Governance Debate 

The debate about “Internet Governance” has continued for more than fifteen 

years and settled into an unsatisfying rut. The established trope is that early 

visionaries (e.g., John Perry Barlow 1) claimed that cyberspace was a new realm of 

freedom, poised to escape from regulation by local governments. Then, later 

“realists” (e.g., Jack Goldsmith and Tim Wu 2) discovered that sovereign 

governments indeed had ways to regulate online speech and even use the 

Internet for surveillance and tyranny. Early idealists envisioned the Internet 

Corporation for Assigned Names and Numbers (charged with setting policy for 

* David Johnson joined New York Law School’s faculty in spring 2004 as a visiting professor 

of law. He is a faculty member of the Institute for Information Law and Policy. 

1 See John Perry Barlow, A Declaration of the Independence of Cyberspace (Feb. 8, 1996), 

https://projects.eff.org/~barlow/Declaration-Final.html [hereinafter A Declaration of 

the Independence of Cyberspace]; John Perry Barlow, Declaring Independence, 4.06 WIRED 121-22 

(June 1996), available at 

http://www.wired.com/wired/archive/4.06/independence.html. 

2 See JACK GOLDSMITH & TIM WU, WHO CONTROLS THE INTERNET?: ILLUSIONS OF A 

BORDERLESS WORLD (2006) [hereinafter WHO CONTROLS THE INTERNET?].


the domain name system and allocating blocks of IP numbers) 3 as a new 

democratic global institution constrained by consensus reached in a global 

community. Later, we observe an expensive bureaucracy imposing complex 

regulations—as one wag has put it: “recapitulating the FCC and doing it 

badly.” 4 Internet Governance as an empowering and liberating democratic ideal 

is a total failure—or so it would seem. 

This debate has missed a fundamental point by asking the wrong question. The 

key question is not “Who will govern the Internet?” Instead, it is “Will the 

global Internet affect the way in which we (the global polity) govern ourselves?” 

One way or another, we act through governments, NGOs, private corporations 

and many other types of groups to collectively set the rules by which we live our 

lives. We all strive for a world in which our own choices determine how we are 

governed. So the more salient way to put the real question at issue is: “Can the 

Internet make society more democratic?” 

The Vision of the Internet’s Founders 

The founders of the Internet (technologists, advocates, policy makers, and 

visionaries) saw its democratic potential. They were not (mostly) seeking to 

create a “lawless frontier.” They were instead seizing a moment of flexibility 

during which new modes of association for community improvement might 

flourish. They opposed rigid regulation and unaccountable power, of course. 

But they also favored collaborative decision-making to establish rules, group 

effort to write empowering and constraining code, and respectful deliberation to 

forge new norms. They favored civility and civic virtue (good netizenship) as 

much as individual liberty. 

We’re not just talking about the founding technologists. Many individuals, 

nonprofit organizations and companies came together, in the late 1980s and 

throughout the 1990s, to create a technology and policy framework to enable a 

democratic Internet: open access, limitations on intermediaries’ powers and 

liabilities, privacy protections for communications, limitations on centralized 

levers of power (like the control over the domain name system) and, at least in 

the United States, establishment of strong First Amendment rights (in contrast 

to regulation of the Internet as a form of mass media). To a considerable 

3 ICANN is generally accepted as the authority for decisions about what top level domains 

will be placed in the “root zone file.” By establishing contractual conditions in connection 

with such additions, it can establish rules that flow down onto registries, registrars and 

registrants. For example, this capability has been used to require registrants in generic Top 

Level Domains to submit to a “Uniform Dispute Resolution Policy” that decides and takes 

action on disputes about “cybersquatting.” 

4 Harold Feld, quoted in Jonathan Weinberg, ICANN, “Internet Stability,” and New Top Level 

Domains 1 n. 1, http://faculty.law.wayne.edu/Weinberg/icannetc.pdf.


degree, from that perspective, the Internet’s founders have, up to this point, 

succeeded. 

I want to re-emphasize that, while the Internet relies on wires and protocols, it 

most fundamentally consists of connections among people. Local governments 

may control who has the right to provide access. Standard setting bodies may 

have some say in what protocols become widely adopted. Law plays a role—and 

the law of local governments does constrain the actions of people over whom 

they can assert jurisdiction. But the “governance” of the Internet is 

fundamentally a question about how we all constrain the manner in which we 

do whatever it is we do in groups online, including establishing new structures 

of society, new forms of social organization, and new roles and rules that 

incentivize our efforts and focus our minds. No government could even hope 

to make the rule set for a global web of relationships involving billions of 

people interacting in complex and ever evolving ways. The governance of the 

social layer of the Internet will be, perforce, decentralized. 

The Democratic Nature of the Internet 

Thus, contrary to Larry Lessig’s suggestion in Code, 5 I submit that the Internet 

of today has a nature: It is inherently democratic. Not inevitably so, in the sense 

that any global communications network would necessarily be democratic. And 

not necessarily so in the future. But historically so and by design—in the sense 

that this Internet, the one we have and the one that scaled globally in no time, 

was successful precisely because it was open, decentralized, tolerant of 

innovation and disagreement, voluntary, and empowering of anyone who cared 

to use it to join with others to improve the world. Every time we address an 

email, or establish a blog, or “agree” to some “terms of service,” we are creating 

the rules for our online society. 

From Wikipedia to PatientsLikeMe, 6 we are continuously learning how to use 

the Internet to come together to share knowledge, improve education, solve 

health care problems, and provide charitable assistance to those in need around 

the world. We use the Internet to participate in local politics and explore new 

ways to make global society energy efficient. In these and countless other ways, 

the Internet is an engine of democratic civic virtue. 

5 See LAWRENCE LESSIG, CODE: AND OTHER LAWS OF CYBERSPACE, VERSION 2.0 (Basic Books 

2006). 

6 For example, PatientsLikeMe allows users to create profiles and then share, interact, and 

learn from the experience of other users on health and wellness issues.


Self-Governance Online 

As noted, and sometimes in defiance of repressive regulations, we choose many 

of the rules under which these collaborations occur simply by logging in to one 

website or platform rather than another. We can rise up together in protest 

when a site like Facebook changes its “terms of service” and “privacy policy” in 

ways we don’t like. And we increasingly accept obligations to spend our 

attention and effort in support of groups we find online with whom we share a 

persistent purpose or goal. 

Even an email listserv involves social duties (not always honored, of course) to 

fellow participants. Open source efforts have evolved complex, hierarchical, yet 

open and democratic (or meritocratic) self-governance structures. We are 

beginning to understand that, at least for those things that can happen online, 

the choices we make about which groups to join have as much impact on 

setting the terms of our relationships with others than any governmental laws or 

regulations. 

Goldsmith and Wu and other “anti-exceptionalists” have cited the Yahoo! 

France case, Australia’s imposition of its libel laws on a New Jersey-based 

publisher, and Italy’s conviction of Google executives for the proposition that 

the Net cannot escape from the “real world” governance of sovereign states. 7 

Their examples, rather than making a case for a bordered Internet, in fact prove 

the opposite: A world in which every local sovereign seeks to control the 

activities of netizens beyond its borders violates the true meaning of selfgovernance 

and democratic sovereignty. 

Attention Governance 

& Global Civic Virtue 

Even though governments still have the guns and have not yet uniformly agreed 

to defer to self-governing online groups, “We the Netizens” are still mostly in 

control of what happens online. The everyday actions of millions of bloggers 

and tweeters (and re-tweeters) and senders of emails and instant messages draw 

the attention of the entire world to new and interesting (though not always 

important) developments every day. The distributed collaboration arising from 

ratings, rankings and reviews disclose and shame, or shine a flattering light on, 

every action of every author, seller, politician, organization and anyone else who 

wields any form of power. We are learning to use the Internet to engage in a 

decentralized form of democracy that might be called “attention governance.” 

Democracy is about decentralization and equalization of power—particularly 

the power to influence the rules under which we live our lives together. This 

7 WHO CONTROLS THE INTERNET?, supra note 2.


requires the absence of centralized, unaccountable power—tyranny—whether 

exercised by government or corporations. It also requires that individuals 

participate in collective action and adopt a frame of mind that asks how to 

improve society rather than only how best to achieve their own private goals. 

That frame of mind is called “civic virtue.” It creates a feedback loop—showing 

us all ourselves in a mirror, thereby enhancing our ethical standards for both 

individual actions and the actions we take together in organizations and groups 

(including via the global corporations in which we invest, serve as employees or 

participate as customers). Perhaps the single most powerful contribution to the 

sovereignty of the people made by the Internet is its ability to direct our 

collective attention. At one time, mass media held that power. Now we all do, in 

potentially equal measure. While online anonymity may (sometimes 

unfortunately) give us the power to act without disclosing our identity, the 

Internet simultaneously makes it virtually impossible for groups of people to act 

together without being confronted with the consequences of their actions and 

the views of others regarding the moral and social value (or lack thereof) of 

those consequences. 

Democracy is not just about how we organize political campaigns or 

governmental institutions. It is about how we self-organize all aspects of 

society—and what we can do in collaboration with others to improve the world. 

It is not just about freedom from arbitrary control by government (or corporate 

tyrants)—it is, rather, about the myriad ways in which we come together to 

construct society. The Internet has had a profound impact on political 

campaigns by making it much easier for individuals to contribute small amounts 

and get involved in local activities. And netizens could and should use their 

newfound collective voice to instruct their local governments to protect the 

Internet and its new freedoms, rather than using it to restrict our freedom. But 

the Internet has done even more to decentralize decision-making—about how 

we spend our attention and effort and how we organize our collaborative 

efforts. Today, anyone can form a purposeful group online on Facebook, 

Yahoo! Groups, Google Groups, Twitter, etc. and attract adherents to a cause 

or even start a new organization. When we can more easily act together, we 

become more powerful and more free—we enjoy what Tocqueville would have 

called “a new equality of condition.” 8 

The interesting thing about attention is that, as long as you are awake, you have 

to spend it! It is a non-renewable resource. You can fritter it away on the 

entertainments of television or tweeting. You can give it away to an employer 

whose goals and ethics don’t match your own. Or, after the basic necessities of 

life have been arranged for, you can apply a higher standard. Through the 

8 ALEXIS DE TOCQUEVILLE, DEMOCRACY IN AMERICA (1835).


Internet, you can join with others in groups that try to make the world a better 

place, and talk together about what that means. 9 

Good Netizenship 

Governing the Internet well fundamentally entails governing ourselves—making 

sure that more of our time, attention and effort is spent in roles that are defined 

in relation to social organizations and purposive groups that make society more 

productive, congruent, ethical, and, yes, interesting, complex and empowering 

for everyone. It involves defending the new civic religion of the Internet, 

including preservation of individual choice and deference to the self-governance 

of online groups. It is no light duty to be a good netizen. 

Goldsmith and Wu have shown that open communications via the Internet can 

indeed be shut down by governments. 10 But they don’t say anything about how 

to preserve and enhance democracy in the context of a global Internet. They 

postulate the extension of local and state power onto people who have no 

opportunity to participate in making these policies. But they fail to take account 

of the possibility that once We the Netizens understand the threat, we could 

refuse to allow that to happen. 

How could we netizens prevent the tyranny of local governments or of 

corporate intermediaries with a new kind of power generated by network 

effects? It may be a struggle at times. But we use our minds online—we direct 

our attention and support to groups we value. It takes a very seriously repressive 

governmental regime to regulate minds rather than behavior. And not even 

governments, much less corporations, can stand forever in opposition to what 

large segments of the people they regulate think—especially when they are 

thinking, and talking, together. As Victor Hugo famously remarked, “One 

resists the invasion of armies; one simply cannot resist the invasion of ideas.” 11 

Even those who purport to want to preserve civil civic dialogue and selfgovernance 

by the people sometimes suggest that the Internet has broken our 

existing (U.S.) democratic institutions—polarizing political factions, eliminating 

the possibility of political compromise, fostering hate speech and inciting the 

9 As Tocqueville observed about America in the 1830s: “In their political associations the Americans, 

of all conditions, minds, and ages, daily acquire a general taste for association and grow accustomed to the use 

of it. There they meet together in large numbers, they converse, they listen to one another, 

and they are mutually stimulated to all sorts of undertakings. They afterwards transfer to civil 

life the notions they have thus acquired and make them subservient to a thousand purposes. Thus it is by the 

enjoyment of a dangerous freedom that the Americans learn the art of rendering the dangers of freedom less 

formidable.” Id. (emphasis added). 

10 WHO CONTROLS THE INTERNET?, supra note 2. 

11 VICTOR HUGO, THE HISTORY OF A CRIME (1877).


mob. 12 Even those who favor civic virtue suggest that the Internet has led to a 

retreat into individualism, mindless narcissism, pornography, intellectual 

distraction and worse. 13 

The democratic potential of the Internet is under threat—as democracy always 

is. Some local sovereigns attempt to limit its freedoms. People lacking the 

requisite civic virtue may lapse into self-indulgent individualism. We might all 

decide to live under the benevolent dictatorship of search engines and online 

app stores that make our lives a little more convenient and secure. We might all 

decide it is too much trouble to help our fellow netizens in foreign countries 

who are fighting repressive local governments. Large corporations with 

monopolies born of network effects might gain enough power to become a new 

aristocracy, purporting to benefit the people, but ruling as they please and giving 

priority to profit. Bad actors might turn this new communications medium into 

a social nightmare. So any theory of the democratic potential (and actual 

achievement) of the Internet must include a view regarding how we can all use 

the Internet itself to preserve and enhance gains achieved up till now in popular 

sovereignty and civic collaboration. 

The Trajectory of Freedom 

I have such a theory—one derived from reflecting on Tocqueville’s views 

regarding the new democracy that he discovered in the America of 1830: The 

Internet establishes a new equality of condition and enables us to exercise liberty to form 

associations to pursue new civic, social, and cultural goals. Such a world can produce 

wonders simply because we become more powerful when we act together in 

groups. Moreover, to paraphrase Oliver Wendell Holmes, “Man’s mind, once 

stretched to a new democratic practice, never regains its original dimensions.” 14 

The theory, then, is that having discovered and exercised new ways to improve 

the world, whatever they mean by “improve,” netizens will collaborate in 

myriad ways to protect their newfound powers. 

The actual state of society may, of course, periodically regress. Some groups will 

adopt definitions of “improvement” that are so intrusive upon and 

unacceptable to other groups that governmental and corporate powers will be 

rightly invited in to constrain such non-congruent actions. (For example, almost 

everyone agrees that the Internet should not provide a safe haven for child 

12 See, e.g., CASS SUNSTEIN, REPUBLIC.COM 2.0 (2007). 

13 See, e.g., NICHOLAS CARR, THE SHALLOWS: WHAT THE INTERNET IS DOING TO OUR BRAINS 

(2010). 

14 The original quotation is as follows: “Man’s mind, once stretched by a new idea, never 

regains its original dimensions.” Oliver Wendell Holmes, quoted in H. JACKSON BROWN, JR., A 

FATHER’S BOOK OF WISDOM (1989).


pornography or terrorism and governmental powers will need to be used to 

address these problems.) 

But the trajectory of freedom and even civic virtue has been, in broad terms, 

over time, constantly upward—because everyone who gets a chance to 

experience an increased level of democratic self-government—a new “equality 

of condition”—a new kind of power that comes from the ability to direct and 

control one’s own attention and combine one’s efforts with those of others— 

comes to share a desire to have a voice in shaping the world for the better (even 

when we don’t all agree on what “better” means). And everyone has now tasted 

an empowering opportunity to join with others, online, to do so. 

Acting together, the founding netizens created a global network, and thus, 

inevitably, a global economy, society and politics. The visionary founders of the 

Internet did not seek to liberate selfish individualism, frontier justice based on 

force, or mere wilderness escape. They were civically virtuous themselves and 

foresaw the creation of great schools and libraries, social services, cultural 

venues, and everything else a prosperous and democratic global township might 

want, online. 15 Perhaps they were naïve, a bit too optimistic that everyone else 

shared their civility. Perhaps they assumed that most online groups would make 

rules and take actions designed to benefit those who were affected by those 

rules and actions. Confronted with criminals or tyrants, these Internet optimists 

would be (and are) as quick as anyone to call for a “rule of law.” But they can 

now envision a “law” consisting in part of globally applicable rule sets and 

globally accessible self-governing organizations that exist only because netizens 

have devoted their time, attention and effort to support or shun new online 

institutions. 

Copyright law won’t disappear, but we now also have Creative Commons. Laws 

against spam will survive, but we also have software filters. Local content 

regulation will persist, but we now have proxy servers. Banks will be regulated, 

but online currencies can also flourish. Governments will still regulate and tax 

the shipment of physical goods, but most long ago gave up trying to establish 

custom houses at their virtual border. Every netizen is still a citizen, subject to 

local regulation. But, increasingly, we can travel online to virtual places that have 

rules no local legislature would adopt. Land use in Second Life will not become 

a subject of any real world government’s zoning laws. Topic moderation in an 

online discussion group is not likely to become a matter of local regulation. 

15 See, e.g., HOWARD RHEINGOLD, THE VIRTUAL COMMUNITY: HOMESTEADING ON THE 

ELECTRONIC FRONTIER (1993); FRED TURNER, FROM COUNTER CULTURE TO 

CYBERCULTURE: STEWARD BRAND, THE WHOLE EARTH NETWORK, AND THE RISE OF 

DIGITAL UTOPIANISM (2006); KATIE HAFNER & MATTHEW LYON, WHERE WIZARDS STAY UP 

LATE: THE ORIGINS OF THE INTERNET (1996).


Online schools will establish their own rules for participation in the classes they 

offer. 

Where we congregate online, most of the relevant rules will originate with our 

shared support of the manner in which the proprietor (the owner of the server, 

the writer of constraining code, a moderator) “governs” that online space. 

Governments will, in general, defer to online spaces that are mostly minding 

their own business, rather than inflicting harms on outsiders. They have enough 

to do in securing our physical safety. So congruent rule sets voluntarily 

“adopted” by willing “users” will become most of the applicable “law” of online 

life. That new law will be fundamentally democratic in character, not because 

we elected representatives to a global legislature, but because we all have new 

powers to decide where to go online and to persuade others to join us. 

Because global scale and interconnection make it easier to find each other, 

valuable online groups can exist even if there are very few who share the 

group’s goals and interests. By the same token, online sites that no one visits 

lose social salience. An online source of destructive code or spam may still 

intrude upon our finite attention, but almost everyone agrees that we can and 

should get better at constraining the actions of those who use the Internet to 

inflict that kind of harm on others. ISPs provide centralized filters but these 

depend in part on user actions to flag spam and malicious code. This same 

dependence on our collective attention happens to apply to governments and 

corporations, the very real “legal fictions” that we call into existence by means 

of a shared act of imagination. They thrive only insofar as we allocate our own 

time, effort and attention to facilitate their purposes. If we refuse to play along, 

governments and corporations are doomed to lose (legitimate) power and 

cannot ultimately impose their will on an unwilling global polity. 

In short, there was never any possibility, or dream, that global online society 

would be a society without some amount of order. Even the most idealistic 

Internet Exceptionalists, like John Perry Barlow, never denied that there might 

be problems in Cyberspace that needed to be solved, only that “We in 

Cyberspace” would solve them by “forming our own Social Contract.” 16 The 

questions have always been: How much order? Where would it come from? 

And, if it came from the online community, how closely would its mechanisms 

approximate the democratic ideal of giving everyone at least a potentially equal 

say in what particular kind of order (rules, norms, incentives, roles) were to be 

established? Who gets to tell who else what they can and cannot do? Will the 

Internet preserve, or even enhance, the sovereignty of the people? Those were 

the key questions from the Internet’s very beginning. 

16 See A Declaration of the Independence of Cyberspace, supra note 1.


Wu, Goldsmith and others triumphantly declare the death of Internet 

Exceptionalism based on the reality that governments seek to regulate the 

Internet—and succeed in doing so to a much greater extent than some Internet 

Exceptionalists might have imagined possible. But this does not mean the 

Internet does not present profoundly new opportunities for self-government—a 

new potential form of democracy. It is up to us to seize that opportunity. If we 

do so, the Internet will have proved itself as exceptional as its founders hoped. 

Cosmopolitan Pluralism 

Because people have many different values (ideas about the social good), the 

global online society will need to be pluralistic, cosmopolitan, and tolerant of 

diverse coexisting groups. As the Internet’s founding netizens urged, online 

society could and should be based on the moral norm that all groups should be 

“conservative in what they send, liberal in what they accept.” 17 Above all, they 

imagined (correctly, in my view) that such a global online society would become 

ever more complex—providing increasingly diverse roles for people to play 

while, simultaneously, preserving connections (causal and communicative) 

among all its parts and the whole—thereby creating vast new wealth (of all 

types) for all to share. All wealth ultimately comes from trade. And trade 

requires two people who value whatever they have to exchange differently—so 

that the bargain makes them both better off. 18 We need our differences. 

Democratic Internet governance can preserve them, by allowing us to tolerate 

diverse groups, rather than seeking to impose a single rule set on everyone. 

If you think democracy is about voting, then the apogee of democratization is 

when everyone has an equal vote. That just doesn’t scale globally. We will 

“vote” online with our clicks, not with ballots. If you think democracy is about 

deliberation and discussion, then the ideal would seem to be a continuous global 

town meeting, with everyone getting an equal turn at the microphone. Please, 

spare us! 

If, however, you think that democracy is about equalization of (potential) power 

to have an influence on how society is structured, on how we will improve the 

world, then you have to ask: What is it that we all have in equal measure, the 

deployment of which can shape our world and its rules. The answer is attention 

and effort. That is why attention governance is inherently democratic. That is 

17 The Internet Society, RFC 4677, The Tao of IETF: A Novice’s Guide to the Internet Engineering 

Task Force 6 (Sept. 2006) (quoting Jon Postel), available at 

http://tools.ietf.org/pdf/rfc4677.pdf. 

18 See generally, ERIC D. BEINHOCKER, THE ORIGIN OF WEALTH: EVOLUTION, COMPLEXITY, AND 

THE RADICAL REMAKING OF ECONOMICS (2006); DAVID WARSH, KNOWLEDGE AND THE 

WEALTH OF NATIONS: A STORY OF ECONOMIC DISCOVERY (2006).


how the Internet can make self-governance real in the context of a new global 

society of netizens. 

Every social organization with power (including governments and corporations) 

depends critically on a collective act of imagination. We create our social 

institutions by going along together with the idea that they exist, adopting roles 

that constrain our time and attention in the service of their goals and investing 

our attention and effort in ways that further empower them. The Internet is, 

centrally, a way for us to deploy our attention and effort together, in a context 

in which we can see the resulting effects and correct or constrain social 

organizations that don’t share our values. It makes us all citizens (and, indeed, 

global netizens) in a new way. It increases our power and, by doing so, our 

responsibilities. 

A New Sovereignty of the People 

Internet governance will not be about voting, or complex governmental 

regulations, or even treaties among states. It will be more mundane, more 

pervasive, and more profoundly important than that. As long as some states 

create havens for freedom (whether the First Amendment and Section 230 of 

the Communications Decency Act 19 in the United States or the Icelandic 

protection of WikiLeaks 20), netizens will find ways to exercise those freedoms. 

Insofar as some states create havens for wrongdoers (think Nigerian phishers, 

Russian botnets), governments, corporations and individuals will all respond, in 

their own ways, to avoid or suppress such evils. It would be folly to suggest that 

no organizations (and, therefore, people in organizational roles) will be more 

powerful than others, or that no one will use power for evil so widely 

condemned that it cannot and should not be “tolerated.” But no collective 

action can persist online over the long term if it is not “tolerated” by those who 

decide where to direct their attention, what products to buy, what Terms of 

Service to accept, what jobs to take, what companies to invest in, and what local 

politicians to send packing. 

America was founded on the idea of sovereignty of the people. The global 

Internet gives the people more tools to exercise that sovereignty and greater 

visibility on when and how and where to do so. Good netizenship isn’t 

effortless. Civic virtue requires, well, virtue. It is about responsibilities, not 

rights. Whether the Internet will realize its democratic potential will ultimately 

depend, of course, on the character of the people—the new global polity. 

19 47 U.S.C. § 230 (providing liability immunity for providers and users of an “interactive 

computer service” who publish third party information). 

20 See Robert Mackey, Victory for WikiLeaks in Iceland’s Parliament, The Lede Blog, N.Y. TIMES, June 

17, 2010, available at http://thelede.blogs.nytimes.com/2010/06/17/victory-forwikileaks-in-icelands-parliament/.


I’m optimistic about the prospects for enhanced self-governance by a global 

polity, for one simple reason: Most of us, whatever our nationality, want to be 

empowered. To communicate. To associate to make the world better for 

ourselves, our children and everyone else. We’ve had a taste of the new 

democracy of the Internet. We’ll never willingly, or for long, go back. We 

certainly shouldn’t turn away from this new opportunity, the Internet founders’ 

shared dream of a more empowering, democratic, global society. If we 

remember the democratic visions of the founders, and commit to continue to 

act as good netizens, then we could and should all “govern the Internet,” 

together, in 2020—and beyond.


Who’s Who in Internet Politics: 

A Taxonomy of Information 

Technology Policy & Politics 

By Robert D. Atkinson * 

Where’s the Internet in the United States going to be in a decade? Given the 

important role of public policy in shaping a host of Internet issues, one way to 

answer this question is to understand the political constellation that now shapes 

U.S. and—to some extent—international Internet policy. 

Debates have erupted over myriad information technology (IT) issues such as 

copyright protection, privacy, open source software procurement, cybersecurity, 

Internet taxation, media ownership, Internet governance, electronic voting, 

broadband deployment and adoption, anti-trust, spectrum reform, net 

neutrality, Internet censorship, and equality of access. These issues raise familiar 

legal and political questions in some unfamiliar contexts, and have given rise to 

a lively, increasingly shrill, and important digital politics. Today, interest groups 

of all kinds, including a host of single-issue advocacy organizations, routinely 

weigh in on a range of Internet and digital economy issues. Vexing policy 

conundrums arise constantly, with each new business model and Internet 

innovation creating a new wrinkle in the fabric of the debate. 

How we resolve these issues will have important implications for what the 

Internet of 2020 looks like. The debate over IT policy issues does not take 

place in a vacuum or only in the corridors of Congress. From think tanks to 

trade associations to single-issue advocacy groups, a proliferation of 

organizations fights to shape digital policy debates. This essay is a field guide to 

help the reader understand the politics of IT. 1 It describes the major groups of 

players in the IT policy debate and discusses how they differ along two key 

* Robert D. Atkinson is the founder and president of the Information Technology and 

Innovation Foundation, a Washington, DC-based technology policy think tank. He is also 

author of the State New Economy Index series and the book, THE PAST AND FUTURE OF 

AMERICA’S ECONOMY: LONG WAVES OF INNOVATION THAT POWER CYCLES OF GROWTH 

(Edward Elgar, 2005). 

1 For other useful attempts at creating Internet policy typologies, see Adam Thierer & Berin 

Szoka, Cyber-Libertarianism: The Case for Real Internet Freedom, THE TECHNOLOGY LIBERATION 

FRONT, Aug. 12, 2009, http://techliberation.com/2009/08/12/cyber-libertarianismthe-case-for-real-internet-freedom/; 

and Adam Thierer, Are You an Internet Optimist or 

Pessimist? The Great Debate over Technology’s Impact on Society, THE TECHNOLOGY LIBERATION 

FRONT, Jan. 31, 2010, http://techliberation.com/2010/01/31/are-you-an-internetoptimist-or-pessimist-the-great-debate-over-technology%E2%80%99s-impact-onsociety.


dimensions shaping policy: individual empowerment vs. societal benefit; and 

laissez-faire vs. government regulation. It then uses four timely and important 

policy cases (privacy, taxation, copyright protection, and net neutrality) to 

illuminate how these politics play out today in the United States. While primarily 

focused on American digital politics, this framework is not entirely unique to 

the United States. 

The Major Players 

The primary players in the IT policy debate fall into eight basic groups: 

1. Internet Exceptionalists: These “Netizens” believe that they launched the 

Internet revolution. Typified by groups such as the Free Software 

Foundation and the Electronic Frontier Foundation, and dedicated readers 

of Wired magazine, they believe “information wants to be free” 2 and that all 

software should be open-source. They think technology itself can solve 

many problems that it might create (if users are only smart enough to 

program software to protect themselves), and that cyberspace should be 

governed by the informally enforced social mores (i.e., “netiquette”) that 

evolved among early users. Like John Perry Barlow in his 1996 Declaration 

of Independence of Cyberspace, 3 they deplore both government 

involvement in the Internet and its widespread commercialization. In their 

view, anyone who suggests that society, through its legitimately elected 

government leaders, might have a role to play in shaping the Internet, 

including defending copyright, “just doesn’t get it.” Internet exceptionalists 

believe the Internet should be governed by its users. Afraid your privacy is 

being violated? Technologically-empowered users are the best solution, as 

they set their Web browser to reject cookies, use anonymizer tools and 

encrypt their web traffic. Worried about the recording industry losing 

money from Internet piracy? Encourage artists to find a new business 

model, like selling T-shirts and putting on more concerts. Worried over 

lackluster IT industry competitiveness in the U.S.? Don’t make waves, 

Government intervention generally makes things worse. After all, Silicon 

Valley didn’t need Washington to get where it is. 

2. Social Engineers: These liberals believe the Internet is empowering but 

they worry that its growth is having unintended and sometimes dire 

consequences for society—whether they invoke the so-called “Digital 

Divide “ (between the “wired” and the “unwired”) the purported loss of 

privacy, net neutrality, or concern that corporations are controlling the use 

2 Stewart Brand, speaking at the first Hacker’s Conference, 1984. Roger Clarke, Information 

Wants to be Free, http://www.rogerclarke.com/II/IWtbF.html. 

3 John Perry Barlow, A Declaration of the Independence of Cyberspace, Feb. 8, 1996, 

https://projects.eff.org/~barlow/Declaration-Final.html.


of digital content. They mistrust both government and corporations, the 

latter especially—particularly large telecommunications companies and 

Internet companies making money from the use of consumer data (to, 

ironically, provide free content and services). A large array of groups and 

individuals carry this mantle, including the Benton Foundation, Center for 

Democracy and Technology (on some issues), Center for Digital 

Democracy, Civil Rights Forum on Communication Policy, Consumer 

Project on Technology, Electronic Privacy Information Center, Free Press, 

Media Access Project, and Public Knowledge, and scholars such as 

Columbia’s Tim Wu and most of those hanging their hats at Harvard’s 

Berkman Center (among them, Larry Lessig and Yochai Benkler). Social 

engineers tend to believe the Internet should serve mainly as an educational 

and communications tool. They fear that its empowering capabilities will be 

taken away by powerful multinational corporations and statist governments 

that will reshape it to serve their own narrow purposes (either to steal our 

privacy, limit our freedom on the Internet, spy on us, or all three). As such, 

they minimize the role of IT as an economic engine, and focus more on the 

impact of IT on social issues, such as privacy, community, access to 

information and content, and civil liberties. 

3. Free Marketers: This group views the digital revolution as the great third 

wave of economic innovation in human history (after the agricultural and 

industrial revolutions). IT reduces transaction costs and facilitates the 

application of markets to many more areas of human activity. Free 

marketers envision a dramatically reduced role for government as the 

Internet empowers people, liberates entrepreneurs, and enables markets. 

Influenced by groups such as the Cato Institute, the Mercatus Center, the 

Pacific Research Institute, the Phoenix Center, The Progress & Freedom 

Foundation, and the Technology Policy Institute, they consider the 

emergence of the Internet as a vehicle for commerce (e.g., exchanging 

goods, services, and information in the marketplace) and a liberating and 

progressive force. They are skeptical of the need for government 

involvement, even government partnering with industry to more rapidly 

digitize the economy. 

4. Moderates: This group is staunchly and unabashedly pro-IT, seeing it as 

this era’s driving force for both economic growth and social progress. 

While they view the Internet as a unique development to which old rules 

and laws may not apply, they believe appropriate guidelines must be 

developed if it is to reach its full potential. Likewise, they argue that while 

rules and regulations should not favor bricks-and-mortar companies (see 

#8 below) over Internet ones, neither should they favor Internet companies 

over bricks-and-mortars. Moreover, they argue that while government 

should “do no harm” to limit IT innovations, it should also “actively do 

good” by adopting policies to promote digital transformation in areas such


as broadband, the smart electric grid, health IT, intelligent transportation 

systems, mobile payments, digital signatures, and others. Examples of 

moderates include the Center for Advanced Studies in Science and 

Technology Policy, the Center for Strategic and International Studies, the 

Information Technology and Innovation Foundation (ITIF), and the 

Stilwell Center. 

5. Moral Conservatives: This group sees the Internet as a dangerous place, a 

virtual den of iniquity, populated by pornographers, gamblers, child 

molesters, terrorists, and other degenerates. Unlike the free marketers, the 

moral conservatives have no qualms about enlisting governments to 

regulate the Internet. They have been the driving force behind the 

Communications Decency Act’s censorship restrictions and Child Online 

Protection Act (both deemed unconstitutional), Internet filtering in 

libraries, and worked to push legislation to ban online gambling. They have 

also joined forces with the liberal social engineers (Group #2) in pushing 

for strong “net neutrality” regulations, fearing that Internet Service 

Providers (ISPs) will somehow discriminate against Christians online. This 

group argues that, because the Internet is a public space, some rules and 

laws are necessary to govern anti-social behavior. They do not believe that 

technology can solve all social problems—on the contrary, they believe that 

the Internet is generally furthering the decline of culture. Yet, in some 

instances they embrace the Internet as a tool, as evidenced by former 

Secretary of Education William Bennett’s K-12 Internet-based home 

schooling project. In general, moral conservatives don’t want individuals 

empowered to engage in antisocial behavior, nor do they want corporations 

to facilitate such behavior. Examples are groups like the Christian Coalition 

and Focus on the Family, and around the world with countries like 

Indonesia, Thailand, Saudi Arabia and other religiously conservative nations 

that seek to limit activity on the Internet. 

6. Old Economy Regulators: This group believes that there is nothing 

inherently unique about the Internet and that it should be regulated in the 

same way that government regulates everything else, including past 

technologies. There is a certain sense of urgency among certain elected 

officials, government bureaucrats, and “public interest” advocates who 

believe that cyberspace is in a state of near-anarchy—a haven for criminals, 

con artists, and rapacious corporations. Exemplars of this group include, 

law enforcement officials seeking to limit use of encryption and other 

innovative technologies, veterans of the telecom regulatory wars that 

preceded the breakup of Ma Bell, legal analysts working for social 

engineering think tanks, as well as government officials seeking to impose 

restrictive regulatory frameworks on the broadband Internet. As far as old 

economy regulators are concerned, the 1934 Communications Act (or 

perhaps its 1996 update) answered all the questions that will ever arise


regarding the Internet. Moreover, European, Chinese and other old 

economy regulators overseas fear that, absent more regulation, their nations 

will be bypassed by the American Internet leviathan. 

7. Tech Companies & Trade Associations: This group encompasses a 

range of organizations from the politically savvy hardware, software and 

communications giants to Internet start-ups. These businesses, from old 

stalwarts like IBM, AT&T, and Hewlett Packard to “teenagers” like Cisco 

Systems and Microsoft, and “youngsters” like Google and Facebook, 

understand that trade, tax, regulatory, and other public policy issues 

increasingly affect their bottom line and competitive position. While the 

players in this group (and in Bricks and Mortars) don’t have the same level 

of ideological cohesion as the above groups, they share a certain set of 

interests which justifies their grouping. They realize that getting one’s way 

in politics takes more than being right: It requires playing the game and 

making one’s case persuasively. From time to time, some tech businesses 

may take the Internet exceptionalist position that the Internet should be left 

free from government intervention. Generally, they do so only to avoid 

regulation that might put them at a competitive disadvantage. On the 

whole, tech companies tend to believe that regulation can be both 

advantageous and detrimental; they do not fight against all regulations and 

are in favor of the right ones for them (and increasingly support the 

“wrong” ones for their competitors). 4 To some extent, they also advocate 

policies that are good for the technology industry or the economy as a 

whole. While communication companies, being in a traditionally regulated 

industry, have long recognized the importance of government, most IT 

companies have ignored government and policy issues, being too busy 

creating the technologies that drive the digital world. But as these 

companies have matured and become aware, often through painful 

experience, of how issues in Washington can affect their bottom line, many 

have evolved into political sophisticates. And while individual tech 

companies can have different views on different issues, these differences 

are largely rooted in business model interests, rather than ideological views 

about the market or government. 

8. Bricks-and-Mortars: This group includes the companies, professional 

groups, and unions that gain their livelihood from old-economy, face-toface 

business transactions. These include both producers (such as 

automobile manufacturers, record companies, and airlines) and distributors 

and middlemen (such as retailers, car dealers, wine wholesalers, pharmacies, 

4 For a discussion of how technology companies view public policy see ACT’s Understanding the 

IT Lobby: An Insider’s Guide, 2008, 

http://actonline.org/publications/2008/08/05/understanding-the-it-lobby-aninsiders-guide/.


optometrists, real estate agents, or unions representing workers in these 

industries). Many of them fear, often correctly, that the Internet is making 

them obsolete, while others have worked to transform their business 

models to take advantage of e-commerce. In recent years, there has been a 

widening rift between the bricks-and-mortar producers and the distributors 

and middlemen (and the unions that represent their workers). Producers 

have begun to realize that they can use the Internet to go directly to their 

consumers, bypassing (or at least minimizing) the role of bricks-and-mortar 

middlemen. The middlemen and unions, working actively to keep this from 

happening or at least to forestall the day of reckoning, are not shy about 

enlisting the aid of government to “level the playing field.” Certainly, the 

long running battle over taxing Internet sales represented a fight between 

bricks-and-mortars and tech companies. Likewise, the grocery store 

workers’ union in California has recently worked to pass legislation making 

it more difficult for stores to use self-service checkout systems. 5 

The Dividing Lines 

The above groups’ attitudes about Internet policy can be placed along two axes: 

Individual Empowerment vs. Societal Benefit 

This line separates groups on the basis of beliefs about the Internet’s overriding 

purpose. In some ways this is a variant on the classic tension between liberty 

and equality. However, it goes beyond this to represent the tension between 

individualism and communitarianism, with the former being a focus on 

individual rights, and the latter invoking community benefits like economic 

growth, security, and improved quality of life. 

Those in the individual empowerment category believe that IT’s chief function 

is to liberate individuals from control by, or dependence on, big organizations. 

For them the Internet is a vast, open global communications medium designed 

principally to enable individuals to freely communicate and access information. 

When debating any issue, they examine it principally through the lens of how it 

affects individuals, not society as a whole. Thus, the issue of net neutrality is 

seen in terms of its effect on individual freedom to act in any way desired on 

broadband networks. Such groups want to put the little guy on the same playing 

field as the big boys, whether this means supporting small ISPs, small media 

outlets, or individual open source coders. 

Those belonging in the societal benefit camp believe IT and the Internet’s main 

job is to increase economic productivity, promote government responsiveness 

5 Robert D. Atkinson, Innovation and Its Army of Opponents, BUSINESSWEEK, Sept. 23, 2010, 

http://search.businessweek.com/Search?searchTerm=innovation+and+its+army+o 

f+opponents&resultsPerPage=20.


and efficiency, and enable the development new and better services for 

consumers as a societal whole. They tend to examine individual IT policy issues 

through the lens of how they affect the communitarian interest and are willing 

to accept tradeoffs to individual liberty or freedom if they boost overall 

economic or societal well-being. For example, they see the actions of ISPs to 

manage their broadband networks as being necessary to help the majority of the 

users, even if it means that a few “bandwidth hogs” have to wait a minute or 

two longer to download their pirated copy of Lord of the Rings. They also believe 

that both government and corporations can serve as proxies for community 

interests, and that what’s good for, say, Cisco, AT&T, Microsoft or Google or 

the federal government can be good for America as whole. Some groups fall in 

between the two extremes and argue that tradeoffs between particular 

individual’s benefit (or harm) and community interests are inevitable. 

Internet exceptionalists and social engineers generally believe the Internet is all 

about individual empowerment. The former resent its commercialization and 

view empowerment as inevitable. The latter, as stated earlier, believe the 

Internet should mainly be an educational and social networking tool and fear its 

empowering capabilities will be taken away by powerful multinational 

corporations and statist governments that will reshape the Internet to serve 

their own narrow purposes (profit in the former, control in the latter). Both see 

hackers and pirates as lone champions standing tall against greedy corporate and 

inept government leviathans. 

Bricks-and-mortars and old economy regulators see IT in instrumental terms as 

designed for commerce and, by extension, for the community benefit. They just 

don’t like how the Internet has evolved, whether it’s competition from Dot- 

Coms or the spread of strong encryption that frustrates government 

surveillance, censorship, and other control. Tech companies also see IT in more 

instrumental terms, arguing that its rules should facilitate robust commerce. 

Moral conservatives don’t want individuals empowered, since this will just 

enable even more antisocial behavior, and they also don’t want corporations to 

facilitate such behavior. 

Moderates and free marketers occupy the middle ground. They believe that the 

digitization of the economy holds great promise for boosting productivity and 

improving society. At the same time, they see the Internet as creating 

communities, boosting education, and giving people more control over their 

lives. Free marketers don’t believe that individual interests should necessarily 

trump corporate interests—they see corporations as persons under the law.


Laissez-Faire vs. Government Regulation 

The groups divide along this line over the degree to which the government 

should impose formal rules on IT and the Internet. 

Internet exceptionalists, and to a lesser degree free marketers, believe the 

Internet should be governed by its users. These groups lie on the laissez-faire side 

of the dividing line. They consider the Internet unique and capable of creating 

spontaneous order, a model for how the rest of society should be organized. 

Free marketers believe the Internet is what allows Coase’s vision of a society 

with low transaction costs and ubiquitous markets to become a reality. 6 

At the other extreme are groups on the government regulation side of the line, 

who see the Internet as a new “Wild West” calling for a man with a badge to 

protect vulnerable citizens against intrusive governments and profit-hungry 

corporations. Moral conservatives, social engineers, and old economy regulators 

tend to hold this view, arguing for an array of government actions to limit what 

companies can do. So do bricks-and-mortars, although less as a matter of 

principle than as a way of clinging to their ever-weakening economic position. 

Moderates and tech companies occupy the middle ground. They believe the 

Internet is unique and generally requires a light regulatory touch if IT 

innovation is to thrive. But in some key areas such as cybersecurity and 

copyright protection, they believe that the Internet needs stronger rules, 

especially to enable law enforcement to go after bad actors. In still other areas, 

such as the privacy of non-sensitive data and net neutrality, they believe that 

self-regulating government/business partnerships are the best way to protect 

consumers while giving companies needed flexibility. 

ITIF was formed to advance a set of pragmatic solutions to the growing 

number of technology-related policy problems. We believe the growth of the 

digital economy and society depends on a synthesis of these views: the correct 

position will tend to lie at the intersection of the two axes. The dichotomy 

between individual empowerment and institutional efficiency is not a zero-sum 

game. Individuals benefit both socially and economically when governments 

and corporations work more efficiently and effectively, and institutions benefit 

when individuals are informed and able to make choices. A light touch on 

regulation is important to maintain the flexibility required to operate in this 

high-speed economy, but government action is also necessary to give businesses 

and consumers confidence that the Internet is not a den of thieves or a market 

tilted against fair competition, and to help speed digital transformation (e.g., the 

ubiquitous use of IT throughout the economy and society). 

6 Economist Ronald Coase postulated that high transaction costs engendered large 

organizations. See, e.g., RONALD COASE, THE FIRM, THE MARKET AND THE LAW (1988).


Ongoing Policy Debates 

Of course, the above typology is imperfect—with many individuals and 

organizations falling into more than one group or no group at all. But as one 

looks at the central political fights about the future of information technology, 

the influence of these competing factions is clear. As case studies, we consider 

the recent debates over four key issues: privacy, taxation, copyright protection, 

and net neutrality. 

Privacy 

While the recent flaps over Facebook and Google Street View are the most 

visible examples, the collection and use of personal information about Internet 

users by corporations and government is the source of many heated and 

emotional debates. Old economy regulators and social engineers want to impose 

sweeping regulations that would give individuals control over “their” personal 

data. And while they tolerate, grudgingly, advertising as the one true business 

model for Internet content and services (they oppose ISPs allowing content or 

application companies to voluntarily pay for prioritized service) they want to 

limit the effectiveness of online advertising, and the revenue it can raise, 

because of privacy fears. 

Many tech companies want complete freedom to collect personal data, provided 

they comply with privacy policies they write themselves. And while some tech 

companies have supported moderate “notice and choice” legislation, most 

companies remain wary of any federal regulation of privacy, even as they 

recognize the need for federal laws to preempt increasingly antsy state 

legislators from passing a patchwork of different Internet privacy bills. 

Internet exceptionalists expect technology to solve the problem. As far as 

they’re concerned, users should take responsibility for their own privacy and 

apply the tools available to protect their personal data. 

Free marketers reject the need for privacy legislation, asserting that the harms 

from regulation would far outweigh the benefits, and that government 

regulation is likely to be an imposition on individual liberty and choice, 

including basic rights of free speech. While moderates worry that overly-strict 

privacy laws would stifle innovation and increase costs for consumers, they also 

believe that, absent any rules, users will not develop the trust needed for the 

digital economy and society to flourish. 

The recent furor over Facebook is a perfect example of how these issues play 

out. This social network company announced two new features in 2010: instant 

personalization, which allows users to share data from their Facebook profile 

with partner websites, and social plug-ins for third party websites, which allow


users to more easily share web pages they like with their social network outside 

of Facebook. 7 

Social engineers howled in protest, demanding restrictive government 

regulations to bar such practices. Some, like Danah Boyd, a fellow at Harvard’s 

Berkman Center for Internet and Society, went so far as to claim that Facebook 

functioned as a public utility and should be regulated like one. 8 

Facebook was slow to react, initially focusing more on highlighting the benefits 

of its innovative new tools. However, it quickly responded more appropriately, 

rolling out a much more user-friendly and transparent system of user privacy 

controls. 

ITIF and other moderates as well as free marketers argue that government 

control over the privacy policies of social networks is not necessary to protect 

consumers and moreover, would be harmful to future innovation. In the heated 

political environment of the privacy debate, government intervention would 

probably become regulatory overkill. At the same time, moderates argue that 

legitimate privacy concerns about personally identifiable data and sensitive data 

(financial or medical information, for example) need to be addressed through 

comprehensive industry-wide codes of self-regulation, enforceable by 

government action (e.g., FTC action against companies that do not live up to 

their own privacy policies for unfair and deceptive trade practices). 

When it comes to the collection and use of data by government, the coalitions 

reconfigure. Here the Internet exceptionalists, social engineers, and free 

marketers make common cause in their crusade against “Big Brother.” It largely 

does not matter whether the goal is to crack down on deadbeat dads, catch red 

light runners, or prevent terrorist attacks: If it involves the government 

collecting more information or using existing information for new purposes, 

these groups will generally oppose it. In protesting against the growing practice 

of cities installing red light cameras, former Republican House majority leader 

Dick Armey railed: “This is a full-scale surveillance system. Do we really want a 

society where one cannot walk down the street without Big Brother tracking our 

every move?” 9 

7 For more see: Daniel Castro, Information Technology and Innovation Foundation, The Right 

to Privacy is Not a Right to Facebook, April 2010), http://itif.org/publications/facebooknot-right; 

and Daniel Castro, Information Technology and Innovation Foundation, Facebook 

is Not the Enemy, 2010, http://itif.org/publications/facebook-not-enemy. 

8 Danah Boyd, Facebook is a utility; utilities get regulated, APOPHENIA, May 15, 2010, 

http://www.zephoria.org/thoughts/archives/2010/05/15/facebook-is-a-utilityutilities-get-regulated.html. 

9 Thomas C. Greene, Cops Using High-Tech Surveillance in Florida, THE REGISTER, July 2, 2001, 

http://www.theregister.co.uk/2001/07/02/cops_using_hightech_surveillance/.


High-tech companies have engaged in the debate over government use of and 

access to data based in large part on their business interests. Technology 

companies with direct business interests in providing government technologies 

to collect information (e.g., smart card and biometrics companies) have been 

strong supporters of particular initiatives. Other technology companies, 

worrying that government access to data can restrict commerce or reduce 

consumer trust in the Internet (e.g., in cloud computing applications where 

consumer data is remotely stored) have called for limitations or procedural 

safeguards on government access to data. 

Whether a middle position in the debate can be found remains an ongoing 

question. Moderates support the adoption of new technologies by government, 

if it is clearly demonstrated that they fulfill an important public mission and if 

potential privacy problems are effectively addressed, especially by designing 

privacy protections into systems. At the same time, they support putting into 

place adequate rules and protections governing the access to that data by 

government. 

Internet Sales Taxes 

Tax policy is controversial in any setting, but perhaps particularly so with regard 

to the Internet. The collection of state and local sales taxes for Internet 

transactions is so controversial that 15 years after it was first raised, the issue 

continues to be debated. Old economy regulators want sales taxes to be 

collected on Internet purchases and want high taxes on telecommunications 

services to maintain their revenue. The state of Colorado has gone so far as to 

require Internet retailers to share the names and purchase information of 

Colorado residents with the state government (so the state can collect a “use” 

tax from Internet shoppers). Bricks and mortar companies want sales taxes 

imposed to maintain their competitive position against pure-play Internet 

retailers. Some social engineers favor not only sales tax collection, but also 

special taxes on broadband use to subsidize access for low-income and rural 

households. 

By contrast, the tech companies involved in selling over the Internet do not 

want the burden of collecting taxes over thousands of jurisdictions, and they do 

not want to lose their price advantage. Likewise, they do not want broadband or 

telephone service unfairly taxed at higher rates. Others—like many free 

marketers and Internet exceptionalists—oppose Internet sales taxes on 

principle. They believe “the fewer taxes the better,” especially when it comes to 

promoting the new digital economy. 

Internet exceptionalists, tech companies, and free marketers will likely continue 

to oppose giving states the right to tax Internet sales to their residents from 

companies outside their borders. State governments will press hard for the right, 

citing their large budget shortfalls. And pragmatists will likely favor state sales


taxes, particularly if they are tied to a quid pro quo deal forcing states to rescind 

laws and regulations that discriminate against e-commerce sellers, and if taxation 

is administered in ways that minimize administrative burden. For now, however, 

the debate continues, with states legally unable to collect sales taxes and most 

states imposing high, discriminatory taxes on telecommunications services. 

Copyright Protection 

As virtually all media have become digital, protecting copyrights has become a 

nightmare. The controversy over the file sharing system Napster almost a 

decade ago was just the beginning. The ubiquity of file-sharing technologies, 

coupled with computers that can rip digital files from CDs or DVDs, and highspeed 

broadband networks that can quickly transfer large files, has meant that 

“digital piracy” has grown like wildfire. Internet exceptionalists argue that the 

Internet Age marks the end of intellectual property rights because enforcing 

copyright protections on digital media is too difficult (hence their mantra 

“information wants to be free.”) These advocates claim that non-commercial 

file “sharing” of copyrighted media is a form of fair use, which they assert is 

legal under copyright law. For example, the Electronic Freedom Forum’s “Let 

the Music Play” campaign protests the music and film industries’ prosecution of 

file copiers. In their ideal world, some rich dot-com entrepreneur would 

establish a separate country on a desert island, linked to the rest of the world by 

high speed fiber-optic cable and hosting a massive computer with a cornucopia 

of pirated digital content, all beyond the reach of national copyright laws. 

Many social engineers side with the Internet exceptionalists, though for very 

different reasons. They fear that technology will let copyright holders exact such 

strict control on content that traditional notions of fair use will become 

obsolete. And they fear that digital rights management (DRM) technologies will 

become so stringent that activities consumers have long enjoyed (like the ability 

to play music files on more than one device) will be prohibited. Both argue 

strongly against any efforts to better control digital copyright theft that may 

impinge on individual liberty or individual rights like free speech (e.g., permitting 

ISPs to filter for illegal content or crafting international treaties like ACTA to 

strengthen and harmonize anti-piracy efforts). And both would love to see the 

Digital Millennium Copyright Act (DMCA) enter the dust bin of IT policy 

history, particularly the academics and engineers who feel the DMCA restricts 

their ability to hack DRM technology in the name of research. 10 

Because of their emphasis on property rights, most free marketers tend to 

strongly support efforts to limit digital copyright theft. But with their focus on 

freedom, a few come all the way around to the left, arguing that because liberty 

10 Michele Boldrin, Against Intellectual Monopoly, Nov. 10, 2008, 

http://www.cato.org/event.php?eventid=5362.


trumps property, the grant of intellectual property rights by government 

amounts to the provision of a state-sanctioned monopoly. 11 In their view, 

individuals should be free to use digital content in ways they want and content 

owners—not others such as digital intermediaries—should be responsible for 

policing the use of their content. 

Moderates also support efforts to limit digital copyright theft, believing that 

such theft is wrong, and that a robust digital ecosystem requires economic 

incentives to produce often expensive digital content. At the same time, 

however, they are not absolutists, and in particular seek to balance the costs and 

benefits of copyright defense, especially through fair use. 

The bricks and mortar companies—including the Recording Industry 

Association of America—initially worked to block the development of new 

technologies that facilitate playing downloaded and possibly pirated music. But 

more than a decade later the content industries are not so much fighting against 

such technologies as they are working to develop and use technologies that can 

counter copyright theft, and going after organizations that enable widespread 

digital content theft (e.g., the Swedish Pirate Bay). 12 And even as they have 

struggled to cope with music and movie piracy, content producers have largely 

come to terms with the realities of the digital era: They have begun providing 

legal, affordable, and consumer-friendly means for consumers to buy or view 

copyright-protected digital content, with Apple’s iTunes music store and Hulu 

being the most prominent examples. 

Although generally sympathetic to the content providers’ copyright concerns, 

many high-tech companies (e.g., ISPs, search engines, social networks) fear that 

the federal government will require them to adjust their businesses to become 

copyright enforcers, either by having to take action against their customers or 

by building in expensive content protection technologies. Once again, the 

question is whether a compromise can be found, ensuring that content holders 

have the legal protections and economic incentives they need to continue 

producing copyrighted materials without imposing overly-large burdens on 

technology companies, and by extension their customers. 

11 See, e.g., Cato Institute, Against Intellectual Monopoly, 2008, 

http://www.cato.org/event.php?eventid=5362. 

12 Eric Pfanner, Music Industry Counts the Cost of Piracy, THE NEW YORK TIMES, Jan. 21, 2010, 

http://www.nytimes.com/2010/01/22/business/global/22music.html?ref=recordin 

g_industry_association_of_america; Eamonn Forde, From Peer to Eternity: Will Jumping the 

Legal Divide Solve Anything?, THE MUSIC NETWORK, Dec. 6, 2010, 

http://www.themusicnetwork.com/music-features/industry/2010/12/06/frompeer-to-eternity-will-jumping-the-legal-divide-solve-anything/.


Net Neutrality 

What has become a highly contentious issue, net neutrality, refers to the idea 

that the individual networks collectively forming the Internet be controlled by 

users rather than by their owners and operators. While network operators are in 

a unique position to manage their resources, proponents of net neutrality 

believe they cannot be trusted to utilize their knowledge for the good of the 

Internet user community. 

Social engineers are the most passionate about net neutrality, but they make 

common cause with the veterans of the old economy regulator group and 

Internet exceptionalists. Indeed, social engineer Tim Wu coined the stillmystifying 

term “net neutrality.” 13 These groups fear that the Internet’s unique 

nature is under threat by the forces of incumbent telecommunications and cable 

companies providing broadband service. If “Big Broadband” gets its way, 

neutralists fear the Internet will go the way of cable TV, the “vast wasteland” 14 

where elitist programming such as The Wire competes with advertisingsupported, 

populist programming such as American Idol. 

Free marketers see net neutrality as one more attack by big government 

regulators on the Internet, the last bastion of freedom from regulation. They 

argue that market forces and consumer choice will always discipline any anticonsumer 

violations of net neutrality, while antitrust or tort law will serve as a 

handy tool to remedy any anti-business violations. 

Tech companies are split on the issue, largely around which side of the network 

they are on. Those tech companies providing network services (e.g., ISPs and 

major equipment makers) are generally against strong regulations in support of 

network neutrality (at least with regard to the network itself) while companies 

whose business model depends on using the network to gain access to 

customers (e.g., content & service providers like Google) are either neutral or in 

favor of a stronger regulatory regime (at least with regard to the infrastructure 

layers, as opposed to other parts of the Internet “stack,” such as applications.) 

13 Tim Wu, Network Neutrality, Broadband Discrimination, 2 JOURNAL OF TELECOMMUNICATIONS 

AND HIGH TECHNOLOGY LAW 141 (2003), 


14 On May 9, 1961, in a speech to the National Association of Broadcasters, newly-appointed 

FCC chairman Newton N. Minow referred to television as a “vast wasteland.” Newton N. 

Minow, “Television and the Public Interest,” address to the National Association of 

Broadcasters, Washington, D.C., May 9, 1961.


However, these differences have begun to blur somewhat, as evidenced by the 

October 2009 joint statement on net neutrality issued by Google and Verizon. 15 

Moderates generally see the Internet as a work-in-progress. Moderates believe it 

is good that network equipment producers are improving the Internet and 

recognize operators as possessing the highly specialized knowledge needed to 

provide equitable access to the Internet’s pool of resources. But moderates 

realize that competition doesn’t operate as efficiently in some network markets 

as it does in the markets for general-purpose consumer goods and services. In 

other words, some network markets are under-competitive (because network 

effects create market power), so markets alone aren’t sufficient to guarantee an 

open Internet for everyone. 16 The role of government in Internet regulation is 

to ensure that all consumers enjoy the fruits of investment and innovation, but 

only in ways that don’t limit continued investment and innovation. 

As these and other issues continue to be fought in legislatures and communities 

around the country, government officials should seek solutions that balance the 

needs of individuals with those of society, and that offer the security of codified 

laws when necessary and the flexibility of informal rules when appropriate. As 

the technology policy debates go on and the various factions push for the 

solutions that fit their ideologies and interests, the policies that promote the 

growth and vitality of the digital economy will not be found at the extremes, but 

instead in the vital center. 

The Future of Digital Politics 

Some might argue that these issues are transitory and will recede in importance 

as the digital economy matures. But there is good reason to believe otherwise: 

The debates that pit online consumers against resistant middlemen are likely to 

continue as new forms of online distribution evolve. The emergence of much 

faster and ubiquitous wired and wireless broadband networks will mean more 

Americans using these networks and more business models developing to take 

advantage of them. Data generated by emerging new technologies such as 

wireless location systems, digital signature systems, intelligent transportation 

systems, the smart electric grid, health IT, and radio frequency identification 

(RFID) devices—some used by government, others by the private sector—will 

drive new privacy concerns among social engineers and their fellow travelers. In 

some ways, the digital revolution has been so successful that many previously 

15 Lowell McAdam, CEO Verizon Wireless & Eric Schmidt, CEO Google, Finding Common 

Ground on an Open Internet, Verizon PolicyBlog, Oct. 21, 2009, http://policyblog. 

verizon.com/BlogPost/675/FindingCommonGroundonanOpenInternet.aspx. 

16 Richard Bennett, Information Technology and Innovation Foundation, ITIF Comments on 

FCC Broadband Reclassifying, August 10, 2010, http://www.itif.org/publications/itifcomments-fcc-broadband-reclassifying.


analog political issues have become digital issues; on the other hand, the 

political issues of the future remain unformed, precisely because the 

technologies are changing so quickly. 

The public policy issues surrounding the IT revolution are no longer sideshows 

or mere theoretical discussions for a handful of technologically savvy people, 

nor are they the royal road to a utopia of untold wealth and perfect freedom. 

The battle lines have been drawn, and the issues are both serious and complex. 

Digital politics, if not the great issue of our age, will be central to the life of our 

nation in the decade ahead—and well beyond.

PART II 

ISSUES & APPLICATIONS 

343

CHAPTER 6 

SHOULD ONLINE INTERMEDIARIES 

BE REQUIRED TO POLICE MORE? 

Trusting (and Verifying) Online Intermediaries’ Policing 347 


Online Liability for Payment Systems 365 


345 

Fuzzy Boundaries: The Potential Impact 

of Vague Secondary Liability Doctrines 

on Technology Innovation 393 

Paul Szynol

346 CHAPTER 6: SHOULD ONLINE INTERMEDIARIES BE REQUIRED TO POLICE MORE?


Trusting (and Verifying) 

Online Intermediaries’ Policing 

By Frank Pasquale * 

Introduction 

Internet Service Providers (ISPs) and search engines have mapped the Web, 

accelerated e-commerce, and empowered new communities. They can also 

enable intellectual property infringement, harassment, stealth marketing, and 

frightening levels of surveillance. As a result, individuals are rapidly losing the 

ability to control their own image on the web, or even to know what data others 

are presented with regarding them. When Web users attempt to find 

information or entertainment, they have little assurance that a carrier or search 

engine is not subtly biasing the presentation of results in accordance with its 

own commercial interests. 1 

None of these problems is readily susceptible to swift legal intervention. 

Instead, intermediaries themselves have begun policing their own virtual 

premises. eBay makes it easy for intellectual property owners to report 

infringing merchandise. A carrier like Comcast has the technical power to slow 

or block traffic to and from a site like BitTorrent, which is often accused of 

infringement. 2 Google’s StopBadware program tries to alert searchers about 

malware-ridden websites, 3 and YouTube employs an indeterminate number of 

people to police copyright infringement, illegal obscenity, and even many 

grotesque or humiliating videos. 4 Reputable social networks do the same for 

their own content. 

* Professor of Law, Seton Hall Law School; Visiting Fellow, Princeton Center for Information 

Technology Policy. 

1 Benjamin Edelman, Hard-Coding Bias in Google “Algorithmic” Search Results, Nov. 15, 2010, 

available at http://www.benedelman.org/hardcoding/ (“I present categories of searches 

for which available evidence indicates Google has “hard-coded” its own links to appear at 

the top of algorithmic search results, and I offer a methodology for detecting certain kinds 

of tampering by comparing Google results for similar searches. I compare Google’s hardcoded 

results with Google’s public statements and promises, including a dozen denials but 

at least one admission.”). 

2 See, e.g., Comcast Corp. v. FCC, No. 08-1291, 2010 U.S. App. LEXIS 7039 (D.C. Cir. April 

6, 2010). 

3 For more information, visit http://stopbadware.org/. 

4 YouTube, YouTube Community Guidelines, 

http://www.youtube.com/t/community_guidelines (“YouTube staff review flagged

348 CHAPTER 6: SHOULD ONLINE INTERMEDIARIES BE REQUIRED TO POLICE MORE? 

Yet all is not well in the land of online self-regulation. However competently 

they police their sites, nagging questions will remain about their fairness and 

objectivity in doing so. Is Comcast blocking BitTorrent to stop infringement, 

or to decrease access to content that competes with its own for viewers? How 

much digital due process does Google need to give a site it accuses of harboring 

malware? If Facebook eliminates a video of war carnage, is that a token of 

respect for the wounded or one more reflexive effort of a major company to 

ingratiate itself with a Washington establishment currently committed to 

indefinite military engagement in the Middle East? 

Questions like these will persist, and erode the legitimacy of intermediary selfpolicing, 

as long as key operations of leading companies are shrouded in 

secrecy. Administrators must develop an institutional competence for 

continually monitoring rapidly-changing business practices. A trusted advisory 

council charged with assisting the Federal Trade Commission (FTC) and 

Federal Communications Commission (FCC) could help courts and agencies 

adjudicate controversies concerning intermediary practices. An Internet 

Intermediary Regulatory Council (IIRC) would spur the development of what 

Christopher Kelty calls a “recursive public”—one that is “vitally concerned with 

the material and practical maintenance and modification of the technical, legal, 

practical, and conceptual means of its own existence as a public.” 5 Questioning 

the power of a dominant intermediary is not just a preoccupation of the 

anxious. Rather, monitoring is a prerequisite for assuring a level playing field 

online. 

Understanding Intermediaries’ Power 

Internet intermediaries govern online life. 6 ISPs and search engines are 

particularly central to the web’s ecology. Users rely on search services to map 

the web for them and use ISPs to connect to one another. Economic 

sociologist David Stark has observed that “search is the watchword of the 

information age.” 7 ISPs are often called “carriers” to reflect the parallel 

videos 24 hours a day, seven days a week to determine whether they violate our Community 

Guidelines. When they do, we remove them.”) 

5 CHRISTOPHER M. KELTY, TWO BITS: THE CULTURAL SIGNIFICANCE OF FREE SOFTWARE 3 

(Duke Univ. Press 2007). 

6 For a definition of intermediary, see Thomas F. Cotter, Some Observations on the Law and 

Economics of Intermediaries, 2006 MICH. ST. L. REV. 67, 68–71 (“[A]n ‘intermediary’ can be any 

entity that enables the communication of information from one party to another. On the 

basis of this definition, any provider of communications services (including telephone 

companies, cable companies, and Internet service providers) qualify as intermediaries.”). 

7 DAVID STARK, THE SENSE OF DISSONANCE: ACCOUNTS OF WORTH IN ECONOMIC LIFE 1 

(Princeton Univ. Press 2009) (“Among the many new information technologies that are 

reshaping work and daily life, perhaps none are more empowering than the new technologies 

of search.”).


between their own services in the new economy and transportation 

infrastructure. Online intermediaries organize and control access to an 

extraordinary variety of digitized content. Content providers aim to be at the 

top of Google Search or Google News results. 8 Services like iTunes, Hulu, and 

YouTube offer audio and video content. Social networks are extending their 

reach into each of these areas. Cable-based ISPs like Comcast have their own 

relationships with content providers. 9 

When an Internet connection is dropped, or a search engine fails to produce a 

result the searcher knows exists somewhere on the web, such failures are 

obvious. However, most web experiences do not unfold in such a binary, pass– 

fail manner. An ISP or search engine can slow down the speed or reduce the 

ranking of a website in ways that are very hard for users to detect. Moreover, 

there are many points of control, or layers, of the Web. 10 Even when users’ 

experience with one layer causes suspicion, it can blame others for the problem. 

The new power of intermediaries over reputation and visibility implicates 

several traditional concerns of the American legal system. 11 Unfortunately, 

Internet intermediaries are presently bound only by weak and inadequate 

enforcement of consumer protection and false advertising statutes, which were 

designed for very different digital infrastructures. 

8 See Deborah Fallows & Lee Rainie, Pew Internet & Am. Life Project, Data Memo: The 

Popularity and Importance of Search Engines 2 (Aug. 2004), 

http://www.pewinternet.org/pdfs/PIP_Data_Memo_Searchengines.pdf (“The 

average visitor scrolled through 1.8 result pages during a typical search.”); Leslie Marable, 

False Oracles: Consumer Reaction to Learning the Truth About How Search Engines Work: Results of an 

Ethnographic Study, CONSUMER WEBWATCH, June 30, 2003, at 5, available at 

http://www.consumerwebwatch.org/pdfs/false-oracles.pdf (“The majority of 

participants never clicked beyond the first page of search results. They trusted search 

engines to present only the best or most accurate, unbiased results on the first page.”). 

9 ROBERT W. MCCHESNEY, RICH MEDIA, POOR DEMOCRACY: COMMUNICATION POLITICS IN 

DUBIOUS TIMES 123 (2000) (describing how convergence of digital technology “eliminates 

the traditional distinctions between media and communications sectors”). 

10 JONATHAN ZITTRAIN, THE FUTURE OF THE INTERNET—AND HOW TO STOP IT 67 (2008) 

(describing a physical layer, the “actual wires or airwaves over which data will flow;” an 

application layer, “representing the tasks people might want to perform on the network;” a 

content layer, “containing actual information exchanged among the network’s users;” and a 

social layer, “where new behaviors and interactions among people are enabled by the 

technologies underneath”). 

11 Yochai Benkler, Communications Infrastructure Regulation and the Distribution of Control over Content, 

22 TELECOMM. POL’Y 183, 185–86 (1998) (describing the power of intermediaries over 

information flow: “technology, institutional framework, and organizational adaptation … 

determine … who can produce information, and who may or must consume, what type of 

information, under what conditions, and to what effect”); Cotter, supra note 6, at 69–71 

(discussing some of the functions of technological intermediaries, including their control of 

information flow from suppliers to consumers).


In the space of a brief essay, I cannot survey the entire range of intermediary 

policing practices. But it is worthwhile to drill down a bit into the tough 

questions raised by one intermediary—the dominant search engine, Google—as 

it decides what is and is not an acceptable practice for search engine optimizers 

who want their clients’ sites to appear higher in the rankings for given queries. 

Search engineers tend to divide the search engine optimization (SEO) business 

into “good guys” and “bad guys,” often calling the former “white hat SEO” and 

the latter “black hat SEO.” 12 Some degree of transparency regarding the search 

engine’s algorithm is required to permit white hat SEO. These rules are 

generally agreed upon as practices that “make the web better;” i.e., have fresh 

content, don’t sell links, don’t “stuff metatags” with extraneous information just 

to get attention. However, if there were complete transparency, “black hat” 

SEOs could unfairly elevate the visibility of their clients’ sites—and even if this 

were only done temporarily, the resulting churn and chaos could severely reduce 

the utility of search results. Moreover, a search engine’s competitors could use 

the trade secrets to enhance its own services. 

This secrecy has led to a growing gray zone of Internet practices with uncertain 

effect on sites’ rankings. Consider some of the distinctions below, based on 

search engine optimization literature: 

White Hat (acceptable) 13 Gray Area (unclear how these 

are treated) 14 

Asking blogs you like to 

link to you, or engaging in 

reciprocal linking between 

your site and other sites in a 

legitimate dialogue. 15 

Paying a blogger or site to link 

to your blog in order to boost 

search results and not just to 

increase traffic. 

Black Hat (unacceptable; can 

lead to down-ranking in Google 

results or even the “Google 

Death Penalty” of De-Indexing) 

Creating a “link farm” of spam 

blogs (splogs) to link to you, or 

linking between multiple sites 

you created (known as link 

farms) to boost search results. 16 

12 Elizabeth van Couvering, Is Relevance Relevant?, 

http://jcmc.indiana.edu/vol12/issue3/vancouvering.html (search engineers’ 

“animosity towards the … guerilla fighters of spamming and hacking, is more direct” than 

their hostility toward direct business competitors); Aaron Wall, Google Thinks YOU Are a 

Black Hat SEO. Should You Trust Them?, SEOBOOK, Apr. 17, 2008, 

http://www.seobook.com/to-google-you-are-a-spammer (claiming that Google 

discriminates against self-identified SEOs). 

13 Phil Craven, ‘Ethical’ Search Engine Optimization Exposed!, WebWorkshop, 

http://www.webworkshop.net/ethical-search-engine-optimization.html (last visited 

Jun. 8, 2009). 

14 Grey Hat SEO, http://greyhatseo.com/ (last visited Jun. 5, 2006) (claiming a Grey Hat 

SEO is someone who uses black hat techniques in an ethical way.) 

15 Link Schemes, GOOGLE WEBMASTER CENTRAL, 

http://www.google.com/support/webmasters/bin/answer.py?answer=66356 (“The




Running human-conducted 

tests of search inquiries 

with permission from the 

search engine. 

Creating non-intentional 

duplicate content (through 

printer-friendly versions, 

pages aimed at mobile 

devices, etc.) 18 

Generating a coherent site 

with original and 

informative material aimed 

at the user 

Doing a few queries to do 

elementary reverse 

engineering. (This may not be 

permitted under the Terms of 

Service). 

Intentionally creating 

permitted duplicate content to 

boost search results 

Creating content or additional 

pages that walk the line 

between useful information 

and “doorway pages” 





Using computer programs to 

send automated search queries 

to gauge the page rank 

generated from various search 

terms (Terms of Service 

prohibit this) 17 

Intentionally creating 

unnecessary duplicate content 

on many pages and domains to 

boost results 

Creating “doorway pages” that 

are geared towards popular 

keywords but that redirect to a 

largely unrelated main site. 19 

best way to get other sites to create relevant links to yours is to create unique, relevant 

content that can quickly gain popularity in the Internet community. The more useful content 

you have, the greater the chances someone else will find that content valuable to their 

readers and link to it.”). 

16 Duncan Riley, Google Declares Jihad On Blog Link Farms, TECHCRUNCH, Oct. 24, 2007, 

http://www.techcrunch.com/2007/10/24/google-declares-jihad-on-blog-linkfarms/. 

17 Automated Queries, GOOGLE WEBMASTER CENTRAL, 

http://www.google.com/support/webmasters/bin/answer.py?answer=66357 

(“Google’s Terms of Service do not allow the sending of automated queries of any sort to 

our system without express permission in advance from Google.”); Google Terms of 

Service: Use of the Services by you, http://www.google.com/accounts/TOS (last visited 

Jun. 4, 2009) (“You agree not to access (or attempt to access) any of the Services by any 

means other than through the interface that is provided by Google, unless you have been 

specifically allowed to do so in a separate agreement with Google.”). 

18 Duplicate Content, GOOGLE WEBMASTER CENTRAL, 

http://www.google.com/support/webmasters/bin/answer.py?answer=66359 

(“Examples of non-malicious duplicate content could include: Discussion forums that can 

generate both regular and stripped-down pages targeted at mobile devices, Store items 

shown or linked via multiple distinct URLs, Printer-only versions of web pages”). 

19 Google Blogoscoped, German BMW Banned From Google, Feb. 4, 2006, 

http://blogoscoped.com/archive/2006-02-04-n60.html; Matt Cutts, Ramping up on 

International Webspam, MATT CUTTS: GADGETS, GOOGLE, AND SEO, Feb. 4, 2006, 

http://www.mattcutts.com/blog/ramping-up-on-international-webspam/ (Google 

employee confirming BMW’s ban).




Targeting an appreciative 

audience 20 

Influencing search engine 

by making pages easier to 

scan by automated bots 23 

Putting random references to 

salacious or celebrity topics on 

a blog primarily devoted to 

discussing current affairs 21 

Creating “hidden pages” when 

there may be a logical reason 

to show one page to search 

engine bots and another page 

to users who type in the page’s 

URL 





Distracting an involuntary 

audience with completely 

misleading indexed content 

(akin to “initial interest 

confusion” in Internet 

trademark law) 22 

Using “hidden pages” to show a 

misleading page to search 

engine bots, and another page 

to users who type in the page’s 

URL 

As these practices show, search engines are referees in the millions of contests 

for attention that take place on the web each day. There are hundreds of 

entities that want to be the top result in response to a query like “sneakers,” 

“restaurant in New York City,” or “best employer to work for.” Any academic 

who writes on an obscure subject wants to be the “go-to” authority when it is 

Googled—and for consultants, a top or tenth-ranked result could be the 

20 Webmaster Guidelines: Design and content guidelines, GOOGLE WEBMASTER CENTRAL, 

http://www.google.com/support/webmasters/bin/answer.py?answer=35769 (last 

visited Jun. 4, 2009) (“Think about the words users would type to find your pages, and make 

sure that your site actually includes those words within it.”). 

21 Daniel Solove, Thanks, Jennifer Aniston (or the Manifold Ways to Do the Same Search), 

CONCURRING OPINIONS, 

http://www.concurringopinions.com/archives/2006/01/thanks_jennifer.html (“One 

of my more popular posts is one entitled Jennifer Aniston Nude Photos and the Anti- 

Paparazzi Act. It seems to be getting a lot of readers interested in learning about the 

workings of the Anti-Paparazzi Act and the law of information privacy. It sure is surprising 

that so many readers are eager to understand this rather technical statute. Anyway, for the 

small part that Jennifer Aniston plays in this, we thank her for the traffic.”); Dan Filler, Coffee 

Or Nude Celebrity Photos: A Tale Of Two Evergreen Posts, THE FACULTY LOUNGE, 

http://www.thefacultylounge.org/2008/04/coffee-or-nude.html (“significant amounts 

of traffic arrived in the form of web surfers seeking out pictures of Jennifer Aniston”). 

22 Jason Preston, Google punishes Squidoo for having too much Spam, BLOG BUSINESS SUMMIT, Jul. 11, 

2007, http://blogbusinesssummit.com/2007/07/google-punishes-squidoo-forhaving-too-much-spam.htm. 

23 Webmaster Guidelines: Design and Content Guidelines, GOOGLE WEBMASTER CENTRAL, 

http://www.google.com/support/webmasters/bin/answer.py?answer=35769 (last 

visited Jun. 4, 2009) (“Create a useful, information-rich site, and write pages that clearly and 

accurately describe your content.”); Id. (“Try to use text instead of images to display 

important names, content, or links. The Google crawler doesn’t recognize text contained in 

images.”).


difference between lucrative gigs and obscurity. The top and right hand sides of 

many search engine pages are open for paid placement; but even there the 

highest bidder may not get a prime spot because a good search engine strives to 

keep even these sections very relevant to searchers. 24 The organic results are 

determined by search engines’ proprietary algorithms, and preliminary evidence 

indicates that searchers (and particularly educated searchers) concentrate 

attention there. Businesses can grow reliant on good Google rankings as a way 

of attracting and keeping customers. 

For example, John Battelle tells the story of the owner of 2bigfeet.com (a seller 

of large-sized men’s shoes), whose site was knocked off the first page of 

Google’s results for terms like “big shoes” by a sudden algorithm shift in 

November 2003, right before the Christmas shopping season. The owner 

attempted to contact Google several times, but said he “never got a response.” 

Google claimed the owner may have hired a search engine optimizer who ran 

afoul of its rules—but it would not say precisely what those rules were. 25 Like 

the IRS’s unwillingness to disclose all of its “audit flags,” the company did not 

24 Steven Levy, Secret of Googlenomics: Data-Fueled Recipe Brews Profitability, WIRED, May 2, 2009, 

http://www.wired.com/culture/culturereviews/magazine/17- 

06/nep_googlenomics (in Google’s AdWords program,”The bids themselves are only a 

part of what ultimately determines the auction winners. The other major determinant is 

something called the quality score. This metric strives to ensure that the ads Google shows 

on its results page are true, high-caliber matches for what users are querying. If they aren’t, 

the whole system suffers and Google makes less money.”); see also Google, What is the Quality 

Score and How is it Calculated, 

http://adwords.google.com/support/aw/bin/answer.py?hl=en&answer=10215 (last 

visited Sept. 1, 2009) (“The AdWords system works best for everybody—advertisers, users, 

publishers, and Google too—when the ads we display match our users’ needs as closely as 

possible.”). 

25 JOHN BATTELLE, THE SEARCH: HOW GOOGLE AND ITS RIVALS REWROTE THE RULES OF 

BUSINESS AND TRANSFORMED OUR CULTURE (Portfolio Trade 2005). See also Joe Nocera, 

Stuck in Google’s Doghouse, N.Y. TIMES, Sept. 13, 2008, 

http://www.nytimes.com/2008/09/13/technology/13nocera.html (“In the summer of 

2006 … Google pulled the rug out from under [web business owner Dan Savage, who had 

come to rely on its referrals to his page, Sourcetool]… . When Mr. Savage asked Google 

executives what the problem was, he was told that Sourcetool’s “landing page quality” was 

low. Google had recently changed the algorithm for choosing advertisements for prominent 

positions on Google search pages, and Mr. Savage’s site had been identified as one that 

didn’t meet the algorithm’s new standards… . Although the company never told Mr. Savage 

what, precisely, was wrong with his landing page quality, it offered some suggestions for 

improvement, including running fewer AdSense ads and manually typing in the addresses 

and phone numbers of the 600,000 companies in his directory, even though their Web sites 

were just a click away. At a cost of several hundred thousand dollars, he made some of the 

changes Google suggested. No improvement.”). Savage filed suit against Google on an 

antitrust theory, which was dismissed in March 2010. See TradeComet, LLC v. Google, Inc., 

2010 U.S. Dist. LEXIS 20154 (S.D. N.Y. March 5, 2010), 

http://www.courthousenews.com/2010/03/08/Google%20opinion.pdf.


want to permit manipulators to gain too great an understanding of how it 

detected their tactics. 

So far, claims like 2bigfeet.com’s have not been fully examined in the judicial 

system, largely because Google has successfully deflected them by claiming that 

its search results embody opinions protected by the First Amendment. Several 

articles have questioned whether blanket First Amendment protection covers all 

search engine actions, and that conclusion has not yet been embraced on the 

appellate level in the United States. 26 The FTC’s guidance to search engines, 

promoting the clear separation of organic and paid results, suggests that search 

engines’ First Amendment shield is not insurmountable. 27 While a creative or 

opportunistic litigant could conceivably advance a First Amendment right to 

promote products or positions without indicating that the promotion has been 

paid for, such a challenge has not yet eliminated false advertising law, and even 

political speakers have been required to reveal their funding sources. 28 

Qualified Transparency for Carrier 

& Search Engine Practices 

Both search engines’ ranking practices and carriers’ network management 

should be transparent to some entity capable of detecting biased policing by 

these intermediaries. 29 There are some institutional precedents for the kind of 

monitoring that would be necessary to accomplish these goals. For example, 

the French Commission Nationale De L’Informatique et des Libertes (CNIL) 

has several prerogatives designed to protect the privacy and reputation of 

26 Frank Pasquale, Rankings, Reductionism, and Responsibility, CLEVELAND ST. L. REV. (2006); 

Frank Pasquale & Oren Bracha, Federal Search Commission, 93 CORNELL L. REV. 1149 (2008); 

Jennifer A. Chandler, A Right to Reach an Audience: An Approach to Intermediary Bias on the 

Internet, 35 HOFSTRA L. REV. 1095, 1109 (2007). 

27 See Bracha & Pasquale, Federal Search Commission, supra note 26 (discussing the implications of 

Ellen Goodman’s work on “stealth marketing” for search engines, and how the Hippsley 

Letter of 2002 inadequately addressed such concerns in the industry.). 

28 In early cases alleging an array of unfair competition and business torts claims against search 

engines, the First Amendment has proven a formidable shield against liability. Search 

engines characterize their results as opinion, and lower courts have been reluctant to penalize 

them for these forms of expression. In other work, I have described why this First 

Amendment barrier to accountability should not be insurmountable. Search engines take 

advantage of a web of governmental immunities that they would be loath to surrender. 

FAIR v. Rumsfeld, 547 U.S. 47 (2006) and cognate cases stand for the proposition that such 

immunities can be conditioned on agreement to certain conditions on an entity’s speech. 

Whatever the federal government’s will, it is within its power to regulate ranking and rating 

entities in some way when they are so deeply dependent on governmental action. Frank 

Pasquale, Asterisk Revisited, 3 J. BUS. & TECH. LAW 61 (2008). 

29 I mean partial in two senses of the word—unduly self-interested, or only partly solving 

problems they claim to be solving.


French citizens, and to enforce standards of fair data practices. 30 CNIL 

“ensure[s] that citizens are in a position to exercise their rights through 

information” by requiring data controllers to “ensure data security and 

confidentiality,” to “accept on-site inspections by the CNIL,” and to “reply to 

any request for information.” 31 CNIL also grants individual persons the right to 

obtain information about the digital dossiers kept on them and the use of this 

information. For example, CNIL explains that French law provides that: 

Every person may, on simple request addressed to the 

organisation in question, have free access to all the information 

concerning him in clear language. 

Every person may directly require from an organisation 

holding information about him that the data be corrected (if 

they are wrong), completed or clarified (if they are incomplete 

or equivocal), or erased (if this information could not legally be 

collected). 

30 Law No. 78-17 of January 6, 1978, J.C.P. 1978, III, No. 44692. English translation of law as 

amended by law of August 6, 2004, and by Law of May 12, 2009, 

http://www.cnil.fr/fileadmin/documents/en/Act78-17VA.pdf; French language text 

modified through Law No. 2009-526 of May 12, 2009, J.O., May 13, 2009, 

http://www.cnil.fr/la-cnil/qui-sommes-nous/; French language consolidated version as 

of May 14, 2009, 

http://www.legifrance.gouv.fr/affichTexte.do?cidTexte=JORFTEXT000000886460 

&fastPos=1&fastReqId=826368234&categorieLien=cid&oldAction=rechTexte. 

Commission Nationale de l’Informatique et des Libertés (CNIL), founded by Law No. 78-17 

of January 6, 1978, supra, is an independent administrative French authority protecting 

privacy and personal data held by government agencies and private entities. Specifically, 

CNIL’s general mission consists of ensuring that the development of information 

technology remains at the service of citizens and does not breach human identity, human 

rights, privacy, or personal or public liberties. 

31 CNIL, Rights and Obligations, http://www.cnil.fr/english/the-cnil/rights-andobligations/ 

(last visited Mar. 12, 2010). Specifically, Chapter 6, Article 44, of the CNILcreating 

Act provides: 

The members of the “Commission nationale de l’informatique et des 

libertés” as well as those officers of the Commission’s operational services 

accredited in accordance with the conditions defined by the last paragraph of 

Article 19 (accreditation by the commission), have access, from 6 a.m to 9 

p.m, for the exercise of their functions, to the places, premises, surroundings, 

equipment or buildings used for the processing of personal data for 

professional purposes, with the exception of the parts of the places, 

premises, surroundings, equipment or buildings used for private purposes. 

Law No. 78-17 of January 6, 1978, J.C.P. 1978, III, No. 44692, ch. 6, art. 44, at 30, 

http://www.cnil.fr/fileadmin/documents/en/Act78-17VA.pdf.


Every person may oppose that information about him is used 

for advertising purposes or for commercial purposes. 32 

While the United States does not have the same tradition of protecting privacy 

prevalent in Europe, 33 CNIL’s aims and commitments could prove worthwhile 

models for U.S. agencies. 

U.S. policymakers may also continue to experiment with public–private 

partnerships to monitor problematic behavior at search engines and carriers. 

For instance, the National Advertising Division (NAD) of the Council of Better 

Business Bureaus is a “voluntary, self-regulating body” that fields complaints 

about allegedly untruthful advertising. 34 The vast majority of companies 

investigated by NAD comply with its recommendations, but can also resist its 

authority and resolve the dispute before the FTC. 35 Rather than overwhelming 

the agency with adjudications, the NAD process provides an initial forum for 

advertisers and their critics to contest the validity of statements. 36 NAD is part 

of a larger association called the National Advertising Review Council (NARC), 

which promulgates procedures for NAD, the Children’s Advertising Review 

Unit (CARU), and the National Advertising Review Board (NARB). 37 

32 CNIL, Rights and Obligations, supra note 31. 

33 James Whitman, The Two Western Cultures of Privacy: Dignity Versus Liberty, 113 YALE L.J. 1151, 

1155 (2004) (comparing U.S. and European privacy law). 

34 Seth Stevenson, How New Is New? How Improved Is Improved? The People Who Keep Advertisers 

Honest, SLATE, July 13, 2009, http://www.slate.com/id/2221968. 

35 Id. (“When an ad is brought to their attention, the NAD’s lawyers review the specific claims 

at issue. The rule is that the advertiser must have substantiated any claims before the ad was 

put on the air, so the NAD will first ask for any substantiating materials the advertiser can 

provide. If the NAD lawyers determine that the claims aren’t valid, they’ll recommend that 

the ad be altered. The compliance rate on this is more than 95 percent. But if the advertiser 

refuses to modify the ad (this is a voluntary, self-regulating body, not a court of law), the 

NAD will refer the matter to the Federal Trade Commission. One such FTC referral 

resulted in an $83 million judgment against a weight-loss company.”). 

36 Id. 

37 NATIONAL ADVERTISING REVIEW COUNCIL, THE ADVERTISING INDUSTRY’S PROCESS OF 

VOLUNTARY SELF-REGULATION: POLICIES AND PROCEDURES § 2.1(a) (July 27, 2009) (“The 

National Advertising Division of the Council of Better Business Bureaus (hereinafter NAD), 

and the Children’s Advertising Review Unit (CARU), shall be responsible for receiving or 

initiating, evaluating, investigating, analyzing (in conjunction with outside experts, if 

warranted, and upon notice to the parties), and holding negotiations with an advertiser, and 

resolving complaints or questions from any source involving the truth or accuracy of 

national advertising.”). Though billed as “self-regulation,” it is difficult to see how the policy 

would have teeth were it not self-regulation in the shadow of an FTC empowered by the 

Lanham Act to aggressively police false advertising. The FTC has several mechanisms by 

which to regulate unfair business practices in commerce. See, e.g., 15 U.S.C. § 45(b) (2006)


Instead of an “Innovation Environment Protection Agency (iEPA)” (the agency 

Lawrence Lessig proposed to supplant the FCC), I would recommend the 

formation of an Internet Intermediary Regulatory Council (IIRC), which would 

assist both the FCC and FTC in carrying out their present missions. 38 Like the 

NARC, the IIRC would follow up on complaints made by competitors, the 

public, or when it determines that a practice deserves investigation. If the selfregulatory 

council failed to reconcile conflicting claims, it could refer complaints 

to the FTC (in the case of search engines, which implicate the FTC’s extant 

expertise in both privacy and advertising) or the FCC (in the case of carriers). 

In either context, an IIRC would need not only lawyers, but also engineers and 

programmers who could fully understand the technology affecting data, ranking, 

and traffic management practices. 

An IIRC would research and issue reports on suspect practices by Internet 

intermediaries, while respecting the intellectual property of the companies it 

investigated. An IIRC could generate official and even public understanding of 

intermediary practices, while keeping crucial proprietary information under the 

control of the companies it monitors. An IIRC could develop a detailed 

description of safeguards for trade secrets, which would prevent anyone outside 

its offices from accessing the information. 39 Another option would be to allow 

IIRC agents to inspect such information without actually obtaining it. An IIRC 

could create “reading rooms” for use by its experts, just as some courts allow 

restrictive protective orders to govern discovery in disputes involving trade 

secrets. The experts would review the information in a group setting (possibly 

during a period of days) to determine whether a given intermediary had engaged 

in practices that could constitute a violation of privacy or consumer protection 

laws. Such review would not require any outside access to sensitive 

information. 

I prefer not to specify at this time whether an IIRC would be a private or public 

entity. Either approach would have distinct costs and benefits explored (in 

part) by a well-developed literature on the role of private entities in Internet 

(giving the commission the authority to register an official complaint against an entity 

engaged in unfair business methods). 

38 It could include a search engine division, an ISP division focusing on carriers, and eventually 

divisions related to social networks or auction sites if their practices begin to raise 

commensurate concerns. 

39 This is the way that the NAD proceeds. It provides specific procedures under which the 

participants can request that certain sensitive information be protected. See NAT’L 

ADVERTISING REVIEW COUNCIL, THE ADVERTISING INDUSTRY’S PROCESS OF VOLUNTARY 

SELF-REGULATION § 2.4(d)–(e), at 4–5 (2009), 

http://www.nadreview.org/07_Procedures.pdf (discussing procedure for confidential 

submission of trade secrets).


governance. 40 Regardless of whether monitoring is done by a governmental 

entity (like CNIL) or an NGO (like NARC), we must begin developing the 

institutional capacity to permit a more rapid understanding of intermediary 

actions than traditional litigation permits. 41 

It is not merely markets and antitrust enforcement that are insufficient to 

constrain problematic intermediary behavior—the common law is also likely to 

fall short. It is hard to imagine any but the wealthiest and most sophisticated 

plaintiffs’ attorneys attempting to understand the tweaks to the Google 

algorithm that might have unfairly diminished their clients’ sites’ salience. Trade 

secrets have been deployed in the context of other litigation to frustrate 

investigations of black box algorithms. 42 Examination of Google’s algorithms 

subject to very restrictive protective orders would amount to a similar barrier to 

accountability. Given its recent string of litigation victories, it is hard to imagine 

rational litigants continuing to take on that risk. Moreover, it makes little sense 

for a court to start from scratch in understanding the complex practices of 

intermediaries when an entity like the IIRC could develop lasting expertise in 

interpreting their actions. 

A status quo of unmonitored intermediary operations is a veritable “ring of 

Gyges,” 43 tempting them to push the envelope with policing practices which 

40 See, e.g., Philip J. Weiser, Internet Governance, Standard Setting, and Self-Regulation, 28 N. KY. L. 

REV. 822, 822 (2001) (examining “in particular the nature and limits of a key private 

regulator of the Internet: standard-setting organizations and their institution of open, 

interoperable standards”). 

41 Google has already recognized the need for some kind of due process in response to 

complaints about its labeling of certain websites as “harmful” (due to the presence of viruses 

or other security threats at the sites) via the StopBadware program. See ZITTRAIN, FUTURE 

OF THE INTERNET, supra note 10, at 171 (“Requests for review—which included pleas for 

help in understanding the problem to begin with—inundated StopBadware researchers, who 

found themselves overwhelmed in a matter of days by appeals from thousands of Web sites 

listed. Until StopBadware could check each site and verify it had been cleaned of bad code, 

the warning page stayed up.”). Google’s cooperation with the Harvard Berkman Center for 

Internet Research to run the StopBadware program could prefigure future intermediary 

cooperation with NGOs to provide “rough justice” to those disadvantaged by certain 

intermediary practices. 

42 See Jessica Ring Amunson & Sam Hirsch, The Case of the Disappearing Votes: Lessons from the 

Jennings v. Buchanan Congressional Election Contest, 17 WM. & MARY BILL RTS. J. 397, 397–98 

(2008) (“[T]he litigation ultimately was utterly inconclusive as to the reason for the 18,000 

electronic undervotes because discovery targeting the defective voting system was thwarted 

when the voting machines’ manufacturer successfully invoked the trade-secret privilege to 

block any investigation of the machines or their software by the litigants.”). 

43 “The Ring of Gyges is a mythical magical artifact mentioned by the philosopher Plato in 

book 2 of his Republic (2.359a–2.360d). It granted its owner the power to become invisible 

at will. Through the story of the ring, Republic discusses whether a typical person would be 

moral if he did not have to fear the consequences of his actions.” Wikipedia, Ring of Gyges, 

http://en.wikipedia.org/wiki/Ring_of_Gyges (last accessed Dec. 1, 2010).


cannot be scrutinized or challenged. Distortions of the public sphere are also 

likely. While a commercially-influenced “fast-tracking” or “up-ranking” of 

some content past others might raise suspicions among its direct (but dispersed) 

victims, the real issues it raises are far broader. If an online ecology of 

information that purports to be based on one mode of ordering is actually based 

on another, it sets an unfair playing field whose biases are largely undetectable 

by lay observers. Stealth marketing generates serious negative externalities that 

menace personal autonomy and cultural authenticity. Moreover, the degree of 

expertise necessary to recognize these externalities in the new online 

environment is likely to be possessed by only the most committed observers. 

This potent combination of expertise and externalities is a classic rationale for 

regulation. As Danny Weitzner’s proposal for “extreme factfinding” (in the 

context of the Google–DoubleClick merger review) recognized, only a 

dedicated group of engineers, social scientists, attorneys, and computer 

scientists are likely to be adept enough at understanding search engine decisions 

as a whole to understand particular complaints about them. 44 Someone needs 

to be able to examine the finer details of the publicly undisclosed operation of 

culturally significant automated ranking systems—that is, to watch those who 

watch and influence us. 45 

44 See generally, Danny Weitzner, What to Do About Google and Doubleclick? Hold Google to It’s Word 

With Some Extreme Factfinding About Privacy Practices, GOOGLE OPEN INTERNET POLICY BLOG, 

Oct. 8, 2007, http://dig.csail.mit.edu/breadcrumbs/node/203: 

In the 1990s, the FTC under Christine Varney’s leadership pushed operators 

of commercial websites to post policies stating how they handle personal 

information. That was an innovative idea at the time, but the power of 

personal information processing has swamped the ability of a static statement 

to capture the privacy impact of sophisticated services, and the level of 

generality at which these policies tend to be written often obscure the real 

privacy impact of the practices described. It’s time for regulators to take the 

next step and assure that both individuals and policy makers have 

information they need. 

Weitzner proposes that “[r]egulators should appoint an independent panel of technical, legal 

and business experts to help them review, on an ongoing basis the privacy practices of 

Google.” Id. The panel would be “made up of those with technical, legal and business 

expertise from around the world.” Id. It would hold “public hearings at which Google 

technical experts are available to answer questions about operational details of personal data 

handling.” Id. There would be “staff support for the panel from participating regulatory 

agencies,” “real-time publication of questions and answers,” and “[a]n annual report 

summarizing what the panel has learned.” Id. 

45 In the meantime, Google has been developing a tool that would help consumers detect if 

their Internet service provider was “running afoul of Net neutrality principles.” Stephanie 

Condon, Google-Backed Tool Detects Net Filtering, Blocking, CNET NEWS, Jan. 28, 2009, 

http://news.cnet.com/8301-13578_3-10152117-38.html (“[The tool, M-Lab,] is running 

three diagnostic tools for consumers: one to determine whether BitTorrent is being blocked 

or throttled, one to diagnose problems that affect last-mile broadband networks, and one to


Why Dominant Search Engines & 

Carriers Deserve More Scrutiny than 

Dominant Auction Sites & Social Networks 

Those skeptical of the administrative state may find this proposal to “watch the 

watchers” problematic. They think of intermediaries as primarily market actors, 

to be disciplined by market constraints. However, the development of 

dominant Web 2.0 intermediaries was itself a product of particular legal choices 

about the extent of intellectual property rights and the responsibilities of 

intermediaries made in legislative and judicial decisions in the 1990s. As 

intermediaries gained power, various entities tried to bring them to heel— 

including content providers, search engine optimizers, trademark owners, and 

consumer advocates. In traditional information law, claims under trademark, 

defamation, and copyright law might have posed serious worries for 

intermediaries. However, revisions of communications and intellectual property 

law in the late 1990s provided safe harbors that can trump legal claims sounding 

in each of these other areas. 46 Some basic reporting responsibilities are a small 

price to pay for continuing enjoyment of such immunities. 

An argument for treating internet intermediaries more like regulated entities 

owes much to the trail-blazing work of legal realists. Among these, Robert 

Hale’s work on utilities remains especially inspirational. 47 Hale developed many 

of the theoretical foundations of the New Deal, focusing on the ways in which 

the common law became inadequate as large business entities began ordering 

diagnose problems limiting speeds.”). It remains to be seen whether Google itself would 

submit to a similar inspection to determine whether it was engaging in stealth marketing or 

other problematic practices. 

46 17 U.S.C. § 512(d) (2000) (Digital Millennium Copyright Act of 1998 safe harbor); 47 U.S.C. 

§ 230(c)(1) (2000) (Communications Decency Act of 1997 safe harbor for intermediaries). 

For critical commentary on the latter, see Michael L. Rustad & Thomas H. Koenig, Rebooting 

Cybertort Law, 80 WASH. L. REV. 335, 371 (2005) (“An activist judiciary, however, has radically 

expanded § 230 by conferring immunity on distributors. Section 230(c)(1) has been 

interpreted to preclude all tort lawsuits against ISPs, websites, and search engines. Courts 

have … haphazardly lump[ed] together web hosts, websites, search engines, and content 

creators into this amorphous category.”). 

47 Ilana Waxman, Note, Hale’s Legacy: Why Private Property is Not a Synonym for Liberty, 57 

HASTINGS L.J. 1009, 1019 (“Hale’s most fundamental insight was that the coercive power 

exerted by private property owners is itself a creature of state power… . By protecting the 

owner’s property right … ‘the government’s function of protecting property serves to 

delegate power to the owners’ over non-owners, so that ‘when the owners are in a position 

to require non-owners to accept conditions as the price of obtaining permission to use the 

property in question, it is the state that is enforcing compliance, by threatening to forbid the 

use of the property unless the owner’s terms are met.’ … .[A]ll property essentially 

constitutes a delegation of state power to the property owner…). For a powerful application 

of these ideas to Internet law, see Julie Cohen, Lochner in Cyberspace: The New Economic 

Orthodoxy of ‘Rights Management,’ 97 MICH. L. REV. 462 (1998).


increasing proportions of the national economy. 48 Hale’s crucial insight was 

that many of the leading businesses of his day were not extraordinary 

innovators that “deserved” all the profits they made; rather, their success was 

dependent on a network of laws and regulation that could easily shift favor 

from one corporate player to another. 49 Hale focused his theoretical work on 

the utilities of his time, expounding an economic and philosophical justification 

for imposing public service obligations on them. Regulatory bodies like state 

utility commissions and the FCC all learned from his work, which showed the 

inadequacy of private law for handling disputes over infrastructural utilities. 

Market advocates may worry that monitoring of search engines and carriers will 

lead to more extensive surveillance of the affairs of other intermediaries, like 

social networks and auction platforms. They may feel that competition is 

working in each of those areas, and should be the foundation of all intermediary 

policy. However, competition is only one of many tools we can use to 

encourage responsible and useful intermediaries. We should rely on 

competition-promotion via markets and antitrust only to the extent that (a) the 

intermediary in question is an economic (as opposed to cultural or political) 

force; (b) the “voice” of the intermediary’s user community is strong; 50 and (c) 

competition is likely to be genuine and not contrived. These criteria help us 

map older debates about platforms onto newer entities. 

For search engines and carriers, each of these factors strongly militates in favor 

of regulatory intervention. Broadband competition has failed to materialize 

beyond duopoly service for most Americans. There are several reasons to 

suspect that Google’s dominance of the general purpose search market will 

continue to grow. 51 Just as past policymakers recognized the need for common 

48 Duncan Kennedy, The Stakes of Law, or, Hale and Foucault, 15 LEGAL STUDIES FORUM 

(4) (1991). 

49 BARBARA FRIED, THE PROGRESSIVE ASSAULT ON LAISSEZ FAIRE: ROBERT HALE AND THE 

FIRST LAW AND ECONOMICS MOVEMENT (1998), available at 

http://www.hup.harvard.edu/catalog/FRIPRA.html. 

50 Competition is designed to provide users an “exit” option; regulation is designed to give 

them more of a “voice” in its governance. Hirschman ALBERT O. HIRSCHMAN, Exit and 

Voice: An Expanding Sphere of Influence, in RIVAL VIEWS OF MARKET SOCIETY AND OTHER 

RECENT ESSAYS 78–80 (1986) (describing “exit” and “voice” as two classic options of 

reform or protest). To the extent exit is unavailable, voice (influence) within the relevant 

intermediary becomes less necessary; to the extent voice is available, exit becomes less 

necessary. 

51 Bracha & Pasquale, Federal Search Commission, supra note 26, at 1179. Section III of the article, 

“Why Can’t Non-Regulatory Alternatives Solve the Problem?,” addresses the many factors 

impeding competition in the search market. Present dominance entrenches future 

dominance as the leading search engine’s expertise on user habits grows to the extent that no 

competitor can match its understanding of how to target ads well. Id. Since that article was 

published, Harvard Business School Professor Ben Edelman has investigated another selfreinforcing 

aspect of Google’s market power: the non-portability of AdSense data, which


carrier obligations for concentrated communications industries, present ones 

will need to recognize carriers’ and search engines’ status as increasingly 

essential facilities for researchers, advertisers, and media outlets. 52 

The parallel is apt because, to use the three dimensions discussed above, carriers 

and dominant general-purpose search engines a) are just as important to culture 

and politics as they are to economic life, b) conceal key aspects of their 

operations, and are essentially credence goods, vitiating user community 

influence, and c) do not presently face many strong competitors, and are 

unlikely to do so in the immediate future. The first point—regarding cultural 

power—should lead scholars away from merely considering economies of scale 

and scope and network effects in evaluating search engines. We need to 

consider all dimensions of network power—the full range of cultural, political, 

and social obstacles to competition that a dominant standard can generate. 53 

Moreover, policymakers must acknowledge that competition itself can drive 

practices with many negative externalities. The bottom line here is that someone 

needs to be able to “look under the hood” of culturally significant automated 

ranking systems. 

What about auction platforms, another important online intermediary? 54 Here, 

a purely economic, antitrust-driven approach to possible problems is more 

appropriate. To use the criteria mentioned above: (a) a site like eBay is a very 

important online marketplace, but has little cultural or political impact and (b) 

the user community at eBay understands its reputation rankings very well, and 

has shown remarkable capacities for cohesion and self-organization to protest 

makes it difficult for Google customers to apply what they have learned about their Internet 

customers to ad campaigns designed for other search engines. Ben Edelman, PPC Platform 

Competition and Google’s ‘May Not Copy’ Restriction, June 27, 2008, 

http://www.benedelman.org/news/062708-1.html. As Edelman shows, Google has 

tried to make the data it gathers for companies “sticky,” inextricable from its own 

proprietary data structures. 

52 TIM WU, THE MASTER SWITCH (Knopf, 2010) (promoting “separations principle” in the 

digital landscape.). 

53 DAVID GREWAL, NETWORK POWER: THE SOCIAL DYNAMICS OF GLOBALIZATION 45 (Yale 

Univ. Press 2008) (“[T]he network power of English isn’t the result of any intrinsic features 

of English (for example, ‘it’s easy to learn’): it’s purely a result of the number of other people 

and other networks you can use it to reach… . The idea of network power … explains how 

the convergence on a set of common global standards is driven by the accretion of 

individual choices that are free and forced at the same time.”). 

54 David S. Evans, Antitrust Issues Raised by the Emerging Global Internet Economy, 102 NW. U. L. 

REV. COLLOQUY 285, 291 (2008) (“European Community law and decisional practice … 

impose special obligations and significant scrutiny on firms that have market shares as low as 

40 percent.”). Evans compiles data demonstrating that some leading auction platforms 

(such as eBay) are well above this market share in Europe and the U.S. Id. (citing comScore, 

MyMetrix qSearch 2.0 Key Measures Report, Dec. 2007, 

http://www.comscore.com/method/method.asp).


(and occasionally overturn) policies it dislikes. These factors overwhelm the 

possibility that (c) competition in the general auction market (as opposed to 

niche auctions) may be unlikely to develop. If real competitors fail to 

materialize due to illicit monopolization, antitrust judgments against Microsoft 

(and parallel requirements of some forms of “operating system neutrality”) can 

guide future litigants seeking online auction platform neutrality. While eBay’s 

user community successfully pressured Disney to end its 2000 specialpreference 

deal with eBay, in the future antitrust judgments or settlements 

might require the full disclosure of (and perhaps put conditions on) such deals. 55 

In social networks, another area where tipping can quickly lead to one or a few 

players’ dominance, 56 the situation is more mixed. While Rebecca Mackinnon 

and danah boyd have compared Facebook to a utility, the famously marketoriented 

Economist magazine has compared it to a country, possibly in need of 

a constitution and formal input from users. Social networks are closer to search 

engines than auction sites with respect to factor a: they are becoming crucial 

hubs of social interactions, cultural distribution and promotion, and political 

organizing. 57 

On the other hand, social networks provide a some leverage to their members 

to police bad behavior, opening up “voice” options, with respect to factor b, far 

more potent than those available to the scattered searchers of Google. A group 

named “Facebook: Stop Invading My Privacy” became very popular within 

Facebook itself, catalyzing opposition to some proposed features of its Beacon 

program in 2008. 58 Facebook’s privacy snafus in early 2009 led the company to 

organize formal user community input on future alterations to the company’s 

terms of service. On the final factor, competitive dynamics, it appears that 

competition is more likely to develop in the social network space than in the 

broadband, search engine, or auction platform industries. There is a more 

55 In 2000, eBay granted special perks to Disney on a platform within its auction site. After 

protest from “the eBay community,” the perks ceased. eBay CEO Meg Whitman said of the 

special Disney deal: “We’ve concluded that eBay has to be a level playing field. That is a 

core part of our DNA, and it has to be going forward.” ADAM COHEN, THE PERFECT 

STORE: INSIDE EBAY 292 (Back Bay Books 2006). 

56 In early 2008, 98% of Brazilian social networkers used Google’s Orkut; 97% of South 

Korean social networkers used CyWorld, and 83% of American social networkers used 

MySpace or Facebook. Evans, supra note 54 at 292. 

57 James Grimmelmann, Saving Facebook, 94 IOWA L. REV. 1137, 52-59 (2009), 

http://works.bepress.com/james_grimmelmann/20/. 

58 William McGeveran, Disclosure, Endorsement, and Identity in Social Marketing, 2009 ILL. L. REV. 

1105, 1120 (2009), 

http://www.law.uiuc.edu/lrev/publications/2000s/2009/2009_4/McGeveran.pdf.


diverse playing field here than in the carrier or search space, with more than 

4,000 social networks in the United States. 59 

Any policy analysis of dominant intermediaries should recognize the sensitive 

cultural and political issues raised by them. The cultural, communal, and 

competitive dynamics surrounding dominant search engines and carriers defy 

easy or stereotyped responses. Qualified transparency will assist policymakers 

and courts that seek to address the cultural, reputational, and political impact of 

dominant intermediaries. 

Conclusion 

As David Brin predicted in The Transparent Society, further disclosure from 

corporate entities needs to accompany the scrutiny we all increasingly suffer as 

individuals. 60 While the FTC and the FCC have articulated principles for 

protecting privacy, they have not engaged in the monitoring necessary to 

enforce these guidelines. This essay promotes institutions designed to develop 

better agency understanding of privacy-eroding practices. Whether public or 

private, such institutions would respect legitimate needs for business 

confidentiality while promoting individuals’ capacity to understand how their 

reputations are shaped by dominant intermediaries. 

59 Evans, supra note 54, at 290. 

60 DAVID BRIN, THE TRANSPARENT SOCIETY: WILL TECHNOLOGY FORCE US TO CHOOSE 

BETWEEN PRIVACY AND FREEDOM? (Basic Books 1999).


Online Liability for 

Payment Systems 

By Mark MacCarthy � 

Introduction 

U.S. policy toward the liability of Internet intermediaries for online harms was 

set in the late 1990s. It consisted of two parts. The first part was Section 230 of 

the 1996 Telecommunications Act, providing a safe harbor from indirect 

liability for online service providers. 1 This safe harbor is an exception from a 

range of normal liabilities that would apply to traditional providers of media 

content such as broadcasters and newspapers. It does not apply to all 

intermediaries or platform providers, but to what might be called “pure” 

Internet intermediaries. That is, it covers intermediaries to the extent they are 

providing services that are somehow intrinsic to the Internet. Under its terms, 

except for requirements of contract law, criminal law, and intellectual property 

law, online entities are not responsible for the content of the material that is 

found on their systems as long as it has been provided by another information 

content provider. 

The second part of the U.S. policy toward Internet intermediary liability was set 

out in 1998 with the Digital Millennium Copyright Act. 2 DMCA allows a 

complete exemption from copyright liability for entities involved in pure 

transmission activities. It also creates a notice-and-takedown regime for web 

hosts and other online service providers. It also allows recipients of these 

notices to challenge them. Upon receipt of a response, the online service 

providers are required to reinstate the allegedly infringing material unless the 

rights holder has filed a legal infringement action. Online service providers are 

exempt from liability for good faith removal of material following a notice. It 

also provides for penalties if a rights holder files a notification that knowingly 

� Mark MacCarthy is Adjunct Professor in the Communications Culture and Technology 

Program at Georgetown University. Formerly, he was Senior Vice President for Public 

Policy at Visa Inc. Substantial portions of this essay were originally published as Mark 

MacCarthy, What Payment Intermediaries Are Doing About Online Liability and Why It Matters, 25 

BERKELEY TECHNOLOGY LAW JOURNAL 1039 (2010). 

1 47 U.S.C. § 230(c)(1) (2006) (“No provider or user of an interactive computer service shall 

be treated as the publisher or speaker of any information provided by another information 

content provider.”). The interpretation of this provision is quite broad. See, e.g., Zeran v. Am. 

Online, Inc., 129 F.3d 327, 330-31 (4th Cir. 1997) (finding that plaintiff’s tort claims of 

defamation were preempted by § 230). The immunity does not extend to criminal law, 

contract law, or intellectual property law. 47 U.S.C. § 230(e)(1)-(4) (2006). 

2 17 U.S.C. § 512 avaiable at http://www4.law.cornell.edu/uscode/17/512.html.


misrepresents that the material is infringing. DMCA also requires online service 

providers to have in place some procedures to respond to “repeat infringers,” 

including termination of accounts in appropriate circumstances. 3 

Many commentators think that DMCA represents a balanced compromise. 4 

However, controversy persists. Content providers have successfully lobbied for 

laws imposing more robust responsibilities for stopping copyright infringement 

on Internet service providers (ISPs). France and the United Kingdom, for 

example, have adopted “graduated response” mechanisms. 5 These liability 

regimes require ISPs to forward copyright infringement notices to alleged 

infringers, and to disconnect alleged repeat infringers. 

On the other hand, defenders of civil liberties and the First Amendment think 

DMCA notice-and-takedown requirements are too strong, arguing that a large 

proportion of the complaints filed under the law are improper, 6 and that they 

contains an inherent imbalance toward takedown, even when First Amendment 

values are implicated. 7 

In another essay in this collection, Brian Holland strongly defends Section 230 

as a modified version of Internet exceptionalism, 8 and as providing the basis for 

the development of innovation on the Internet. 9 However, it, too, has been 

3 17 U.S.C. § 512(i) conditions the eligibility of the safe harbor. It applies only if the service 

provider “has adopted and reasonably implemented, and informs subscribers and account 

holders of the service provider’s system or network of, a policy that provides for the 

termination in appropriate circumstances of subscribers and account holders of the service 

provider’s system or network who are repeat infringers.” Intermediaries such as Google, 

YouTube and AT&T appear to have established termination policies for copyright 

infringement. 

4 See, for example, “United States and Canada Overview,” in RONALD DEIBERT, JOHN 

PALFREY, RAFAL ROHOZINSKI, AND JONATHAN ZITTRAIN, ACCESS CONTROLLED 378 ( 

2010) and JONATHAN ZITTRAIN, THE FUTURE OF THE INTERNET AND HOW TO STOP IT 119 

(2008). 

5 Eric Pfanner, U.K. Approves Crackdown on Internet Pirates, NEW YORK TIMES, April 8, 2010 at 

http://www.nytimes.com/2010/04/09/technology/09piracy.html?scp=1&sq=digital 

%20economy%20bill%20uk&st=cse. Eric Pfanner, France Approves Wide Crackdown on Net 

Piracy, NEW YORK TIMES, Oct. 22, 2009, 

http://www.nytimes.com/2009/10/23/technology/23net.html?_r=1. 

6 Jennifer M. Urban & Laura Quilter, Efficient Process or ‘Chilling Effects’? Take-down Notices Under 

Section 512 of the Digital. Millennium Copyright Act, 22 SANTA CLARA HIGH TECH LJ 621 (2006). 

7 Wendy Seltzer, Free Speech Unmoored in Copyright’s Safe Harbor: Chilling Effects of the DMCA on 

the First Amendment, BERKMAN CENTER RESEARCH PUBLICATION NO. 2010-3, p. 16, March 

2010, available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1577785. 

8 See chapter 3, see also H. Brian Holland, In Defense of Online Intermediary Immunity: Facilitating 

Communities of Modified Exceptionalism, 56 U. KAN. L. REV. 369, 397 (2007). 

9 Remarks by Lawrence Strickling, Assistant Secretary of Commerce for Communications and 

Information, to Internet Society’s INET Series: Internet 2020: The Next Billion Users, April 29,


controversial. Some argue that it allows ISPs to avoid making socially-desirable 

investments necessary to provide security on their networks. 10 Others think it 

allows hosting sites to escape their responsibility for defamation and other 

harms caused by people who use their sites to spread false and damaging 

information. 11 

Commentary on the controversies involved in these two pillars of the U.S. 

policy toward online liability is growing. 12 Work on whether to revise the 

consensus position on intermediary liability is underway. 13 

This essay attempts to contribute to this debate by looking at what payment 

systems have been doing about online liability. This will provide an illuminating 

perspective on the debate for a very straightforward reason: Payment systems 

have operated outside this framework for online liability. They are not covered 

by Section 230 and they are not subject to the notice-and-takedown provisions 

of the DMCA. How have they handled issues relating to the use of payment 

systems for illegal activity online? This essay explores this question through an 

examination of two cases in which they have been called upon to take steps to 

control illegal activity involving their payment systems: Internet gambling and 

copyright infringement. 

Some have argued that payment systems should have legal responsibility for 

keeping their systems free of illegal online activity. 14 Payment systems can keep 

track of those who use their system online – both merchants and cardholders 

have contracts with financial. The online transactions using payment systems 

can be tracked electronically by type. Governments and aggrieved parties might 

not be able to find wrong-doers who use payment systems for illegal online 

activity, but the payment system providers can. They are the “least-cost” 

2010, available at 

http://www.ntia.doc.gov/presentations/2010/InternetSociety_04292010.html. 

10 Doug Lichtman & Eric Posner, Holding Internet Service Providers Accountable, 14 U. CHI. SUP. 

CT. ECON. REV. 221 (2006). 

11 JOHN PALFREY AND URS GASSER, BORN DIGITAL 106 (2008), and DANIEL SOLOVE, THE 

FUTURE OF REPUTATION 125-160 (2007). 

12 See, Adam Thierer, Dialogue: The Future of Online Obscenity and Social Networks, ARS TECHNICA, 

March 5, 2009, http://arstechnica.com/tech-policy/news/2009/03/a-friendlyexchange-about-the-future-of-online-liability.ars. 

13 See, for instance, Organization for Economic Cooperation and Development, The Economic 

and Social Role of Internet Intermediaries, April 2010, 

http://www.oecd.org/dataoecd/49/4/44949023.pdf. 

14 Ronald J. Mann & Seth R. Belzley, The Promise of Internet Intermediary Liability, 47 WM. & MARY 

L. REV. 239, 249-50 (2005), available at 

http://scholarship.law.wm.edu/cgi/viewcontent.cgi?article=1225&context=wmlr.


avoider of the damage done by this illegal online activity and so should bear the 

burden of controlling it. 

This perspective seems wrong to me. Still, payment system practices toward 

illegal online activity on their systems suggest several lessons. First, regardless of 

the precise legal liabilities, intermediaries have a general responsibility to keep 

their systems free of illegal transactions and they are taking steps to satisfy that 

obligation. Second, the decision to impose legal responsibilities on 

intermediaries should not be based on the least cost avoider principle. 

Assessments of intermediary liability must take into account market failures, as 

well as an analysis of costs, benefits and equities. Third, if intermediaries are 

shouldered with responsibilities to control illegal online activity, these 

responsibilities needed to be clearly spelled out. Fourth, if governments are 

going to use intermediaries to enforce local laws, they must harmonize these 

local laws. 

Part II of this essay outlines a framework for the analysis of intermediary 

liability. This framework calls for a thorough analysis, including an assessment 

of market failure and an analysis of the costs, benefits, and equities involved in 

imposing intermediary liability. Part III applies this framework to the policies 

and practices of payment intermediaries in the areas of Internet gambling and 

online copyright infringement. Part IV draws some conclusions from these 

experiences. 

Indirect Intermediary 

Liability Regimes 

Most legal regimes hold parties liable for their own misconduct. In contrast, an 

indirect liability regime holds a person responsible for the wrongs committed by 

another. There are usually several parties involved in an indirect liability regime: 

the bad actor, the wronged party and a third party. The bad actor is the person 

directly involved in causing the harm to the wronged party. A third party, 

neither the bad actor nor the wronged party, is assigned responsibility in an 

attempt to prevent the harmful conduct of the bad actor or to compensate the 

wronged party for the harm. In the case of copyright infringement, for example, 

the bad actor would be the infringer, the wronged party would be the record 

company that owned the music copyrights, and the third party would be an ISP 

or a payment system that facilitates the infringement. Indirect liability can be 

imposed through a variety of legal mechanisms. 15 

15 See, e.g., Douglas Lichtman, Holding Internet Service Providers Accountable, 27 REG. 54, 59 (2004) 

(proposing that ISP liability for cyber security issues could be established in a regime of 

“negligence or strict liability, whether it is best implemented by statute or via gradual 

common law development”); Mann & Belzley, supra note 14, at 269-72 (suggesting three 

possible regimes: traditional tort regime, a takedown requirement, and a hot list).


A Framework for Analysis 

Indirect liability holds a party responsible for wrongs committed by another 

person. Why should there be any such rule? Why not simply hold the bad actor 

responsible? The economic analysis of indirect liability attempts to answer this 

question using some standard economic tools and concepts. 16 A standard 

economic framework considers issues of market failure, costs and benefits, and 

equity to assess the need for an indirect liability regime in specific cases. 17 

Market Failure Analysis 

Before imposing an indirect liability regime, economic analysis asks whether 

there is really any market failure. If there is no market failure, there is no need 

for an indirect liability rule. In particular, there need not be an indirect liability 

rule when the law or the wronged party can effectively reach the bad actor 

directly 18 and transaction costs are not significant. 

Even if the wronged party cannot easily reach a bad actor that a third party can 

reach, it is still not necessary to impose liability on the third party. When the 

wronged party and the intermediary can easily negotiate an arrangement, 

efficiency will guide the third party to undertake enforcement efforts on behalf 

of the wronged party. This is a key aspect of a market failure analysis. Unless 

transaction costs interfere with contracting, affected parties can allocate liability 

efficiently through contractual design. 19 

16 See generally Lichtman & Posner, supra note 10 (summarizing this perspective); Douglas 

Lichtman & William Landes, Indirect Liability for Copyright Infringement: An Economic Perspective, 

16 HARV. J.L. & TECH. 395, 396-99 (2003). 

17 See Lichtman & Posner, supra note 10, at 228-33. 

18 The effective reach condition is evaluated prior to an assessment of the ability of a third 

party to effectively control the bad activity. See id. at 230-31. If the law or the wronged party 

can easily reach the bad actor, then why even consider whether to impose a duty on a third 

party? Of course, the bad actors are never totally out of reach of the law or wronged parties. 

With some finite expenditure of resources, perhaps very large, the direct bad actors could be 

brought to justice or harms could be prevented. The real economic question is whether 

those costs are larger than the costs of assigning that enforcement role to a third party. And 

this means that the effective reach condition collapses into the control factor discussed, 

infra. Landes and Lichtman put the comparative point accurately, applied to the specific case 

of contributory copyright liability: “Holding all else equal, contributory liability is more 

attractive … the greater the extent to which indirect liability reduces the costs of copyright 

enforcement as—compared to a system that allows only direct liability.” Lichtman & 

Landes, supra note 16, at 398. 

19 Lichtman & Posner, supra note 10, at 235. Lichtman and Posner also focus on what the 

parties might do: “The right thought experiment is to imagine that all the relevant entities 

and all the victims and all the bad actors can efficiently contract one to another and then to 

ask how the parties would in that situation allocate responsibility for detecting and deterring 

bad acts.” Id. at 257. But there is no need to conduct this thought experiment in the abstract. 

Free, equal, and rational parties can bargain to allocate responsibility and so we can answer


Cost Benefit Analysis 

Some arguments for indirect liability follow a least cost analysis. A “least cost” 

perspective puts the burden of enforcing the law on the party that can stop the 

illegal transactions at the lowest cost. Focusing on costs is desirable in order to 

create an efficient enforcement regime. In a “least cost” framework, the cost to 

the intermediary itself and to the direct customers of the intermediary must be 

taken into account. If ISPs or payment systems have to incur costs to monitor 

their system for illegal content, those costs will be passed down to their direct 

customers. With the price increase, some customers stop using the service or 

reduce their usage of it. If the service provided is a network service, then the 

external network effects on other users of the service from an overall reduction 

in use also have to be counted. 20According to the least cost idea, when these 

costs are less than the cost of enforcement activity by the wronged party or by 

enforcement officials, then liability rests with the intermediary. 

This least-cost analysis is limited. It ignores the size of the harms that can be 

avoided by intermediary action. The mistake is to think that if efforts by third 

parties provide more enforcement than efforts by the wronged parties then it 

must be worthwhile for the third parties to take these enforcement steps. 

Similarly, it is sometimes thought that if third parties can more easily reach bad 

actors than the wronged parties, then they should be required to do so. But this 

is wrong. It is almost always possible to spend more on enforcement and obtain 

some return. From an economic point of view, the question is whether that 

extra spending provides commensurate reductions in damages. Therefore, the 

least cost rule is not the right decision rule, even in a strictly economic analysis. 

Instead, a full cost-benefit analysis is more appropriate. 21 

the question of what the parties would do in this thought experiment by looking at what they 

actually do. The relevant inquiry is whether the bargaining situation is free of significant 

transaction costs or other obstacles to reaching an agreement. 

20 If there are fewer Internet subscribers then the service is less valuable to e-commerce 

merchants as well since there are fewer potential customers. See Matthew Schruers, The 

History and Economics of ISP Liability for Third Party Content, 88 VA. L. REV. 205, 250-52 (2002); 

see also Lichtman & Posner, supra note 10, at 241-43 (seeming to minimize the importance of 

these external, network effects in assessing liability regimes: “Immunizing ISPs from liability 

is not the correct mechanism for encouraging them to provide positive externalities.” Id. at 

243). However, the loss of the ISP-generated external benefits is a potential cost of assigning 

liability that has to be taken into account when assessing whether to assign liability. Mann 

and Belzley’s article gets the overall point right, noting: “To the extent the regulation affects 

conduct with positive social value, as is likely in at least some of the contexts this essay 

discusses, the direct and indirect effects on that conduct must be counted as costs of any 

regulatory initiative.” Mann & Belzley, supra note 14, at 274. 

21 The least-cost analysis seems to function like a cost effectiveness analysis, where a given 

level of enforcement is assumed and the question is how that goal can be reached at the 

lowest cost. See Mann & Belzley, supra note 14, at 250 (adopting that perspective as “a 

mature scheme of regulation that limits the social costs of illegal Internet conduct in the


There is a difference between the costs and benefits to private parties involved 

and the costs and benefits to society. The costs and benefits of third party 

enforcement efforts fall on different parties. A wronged party benefits from 

third party enforcement efforts and the third party pays the costs. The wronged 

party has a natural incentive to have the third party do as much as possible in 

the way of enforcement—even past the point where there is a corresponding 

reduction in damages—because the wronged party appropriates the damage 

reduction but pays no costs. From an economic efficiency point of view, 

enforcement efforts that do not yield a commensurate reduction in damages are 

wasted. Private benefits may not be worth it from a social point of view when 

balanced against the costs to other parties. 

Equity Analysis 

The cost benefit framework just described lacks a normative dimension. It does 

not take into account questions of fairness, rights, and justice. And it does not 

consider who deserves the benefit of protection from harm or who is at fault or 

blameworthy for failing to take preventive measures. 

The view that an economic efficiency standard, by itself, is sufficient to create 

indirect liability is too strong. The focus on parties who had no part in creating 

the problem and who are not responsible for the illegal activity puts a burden 

on people who are innocent of any wrong-doing. Burdening innocent people 

seems unfair, and arguments that justify this approach on grounds that it is 

good for society as a whole violate widely accepted moral principles and are 

unlikely to withstand public scrutiny. 22 

We should require a person to right the wrongs committed by others only if we 

think that person is somehow responsible for those wrongs. Determining who 

is responsible for righting wrongs committed by others is controversial in both 

moral and political philosophy. 23 Libertarians generally maintain that people 

need to fix only the problems that they themselves directly created. 24 Without 

most cost-effective manner”). But a full cost-benefit analysis gives up the assumption of a 

fixed benefit goal and takes the value of benefits into account as well. 

22 See, e.g., JONATHAN WOLFF, AN INTRODUCTION TO POLITICAL PHILOSOPHY 57 (1996) 

(stating that “utilitarianism will permit enormous injustice in the pursuit of the general 

happiness”). A more sophisticated indirect or rule utilitarian approach can attempt to meet 

this difficulty, but that approach is subject to difficulties of its own. See generally JOHN RAWLS, 

A THEORY OF JUSTICE (1971) (critiquing utilitarianism). The underlying intuition behind this 

alternative account of social justice is that “[e]ach person possesses an inviolability founded 

on justice that even the welfare of society as a whole cannot override.” Id. at 3 

23 See infra notes 24-27 and accompanying text. 

24 See Jim Harper, Against ISP Liability, 28 REG. 30, 30-31 (2005) (arguing that ISPs should be 

liable for harms to third parties only if they have a duty to these parties and that “efficiency” 

considerations do not override the lack of such a duty founded on justice). Libertarians


this limitation, it is hard not to slide into a doctrine that requires all actors to 

stop misconduct whenever they can. 25 Others think that one has a duty to 

correct injustices to the extent that one participates in an institutional 

framework which produces injustice. 26 Still others believe in general positive 

duties to eliminate harms even when one has no direct role in causing them. 27 

Ultimately, the analysis of indirect liability cannot avoid considerations of 

fairness, rights, and justice. The key factors in this assessment will be those that 

have been used traditionally: directness of the involvement by third parties in 

activities that lead to harm to another person, an assessment of the degree of 

harm involved, the knowledge that third parties have or should have about the 

specific harm involved, what their intentions are, whether they are consciously 

acting in furtherance of a crime or other illegal act, and other similar 

considerations. 28 These complicated normative and empirical questions cannot 

be avoided by a single principle that purports to look at costs and benefits 

alone. 29 

generally reject the idea that we have positive duties to ameliorate harms we did not cause. 

E.g., id. 

25 Mann & Belzley, supra note 14, at 272 (noting that the principle that liability should be 

assigned regardless of blameworthiness “easily could shade into judicial doctrines that would 

obligate all actors to stop all misconduct whenever possible” and thinking that this 

“unbounded principle” is “unduly disruptive”). But it is hard to see how their proposal to 

implement indirect liability through regulation whenever it would be less expensive than 

leaving liability with the wronged party would be less disruptive. 

26 See, e.g., THOMAS W. POGGE, WORLD POVERTY AND HUMAN RIGHTS 172 (2002) (arguing 

that those involved in an institutional order that authorizes and upholds slavery have a duty 

to protect slaves or to promote institutional reform, even if they do not own slaves 

themselves). 

27 See, e.g., David Luban, Just War and Human Rights, in INTERNATIONAL ETHICS 195, 209 

(Charles R. Beitz et al. eds., 1985) (stating that “all humans in a position to effect” a human 

right have an obligation to do so). 

28 Mann and Belzley criticize the “myopic focus on the idea that the inherent passivity of 

Internet intermediaries makes it normatively inappropriate to impose responsibility on them 

for the conduct of primary malfeasors.” Mann & Belzley, supra note 14, at 261-62. But 

passivity is relevant to the knowledge and control factors needed to assess liability from an 

equity point of view. Lichtman and Landes seem to criticize the focus of current law on 

“knowledge, control, the extent of any non-infringing uses, and other factors” because they 

are not “particularly clear as to why those issues are central.” Lichtman & Landes, supra note 

16, at 405. But these factors are crucial because they relate to the way in which the equity 

issues can be resolved. 

29 These equity considerations can interact with the cost analysis. Consider the following: 

suppose transaction costs make it impossible for the wronged parties to negotiate 

enforcement deals with a third party–they are too numerous or lack the resources to 

compensate the third party. Suppose further it is possible that the cost savings involved in 

assigning liability to a third party are substantial. And finally stipulate that the third party’s 

involvement in the harm is so remote that assigning blame is a mistake. We might in that


An economic framework, broadly construed and supplemented with suitable 

considerations of equity, can be a useful way to assess the need for indirect 

liability for intermediaries in specific cases. The elements of the framework are 

as follows: 

� Market Failure Analysis: Are there substantial transaction costs? Can 

enforcement be achieved without an indirect liability rule? Can private 

parties work out enforcement arrangements among themselves? Can 

third parties effectively work with law enforcement without an indirect 

liability mandate? 

� Cost-Benefit Analysis: Does the burden on the wronged party or on 

law enforcement to take enforcement steps exceed the burden on the 

third parties? Are the costs of enforcement efforts reasonable in light 

of the reduction in harm? Are there longer-term or dynamic 

considerations to take into account? 

� Equity Analysis: Do third parties exercise such close control over the 

harm that they should be held responsible for its mitigation or 

elimination? Are they blameworthy for not taking steps against it? Is 

the harm particularly egregious? 

Applying the Framework 

to Payment Intermediaries 

Payment intermediaries have developed and refined policies and practices to 

deal with illegal Internet transactions in their payment networks. Two general 

conclusions can be drawn from an analysis of these policies and practices. 

The first is that payment intermediary action has been effective. As the 

following discussions demonstrate, Internet gambling websites have been 

denied access to the U.S. market, and their current and projected revenues are 

in decline. As a result of the payment system action in the Allofmp3.com 

copyright infringement case, Allofmp3.com was confined to a domestic market 

and experienced a dramatic reduction in the volume of activity at its website. 

The second conclusion is that the widespread assumption that payment system 

action in this area is simple and almost cost-free deserves more careful 

consideration. 30 The discussion of payment intermediaries’ activities to control 

circumstance nevertheless assign liability to the third party. The gains to the rest of us are 

just too great. However, should we not compensate the third party for taking the 

enforcement steps he is required to take? Assigning indirect liability when there is not this 

level of control or fault to justify blameworthiness might be so efficient under a cost analysis 

that it is worth considering, but in that case the use of compensation mechanisms should 

also be considered. 

30 See, e.g., Perfect 10, 494 F.3d at 824 (Kozinski, J., dissenting).


illegal activity on their systems reveals substantial costs that should give policy 

makers pause before moving ahead with the imposition of an indirect liability 

scheme for payment providers. These include: 

� The cost to maintain and enforce an Internet gambling coding and 

blocking scheme that is entirely manual and cannot be automated; 

� The cost from over-blocking legal transactions; 

� The cost to screen and check the business activity of merchants 

participating in the payment systems; 

� The cost to monitor the use of payment systems for specific illegal 

activity, where the payment systems are in no better position than 

anyone else to conduct this monitoring activity; 

� The cost to assess complaints of illegality, where the intermediary has 

no special expertise and is often less familiar with the legal and factual 

issues than the wronged party and the allegedly bad actor; 

� The cost to defend against legal challenges to enforcement actions, 

where the challenge typically comes in an off-shore jurisdiction; and 

� Longer-term costs to the United States from taking unilateral action in 

this area, including the encouragement of copycat regimes in other 

areas of law and in other jurisdictions. 

The reasonableness of these costs in light of the benefits achieved has not yet 

been seriously studied. Instead, it seems to be assumed that small compliance 

costs are justified by large enforcement benefits. Although precision in the 

estimates of costs and benefits is unlikely in this area, a more disciplined 

qualitative analysis is required. 

Internet Gambling Legislation 

The development of the Internet as a commercial medium presented a 

challenge to local gambling laws. With access to the Internet, individuals could 

reach gambling services from their homes, without the need to travel to a 

gambling merchant’s physical operation. The Internet provided a way for 

gambling merchants who were legal in their own jurisdictions to provide service 

to customers in different jurisdictions where gambling was not allowed. 

The United States Congress began its consideration of how to react to illegal 

Internet gambling in the late 1990s. 31 One early proposal was to put an 

31 See General Accounting Office, Internet Gambling, An Overview of the Issues, Dec. 2002, 

http://www.gao.gov/new.items/d0389.pdf. Many state laws made Internet gambling 

illegal and Federal law also appeared to outlaw at least some forms of it in interstate 

commerce. But the legal situation was ambiguous with respect to some forms of Internet 

gambling. The Interstate Wire Act of 1961 applied to Internet gambling and appeared to 

prohibit the use of the Internet for the “placing of bets or wages on any sporting event or 

contest.” See The Interstate Wire Act (18 U.S.C. § 1084) at


enforcement burden on ISPs. It would have required ISPs to terminate 

domestic Internet gambling merchants and to block foreign Internet gambling 

merchants upon request of law enforcement. 32 This initial effort failed to pass, 

in part because of concerns about the effectiveness and appropriateness of 

putting an enforcement burden on ISPs. 33 

In 2006, Congress passed the Unlawful Internet Gambling Enforcement Act 

(UIGEA), which imposed a system of indirect liability on financial institutions 

for the purpose of preventing illegal Internet gambling transactions. 34 Prior to 

the passage of UIGEA, payment card networks devised a coding and blocking 

system in order to manage the risks of Internet gambling. 35 Each merchant in 

the payment system is normally required to identify its major line of business 

and to include a four digit “merchant category code” in each authorization 

message. 36 For gambling, this merchant category code was 7995. 37 In addition, 

merchants were required to use an electronic commerce indicator when an 

Internet transaction was involved. 38 Together, these two pieces of information 

http://www.law.cornell.edu/uscode/18/usc_sec_18_00001084----000-.html. The U.S. 

Fifth Circuit Court of Appeals ruled in 2002 that the Wire Act applied only to sports betting 

and not to other types of online gambling. See In re MasterCard, 313 F.3d 257 (5th Cir. 

2002). The status of horseracing was similarly unclear. The Interstate Horse Racing Act 

appeared to allow the electronic transmission of interstate bets. It was amended in 

December 2000 to explicitly include wagers through the telephone or other electronic media. 

See the Interstate Horse Racing Act (15 U.S.C. §§ 3001-3007) at 

http://www.law.cornell.edu/uscode/15/usc_sup_01_15_10_57.html. These statutes 

appeared to allow the Internet to be used for both non-sports gambling and for gambling on 

horse races. The U.S. Department of Justice, however, thought, and still thinks, that existing 

statutes bar all forms of Internet gambling. See Letter from William E. Moschella, Assistant 

Attorney General to Rep. John Conyers Jr., July 14, 2003 at 

http://www.igamingnews.com/articles/files/DOJ_letter-031714.pdf (“The 

Department of Justice believes that current federal law, including 18 U.S.C. §§ 1084, 1952, 

and 1955, prohibits all types of gambling over the Internet.”). 

32 H.R. 3125 at http://thomas.loc.gov/cgi-bin/query/D?c106:2:./temp/~c106mktqmw 

33 See the floor debate on H.R. 3125, CR H6057-6068, July 17, 2000, 

http://thomas.loc.gov/cgi-bin/query/R?r106:FLD001:H56058. 

34 Unlawful Internet Gambling Enforcement Act of 2006, Pub. L. No. 109–347, 120 Stat. 1884 

(codified at 31 U.S.C. §§ 5361–5367 (2006)). 

35 Financial Aspects of Internet Gaming: Good Gamble or Bad Bet?: Hearing Before the 

Subcomm. on Oversight and Investigations of the H. Comm. on Financial Servs., 107th 

Cong. 25-27, 34-35 (2001) [hereinafter Financial Aspects of Internet Gaming Hearing] 

(statement and testimony of Mark MacCarthy, Senior Vice President, Public Policy, Visa, 

U.S.A., Inc.) (describing this system of coding and blocking Internet gambling transactions); 

U.S. GEN. ACCOUNTING OFFICE, supra note 31, at 20-25. 

36 U.S. GEN. ACCOUNTING OFFICE, supra note 31, at 22. 

37 VISA MERCHANT CATEGORY CLASSIFICATION (MCC) CODES DIRECTORY, available at 

http://www.da.usda.gov/procurement/card/card_x/mcc.pdf. 

38 U.S. GEN. ACCOUNTING OFFICE, supra note 31, at 22.


in the authorization message allowed payment networks or issuing banks to 

identify transactions involving Internet gambling merchants. 39 

Given this system, it was entirely feasible for the issuing bank or the payment 

network to block Internet gambling transactions. The system could 

accommodate conflicting laws in different jurisdictions in the following way: If 

it was illegal in one country, such as the United States, for cardholders to engage 

in Internet gambling, then the issuing banks based in that country could decline 

authorization requests for all properly coded Internet gambling transactions. 

This would effectively block these transactions. However, the banks in other 

countries who permit Internet gambling, such as the United Kingdom, could 

allow the use of their cards for Internet gambling by not declining properly 

coded Internet gambling transactions. 

The system was limited in detecting nuances in illegal versus legal Internet 

gambling. If a jurisdiction recognized some Internet gambling transactions as 

legal and others as illegal, the system would not detect it. 40 The merchant 

category code described a type of business, not the legal status of the 

transaction involved. 41 If a particular jurisdiction allowed casino gambling, but 

not sports betting, both transactions would nevertheless be labeled 7995. And if 

the system was set up to block these coded transactions, then both transactions, 

legal and illegal, would be blocked. 42 

Another weakness in the system was enforcement. If an Internet gambling 

merchant realized that his transactions would be blocked in a large jurisdiction 

such as the United States, then he would have every incentive to hide. 43 Instead 

of describing itself as a gambling merchant, it would just code itself as a T-shirt 

sales site or some other legal merchant. Without the proper merchant category 

code, the system was blind and could not effectively block the merchant’s 

transactions. 44 

The payment networks addressed this enforcement issue with a special program 

to verify that Internet gambling merchants coded their transactions correctly. 45 

39 Id. 


41 See VISA MERCHANT CATEGORY CLASSIFICATION (MCC) CODES DIRECTORY, supra note 37 

(listing all the MCC codes by “merchant type”). 


43 Id. at 26. 

44 Id. 

45 Id. at 31-32. The fines for incorrectly identifying authorization requests for online gambling 

transactions are set out at page 557 of the Visa International Operating Regulations. VISA, 

VISA INTERNATIONAL OPERATING REGULATIONS (April 2010),


Payment network personnel would test transactions at popular Internet 

gambling sites. They would enter a transaction at the web site and track the 

transaction through the payment system. They would be able to tell whether the 

transaction was coded properly or not after they identified the transaction in the 

system. If the transaction was not properly coded, the network would contact 

the bank that worked with the merchant and tell the bank that its merchant was 

out of compliance with the coding rule. The payment network would ask the 

bank to take steps to bring the merchant into compliance. Finally, the network 

would retest the site for proper coding. 46 

The UIGEA required payment systems to have policies and procedures 

reasonably designed to stop illegal Internet gambling transactions. 47 The statute 

creates a safe harbor for payment systems that adopt a coding and blocking 

scheme. 48 The Federal Reserve Board and the Department of the Treasury 

implemented this safe harbor with a non-exclusive description of one way in 

which a payment system can demonstrate that its policies and practices are 

reasonably designed to stop illegal Internet gambling transactions. 49 This nonexclusive 

description tracked the existing industry practices. 

http://usa.visa.com/download/merchants/visa-international-operating-regulationsmain.pdf#557. 

In addition, Visa requires online gambling merchants to post certain notices: 

“a Website for an Online Gambling Merchant must contain … [t]he statement ʻ Internet 

Gambling may be illegal in the jurisdiction in which you are located; if so, you are not 

authorized to use your payment card to complete this transaction.’” Id. at 594. 

46 U.S GEN. ACCOUNTING OFFICE, supra note 31, at 32. 

47 Unlawful Internet Gambling Enforcement Act of 2006, Pub. L. No. 109–347, 120 Stat. 1884 

(codified at 31 U.S.C. §§ 5361–5367 (2006)). 

48 12 C.F.R. § 233.6(d)(1)(ii) (2009). 

49 The code’s relevant section reads: 

(ii) Implementation of a code system, such as transaction codes and 

merchant/business category codes, that are required to accompany the 

authorization request for a transaction, including— 

(A) The operational functionality to enable the card system operator or the 

card issuer to reasonably identify and deny authorization for a transaction 

that the coding procedure indicates may be a restricted transaction; and 

(B) Procedures for ongoing monitoring or testing by the card system 

operator to detect potential restricted transactions, including— 

(1) Conducting testing to ascertain whether transaction authorization 

requests are coded correctly; and 

(2) Monitoring and analyzing payment patterns to detect suspicious payment 

volumes from a merchant customer … . 

Id.


Implementation Challenges with 

the Internet Gambling Act 

UIGEA defines illegal Internet gambling as whatever is illegal under current 

U.S. state and Federal law. It therefore continues the uncertainty regarding the 

illegality of some Internet gambling activities. 50 Financial intermediaries have 

the discretion to block or not block these transactions based upon their own 

judgment and the strength of the legal arguments presented to them. UIGEA 

also provides them with protection from liability if they over-block Internet 

gambling sites that turn out to be legal. The current law thereby allows 

substantial over-blocking and puts substantial discretion in the hands of the 

payment companies. 

Impact of UIGEA 

A large percentage of non-U.S. companies that derived extensive revenues from 

their operations in the United States left the market after the passage of 

UIGEA. All European companies that had been active in the U.S. market left it 

after the passage of UIGEA. 51 By December 2008, all the publicly-trade online 

gambling firms had left the U.S. market, even though most of the private firms 

remained. 52 

Three major European online gambling merchants lost $3 billion in 2006 from 

this withdrawal from the U.S market. 53 Measured traffic at particular sites 

declined as well. In September 2006, Party Poker, for example, which derived 

much of its traffic from the United States, had an average of about 12,000 active 

players. By November 2006, that number had dropped to about 4,000. 54 

50 These uncertainties affect several types of gambling, including horse racing, state lotteries, 

Indian gaming, and games of skill. 

51 European Commission Directorate-General for Trade, Examination Procedure Concerning an 

Obstacle to Trade, Within the Meaning of Council Regulation (EC) No 3286/94, Consisting of Measures 

Adopted by the United States of America Affecting Trade in Remote Gambling Services Complaint, Report 

to the Trade Barriers Regulation Committee (Commission Staff Working Paper) 59, June 10, 2009, 

available at http://trade.ec.europa.eu/doclib/docs/2009/june/tradoc_143405.pdf 

[hereinafter EC Gambling Report]. 

52 Casino City, Online Gambling in the United States Jurisdiction, 2009, 

http://online.casinocity.com/jurisdictions/united-states/. 

53 See EC Gambling Report, supra note 51, at 79 (“the direct losses in revenue due to the loss of 

the US market for just these three companies were above $3 billion in 2006.”). 

54 See WhichPoker.com, UIEGA Effects, 

http://www.whichpoker.com/stats/UIGEAEffects (last accessed Oct. 18, 2010). 

WhichPoker attributes the departure of the biggest publicly-traded online poker sites from 

the US market to stock market rules.


Shortly after UIGEA was signed into law in October 2006 analysts estimated 

that the value of British Internet gambling stocks declined by $7.6 billion. 55 In 

the 9 months between January 1, 2006 and November 1, 2006, just after the 

passage of UIGEA, three major European online gambling firms lost an 

estimated 75% of their value, totaling approximately 8.3 billion euros. 56 

An estimate by the European Commission of the likely evolution of the U.S. 

market in the absence of the specific restrictions imposed in 2006, based on an 

assumption of a 3% yearly growth, show U.S. Internet gambling accounting for 

about $5.8 billion per year in gross revenue in 2006, and reaching almost $14.5 

billion in 2012. Following the passage of UIGEA, the annual figure declined to 

about $4.0 billion in 2006, and by 2012 was estimated to be at only $4.6 

billion. 57 UIGEA reduced thus the size of the U.S. market well below what it 

would otherwise have been. 

Internet Gambling Assessment 

On equity grounds, it seems that the payment system connection to Internet 

gambling is too passive to justify imposing legal responsibility for blocking 

illegal Internet gambling. Payment intermediaries are not to blame when others 

use their system for Internet gambling because these intermediaries have no 

specific connection to the activity other than operating a general purpose 

payment system. They do not reap extra profits through special arrangements 

with the Internet gambling merchants. Internet gambling transactions are no 

different from any other payment card transaction. On pure equity grounds 

alone, then, there is no reason to single out these transactions and impose 

special legal responsibilities. 

A market analysis indicates that there are still some feasible enforcement 

arrangements that were not established prior to the passage of the UIGEA. 

Although intermediaries may not be responsible for their customers’ gambling, 

many of them are concerned about the social ills connected with the activity and 

want to reduce its prevalence. 58 U.S. financial intermediaries had already refused 

to sign up domestic Internet merchants because these merchants were not 

55 Eric Pfanner and Heather Timmons, U.K. Seeks Global Rules for Online Gambling, 

INTERNATIONAL HERALD TRIBUNE, Nov. 2, 2006, at 14, available at 

http://www.nytimes.com/iht/2006/11/02/technology/IHT-02gamble.html. The 

basis for this decline in share value was the withdrawal of these firms from the lucrative US 

market and the perception that they would not be able to recover the revenue lost from non- 

U.S. customers. 

56 EC Gambling Report, supra note 51, at 83. 

57 Id. at 19. 

58 See Financial Aspects of Internet Gaming Hearing, supra note 39, at 25-26 (statement of Mark 

MacCarthy, Senior Vice President, Public Policy, Visa U.S.A., Inc.).


authorized to act legally in the United States. 59 Some state attorneys general 

requested the intermediaries to block offshore gambling activities, and many 

cooperated. 60 These agreements did not extend to all financial institutions and 

did not cover all states, but they could have been extended without imposing a 

legislative requirement. 

A cost-benefit analysis of the UIGEA starts with an estimate of its effect on the 

amount of illegal Internet gambling activity. As we have seen, the legislation did 

not eliminate Internet gambling in the United States, but it did reduce it 

substantially below what it would otherwise have been. 

The costs associated with the payment systems’ compliance with the legislation 

include the costs of maintaining and enforcing an Internet gambling coding and 

blocking scheme, which is entirely manual and cannot be automated, as noted 

above. 

Another cost is the over-blocking problem created by the way in which 

payment intermediaries comply with UIGEA. Perfectly legal transactions will 

likely be blocked because payment intermediaries cannot distinguish them from 

illegal transactions. This example illustrates that intermediaries are usually better 

than others at monitoring their own systems for business activity of a certain 

type, but not at detecting the illegality of activity on their systems. 61 The point 

arises in Internet gambling because the codes used by financial institutions 

reflect the business activity of gambling, not its status as legal or illegal. As a 

result, the payment systems’ policies and procedures, which were adopted to 

comply with the Act and which have been accepted by the implementing 

regulations, over-block and prevent perfectly legal activity from taking place. 62 

59 Id. at 26; U.S. GEN. ACCOUNTING OFFICE, supra note 31, at 20. 

60 See, e.g., JACK GOLDSMITH & TIM WU, WHO CONTROLS THE INTERNET?: 

ILLUSIONS OF A BORDERLESS WORLD 82 (2006) (discussing Spitzer’s efforts “to 

convince every major American credit card provider and online payment system to stop 

honoring web gambling transactions.”). 

61 See Mann & Belzley, supra note 14, at 278 (“Surely eBay is more adept at searching and 

monitoring its marketplace than Tiffany & Co., while eBay probably is not as effective as 

Tiffany & Co. in distinguishing bona fide Tiffany products from counterfeits.”); see also 

Schruers, supra note 20, at 252 (“[T]he ISP is not the least-cost avoider when it comes to 

discovering [illegal] content; it is only well suited for cost avoidance after it is apprized of the 

problem.”). Schruers adds that in this case, the wronged party may be better suited to the 

task of locating the offending content. Id. at 252. 

62 Mann & Belzley, supra note 14, at 294. Mann and Belzley have a useful discussion of this 

over-blocking issue: 

[A] risk always exists that imposing additional burdens on intermediaries will 

chill the provision of valuable goods and services. That will be especially 

problematic in cases where considerable risk of chilling legal conduct that is


Alternatives to UIGEA 

In light of this difficulty, there might be more effective ways of assigning 

liability. The new law creates unnecessary confusion by failing to define the 

term “unlawful Internet gambling.” Congressman Barney Frank has introduced 

legislation to license and regulate Internet gambling merchants. 63 The lack of 

clarity about which merchants are legal would be resolved through a licensing 

process. At best, the system would rely on a list of approved gambling entities 

that the payment networks could check before approving gambling transactions 

from particular Internet merchants. 64 

The new licensing regime proposed in Congressman Frank’s legislation would 

be an improvement over the existing system in the short term. But over time, 

the only way payment systems can operate is through a reduction in the 

diversity of the laws they must accommodate. The U.S. government must either 

find other ways to enforce its laws abroad or begin harmonizing its laws with 

those of other countries. One solution is an international agreement that would 

recognize licensing arrangements in different countries as long as they satisfied 

certain agreed-upon minimum standards. 

adjacent to the targeted conduct exists. As discussed below, that might tend 

to make the use of intermediaries less plausible in file-sharing contexts where 

determining whether any particular act of file-sharing is illegal is difficult, and 

much more plausible in the gambling context where in many cases 

substantially all traffic to a particular site likely involves illegal conduct. 

Requiring intermediaries to make those kind [sic] of subjective decisions 

imposes costs not only on the intermediaries (that must make those 

decisions), but also on the underlying actors whose conduct might be filtered 

incorrectly. 

Id. at 274. The Internet gambling case illustrates that determining when a website is engaged 

in illegal gambling is not a simple task. It is fraught with the kind of “subjective decisions” 

that Mann and Belzley are properly concerned about. Payment systems faced with this 

difficulty do not to make these subjective decisions, instead blocking all gambling activity, 

including legal gambling transactions. 

63 Internet Gambling Regulation, Consumer Protection, and Enforcement Act, H.R. 2267, 

111th Cong. (2009). 

64 See text of H.R. 2267 and discussion at 

http://financialservices.house.gov/press/PRArticle.aspx?NewsID=495. The House 

Financial Services Committee approved the measure on July 29, 2010. See Sewell Chan, 

Congress Rethinks its Ban on Internet Gambling, NEW YORK TIMES, July 29, 2010 available at 

http://www.nytimes.com/2010/07/29/us/politics/29gamble.html. The revised 

legislation contains a ban on the use of credit cards for any Internet gambling, even the 

newly-legalized merchants, but debit cards can be used at the licensed sites. The text of the 

revised legislation is available at 

http://financialservices.house.gov/Media/file/markups/7_28_2010/Amendments-- 

HR%202267/Frank12.pdf


Online Copyright Infringement 

The ideal copyright enforcement mechanism would be for content owners to 

sue direct infringers. But often, direct infringers are too ubiquitous, too small, 

and too difficult to find. The result is well-developed notions of secondary 

liability for copyright infringement that involve intermediaries—as Paul Szynol 

dicusses in another essay in this collection. These doctrines of secondary liability 

have evolved substantially over the past decades. 

Legal Context for Intermediary 

Liability in Copyright Infringement 

Court cases and federal statute define some indirect responsibilities of 

intermediaries regarding copyright. The 1984 Supreme Court decision in Sony 

Corp. of America v. Universal City Studios, Inc. 65 established a standard for assessing 

third party liability. Providers of a technology that can be used for infringing 

activities are not liable when there are “substantial non-infringing uses” of the 

technology. 66 The Digital Millennium Copyright Act of 1998 enabled copyright 

owners to enforce their existing rights in the Internet context by enlisting the 

help of Internet intermediaries. 67 The key mechanism for gaining the 

cooperation of intermediaries is a safe harbor from secondary liability. ISPs are 

given an exemption from secondary liability so long as they act as a pure 

conduit, providing only transitory communications and system caching. 68 Web 

hosts and search engines also receive a safe harbor, provided they comply with a 

specific notice-and-takedown procedure. 69 Upon receiving notification of 

claimed infringement, the provider must expeditiously take down or block 

access to the material. 70 

Successful litigation against peer-to-peer networks in the digital music area also 

increased the ability of copyright owners to use third parties to combat 

copyright infringement where the third party is affirmatively involved in 

fostering the infringement. In an early file-sharing case, the Ninth Circuit found 

that the peer-to-peer service Napster was liable for secondary infringement 

based on its control and facilitation of its users’ infringement of music 

copyrights; 71 The company subsequently went out of business in its original 

65 464 U.S. 417 (1984). 

66 Id. at 442. 

67 17 U.S.C. § 512 (2006). 

68 17 U.S.C. § 512(a). 

69 17 U.S.C. § 512(b). 

70 Id. 

71 A&M Records, Inc. v. Napster, Inc., 239 F.3d 1004 (9th Cir. 2001).


form. 72 More recently, the Supreme Court found that another peer-to-peer 

service, Grokster, violated federal copyright law when it took “affirmative steps 

taken to foster infringement … by third parties,” such as advertising an 

infringing use or instructing how to engage in an infringing use. 73 

Against this background arose a question regarding payment systems: Are they 

liable for secondary infringement when they are used for direct infringement? In 

Perfect 10 v. Visa International Service Ass’n, 74 a subscription-based adult content 

website alleged that numerous websites based in several countries had stolen its 

proprietary images, altered them, and illegally offered them for sale online. 75 In 

response to complaints, Visa did not deny payment services to the allegedly 

infringing sites, and Perfect 10 brought a contributory and vicarious 

infringement action against Visa. The Ninth Circuit affirmed the district court’s 

rejection of liability for Visa. 76 

In Perfect 10, the Ninth Circuit dismissed the charge of contributory 

infringement by focusing on whether the credit card companies “materially 

contributed” to the infringement. 77 The court said the credit card companies did 

not materially contribute to the infringement because they had no “direct 

connection” to the infringement. 78 To have direct connection to the 

infringement they would have had to reproduce, display, or distribute the 

allegedly infringing works, which they did not do. 79 Payment services might 

make it more profitable to infringe, but they are too far removed in the causal 

72 Benny Evangelista, Napster Runs Out of Lives – Judge Rules Against Sale, S.F. CHRONICLE, Sept. 

4, 2002, at B1. 

73 MGM Studios, Inc. v. Grokster, Ltd., 545 U.S. 913, 919 (2005). 

74 494 F.3d 788 (9th Cir. 2007); see Jonathan Band, The Perfect 10 Trilogy, 5 COMPUTER L. REV. 

INT’L 142 (2007) (discussing Perfect 10 v. Visa International Service Ass’n and its relationship to 

similar secondary liability cases). Band summarizes the Visa case: 

Here the Ninth Circuit rejected what would have represented a significant 

expansion of secondary liability to actors far removed from the infringing 

activity. However, unlike the other cases, this case provoked a strong dissent 

by respected jurist Alex Kozinski. This dissent suggests that the outer edges 

of secondary liability remain to be defined. 

Id. at 14. Judge Kozinski’s dissent is indeed stinging, but it also underestimates the burden 

that secondary liability would place on intermediaries. Id. 

75 Perfect 10, 494 F.3d at 793. 

76 Id. 

77 Id. at 796. 

78 Id. 

79 Id.


chain that leads to the actual infringing acts for them to be described as making 

a material contribution. 80 

The court made a similar point about vicarious liability, finding that the card 

companies had no practical ability or right to prevent the infringing activity. 81 

While credit card services can exert financial pressure on the infringing 

websites, they cannot stop the actual reproduction or distribution of the 

infringing images. 82 

In his dissent, Judge Kozinski rejected both arguments. 83 According to Judge 

Kozinski, the card companies were directly connected to the infringement 

because they provided payment services. 84 Without these payment services there 

would be no infringement. 85 The card companies had the contractual right to 

terminate illegal activity on their systems, as well as the practical ability to exert 

financial pressure to stop or limit the infringing activity. 86 

This dissent apparently played a role in a more recent case in which a district 

court found payment processors liable for trademark infringement for failing to 

take down allegedly infringing content. A key element in this case was the 

knowledge imputed to the payment processor of infringing activity that should 

have been apparent from an analysis of chargeback claims. 87 

Payment System Complaint Program 

Even though payment intermediaries may not be required to take steps against 

online copyright infringement, they have chosen to do so. 88 Payment systems 

cannot monitor their networks for copyright law violations. They do not have 

the factual basis to conclude that a particular sale of a product is a violation of 

80 Id. at 797. 

81 Id. at 803. 

82 Id. at 804. 

83 Id. at 810-11 (Kozinski, J., dissenting). 

84 Id. at 811-12. 

85 Id. 

86 Id. at 816-17. 

87 Gucci v. Frontline No. 9 Civ.6925 (HB) U.S. District Court for the Southern District, June 

23, 2010 

88 See generally International Piracy: The Challenges of Protecting Intellectual Property in the 

21st Century: Hearing Before the Subcomm. on Courts, the Internet, and Intellectual 

Property of the H. Comm. on the Judiciary, 110th Cong. 73-82 (2007) [hereinafter 

International Piracy Hearing] (statement of Mark MacCarthy, Senior Vice President for 

Global Public Policy, Visa Inc.) (providing this account of payment intermediaries and 

intellectual property).


someone’s copyright. 89 Many music downloads are perfectly legal transactions, 

but some are not. Distinguishing the two is often a complex factual and legal 

question which payment intermediaries do not have the expertise or ability to 

resolve. 

The payment systems have no way of knowing whether a transaction involves 

copyright infringement without a complaint. The payment networks have thus 

developed policies and procedures to handle these complaints. 90 

The complaint process starts when a business entity approaches a payment 

system with clear, documented evidence of illegal activity and adequately 

identifies the infringing Internet merchant. 91 The business entity must provide 

substantiation that the activity is illegal and documentation that payment cards 

are actually being used for this illegal activity. 92 

The next step is to assess legality, which can be complex in cross-border 

situations. 93 After wrestling with these issues, the payment networks developed 

a policy for cross-border transactions: If a transaction would be illegal in either 

the jurisdiction of the merchant or the jurisdiction of the cardholder, the 

transactions should not be in the payment system. 94 In cases like copyright 

infringement, this means that merchants are responsible for making sure that 

the transactions they submit to the payment system are legal in both their 

operating jurisdiction and the jurisdiction in which their customer is located. 

This assessment of legality requires the payment network to determine whether 

the type of transaction would be illegal in either jurisdiction. 95 Since the facts 

and law involved are often complex, the payment networks are willing to take 

on only the clearest cases of copyright violation. Once they determine illegality, 

the payment providers do what they reasonably can to assist the complaining 

party. Since payment networks do not work directly with merchants, they 

typically try to locate the bank that has the merchant account and provide the 

complaint to the bank involved, which usually resolves the issue. 96 In most 

cases, either the bank does not want the business and terminates the merchant 

89 Id. at 76. 

90 Id. at 77. 

91 This Section describes the process at Visa, but other payment networks use a similar process. 

See id. at 85. 

92 Id. 

93 Id. at 77-78. 

94 Id. at 78. 

95 Id. 

96 Id.


or takes other action to bring the merchant into compliance. 97 If the bank does 

not take action, the payment networks can take further enforcement action 

against the bank. 98 

Allofmp3.com 

In some instances, the merchant resists the enforcement efforts of payment 

systems, insists on the legality of the underlying activity, and goes to a local 

court to vindicate its perceived rights under local law. This is what occurred in 

the Allofmp3.com case. 

In 2005, Visa received a documented complaint from International Federation 

of the Phonographic Industry (IFPI), which represents copyright owners based 

in more than seventy countries. 99 The complaint alleged that Allofmp3.com, a 

website located in Russia, was infringing on the copyrights of IFPI’s members 

by allowing unauthorized downloads of music. 100 Visa assessed the legal 

situation, in part by obtaining a review by outside counsel, and concluded that 

the transactions were illegal under local Russian law. 101 They were also illegal 

under the laws of the vast majority of the merchant’s customers who were 

located primarily in the United Kingdom and the United States. 102 In October 

2005, the Italian authorities shut down a localized version of Allofmp3.com, 

allofmp3.it, and began a criminal investigation of the Italian site. 103 In addition, 

97 Id. 

98 Id. Payment systems have a voluntary program such as this in place for counterfeiting 

complaints as well. This program includes include having a process in place to respond to 

complaints of the use of a payment brand for sales of counterfeit goods. Trademark owners 

would provide information such as a description of the allegedly counterfeit transaction and 

evidence that the payment system brand was involved, and the payment system would look 

into the allegation and take action in according with a publicly stated policy, which could 

include suspension of the merchant involved. Trademark owners would agree to indemnify 

payment systems for steps taken and for legal risk. This system is described by INTA in 

“Addressing the Sale of Counterfeits on the Internet,” September 2009 available as 

attachment 3 in the INTA Submission On The Request For Public Comment Regarding The 

Joint Strategic Plan For IP Enforcement, for the Office of the Intellectual Property 

Enforcement Coordinator (IPEC) through the Office of Management and Budget , March 

24, 2010 available at 

http://www.whitehouse.gov/omb/IPEC/frn_comments/InternationalTrademarkA 

ssociation.pdf 

99 Id. 

100 Id. (discussing IFPI’s role); Nate Anderson, Music Industry Encouraged Visa to Pull the Plug on 

AllofMP3.com, ARSTECHNICA, Oct. 19, 2006, 

http://arstechnica.com/business/news/2006/10/8029.ars. 

101 International Piracy Hearing, supra note 88, at 79 (statement of Mark MacCarthy, Senior Vice 

President for Global Public Policy, Visa Inc.). 

102 Id.


the United States Trade Representative intervened with the Russian government 

to urge them to shut down Allofmp3.com. 104 

At the beginning of September 2006, after appropriate notice, the Russian bank 

working with Allofmp3.com stopped processing Visa transactions for 

Allofmp3.com. 105 At the end of September 2006, the bank also stopped 

processing transactions from an affiliated site called allTunes. 106 After these Visa 

transactions ended, further confirmation of the site’s illegality was forthcoming; 

a Danish court ordered the Internet provider Tele2 to block its subscribers’ 

access to allofmp3.com, thereby making it harder for potential customers in 

Denmark to access the site. 107 MasterCard also cut off payment services to 

allofmp3.com. 108 By May of 2007, the site’s popularity had plummeted. 109 

The company was all but out of business, but the legal process was just starting. 

The owner of allTunes sued the bank that had stopped processing its Visa 

transactions in a Russian court. 110 Visa was a party to that litigation on the side 

of the bank. 111 In June 2007, the owner won a judgment that the bank had 

violated its contract with the merchant, and the judgment required the bank to 

continue to provide processing services. 112 In response to the bank’s claim that 

the merchant was acting illegally, the court determined that there were no 

rulings in Russia establishing that allTunes was making illegal use of exclusive 

rights belonging to rights holders. 113 

103 Press Release, IFPI, Allofmp3.com: Setting the Record Straight, June 2, 2006, 

http://www.ifpi.org/content/section_news/20060601.html. 

104 See International Piracy Hearing, supra note 88, at 26 (testimony of Victoria A. Espinel, Assistant 

U.S. Rep. for Intellectual Property and Innovation, Office of the U.S. Trade Rep.) (“We will 

continue to press Russia to shut down and prosecute the operators of illegal Web sites 

operating in Russia, including the successors to the infamous AllOfMP3.com.”). 

105 Id. at 79 (statement of Mark MacCarthy, Senior Vice President for Global Public Policy, Visa 

Inc.). 

106 Id. 

107 Press Release, IFPI, New Court Setback for Allofmp3.com, Oct. 26, 2006, 


108 BBC News, MP3 site’s voucher system closes, May 21, 2007, 

http://news.bbc.co.uk/2/hi/entertainment/6677265.stm. 

109 IFPI reported in May 2007 that Allofmp3 “rated outside the top 2000 websites.” Press 

Release, IFPI, Police Dawn Raid Stops Allofmp3.com Pirate Vouchers Scheme, May 21, 2007, 


110 Arbitration Court of Moscow 2007, A40-70411/06-67-500. 

111 Id. at 1. 

112 Id. at 5. 

113 Id. The court stated:


In August 2007, another Russian court issued a ruling in a different case, 

relating to criminal copyright infringement initiated by IFPI against the owner 

of Allofmp3.com. 114 This ruling stated that there had not been sufficient 

confirmation of any illegal activity by the site’s owner. 115 Even though the 

copyright owners had not given permission to distribute their recorded material, 

a Russian collective rights society (the Russian Multimedia and Internet Society, 

or ROMS by its initials in Russian) was deemed to be operating legitimately 

under Russian law. 116 The court implied that Allofmp3.com and similar sites 

would be in compliance with Russian law to the extent that they paid for rights 

from this Russian collective rights society. 117 

These court cases created a challenge for Visa because the payment system had 

responded to a documented complaint of copyright infringement. 118 Despite an 

outside review that seemed to establish illegality in the local jurisdiction, a local 

court ordered a local bank to continue to provide payment services. 119 Yet these 

transactions would still be illegal in virtually every other country in the world. 

To preserve its cross-border policy, Visa decided to allow the local bank to 

provide only domestic service to the site involved in the court case. 120 

Transactions from customers in other countries would not be allowed. 121 

According to Article 49 of the Russian Federation Law “On Copyright and 

Allied Rights,” it is only the Court that can execute actions in connection 

with illegal use of copyrights and allied rights, if there is a lawsuit filed by 

exclusive right holders, which the Defendants, VISA and IFPI are not, while 

in this case there are no court rulings with the force of res judicata establishing 

the Plaintiff’s illegal use of exclusive rights belonging to some right holders. 

Id. The Defendant was Rosbank, the Russian financial institution licensed by Visa to 

authorize merchants in Russia to accept Visa. Id. 

114 Cheremushkinsky [District Court of Moscow], 2007, No. 1-151-07. 

115 Id. at 4. 

116 Id. at 5. 

117 Id.; see also International Piracy Hearing, supra note 88, at 99 (testimony of Victoria A. Espinel, 

Assistant U.S. Rep. for Intellectual Property and Innovation, Office of the U.S. Trade Rep.) 

(“My understanding of the case is that Media Services, the company that operated allTunes, 

was able to successfully argue in Russian court that it was not acting illegally because it was 

paying royalties to collecting societies, collecting societies that were not authorized by the 

rights holders.”). 

118 Id. at 80 (statement of Mark MacCarthy, Senior Vice President for Global Public Policy, Visa 

Inc.). 

119 Id. at 80-81. 

120 Id. 

121 Id. at 81.


Assessment of Payment System Actions 

on Online Copyright Infringement 

Are payment systems doing enough on their own to respond to online 

copyright infringement? Does there need to be a system of legal liability for 

them to control online copyright infringement using their payment systems? 

First, Perfect 10 properly rejected indirect liability for payment intermediaries. 122 

The involvement of payment networks in copyright violations is attenuated and 

entirely passive. On control grounds, there is simply no way to draw a line 

between payment network involvement in allegedly infringing transactions and 

involvement in a wide range of other potentially illegal activities. If they are 

liable in this case, why wouldn’t they be liable for all cases of illegal activity on 

their payment systems? Unintentionally, Judge Kozinski’s dissent brought out 

this implication. 123 

But the actual experience of payment intermediaries reveals that things are 

never as simple as removing infringing material. At best, there is a welldocumented 

assertion of infringement under the laws of a particular 

jurisdiction. Judge Kozinski appears to favor a notice-and-takedown approach, 

so that payment intermediaries are not responsible for illegal conduct of which 

they are unaware. 124 But as Visa found in Allofmp3.com, payment card services 

and their associated financial service partners can be liable for wrongful 

termination of services in those jurisdictions if they react to an allegation of 

infringement by “kick[ing] the pirates off their payment networks.” 125 

122 Perfect 10, Inc. v. Visa Int’l Serv. Ass’n, 494 F.3d 788, 798 (9th Cir. 2007). For analysis, see 

Band, supra note 74. 

123 See id. at 824 (Kozinski, J., dissenting) (“Credit cards already have the tools to police the 

activities of their merchants, which is why we don’t see credit card sales of illegal drugs or 

child pornography.”). Of course, card companies use different tools in the case of illegal 

drugs and child pornography, namely, proactive monitoring, but it is hard to see on 

Kozinski’s analysis why card companies shouldn’t use whatever tools they can to stop illegal 

activity in all cases. See id. (“Plaintiff is not asking for a huge change in the way credit cards 

do business; they ask only that defendants abide by their own rules and stop doing business 

with crooks. Granting plaintiff the relief it seeks would not … be the end of Capitalism as 

we know it.”). But it might be the end of payment systems as we know them if indirect 

liability for them means an obligation to stop doing business with everyone who might be 

involved with illegality anywhere. Kozinski attempts to limit his analysis to those cases where 

there are special arrangements between bad actors and the payment system, id. at 819-20, but 

nothing in his analysis turns on these special arrangements. These special arrangements turn 

out to be risk-based pricing for adult content websites. Would he really have voted with the 

majority if the price that adult content merchants face for accepting cards was the same as 

the price set for less risky merchants? 

124 Perfect 10, 494 F.3d at 824 (Kozinski, J., dissenting). 

125 Id. at 817.


Second, there is no market failure in this situation that would justify imposing 

intermediary liability on payment systems. There are available arrangements 

between payment intermediaries and copyright owners that can reduce the 

amount of copyright infringement on the Internet. These arrangements are 

informal, but expanding. They rely on complaints by copyright owners, 

followed by investigation and action by intermediaries. They seem to strike a 

cost-based balance by putting the burden of discovering infringement on the 

copyright owner and triggering action by the third party only after notification. 

The arrangements may involve compensating payment intermediaries for 

performing enforcement services, but if this enables copyright owners to reduce 

the harm of copyright infringement, they might very well pay. If there are extra 

efforts, above and beyond standard practices, that a particular copyright owner 

would like payment intermediaries to make, those efforts should be open to 

negotiation. There do not seem to be any transaction costs that would prevent 

the parties from negotiating adjustments to these arrangements over time. And 

there appears to be no market failure that would justify not relying on private 

sector enforcement arrangements. 

Third, given the legal risks involved, copyright owners should be willing to 

indemnify payment intermediaries for damages resulting from enforcement 

actions against alleged infringers. Allofmp3.com indicates that these legal risks are 

not hypothetical. If the copyright owner believes in the legal soundness of his 

case, he should be prepared to assume the risk. It might be one way to assure 

that only strong complaints are brought to the attention of the payment 

intermediary. An additional mechanism might be to require the presence of a 

court or governmental agency that holds that the activity involved is infringing. 

A statute could potentially help provide legal immunity to payment 

intermediaries when they take good-faith action against alleged infringers. But 

U.S. law cannot provide immunity in other jurisdictions, which is where the aid 

of global payment intermediaries is needed. 

Fourth, this case illustrates the need for greater clarity in the legal environment 

in which intermediaries operate. Intermediaries cannot be in the position of 

creating new global law through their own interpretation of current statutes. 

Again Allofmp3.com suggests the need for even greater harmonization of local 

laws that intermediaries are expected to enforce. 

In sum, the experience of payment intermediaries indicates that some efforts on 

their part to respond to legitimate complaints would be justified. It is not 

appropriate to do nothing in response to allegations of copyright infringement. 

The current complaint procedure and case-by-case response is reasonable. It 

could be improved through further discussions among the parties, further 

recourse to court judgments of infringement, and harmonization of current 

international standards.


Conclusion 

The question remains: Should the government place an enforcement burden on 

payment intermediaries? The standard least cost analysis suggests that the 

advantages of government intervention sometimes appear to be substantial, but 

nothing in the analysis suggests that Internet intermediaries are always the best 

vehicle for government control. The costs, benefits and equities involved in 

specific cases have not been adequately assessed. Intermediaries are often in a 

position to voluntarily police their own communities and have taken steps to do 

this without explicit government requirements. The equities set out in current 

law establish a regime that works tolerably well. Even when government 

requirements are explicit, as in the case of Internet gambling, they are often 

crafted to fit the architecture and structure of the intermediaries themselves. 

While some adjustments would improve these legal regimes, nothing suggests 

that more liability imposed unilaterally by local governments would be an 

improvement. 

Greater government coordination on the rules that intermediaries must follow 

on the Internet would be an improvement. To avoid legal liability and to 

comply with local laws, payment intermediaries are moving toward accepting 

the laws of all jurisdictions. They also have wide discretion on what activities to 

allow on their systems. But this situation is problematic. Intermediaries are not 

the best- situated to decide which rules to follow. Also, no laws are selfinterpreting. 

They often apply to particular situations in obscure and heavily 

fact-dependent ways. Intermediaries’ flexibility in adjudication leaves room for 

private, strategic, and unaccountable decisions that affect the shape and 

direction of online activity. Coordinated government rules are best for an 

additional reason: The intermediary role does not scale well in a world of 

multiple, overlapping, and conflicting rules. If governments are going to use 

intermediaries to regulate the Internet, they need to coordinate their own laws 

to make that role possible.



Fuzzy Boundaries: 

The Potential Impact of Vague 

Secondary Liability Doctrines 

on Technology Innovation 

By Paul Szynol * 

Last year, Ninth Circuit Judge Alex Kozinski, with Josh Goldfoot from the 

Department of Justice’s Criminal Division, published an article in the Columbia 

Journal of Law and the Arts entitled “A Declaration of the Dependence of 

Cyberspace.” Its title is a play on the title of John Perry Barlow’s 1996 “A 

Declaration of the Independence of Cyberspace”; its content, as the title 

suggests, is something of an attack on Barlow’s philosophies, and, more 

generally, on the idea that the Internet is a unique entity that requires custom 

legal treatment. The authors make several key claims about the law’s 

relationship to the internet, but the central argument focuses on secondary 

liability—the copyright doctrine that makes makers of multi-use technologies 

legally liable for other people’s infringing uses of their technology. 

Broadly stated, the rationale at the heart of the secondary liability doctrine is 

this: An entity that knowingly helps to facilitate the commission of an illegal act 

(such as copyright infringement, for example) should be penalized for its 

contribution to the illegal activity. 1 If a technology company induces its 

customers to use its product for infringing purposes, for instance, both the 

users and the company should be liable for such infringement—the users for 

direct infringement and the company for contributory infringement, which is a 

species of secondary liability. 

The doctrine is appealing as a practical solution to widespread infringement 

because it targets the entities that enable illegal behavior—e.g., the Napsters and 

Groksters of the world—and thus eradicates the distribution mechanism that 

enables infringement in the first place. Judge Kozinski and Mr. Goldfoot (I’ll 

generally refer to them as “the authors” from here on), like the movie and 

music industries, certainly believe that the doctrine of secondary liability should 

be readily used as a handy and effective tool for weeding out copyright 

* Paul Szynol graduated from Columbia University, where he studied history and philosophy, 

and Yale University, where he studied intellectual property law. 

1 The specific theories of secondary liability have more nuanced elements, such as the 

requirements of materiality for contributory infringement and direct financial benefit for 

vicarious infringement. Since these elements are not critical to the essay’s main thesis, I’ve 

avoided spelling them out in detail.


infringement. According to the authors, people “who provide powerful tools 

that can be used for good or evil have some responsibility to make sure that 

those tools are used responsibly.” Put more bluntly, however, if you outlaw the 

tool, you needn’t chase after the users, so in practice it’s less a question of ethics 

and more a question of convenience and efficiency. 

One of the principal problems with this approach, however, is the fact that the 

boundaries of secondary liability are not precisely set, and, short of extreme 

cases, it is not at all clear under what circumstances a product manufacturer will 

be liable for secondary infringement. Such wholesale endorsement for 

secondary liability doctrines should therefore give us some pause. For example, 

at what point does a software company that develops a peer-to-peer application 

utilized by end users to exchange copyrighted materials begin to “contribute” to 

the infringement and become secondarily liable? Does the company contribute 

simply by writing software that is merely capable of infringing uses? 2 Or does 

the company contribute only if the software’s primary use is, by design, 

infringing? Or, further yet, does the company contribute only if a substantial 

portion of the end-users utilize the technology for infringing purposes? If so, 

how much of the user base must engage in infringing activity for it to be a 

substantial portion? 3 Or, as yet another option, does the company “contribute” 

only if it promotes infringing uses of its software? And, if that’s the case, how 

much promotion is too much promotion? For example, is the advertising 

slogan “Rip. Mix. Burn.” too much of an inducement to make infringing 

copies of music? 4 

These are fundamental, starting-point questions about the secondary liability 

doctrine, and one would expect that case law or legislation provides a clear 

answer to each. Yet the law is ambiguous (and the authors are altogether silent) 

on these points. Outside of extreme cases, no one knows with certainty— 

including lawyers, judges, company officers, engineers and academics—when 

secondary liability might attach to a product that facilitates the transmission of 

copyrighted materials. The legal system’s failure to provide clear guidelines is 

2 An argument that the Supreme Court famously rejected in its 1984 “Betamax” decision. 

Sony Corp. of America v. Universal City Studios, Inc., 464 U.S. 417 (1984). 

3 See, for example, the Napster litigation. According to the District Court’s opinion, 87% of 

the content on Napster was copyrighted, and “virtually all Napster users” transferred 

copyrighted content. A & M Records, Inc. v. Napster, Inc., 114 F. Supp. 2d 896 (N.D.Ca. 

2000). A decade later, a critical question remains essentially unanswered: How much lower 

would those percentages have to be for a manufacturer to be safe from secondary liability? 

4 The standard introduced in Grokster is “clear expression”, which is not much of a lodestar 

for someone seeking to gauge risk with any degree of precision. Metro-Goldwyn-Mayer Studios 

Inc. v. Grokster, Ltd., 545 U.S. 913, 914 (2005). One could persuasively argue that Apple’s 

very large, very prominent and very ubiquitous “Rip. Mix. Burn.” billboards amounted to 

“clear expression.”


the equivalent of posting a sign on a freeway that says “obey the speed limit” 

without giving an actual speed. 

The effect is potentially detrimental to the entire technology sector. A clear rule 

is a predictable rule, and a predictable rule is one on which innovators can rely 

when developing a product. Without clear guidance from the legal system, tech 

companies are forced to engage in a “fingers crossed” product design process, 

and, subsequently, face a market that can be an explosive landmine of 

infringement liability. The potential economic damage to a company found 

guilty of secondary liability can be substantial, to say the least. Since statutory 

damages for copyright infringement range from $750 to $150,000 per 

infringement, a maker of a multi-use technology may confront liabilities on a 

scale that can threaten the viability of even the wealthiest corporations. The risk 

is further exacerbated by the recent trend of unpredictable and often very 

bloated damage awards granted to copyright plaintiffs. Such risk can dissuade 

even the most resolute investors from marketing their invention—and it can 

literally bankrupt the braver among them. The loss of a robust distribution tool 

harms the content sector, too, since a powerful method for distributing content 

to end users will not be brought to market. 

Judge Kozinski and Mr. Goldfoot are not concerned with the chilling effect that 

the legal system’s ambiguity can have on technology innovation. In fact, they 

reject the proposition, and confidently point to the pharmaceutical and auto 

industries as counter-examples: Both industries have to comply with legal 

regulation yet manufacturers in both industries nevertheless innovate. 

It’s not a very persuasive comparison. First, the auto industry is hardly a hotbed 

of innovation. We might really like power windows and power steering, but, as 

advancements over prior art, these innovations are an order of magnitude 

smaller than the innovation we’ve seen on the internet. Second, the players in 

the auto and pharmaceutical industries are frequently different from the players 

in the technology sector. It is rare, after all, if not unheard of, that a single 

person invents valuable medicine—the medical R&D process takes place in the 

laboratories of some of the wealthiest companies under the sun. In addition, 

medical innovation is subject to review and approval by government regulatory 

agencies, so by the time a medicine reaches the market, it has already been 

approved by the government. Innovation in information technology, in 

contrast, is often the result of the proverbial garage inventor who releases the 

technology entirely on its own. Think of eBay, Napster, Apple, Google and 

Microsoft, each of which had a modest start in someone’s home or garage at the 

hands of one or two people (and many subsequently acquired similarly 

independent garage innovations). The distinction between a multinational 

company and a garage inventor is critical. First, there is no government 

imprimatur for multi-use technologies. Second, in contrast to wealthy 

companies that can afford sophisticated legal teams, garage inventors typically


lack the economic resources necessary to pay for a comprehensive legal review 

of product design prior to the product’s release. That inability increases the 

likelihood that the garage inventor will—unwittingly—design its product in a 

way that leads to legal liability, or the likelihood that, after releasing the product 

and receiving angry threats of litigation, the garage inventor will have to 

backtrack and redesign the product in order to avoid liability. These are very 

expensive measures. If the inventor can afford them, the inventor will have 

spent money that it would have saved had the law simply been clearer in the 

first place; if the inventor cannot afford them, the outcome is even worse: the 

start-up will simply fold, thus wasting its investment costs, while consumers will 

miss out on the product altogether. 

That outcome is bad enough, but it’s the third reason for the comparison’s 

inadequacy that should give all of us some pause: Because the legal landscape 

around copyright secondary liability is so unclear, even if the would-be inventor 

did have the resources to hire outside counsel, lack of clarity in the law means 

that, unless the product clearly crosses a line, lawyers—no matter how high 

their hourly rates—won’t be able to confidently provide the inventor with a 

legal imprimatur. In other words, no matter how much a company tries, lack of 

clear standards means that its lawyers might “get it wrong,” and the company 

may face infringement liability if it releases the product, or incur the costs of 

post-release redesign, or both. That is a very expensive proposition, and its 

corollary is clear: Faced with potential liability exposure and potential redesign 

costs, each of which could figure in the millions or even billions of dollars, 5 

some would-be inventors and investors will, as rational economic actors, forego 

the whole enterprise—not because they analyzed the risk and found it 

potentially too costly, but because the law’s ambiguity meant they simply couldn’t 

properly analyze the risk in the first place. Notably, the foregoing outcome will 

apply to garage inventors and big companies alike. The garage inventor whose 

coffers won’t be able to withstand the potential cost will retreat to the sound of 

a distant death knell; the big company will retreat because it knows that its deep 

pockets makes it an attractive target for a lawsuit and therefore may well decide 

that the potential litigation and licensing costs, even if not fatal, just aren’t worth 

it. Again, consumers will miss out on a new product. 

An ambiguous secondary liability doctrine also disadvantages American 

products in a global market: U.S. companies will have to worry about drowning 

in the unpredictable and poorly charted quicksand of secondary liability, while 

their international competitors will have clear legal rules to guide them. The 

5 It’s worth emphasizing that the billion dollar figure is not hyperbole—just ask SAP, which 

recently lost its legal dispute with Oracle and was ordered to pay $1.3 billion in damages. 

See Sam Diaz, Jury: SAP Owes Oracle $1.3 Billion for Copyright Infringement, ZDNET, Nov. 23, 

2010. The facts of that case are quite different from the examples given here, of course, but 

the award is a very conspicuous reminder that such astronomical damage awards are a 

startling reality of present day copyright litigation.


domestic market suffers as well: By creating barriers to entry (high and 

unreliable due diligence costs as well as post-release redesign costs), the 

ambiguity favors entrenched entities over newcomers. Advocating secondary 

liability without removing the ambiguity also contradicts the authors’ claim that 

the same set of laws should apply to offline and online worlds: The fuzzy 

secondary liability doctrine which they so strongly espouse in connection with 

technology wouldn’t fly in the physical world. For example, should a car 

company be held liable for drivers who speed? After all, it would be easy 

enough to add a “speed limit compliance chip.” Yet auto manufacturers are not 

forced to pay any portion of a speeding driver’s ticket. Offline, in other words, 

bad actors—the users of technology—are punished for their own transgressions. 

Online, however, the law chases the manufacturers—and applies ad-hoc, 

ambiguous standards to their products. It would seem that the authors want 

Internet-specific laws after all. 

None of this sounds like wise intellectual property policy. The legal system has 

a constitutional imperative to incentivize inventors, after all, and it achieves this 

objective in part by providing both content producers 6 and innovators with a 

stable and predictable legal climate, such as the “bright line” rule devised by the 

Supreme Court in its 1984 Sony ruling. 7 In its current state, the law threatens to 

punish rather than reward those who have the courage to release an innovative 

technology if that technology may be misused by its adopters and if that 

technology has yet to be contemplated and cleared by the judiciary or 

legislature. That is not an environment that encourages innovation. If the 

intent of the judiciary and the Department of Justice is indeed to mightily wield 

the secondary liability sword across the technology sector, the doctrine must be 

clearly defined, so that the rules of engagement are clearly stated and U.S. 

innovators can design their products with confidence—not in fear. 

6 In Community for Creative Non-Violence v. Reid, the Supreme Court acknowledged “Congress’ 

paramount goal in revising the 1976 Act of enhancing predictability and certainty of 

copyright ownership.” 490 U.S. 730, 749 (1989). 

7 Sony v. Universal City Studios, supra note 2.


CHAPTER 7 

IS SEARCH NOW AN 

“ESSENTIAL FACILITY?” 

399 


An Essential Cultural & Political Facility 401 


The Problem of Search Engines as Essential Facilities: 

An Economic & Legal Assessment 419 

Geoffrey A. Manne 

Some Skepticism About Search Neutrality 435 


Search Engine Bias & the Demise 

of Search Engine Utopianism 461 

Eric Goldman

400 CHAPTER 7: IS SEARCH NOW AN “ESSENTIAL FACILITY?”



An Essential Cultural 

& Political Facility 

By Frank Pasquale * 

Many worry about search engines’ growing power. How are worldviews being 

biased by them? Do search engines have an interest in getting certain 

information prioritized or occluded? 1 Dominant search engines (“DSEs”) 2 are a 

key hub of Internet traffic. They provide an ever-expanding array of services. 

Google, for instance, just announced its intention to go into travel shopping. As 

they amass information about their users, calls for regulation have focused on 

the threats to privacy they generate. Some of these efforts have been successful; 

others look more doubtful. One thing is certain: They are only the beginning of 

a struggle over the rights and responsibilities of key intermediaries. Some hope 

that competition law—and particularly the doctrine of “essential facilities”—will 

lead policymakers to scrutinize search engines actions. 

When American lawyers talk about “essential facilities,” they are referring to 

antitrust doctrine that has tried, at various points, to make certain “bottlenecks” 

in the economy provide access on fair and nondiscriminatory terms to all 

comers. As robust American competition law fades into a secluded corner of 

legal history, 3 “essential facilities” doctrine still remains, for some scholars, a ray 

of hope for intermediary responsibility. 4 Oren Bracha and I helped fuel this 

* Professor of Law, Seton Hall Law School; Visiting Fellow, Princeton Center for Information 

Technology Policy. 

1 ALEX HALAVAIS, SEARCH ENGINE SOCIETY 85 (Polity 2008) (“In the process of ranking 

results, search engines effectively create winners and losers on the web as a whole. Now that 

search engines are moving into other realms, this often opaque technology of ranking 

becomes kingmaker in new venues.”); Chi-Chu Tschang, The Squeeze at China’s Baidu, 

BUSINESSWEEK, Dec. 31, 2008, at 

www.businessweek.com/magazine/content/09_02/b4115021710265.htm (“Salespeople 

working for Baidu drop sites from results to bully companies into buying sponsored links [a 

form of paid advertising], say some who have been approached.”). 

2 We can provisionally define a dominant search engine (“DSE”) as one with more than 40 

percent market share. Google clearly satisfies this criterion in the United States and Europe. 

See David S. Evans, Antitrust Issues Raised by the Emerging Global Internet Economy, 102 NW. U. L. 

REV. COLLOQUY 285 (2008) (reporting market shares for leading internet intermediaries). 

3 BARRY LYNN, CORNERED: THE NEW MONOPOLY CAPITALISM AND THE ECONOMICS OF 

DESTRUCTION (John Wiley & Sons, Inc. 2010) (describing the declining impact of American 

antitrust law). 

4 Brett Frischmann & Spencer Weber Waller, Revitalizing Essential Facilities, 75 ANTITRUST L.J. 1, 

2 (2008) (“infrastructure subject to substantial access and nondiscrimination norms [has] … 

been heavily regulated.”).

402 CHAPTER 7: IS SEARCH NOW AN “ESSENTIAL FACILITY?” 

hope in our 2008 article Federal Search Commission, which compared dominant 

search engines to railroads and common carriers in the hope that they would be 

recognized as infrastructural foundations of the information economy. 5 But I 

now see that Federal Search Commission, like many other parts of the search engine 

accountability literature, tried too hard to shoehorn a wide variety of social 

concerns about search engines into the economic language of antitrust policy. 6 

It is now time for scholars and activists to move beyond the crabbed vocabulary 

of competition law to develop a richer normative critique of search engine 

dominance. 

This will not be an easy sell in cyberlaw, which tends to uncritically promote 

competition and innovation as the highest aims of Internet policy. If a 

dominant search engine is abusing its position, market-oriented scholars say, 

market forces will usually solve the problem, and antitrust law can step in when 

they fail to do so. Even those who favor net neutrality rules for carriers are 

wary of applying them to other intermediaries, like search engines. All tend to 

assume that the more “innovation” happens on the Internet, the more choices 

users will have and the more efficient the market will become. Yet these 

scholars have not paid enough attention to the kind of innovation that is best 

for society, and whether the uncoordinated preferences of millions of web users 

for low-cost convenience are likely to address the cultural and political concerns 

that dominant search engines raise. 

In this article, I hope to demonstrate two points. First, antitrust law terms (like 

“essential facility”) cannot hope to capture the complexity of concerns raised by 

an information landscape where one company serves as the predominant map 

of the web, and simultaneously attempts to exploit that dominance by endlessly 

expanding into adjoining fields. Second, I hope to point the way toward a new 

concept of “essential cultural and political facility,” which can help policymakers 

realize the situations where a bottleneck has become important enough that 

special scrutiny is warranted. This scrutiny may not always lead to regulation— 

which the First Amendment renders a dicey enterprise in any corner of the 

information economy. However, it could lead us to recognize the importance 

of publicly funded alternatives to the concentrated conduits and contentproviders 

colonizing the web. 

5 Oren Bracha & Frank Pasquale, Federal Search Commission: Fairness, Access, and Accountability in 

the Law of Search, 93 CORNELL L. REV. 1193 (2008). 

6 RICHARD POSNER, ANTITRUST LAW, at ix (2d ed. 2001) (“Almost everyone professionally 

involved in antitrust today—whether as litigator, prosecutor, judge, academic, or informed 

observer—not only agrees that the only goal of the antitrust laws should be to promote 

economic welfare, but also agrees on the essential tenets of economic theory that should be 

used to determine the consistency of specific business practices with that goal.”).


The Limits of Antitrust as Search Policy 

Antitrust cases tend to consume a great deal of time, in part because economic 

conduct is subject to many different interpretations. 7 One person’s 

anticompetitive conduct is another’s effective business strategy. The same 

unending (and indeterminate) arguments threaten to stall discourse on search 

policy. For example, the Federal Trade Commission’s (FTC) review of the 

Google–DoubleClick merger focused almost entirely on the economic effects 

of the proposed combination, rather than the threats to privacy it posed. 8 

Search engines are among the most innovative services in the global economy. 

They provide extraordinary efficiencies for advertisers and consumers by 

targeting messages to viewers who are most likely to want to receive them. In 

order to attract more users, search engines use revenues from advertising to 

organize and index a great deal of content on the Internet. Like the major 

broadcast networks, search engines are now beginning to displace. They provide 

opportunities to view content (organic search results) in order to sell advertising 

(paid search results). 9 Search engines have provoked antitrust scrutiny because 

proposed deals between major search engines (and between search engines and 

content providers) suggest undue coordination of competitors in an already 

concentrated industry. 10 

7 See Jonathan Zittrain, The Un-Microsoft Un-Remedy: Law Can Prevent the Problem that It Can’t 

Patch Later, 31 CONN. L. REV. 1361, 1361–62 (1999) (“The main concern in finding a remedy 

for [‘bad monopolist behaviors’] may be time: The technology environment moves at a 

lightning pace, and by the time a federal case has been made out of a problem, the problem 

is proven, a remedy fashioned, and appeals exhausted, the damage may already be 

irreversible.”). 

8 News Release, FTC, Federal Trade Commission Closes Google/DoubleClick Investigation 

(Dec. 20, 2007), available at www.ftc.gov/opa/2007/12/googledc.shtm (“The Commissioners 

... wrote that ‘as the sole purpose of federal antitrust review of mergers and acquisitions is 

to identify and remedy transactions that harm competition,’ the FTC lacks the legal authority 

to block the transaction on grounds, or require conditions to this transaction, that do not 

relate to antitrust. Adding, however, that it takes consumer privacy issues very seriously, the 

Commission cross-referenced its release of a set of proposed behavioral marketing 

principles that were also announced today.”). 

9 According to the Google corporate home page, “[W]e distinguish ads from search results or 

other content on a page by labeling them as ‘sponsored links’ or ‘Ads by Google.’ We don’t 

sell ad placement in our search results, nor do we allow people to pay for a higher ranking 

there.” Google, Inc., Corporate Information: Company Overview, 

www.google.com/corporate/ (last visited Mar. 12, 2010). 

10 For example, the deal reached between Microsoft and Yahoo! that would have Microsoft’s 

Bing search engine deliver results for searches on Yahoo! has provoked antitrust concerns 

both domestically and internationally. See Christopher S. Rugaber, Microsoft–Yahoo Deal to Face 

Tough Antitrust Probe, ABCNEWS, July 29, 2009, 

http://seattletimes.nwsource.com/html/localnews/2009563654_apusmicrosoftyaho 

oantitrust.html.


Those opposed to regulation often claim that antitrust law offers a more 

targeted and efficient response to abuses. As Justice Breyer explained in his 

classic work Regulation and Its Reform: 

[T]he antitrust laws differ from classical regulation both in their 

aims and in their methods … . [T]hey act negatively, through a 

few highly general provisions prohibiting certain forms of private 

conduct. They do not affirmatively order firms to behave in 

specified ways; for the most part, they tell private firms what 

not to do … . Only rarely do the antitrust enforcement 

agencies create the detailed web of affirmative legal obligations 

that characterizes classical regulation. 11 

Given the lack of search engine regulation in the U.S., actual and threatened 

antitrust investigations have been a primary government influence on Google’s 

business practices as its dominance in search grows. Many believe that the 

Department of Justice’s (DOJ) suspicion of the company’s proposed joint 

venture with Yahoo! in the search advertising field effectively scuttled the deal 

by late 2008. 12 However, antitrust enforcement appears less promising in other 

aspects of search. 13 This section discusses the limits of antitrust in addressing 

the cultural and political dilemmas raised by Google’s proposed Book Search 

deal with publishers, 14 and its dominance of online advertising. 

11 STEPHEN BREYER, REGULATION AND ITS REFORM 156–57 (Harvard Univ. Press 1982). But 

see A. Douglas Melamed, Antitrust: The New Regulation, 10 ANTITRUST 13, 13 (1995) 

(describing “two paradigms,” the law enforcement model and the regulatory model, and the 

shift of antitrust law from the former to the latter). 

12 Nicholas Thompson & Fred Vogelstein, The Plot to Kill Google, WIRED, Jan. 19, 2009, at 88, 

available at www.wired.com/techbiz/it/magazine/17-02/ff_killgoogle (noting that antitrust 

scrutiny culminated in a hearing in which the DOJ threatened to bring an antitrust case 

against Google and that one prominent DOJ attorney expressed the view that Google 

already is a monopoly). 

13 Daniel Rubinfeld, Foundations of Antitrust Law and Economics, in HOW THE CHICAGO SCHOOL 

OVERSHOT THE MARK: THE EFFECT OF CONSERVATIVE ECONOMIC ANALYSIS ON U.S. 

ANTITRUST 51, 57 (Robert Pitofsky ed., 2008) (describing how “conservative economics has 

fostered a tendency to downplay enforcement in dynamic technological industries in which 

innovation issues play a significant role”). 

14 Despite the DOJ’s intervention to affect the terms of the proposed settlement, many leading 

antitrust experts have argued that the settlement would not violate the antitrust laws. See, e.g., 

Einer Elhauge, Why the Google Books Settlement Is Pro-Competitive 58 (Harvard Law Sch., Law & 

Econ. Discussion Paper No. 646, Harvard Law Sch., Pub. Law & Theory Research Paper 

No. 09-45, 2009), available at 

http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1459028 (“The settlement does 

not raise rival barriers to offering [many] books, but to the contrary lowers them. The 

output expansion is particularly dramatic for out-of-print books, for which there is currently 

no new output at all.”).


Privacy concerns are nearly impossible to address within the economic models 

of contemporary competition law. Antitrust scrutiny did little to address the 

privacy concerns raised when Google proposed to merge with the web 

advertising firm DoubleClick. 15 The proposed deal provoked a complaint from 

the Electronic Privacy Information Center (EPIC). EPIC claimed that Google’s 

modus operandi amounts to a “deceptive trade practice”: 

Upon arriving at the Google homepage, a Google user is not 

informed of Google’s data collection practices until he or she 

clicks through four links. Most users will not reach this page 

… . Google collects user search terms in connection with his 

or her IP address without adequate notice to the user. 

Therefore, Google’s representations concerning its data 

retention practices were, and are, deceptive practices. 16 

One key question raised by the proposed merger was whether privacy and 

consumer protection concerns like these can be addressed by traditional 

antitrust analysis. 17 Privacy law expert Peter Swire argued that they can, because 

“privacy harms reduce consumer welfare … [and] lead to a reduction in the 

quality of a good or service.” 18 Swire believed that consumers would be worse 

off after the merger because of the unparalleled digital dossiers the combined 

entity could generate: 

Google often has “deep” information about an individual’s 

actions, such as detailed information about search terms. 

Currently, DoubleClick sets one or more cookies on an 

individual’s computers, and receives detailed information about 

which sites the person visits while surfing. DoubleClick has 

15 Dawn Kawamoto & Anne Broache, FTC Allows Google–DoubleClick Merger to Proceed, CNET 

NEWS, Dec. 20, 2007, http://news.cnet.com/FTC-allows-Google-DoubleClickmerger-to-proceed/2100-1024_3-6223631.html 

(describing U.S. authorities’ blessing of the 

proposed deal). 

16 See Complaint and Request for Injunction, Request for Investigation and for Other Relief, In 

re Google Inc. and DoubleClick, Inc., No. 071-0170 (FTC Apr. 20, 2007), available at 

http://epic.org/privacy/ftc/google/epic_complaint.pdf at 9 [hereinafter Google, Inc. 

and DoubleClick Complaint]. 

17 See Siva Vaidhyanathan, The Googlization of Everything, Google and DoubleClick: A 

Bigger Antitrust Problem than I Had Imagined, 

www.googlizationofeverything.com/2007/10/google_and_doubleclick_a_bigge.php (Oct. 

21, 2007, 16:05 EST). 

18 Peter Swire, Protecting Consumers: Privacy Matters in Antitrust Analysis, CTR. FOR AM. PROGRESS, 

Oct. 19, 2007, www.americanprogress.org/issues/2007/10/privacy.html (italics omitted).


“broad” information about an individual’s actions, with its 

leading ability to pinpoint where a person surfs. 19 

Initial points of contention include (a) the definition of the products at issue, 

and (b) how to weigh the costs and benefits of a merger. The combined 

company would have different segments of “customers” in a two-sided 

market: 20 (1) searchers trying to find sites, and (2) ad buyers trying to reach 

searchers. Swire contends that many people care about privacy, and “[i]t would 

be illogical to count the harms to consumers from higher prices while excluding 

the harms from privacy invasions—both sorts of harms reduce consumer 

surplus and consumer welfare in the relevant market.” 21 

However, the web searcher category not only consists of consumers who care 

about privacy, but also includes many people who do not highly value it or who 

actively seek to expose their information in order to receive more targeted 

solicitations. According to Eric Goldman’s work on personalized search, some 

may even consider the gathering of data about them to be a service. 22 The more 

information is gathered about them, the better intermediaries are able to serve 

them relevant ads. Many economic models of web publication assume that 

users “pay” for content by viewing ads; 23 they may effectively pay less if the 

advertisements they view bear some relation to things they want to buy. So 

while Swire models advertising and data collection as a cost to be endured, 

19 Id. According to Swire, “[i]f the merger is approved, then individuals using the market 

leader in search may face a search product that has both ‘deep’ and ‘broad’ collection of 

information. For the many millions of individuals with high privacy preferences, this may be 

a significant reduction in the quality of the search product—search previously was 

conducted without the combined deep and broad tracking, and now the combination will 

exist.” Id. 

20 For a definition of two-sided market, see Nicholas Economides & Joacim Tåg, Net Neutrality 

on the Internet: A Two-Sided Market Analysis 1 (NET Inst., Working Paper No. 07-45, N.Y. Univ. 

Law and Econ., Research Paper No. 07-40, 2007), 

http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1019121 (“[P]latforms sell 

Internet access services to consumers and may set fees to content and applications providers 

‘on the other side’ of the Internet.”). In the search engine context, consumers “pay” by 

attending to ads, and ad-purchasers pay Google for the chance to get ad viewers’ attention. 

21 Swire, supra note 18. 

22 Eric Goldman, A Coasean Analysis of Marketing, 2006 WIS. L. REV. 1151, 1162–64 (“Three 

components determine an individual consumer’s utility from a marketing exposure: (1) the 

consumer’s substantive interest in the marketing, (2) the consumer’s nonsubstantive reactions 

to the marketing exposure, and (3) the attention consumed by evaluating and sorting the 

marketing. … [A] consumer may derive utility from the rote act of being contacted by 

marketers or exposed to the marketing, regardless of the marketing content.”). 

23 David S. Evans, The Economics of the Online Advertising Industry, 7 REV. NETWORK ECON. 359, 

359 (2008), available at www.bepress.com/rne/vol7/iss3/2 (describing how many of the top 

websites have adopted the “free-tv” model where the publisher generates traffic by not 

charging for readers but then sell that traffic to advertisers).


Google and DoubleClick argue that the resulting personalized ads serve 

customers. Their arguments prevailed, and Google officially acquired 

DoubleClick in 2008. 24 

Antitrust law is ill prepared to handle a “market” where some percentage of 

consumers consider loss of privacy a gain and others consider it a loss. 

Economic reasoning in general falters in the face of externalities, but usually we 

can all agree that, say, pollution is a harm (or negative externality) and flowers 

are a boon (or positive externality). Privacy preferences are much more 

idiosyncratic. 

Critics of the merger do have a response to this problem of diverse 

preferences—they can shift from characterizing lost privacy as a cost of web 

searching to describing it as a reduction in the quality of the services offered by 

the merging entities. 25 Douglas Kysar’s work on the product–process 

distinction is encouraging here. Kysar has claimed that consumers should have 

a right to make choices of products based on how the products are made, not 

just how well they work. 26 Kysar argues “in favor of acknowledging and 

accommodating [consumer] process preferences within policy analysis, given 

the potential significance that such preferences may serve in the future as 

outlets for public-minded behavior.” 27 Nevertheless, the valuation problems 

here are daunting. How are we to determine how much consumers are willing 

to pay to avoid privacy-eroding companies? 28 

Perhaps, as Lisa Heinzerling and Frank Ackerman suggest in their book Priceless, 

we should stop even trying to pretend that these decisions can be made on 

24 See Press Release, Google Inc., Google Closes Acquisition of DoubleClick (Mar. 11, 2008), 

available at www.google.com/intl/en/press/pressrel/20080311_doubleclick.html. 

25 Both Supreme Court precedent and DOJ guidelines support this approach. See Nat’l Soc’y 

of Prof ’l Eng’rs v. United States, 435 U.S. 679, 695 (1978) (“The assumption that 

competition is the best method of allocating resources in a free market recognizes that all 

elements of a bargain—quality, service, safety, and durability—and not just the immediate 

cost, are favorably affected by the free opportunity to select among alternative offers.”); U.S. 

DEP’T OF JUSTICE, HORIZONTAL MERGER GUIDELINES § 4, at 30–32 (1997) (efficient market 

behavior is indicated by lower prices, new products, and “improved quality”). 

26 Douglas A. Kysar, Preferences for Processes: The Process/Product Distinction and the Regulation of 

Consumer Choice, 118 HARV. L. REV. 526, 529 (2004) (“[C]onsumer preferences may be heavily 

influenced by information regarding the manner in which goods are produced.”). 

27 Id. at 534. 

28 Christopher Yoo has demanded this kind of accounting in the context of net neutrality. See 

Christopher Yoo, Beyond Network Neutrality, 19 HARV. J.L. & TECH. 1, 54 (2005) (“There is 

nothing incoherent about imposing regulation to promote values other than economic 

welfare. … [but] such a theory must provide a basis for quantifying the noneconomic 

benefits and for determining when those benefits justify the economic costs.”).


anything approaching a purely economic basis. 29 Engaging in a cost–benefit 

analysis diminishes privacy’s status as a right. Though many scholars have 

compellingly argued for broader foundations for competition law, the 

mainstream of contemporary antitrust policy in the United States cannot 

accommodate such concerns. Antitrust’s summum bonum is the maximization of 

“consumer welfare,” and this measure of efficiency is notoriously narrow. 30 For 

example, the DOJ was hard pressed to adequately factor in a basic democratic 

commitment to diverse communicative channels during many media mergers. 31 

Given antitrust doctrine’s pronounced tendency to suppress or elide the cultural 

and political consequences of concentrated corporate power, the Bureau of 

Competition and the Bureau of Economics within the FTC are ill-equipped to 

respond to the most compelling issues raised by search engines. 32 The Google– 

Doubleclick merger proceedings ultimately ended with an overwhelming win 

for Google at the FTC. 33 This outcome was all but inevitable given the 

foundations of contemporary antitrust doctrine, 34 and is the logical outgrowth 

of overreliance on legal economic theory that uncritically privileges market 

29 Frank Ackerman & Lisa Heinzerling, Priceless: On Knowing the Price of Everything and 

the Value of Nothing 8–9 (The New Press 2004). 

30 See Maurice E. Stucke, Better Competition Advocacy, 82 ST. JOHN’S L. REV. 951, 1001 (2008) 

(observing the primacy of allocative efficiency in antitrust analysis). Stucke notes that 

“[b]ehind allocative efficiency’s façade of positivism lie [many] moral questions … .” Id. See 

also Julie E. Cohen, Network Stories, 70 LAW & CONTEMP. PROBS. 91, 92 (2007) (“What makes 

the network good can only be defined by generating richly detailed ethnographies of the 

experiences the network enables and the activities it supports, and articulating a normative 

theory to explain what is good, and worth preserving, about those experiences and 

activities.”). 

31 See C. Edwin Baker, Media Concentration: Giving Up on Democracy, 54 FLA. L. REV. 839, 857 

(2002) (“[T]he dominant antitrust focus on power over pricing can be distinguished from power 

over the content available for consumer choice. In the currently dominant paradigm, a merger that 

dramatically reduced the number of independent suppliers of a particular category of 

content—say, news or local news or Black activist news—creates no antitrust problem if, as 

likely, it does not lead to power to raise prices.”). 

32 See STATEMENT OF THE FEDERAL TRADE COMMISSION CONCERNING 

GOOGLE/DOUBLECLICK, FTC File No. 071-0170 (FTC Dec. 20, 2007), available at 

http://www.ftc.gov/os/caselist/0710170/071220statement.pdf [hereinafter STATEMENT 

OF FTC CONCERNING GOOGLE/DOUBLECLICK] (“Although [privacy concerns] may present 

important policy questions for the Nation, the sole purpose of federal antitrust review of 

mergers and acquisitions is to identify and remedy transactions that harm competition.”). 

33 Id. 

34 Maurice Stucke describes and critiques this bias in some detail. See Stucke, supra note 30, at 

1031 (describing a “mishmash of neoclassical economic theory, vignettes of zero-sum 

competition, and normative weighing of the anticompetitive ethereal—deadweight welfare 

loss—against the conjectures of procompetitive efficiencies” at the core of too much 

antitrust law and theory). Among his many important contributions to the literature, Stucke 

makes it clear that competition policy includes far more goals and tactics than antitrust 

enforcement alone. Id. at 987–1008.


outcomes. 35 As long as contemporary doctrine holds that antitrust is singularly 

focused on the “consumer welfare” a proposed transaction will generate, 36 

antitrust policymakers will be unable to address the cultural and political 

consequences of consolidation in the search industry. 

Antitrust challenges to the proposed settlement of a copyright lawsuit by 

authors and publishers against Google’s Book Search program are likely to be 

similarly constrained. 37 As in the Google-Doubleclick merger, the privacy 

implications of Google’s proposed deal with publishers are profound. 38 Anyone 

who cares about public input into the future of access to knowledge should 

approach the potential deal here warily, even if the prospect of constructing a 

digital Library of Alexandria tempts scholars. 39 As Harvard librarian Robert 

Darnton has argued, only a naive optimist could ignore the perils of having one 

profit-driven company effectively entrusted with a comprehensive collection of 

the world’s books. 40 

When publishers challenged Google’s book scanning in 2007, many hoped that 

public interest groups could leverage copyright challenges to Google’s book 

35 Reza Dibadj, Beyond Facile Assumptions and Radical Assertions: A Case for “Critical Legal 

Economics,” 2003 UTAH L. REV. 1155, 1161 (“[T]hree of the most basic assumptions to the 

popular [law & economics] enterprise—that people are rational, that ability to pay 

determines value, and that the common law is efficient—while couched in the metaphors of 

science, remain unsubstantiated.”). But see JAMES R. HACKNEY, JR., UNDER COVER OF 

SCIENCE: AMERICAN LEGAL–ECONOMIC THEORY AND THE QUEST FOR OBJECTIVITY 164–66 

(Duke Univ. Press 2007) (describing the “notable movement to broaden the scope of legal– 

economic theory under the rubric of socioeconomics”). 

36 See Leegin Creative Leather Prods., Inc. v. PSKS, Inc., 551 U.S. 877, 906 (2007) 

(acknowledging the economic foundations of U.S. antitrust law). 

37 Motoko Rich, Google and Authors Win Extension for Book Settlement, N.Y. TIMES, Nov. 9, 2009, at 

B3, available at 

www.nytimes.com/2009/11/10/technology/companies/10gbooks.html?_r=1. The 

DOJ expressed dissatisfaction with the parties’ most recent proposed settlement, as well. See 

Cecilia Kang, Judge Puts Off Ruling on Google’s Proposed Digital Book Settlement, WASH. POST, Feb. 

19, 2010, available at www.washingtonpost.com/wpdyn/content/article/2010/02/18/AR2010021800944.html?hpid=moreheadlines. 

38 Electronic Frontier Foundation, Google Book Search Settlement and Reader Privacy, 

available at www.eff.org/issues/privacy/google-book-search-settlement (last visited 

July 11, 2010). As author Michael Chabon argues, “if there is no privacy of thought — 

which includes implicitly the right to read what one wants, without the approval, consent or 

knowledge of others — then there is no privacy, period.” Id. 

39 See, e.g., Diane Leenheer Zimmerman, Can Our Culture Be Saved? The Future of Digital 

Archiving, 91 MINN. L. REV. 989, 990–91 (2007) (looking at the Google Book Search project 

as a means of saving culture and “explor[ing] whether saving culture and saving copyright 

can be made compatible goals”). 

40 Robert Darnton, The Library in the New Age, 55 N.Y. REV. BOOKS, June 12, 2008, at 39, 

available at www.nybooks.com/articles/21514.


search program to promote the public interest. Courts could condition a pro- 

Google fair use finding on universal access to the contents of the resulting 

database. Landmark cases like Sony v. Universal 41 set a precedent for taking such 

broad public interests into account in the course of copyright litigation. 42 Those 

who opt out of the settlement may be able to fight for such concessions, but for 

now the battle centers on challenges to the settlement itself. 

Both James Grimmelmann and Pamela Samuelson have suggested several 

principles and recommendations to guide judicial deliberations on the proposed 

settlement. 43 Grimmelmann’s work has focused primarily on antitrust issues, 44 

while Samuelson has concentrated on the concerns of academic authors. 45 

Grimmelmann has succinctly summarized the settlement’s potential threats to 

innovation and competition in the market for book indices, and books 

themselves: 

The antitrust danger here is that the settlement puts Google in 

a highly privileged position for book search and book sales. … 

The authors and publishers settled voluntarily with Google, but 

there’s no guarantee they’ll offer similar terms, or any terms at 

all, to anyone else. … [They] could unilaterally decide only to 

talk to Google. 46 

41 Sony Corp. of Am. v. Universal City Studios, Inc., 464 U.S. 417, 442 (1984). 

42 Frank Pasquale, Breaking the Vicious Circularity: Sony’s Contribution to the Fair Use 

Doctrine, 55 CASE W. RES. L. REV. 777, 790 (2005). 

43 See Pamela Samuelson, Google Book Search and the Future of Books in Cyberspace, 94 MINN. L. 

REV. (forthcoming, 2010), available at http://digitalscholarship.org/digitalkoans/2010/01/13/google-book-search-and-the-future-ofbooks-in-cyberspace/ 

(discussing the “six categories of serious reservations that have 

emerged about the settlement … reflected in the hundreds of objections and numerous 

amicus curiae briefs filed with the court responsible for determining whether to approve the 

settlement.”). 

44 See generally James Grimmelmann, How to Fix the Google Book Search Settlement, 12 J. INTERNET 

L., Apr. 2009, at 1 (arguing that the Google Book Search antitrust case settlement should be 

approved with additional measures designed to promote competition and protect 

consumers) [hereinafter Grimmelmann, Google Book Search Settlement]. 

45 Letter from Pamela Samuelson, Richard M. Sherman Distinguished Professor of Law, 

University of California, Berkeley School of Law, to Hon. Denny Chin, Judge, S.D.N.Y. 

(Sept. 3, 2009), available at www.scribd.com/doc/19409346/Academic-Author-Letter-090309 

(urging the judge to condition “approval of the Settlement Agreement on modification of 

various terms identified herein so that the Agreement will be fairer and more adequate 

toward academic authors.”). 

46 James Grimmelmann, In Google We Antitrust, TPMCAFÉ BOOK CLUB, Jan. 15, 2009, 

http://tpmcafe.talkingpointsmemo.com/2009/01/15/in_google_we_antitrust.


Grimmelmann proposes several methods of assuring that the publishers will 

deal with other book search services. 47 Grimmelmann suggests an “[a]ntitrust 

consent decree” and “[n]ondiscrimination among copyright owners” as 

potential responses to the issues raised by the settlement. 48 Most of his 

proposal reflects a policy consensus that presumes competition is the ideal 

solution to abuses of power online. 49 

Yet there are many reasons why competition is unlikely to arise in book search 

services, even if the settlement is altered in order to promote it. 50 Licensing 

costs are likely to be a substantial barrier to entry. A key to competition in the 

search market is having a comprehensive database of searchable materials; the 

more these materials need to be licensed, the less likely it is that a second comer 

can set up its own book archive. As scholars have demonstrated, deals like 

Google’s proposed settlement help entrench copyright holders’ claims for 

licensing revenue. 51 Moreover, innovation in search is heavily dependent on 

having an installed base of users that effectively “train” the search engine to be 

responsive. 52 The more search queries an engine gets, the better able it is to 

sharpen and perfect its algorithm. 53 Each additional user tends to decrease the 

cost of a better quality service for all subsequent users by contributing activity 

that helps the search engine differentiate between high and low quality 

organizational strategies. 54 Thus, incumbents with large numbers of users enjoy 

47 Id. 

48 Grimmelmann, Google Book Search Settlement, supra note 44, at 15. 

49 Grimmelmann does also propose some revised terms that would not be primarily designed 

to incentivize the development of new alternatives to Google Book Search; for example, he 

proposes “[l]ibrary and reader representation at the [Book Rights R]egistry” that would 

administer many aspects of the settlement. Id. 

50 See Bracha & Pasquale, supra note 5, at 1152 (“Though the market choices of users and 

technological developments constrain search engine abuse to some extent, they are unlikely 

to vindicate [certain social] values … .”); Frank Pasquale, Seven Reasons to Doubt Competition in 

the General Search Engine Market, MADISONIAN, Mar. 18, 2009, 

http://madisonian.net/2009/03/18/seven-reasons-to-doubt-competition-in-thegeneral-search-engine-market. 

51 See James Gibson, Risk Aversion and Rights Accretion in Intellectual Property Law, 116 YALE L.J. 

882, 884 (2007) (describing how the decision as to whether to fight for fair use or license a 

copyrighted work can be difficult “because the penalties for infringement typically include 

supracompensatory damages and injunctive relief ”). 

52 James Pitkow et al., Personalized Search, 45 COMMS. ACM, Sept. 2002, at 50 (discussing 

methods of personalizing search systems). 

53 For example, if 100 people search for “alternatives to Microsoft Word software” on a search 

engine on a given day and all pick the third-ranked result, the search algorithm may adjust 

itself and put the third-ranked result as the first result the next day. The most-used search 

engine will have more data to tweak its algorithms than its less-used rivals. 

54 Oren Bracha & Frank Pasquale, Federal Search Commission: Fairness, Access, and Accountability in 

the Law of Search, 93 CORNELL L. REV. 1141, 1181 (2008); David A. Vise & Mark Malseed,


substantial advantages over smaller entrants. Restrictive terms of service also 

deter competitors who aspire to reverse engineer and develop better versions of 

such services. 55 In general purpose search, users cannot reproduce, copy, or 

resell any Google service for any reason, even if the behavior is manual and nondisruptive. 

56 Another section proscribes “creat[ing] a derivative work of … the 

Software.” 57 Advertisers face other restrictions, as Google’s AdWords 

Application Programming Interface (API) Terms & Conditions “impede 

advertisers’ efforts to efficiently copy their ad campaigns to other providers.” 58 

All of these factors militate against robust competition in the comprehensive 

book search field. 

Quantum leaps in technology capable of overcoming these brute disadvantages 

are unlikely, particularly because search is as much about personalized service as 

it is about technical principles of information organization and retrieval. 59 

Current advantage in search is likely to be self-reinforcing, especially given that 

so many more people are using the services now than when Google overtook 

other search engines in the early 2000s. 60 

What does an online world featuring an entrenched Google Book Search as 

gatekeeper look like? Initially, it will prove a vast improvement on the status 

The Google Story 215 (2005) (noting that the most-used search engine will have more data 

to tweak its algorithms than its less-used rivals. ). ( 

55 Though the precise terms of service of Google Book Search have not been finalized, 

Google’s more general terms of service are not promising. Google’s terms of service 

prohibit any action that “interferes with or disrupts” Google’s services, networks, or 

computers. Google Inc., Terms of Service § 5.4 (Apr. 16, 2007), 

www.google.com/accounts/TOS. Repeated queries to the service necessary to gather data 

on its operations may well violate these terms. 

56 Id. § 5.5. 

57 Id. § 10.2. Section 5.3 would proscribe both the automatic data collection and the use of a 

nonapproved “interface” for accessing Google’s database, regardless of the exact means. Id. 

§ 5.3. 

58 Ben Edelman, PPC Platform Competition and Google’s ‘May Not Copy’ Restriction, June 27, 2008, 

http://www.benedelman.org/news/062708-1.html (arguing that “Google’s restrictions 

on export and copying of advertisers’ campaigns … hinder competition in Internet 

advertising”). Though the hearing at which Professor Edelman was to testify was cancelled, 

he has documented these problems in some detail at his website, www.benedelman.org. 

59 John Battelle, THE SEARCH: HOW GOOGLE AND ITS RIVALS REWROTE THE RULES OF 

BUSINESS AND TRANSFORMED OUR CULTURE 8 (2005).at 8 (describing how personalized 

search enhances the value of search engines to both users and advertisers). Due to trade 

secrecy, it is impossible for policymakers to discover how much of the intermediary’s success 

is due to its employees’ inventive genius, and how much is due to the collective contributions 

of millions of users to the training of the intermediary’s computers. 

60 See Randall Stross, Planet Google: One Company’s Audacious Plan to Organize Everything 

We Know 98 (Free Press 2008) (describing success of YouTube, a subsidiary of Google).


quo of bulky, hard-to-acquire, physical copies of books. But when we consider 

the ways in which knowledge can be rationed for profit, or structured to 

promote political ends, some worries arise. Google plans to monetize the book 

search corpus, and one predictable way of increasing its value is to make parts 

of it unavailable to those unwilling to pay high licensing fees. If the settlement 

allowed Google to charge such fees in an unconstrained manner, unmoored 

from the underlying costs of operating the project, the company would 

essentially be exploiting a public easement (to copy books) for unlimited private 

gain. 61 The Open Content Alliance has questioned the restrictive terms of the 

contracts that Google strikes when it agrees to scan and create a digital database 

of a library’s books. 62 Those restrictive terms foreshadow potential future 

restrictions on book search services. The proposed deal raises fundamental 

questions about the proper scope of private initiative in organizing and 

rationing access to knowledge. 

Well-funded libraries may pay a premium to gain access to all sources; lesser 

institutions may be granted inferior access. If permitted to become prevalent, 

such tiered access to information could rigidify and reinforce existing 

inequalities in access to knowledge. 63 Information tiering inequitably 

disadvantages many groups, promoting the leveraging of wealth into status, 

educational, or other occupational advantage. Information is not only 

intrinsically valuable, but also can be a positional good, useful for scoring 

advantages over others. 64 

61 Writers’ Reps and Richard A. Epstein Objection filed with the Southern District of New 

York in re Google Book Search, available at 

http://www.writersreps.com/feature.aspx?FeatureID=172 (arguing that the Google 

Book Search Settlement “would accomplish[] orphan legislation—but just for Google. … If 

[Google] is to be handed exclusive possession after stealing the scans to begin with, then it 

should be required to share those scans.”). 

62 See Open Content Alliance, Let’s Not Settle for This Settlement, 

www.opencontentalliance.org/2008/11/05/lets-not-settle-for-this-settlement (last visited 

Mar. 12, 2010) (“At its heart, the settlement agreement grants Google an effective monopoly 

on an entirely new commercial model for accessing books. It re-conceives reading as a 

billable event. This reading event is therefore controllable and trackable. It also forces 

libraries into financing a vending service that requires they perpetually buy back what they 

have already paid for over many years of careful collection.”). 

63 Frank Pasquale, Technology, Competition, and Values, 8 MINN. J. L. SCI. & TECH. 607, 608 (2007) 

(explaining how “much technology is used not just simply to improve its user’s life, but also 

to help its user gain advantage over others”). For example, “[t]est-preparation technologies 

… creat[e] inequalities; students able to afford test-preparation courses, such as those 

offered by Kaplan, have a definite advantage over those who do not have access to such 

courses.” Id. at 615 (internal citation omitted). 

64 Harry Brighouse & Adam Swift, Equality, Priority, and Positional Goods, 116 ETHICS 471, 472 

(2006) (“[Positional goods] are goods with the property that one’s relative place in the 

distribution of the good affects one’s absolute position with respect to its value. The very 

fact that one is worse off than others with respect to a positional good means that one is


Admittedly, Google Book Search has so far proven a great resource for 

scholars. It has made “book learning accessible on a new, worldwide scale, 

despite the great digital divide that separates the poor from the 

computerized.” 65 Current access to knowledge is stratified in many troubling 

ways; the works of John Willinsky 66 and Peter Suber 67 identify many troubling 

current forms of tiering that pale before the present impact of Google Book 

Search. 68 Given the aggressive pricing strategies of many publishers and 

content owners, Google Book Search is a vital alternative for scholars. 

Nevertheless, there is no guarantee in the current version of the settlement that 

Google Book Search will preserve its public-regarding features. 69 It may well 

end up like the powerful “group purchasing organizations” in the American 

health care system that started promisingly, but have evolved to exploit their 

intermediary role in troubling ways. 70 Google is more than just one among 

many online service providers jostling for a competitive edge on the web. It is 

likely to be the key private entity capable of competing or cooperating with 

academic publishers and other content providers. Dedicated monitoring and 

regulation of the settlement terms now could help ensure that book digitization 

worse off, in some respect, than one would be if that good were distributed equally. So 

while it might indeed be perverse to advocate leveling down all things considered, leveling 

down with respect to positional goods benefits absolutely, in some respect, those who would 

otherwise have less than others. 

65 Darnton, supra note 40, at 76. 

66 JOHN WILLINSKY, THE ACCESS PRINCIPLE: THE CASE FOR OPEN ACCESS TO RESEARCH AND 

SCHOLARSHIP 5 (The MIT Press 2005) (describing extreme “digital divide” between those 

most connected to information resources and those cut off from them). 

67 See generally Peter Suber, Open Access News, www.earlham.edu/~peters/fos/fosblog.html. 

Suber is a leader of the open access movement, which aims to “[p]ut[] peer-reviewed 

scientific and scholarly literature on the internet[,] [m]ak[e] it available free of charge and 

free of most copyright and licensing restrictions[,] [and] remov[e] the barriers to serious 

research.” Id. 

68 See id. (chronicling on a daily basis news and controversies related to open access to scholarly 

materials on the Internet). 

69 Siva Vaidhyanathan, Baidu.com Accused of Rigging Search, The Googlization of Everything, 

Global Google, Jan. 2009, 

http://www.googlizationofeverything.com/2009/01/baiducom_accused_of_rigging 

_se.php (Feb. 19, 2009, 14:20 EST) (“‘Public failure’ [is a] phenomenon in which a private 

firm steps into a vacuum created by incompetent or gutted public institutions. A firm does 

this not for immediate rent seeking or even revenue generation. It does so to enhance 

presence, reputation, or to build a platform on which to generate revenue later or elsewhere. 

It’s the opposite of ‘market failure.’ And it explains a lot of what Google does.”). 

70 For background on group purchasing organizations, see S. PRAKASH SETHI, GROUP 

PURCHASING ORGANIZATIONS: AN UNDISCLOSED SCANDAL IN THE U.S. HEALTH CARE 

INDUSTRY 122 (Palgrave MacMillan 2009) (“The benefits of combined purchases would be 

greatly reduced in conditions where the middlemen … control the entire process through 

restrictive arrangements with suppliers and customers.”).


protects privacy, diverse stakeholder interests, and fair pricing of access to 

knowledge. Alliances between Google Book Search and publishers deserve 

public scrutiny because they permit private parties to take on what have often 

been public functions of determining access to and pricing of information. 

Where “regulatory copyright” 71 has answered such questions with compulsory 

licenses, 72 the new alliances aspire to put into place a regime of crosssubsidization 

resistant to public scrutiny or input. 73 Given the vital public 

interests at stake in the development of this information infrastructure, 

monitoring is vital. 74 Extant law provides little assurance that it will actually 

occur. 

A Public Alternative? 

In other work, I have proposed a number of regulations that would permit 

either government or public accountability groups to monitor search engines to 

detect abuses of their dominant position. To conclude this piece, I would like 

to raise one other alternative: a publicly funded search engine. 

To the extent that search engines resist monitoring and accountability, 

governments should consider establishing public alternatives to them. Here, 

lessons from recent debates over health insurance may be instructive. There 

are structural parallels between the intermediary role of private health insurers 

(which stand as a gatekeeper between patients and providers of health products 

and services) and that of search engines (which stand between searchers and 

71 See Joseph P. Liu, Regulatory Copyright, 83 N.C. L. REV. 87, 91 (2004) (describing the growth 

and scope of compulsory licensing statutes that provide for compensation for copyright 

holders while denying them the right to veto particular uses of their work). 

72 Marybeth Peters, the U.S. Register of Copyrights, has objected to the proposed Google 

Books Settlement on the grounds that it would violate traditional norms of separation of 

powers in copyright policy. See Hearing on Competition and Commerce in Digital Books: The 

Proposed Google Book Settlement Before the House Comm. on the Judiciary, 111th Cong. (2009) 

(statement of Marybeth Peters, Register of Copyrights), available at 

http://judiciary.house.gov/hearings/pdf/Peters090910.pdf, at 2 (“In the view of the 

Copyright Office, the settlement proposed by the parties would encroach on responsibility 

for copyright policy that traditionally has been the domain of Congress. … We are greatly 

concerned by the parties’ end run around legislative process and prerogatives, and we submit 

that this Committee should be equally concerned.”). 

73 Google considers its pricing and ranking decisions a closely held trade secret—an assertion 

that would seem very strange if it came from a public library. See Pamela Samuelson, Google 

Books Is Not a Library, THE HUFFINGTON POST, Oct. 13, 2009, 

www.huffingtonpost.com/pamela-samuelson/google-books-is-not-alib_b_317518.html 

(“Libraries everywhere are terrified that Google will engage in pricegouging 

when setting prices for institutional subscriptions to [Google Book Search] 

contents.”). 

74 Frank Pasquale, Beyond Competition and Innovation: The Need for Qualified Transparency in Internet 

Intermediaries, 104 Nw. U. L. REV. 105 (2010) (offering proposals for monitoring internet 

intermediaries).


providers of information). The 1965 decision to establish Medicare as a public 

option for an elderly population ill-served by private providers and insurers may 

prove a model for an information economy plagued by persistent digital divides. 

As the United States debated health reform from 2009 to 2010, there was a 

tension between regulation-focused approaches (which would require revelation 

and alteration of private insurers’ unfair practices) and a public option that 

would compete with existing insurers. Democrats ultimately gave up on 

pushing the public option, but the debate exposed the many positive aspects a 

state-sponsored alternative can provide in certain markets. A public option 

could play a role in search parallel to the role that Medicare plays in the health 

system: guaranteeing some baseline of transparency in pricing and evaluation. 75 

The recent Google Book Search settlement negotiations have led Siva 

Vaidhyanathan to characterize Google’s archive project as evidence of a “public 

failure.” 76 Whereas government intervention is often necessary in cases of 

“market failure,” Vaidhyanathan argues that the reverse can occur: market 

actors can step into a vacuum where government should have been. In the case 

of digitized books, the problem is presented starkly: Why has the Library of 

Congress failed to require digital deposit of books, instead of merely accepting 

paper copies? We can debate when such a requirement became plausible; 

however, had the government required such deposit as soon as it became 

feasible, the problematic possibility of a Google monopoly here would be much 

less troubling. If digital deposit ever is adopted, the government could license 

its corpus to alternative search services. There is no good reason why the 

company that is best capable of reproducing books (and settling lawsuits based 

on that reproduction) should have a monopoly on search technologies used to 

organize and distribute them. 

More ambitiously, an NGO or quasi-administrative NGO could undertake to 

index and archive the web, licensing opportunities to search and organize it to 

various entities that promise to maintain open standards for ranking and rating 

websites and other Internet presences. 77 Wikipedia, Slashdot, and eBay all 

75 For more on the role of public options like Medicare in the modern medical sector, see Frank 

Pasquale, Making the Case for the Public Plan, Part II: Public Option as Private Benchmark, July 15, 

2009, available at http://balkin.blogspot.com/2009/06/making-case-for-public-planpart-ii.html. 

76 Vaidhyanathan, supra note 69. (“ ‘Public failure’ [is a] phenomenon in which a private firm 

steps into a vacuum created by incompetent or gutted public institutions. A firm does this 

not for immediate rent seeking or even revenue generation. It does so to enhance presence, 

reputation, or to build a platform on which to generate revenue later or elsewhere. It’s the 

opposite of ‘market failure.’ And it explains a lot of what Google does.”). 

77 For a cultural case for government intervention here, see Mário J. Silva, The Case for a 

Portuguese Web Search Engine, 

http://www.google.com/url?sa=t&source=web&cd=1&ved=0CBUQFjAA&url=htt 

p%3A%2F%2Fciteseerx.ist.psu.edu%2Fviewdoc%2Fdownload%3Fdoi%3D10.1.1.106.


suggest methods of evaluating relevance and authority that could be employed 

by public, open search engines. If such a search engine became at least 

somewhat popular (or popular within a given niche), it could provide an 

important alternative source of information and metadata on ranking processes. 

The need for a public option in search becomes even more apparent when we 

consider the waste and inefficiency caused by opaque intermediaries in other 

fields. Like private health insurers, Google is a middleman, standing between 

consumers and producers of knowledge. In programs like Book Search, it will 

effectively collaborate with copyright owners to determine what access people 

get, how much they have to pay, and on what terms. In the health field, 

providers and private insurers are both very concentrated in the U.S., and 

consumers (i.e., the businesses and individuals who buy insurance plans) are not. 

Insurers and providers also jealously guard the secrecy of many pricing 

decisions. 78 That is one key reason why the U.S. spends so much more on 

health care than other industrialized nations, without getting consistently better 

results, access, or quality. 

Health care reformers often split into two camps: those who believe that 

regulation of middlemen like insurers can bring about fair results, and those 

who believe that only a public option can serve as a benchmark for judging the 

behavior of private insurers. The Patient Protection and Affordable Care Act 

(PPACA) of 2010 decisively opted for the regulatory option, and the early stages 

of its implementation have been rocky. The constitutional challenges to search 

engine regulation would likely prove more serious than the many lawsuits now 

attacking PPACA. Therefore, even if the public option in health care is off the 

table now, it should inspire future proposals in information policy, where 

regulation of intermediaries may be even more difficult than it has proven to be 

in health care. If search engines consistently block or frustrate measures to 

increase their accountability, public alternatives could prove to be an 

indispensable foundation of a fair, just, and open information environment. 

5334%26rep%3Drep1%26type%3Dpdf&ei=IWZYTJbaCoKC8gapvY2xCw&usg=AFQ 

jCNHdTPpTBUuNHZhTOZtGaRiVKP6C4g&sig2=9aoaKLXiXOOUuYHMewopV 

Q (describing the value of a Portuguese-oriented search engine); JEAN NOEL JENNENY, 

GOOGLE AND THE MYTH OF UNIVERSAL KNOWLEDGE: A VIEW FROM EUROPE (Univ. of 

Chicago Press 2007). Whereas these authors believe that English-language bias is a 

particularly problematic aspect of Google’s hegemony in the field, I argue that the possibility 

of many kinds of hidden bias counsel in favor of at least one robust, publicly funded 

alternative. 

78 See, e.g., Uwe Reinhart, The Pricing of U.S. Hospital Services: Chaos Behind a Veil of Secrecy, at 

http://healthaff.highwire.org/cgi/content/abstract/25/1/57; Annemarie Bridy, Trade 

Secret Prices and High-Tech Devices: How Medical Device Manufacturers are Seeking to Sustain Profits by 

Propertizing Prices, 17 TEX. INTELL. PROP. L.J. 187 (2009) (discussing “recent claims by the 

medical device manufacturer Guidant that the actual prices its hospital customers pay for 

implantable devices, including cardiac pacemakers and defibrillators, are protectable as trade 

secrets under the Uniform Trade Secrets Act.”).



The Problem of Search Engines 

as Essential Facilities: An 

Economic & Legal Assessment 

By Geoffrey A. Manne * 

What is wrong with calls for search neutrality, especially those rooted in the 

notion of Internet search (or, more accurately, Google, the policy scolds’ bête 

noir of the day) as an “essential facility,” and necessitating governmentmandated 

access? As others have noted, the basic concept of neutrality in 

search is, at root, farcical. 1 The idea that a search engine, which offers its users 

edited access to the most relevant websites based on the search engine’s 

assessment of the user’s intent, 2 should do so “neutrally” implies that the search 

engine’s efforts to ensure relevance should be cabined by an almost-limitless 

range of ancillary concerns. 3 

Nevertheless, proponents of this view have begun to adduce increasingly detailladen 

and complex arguments in favor of their positions, and the European 

Commission has even opened a formal investigation into Google’s practices, 

based largely on various claims that it has systematically denied access to its top 

search results (in some cases paid results, in others organic results) by 

competing services, 4 especially vertical search engines. 5 To my knowledge, no 

* Executive Director, International Center for Law & Economics and Lecturer in Law, Lewis 

& Clark Law School. www.laweconcenter.org; 

www.lclark.edu/law/faculty/geoffrey_manne. 

1 See, e.g., Danny Sullivan, The Incredible Stupidity of Investigating Google for Acting Like a Search 

Engine, SEARCH ENGINE LAND, http://searchengineland.com/the-incredible-stupidityof-investigating-google-for-acting-like-a-search-engine-57268 

(“A search engine’s job is 

to point you to destination sites that have the information you are seeking, not to send you 

to other search engines. Getting upset that Google doesn’t point to other search engines is 

like getting upset that the New York Times doesn’t simply have headlines followed by a 

single paragraph of text that says ‘read about this story in the Wall Street Journal.’”). 

2 A remarkable feat, given that this intent must be inferred from simple, context-less search 

terms. 

3 Perfectly demonstrated by Frank Pasquale’s call, elsewhere in this volume, for identifying 

search engines as “essential cultural and political facilities,” thereby mandating incorporation 

into their structure whatever “cultural” and “political” preferences any sufficiently-influential 

politician (or law professors) happens to deem appropriate. 

4 Competing services include, for example, MapQuest (www.mapquest.com) (competing 

with Google Maps), Veoh (www.veoh.com) (competing with You Tube) and Bing Shopping 

(www.bing.com/shopping) (competing with Google Products). 

5 Vertical search engines are search engines that focus on a particular category of products, or 

on a particular type of search. Examples include Kayak (www.kayak.com) (travel search),


one has yet claimed that Google should offer up links to competing general 

search engines as a remedy for its perceived market foreclosure, but Microsoft’s 

experience with the “Browser Choice Screen” it has now agreed to offer as a 

consequence of the European Commission’s successful competition case 

against the company is not encouraging. 6 These more superficially sophisticated 

claims are rooted in the notion of Internet search as an “essential facility”—a 

bottleneck limiting effective competition. 

These claims, as well as the more fundamental harm-to-competitor claims, are 

difficult to sustain on any economically-reasonable grounds. To understand this 

requires some basic understanding of the economics of essential facilities, of 

Internet search, and of the relevant product markets in which Internet search 

operates. 

The Basic Law & Economics 

of Essential Facilities 

There are two ways to deal with a problematic bottleneck: Remove the 

bottleneck or regulate access to it. The latter is the more common course 

adopted in the U.S. and elsewhere. Complex, Byzantine and often counterproductive 

regulatory apparatuses are required to set and monitor the terms of 

access. Among other things, this paves the way for either intensely-problematic 

judicial oversight of court-imposed remedies or else the creation of sectorspecific 

regulatory agencies subject to capture, political influence, bureaucratic 

inefficiency, and inefficient longevity. The Interstate Commerce Commission 

(and its successor agencies within the Department of Transportation) and the 

Federal Communications Commission (and its implementation beginning in 

1996 of the monstrous Telecommunications Act) in the U.S. are paradigmatic 

examples of these costly effects, and it is certainly questionable whether the 

disease is worse than the cure. 7 

Obviously, an essential facility must be essential. Efforts over the years to 

shoehorn various markets into this category have sometimes strained credulity, 

as it has variously been claimed that Aspen, Colorado ski hills, 8 local voice mail 

SourceTool (www.sourcetool.com) (business input sourcing), and Foundem 

(www.foundem.com) (retail product search and price comparison). 

6 See European Commission, Web browser choice for European consumers, 

http://ec.europa.eu/competition/consumers/web_browsers_choice_en.html (last 


7 Oren Bracha and Frank Pasquale’s call for a “Federal Search Commission” modeled on the 

Federal Trade Commission is in fact an embrace of the need for a bureaucratic apparatus to 

regulate the forced access called for by search neutrality proponents. See Oren Bracha and 

Frank Pasquale, Federal Search Commission: Fairness, Access, and Accountability in the Law of Search, 

93 CORNELL L. REV. 1193 (2008). 

8 Aspen Highlands Skiing Corp. v. Aspen Skiing Co., 738 F.2d 1509 (10th Cir. 1984), aff ’d on 

other grounds, 472 U.S. 585 (1985).


services, 9 soft drinks 10 and direct freight flights between New York and San 

Juan 11 (among many other things) were essential facilities necessitating 

mandated access under the antitrust laws. 12 In these and many other cases, 

myriad alternatives to the allegedly-monopolized market exist and it is arguable 

that there was nothing whatsoever “essential” about these markets. 

In antitrust literature and jurisprudence, a plaintiff would need to prove the 

following to prevail in a monopolization case rooted in the essential facilities 

doctrine: 

1. Control of the essential facility by a monopolist; 

2. A competitor’s inability practically or reasonably to duplicate the 

essential facility; 

3. The denial of the use of the facility to a competitor; and 

4. The feasibility of providing the facility to competitors. 13 

Arguably, since the Supreme Court’s 2004 Trinko decision, 14 a plaintiff would 

also need to demonstrate the absence of federal regulation governing access. 

The Trinko decision significantly circumscribed the area subject to essential 

facilities arguments, limiting such claims to instances where, as in the Aspen 

Skiing case, a competitor refuses to deal on reasonable terms with another 

competitor with whom it has, in fact, dealt in the past. 15 

A key problem with many essential facilities cases is the non-essentiality of the 

relevant facility. While there can be no doubt that to particular competitors, 

particularly those constrained to only one avenue of access to consumers by 

geography or natural monopoly, a facility may indeed seem essential, the 

touchstone of U.S. antitrust law has long been consumer, not competitor, 

welfare. So while, indeed, Aspen Highlands may have had difficulty competing 

with the Aspen Ski Company for consumers who had already chosen to ski in 

Aspen, consumers nonetheless had unfettered access to a wide range of 

alternative ski (and other vacation) destinations, such that the likelihood of the 

9 CTC Communications Corp. v. Bell Atlantic Corp., 77 F. Supp. 2d 124 (D. Me. 1999). 

10 Sun Dun v. Coca-Cola Co., 740 F. Supp. 381 (D. Md. 1990). 

11 Century Air Freight, Inc. v. American Airlines, Inc., 597 F. Supp. 564 (S.D.N.Y. 1984). 

12 For a more complete list of essential facilities (and attempted essential facilities) cases, as 

well as an important treatment of the essential facilities doctrine in US antitrust law, see 

Abbott B. Lipsky, Jr. & J. Gregory Sidak, Essential Facilities, 51 STAN. L. REV. 1187 (1999). 

13 MCI Comm’ns Corp. v. American Tel. & Tel. Co., 708 F.2d 1081 (7th Cir. 1982). 

14 Verizon Comm’ns. v. Law Offices of Curtis V. Trinko LLP, 540 U.S. 398 (2004). 

15 See, e.g., PHILLIP E. AREEDA & HERBERT HOVENKAMP, ANTITRUST LAW (2004 supp.) at 199.


monopolization of Aspen’s ski hills affecting overall consumer welfare was 

essentially non-existent. 16 In such a circumstance, should it matter if a particular 

competitor is harmed? Is that a function of antitrust-relevant conduct on the 

part of another firm, or an unfortunate set of business decisions on the part of 

the first firm? 

As Phillip Areeda and Herbert Hovenkamp have famously said of the essential 

facilities doctrine, “[it] is both harmful and unnecessary and should be 

abandoned.” 17 As another antitrust expert has described it: 

At bottom, a plaintiff making an essential facilities argument is saying 

that the defendant has a valuable facility that it would be difficult to 

reproduce, and suggesting that is a reason for a court to intervene and 

impose a sharing duty. But at least in the vast majority of the cases, the 

fact that the defendant has a highly valued facility is a reason to reject 

sharing, not to require it, since forced sharing “may lessen the incentive 

for the monopolist, the rival, or both to invest in those economically 

beneficial facilities.” 18 

This perennial problem—antitrust laws being used to protect competitors rather 

than consumers—lies at the heart of claims surrounding Internet search as an 

essential facility. 

There is much tied up in the argument, and proponents have often been careful 

to at least go through the motions of drawing the rhetorical line back to 

consumers. In its fullest expression, it is claimed that harm to competitors now 

will mean the absence of competitors later and thus an unfettered monopoly 

with the intent and power to harm consumers. 19 It is also often argued that 

consumers (in this case Internet users searching for certain websites or the 

products they sell) are intrinsically harmed by the unavailability of access to the 

information contained in sites that are denied access to the search engine’s 

“essential facility.” 20 

16 The courts, however, did not agree. 

17 3A AREEDA & HOVENKAMP, ANTITRUST LAW 771c, at 173 (2002). 

18 R. Hewitt Pate, Refusals to Deal and Essential Facilities, Testimony Submitted to DOJ/FTC 

Hearings on Single Firm Conduct, Jul. 18, 2006, available at 

http://www.justice.gov/atr/public/hearings/single_firm/docs/218649.htm (quoting 

Trinko, 540 U.S.. at 408). 

19 See, e.g., European Commission Launches Antitrust Investigation of Google, SEARCH 

NEUTRALITY.ORG, Nov. 30, 2010, http://www.searchneutrality.org (“Google is exploiting 

its dominance of search in ways that stifle innovation, suppress competition, and erode 

consumer choice.”). Meanwhile, complainants have gone to Europe where a showing of 

consumer harm is not necessary to prevail under its competition laws. 

20 As Oren Bracha and Frank Pasquale put it, “Search engines, in other words, often function 

not as mere satisfiers of predetermined preferences, but as shapers of preferences,” Federal 

Search Commission, 93 CORNELL L. REV. at 1185. Bracha and Pasquale also claim that “Market 

participants need information about products and services to make informed economic


The basic essential facilities case against Google is that it controls a bottleneck 

for the Internet—it is the access point for most consumers, and search results 

on Google determine which websites are successful and which end up in 

oblivion. 21 More particularly, it is argued that Google has used its control over 

this bottleneck to deny access by competitors to Google’s users. To understand 

this requires a brief discussion of the economics relevant to Internet Search and 

its relevant market. 

The Basic Economics of Internet Search 

Implicit in claims that Google controls access to an essential facility is that 

access by some relevant set of consumers (or competitors) to relevant content is 

accessible only (or virtually only) through Google. It is necessary, then, to 

assess whether Google’s search results pages are, in fact, without significant 

competition for the economic activity at their heart. Of course the economic 

activity at their heart is advertising. 22 

It is hard to conceive of Internet search—let alone Google’s website—as the 

only means of reducing search costs for potential consumers (Internet 

searchers) and prospective sellers. Leaving aside the incredible range of 

alternative sources to the Internet for commerce, 23 off the top of my head, I can 

imagine Google’s competitor websites finding access to users by 1) advertising 

in print publications and TV; 2) using social networking sites to promote their 

sites, 3) being linked to by other websites including sites specializing in rating 

websites, online magazines, review sites, and the like; 4) implementing affiliate 

programs or other creative marketing schemes; 5) purchasing paid advertising, 

both in Google’s own paid search results, as well as on other, heavily-trafficked 

websites; and 6) securing access via Google’s general search competitors like 

Yahoo! and Bing. Competitors denied access to the top few search results at 

decisions. … [A]attaining visibility and access to users is critical to competition and 

cooperation online. Centralized control or manipulation by search engines may stifle 

innovation by firms relegated to obscurity.” Id. at 1173-74. 

21 Id. at 1173 (“Concentrated control over the flow of information, coupled with the ability to 

manipulate this flow, may reduce economic efficiency by stifling competition.”). 

22 See KEN AULETTA, GOOGLED: THE END OF THE WORLD AS WE KNOW IT 16 (2009) (quoting 

Google CEO Eric Schmidt as saying, “We are in the advertising business”). 

23 There is a tendency for Web sites to view their Internet enterprises as different than their 

offline counterparts’, but, at root, most Internet sites (other than branded ones attached 

directly to offline stores) are founded by entrepreneurs who made a simple business decision 

to ply their trade online rather than off. That this decision may have foreclosed easy access 

to certain offline customers, or put the entrepreneur in a position where access to customers 

could be frustrated by certain competitive disadvantages specific to the Internet, does not 

convert these competitive disadvantages into special problems deserving of antitrust 

treatment. To do so would be to inappropriately and inefficiently insulate the online/offline 

business decision from the healthy effects of Schumpeter’s “perennial gale of creative 

destruction.” JOSEPH SCHUMPETER, THE PROCESS OF CREATIVE DESTRUCTION (1942).


Google’s site are still able to advertise their existence and attract users through a 

wide range of other advertising outlets—extremely wide, in fact: According to 

one estimate Google was responsible in 2007 for only about 7.5% of the 

world’s advertising. 24 

For Google to profit from its business—whether as a monopolist or not—it 

must deliver up to its advertisers a set of users. Interestingly, users of Google’s 

general search engine are mostly uninterested in the paid results. They click 

through the unpaid or “organic” search results by a wide margin ahead of paid 

results. 25 There is thus an asymmetry. On one side of its platform are 

advertisers who care about the quantity and quality (the likelihood that users 

who see an ad will click through to advertisers’ sites and purchase something 

while there) of the users on the other side. Meanwhile, users care very little 

about the quantity of advertisers and care only somewhat about the quality of 

advertisers (preferring greater relevance to lesser, but frequently ignoring paid 

results anyway). Nevertheless, the core of this enterprise is search result 

relevance. Greater relevance improves the quality of searchers from the 

advertisers’ point of view, ensuring that advertisers’ paid results are clicked on 

by the users most likely to find the advertiser’s site of interest and to purchase 

something there. 

But there are problems inherent in the ambiguity of search terms and the ability 

to “game the system” that prevent even the most sophisticated algorithms from 

offering up perfect relevance. First, search terms are often context-less, and a 

user searching for “jaguar” may be searching for information on the car 

company, the operating system, the big cat, or something else. 26 Along a 

different dimension, a user searching for “Nikon camera” might be looking to 

buy a Nikon camera or might be looking for a picture of a Nikon camera to post 

on his blog. Obviously advertisers care very much which of these users clicks 

on their paid result. At the same time, many undesirable websites (spam sites 

and the like) can and do take advantage of predictable search results to occupy 

desirable search result real estate to the detriment of the search engine, its users 

and its advertisers. Efforts to keep these sites out of the top results and to 

ensure maximum relevance from ambiguous search terms require a host of 

algorithm tweaks and even human interventions. That these may (intentionally 

or inadvertently) harm some websites’ rank in certain search results is consistent 

with a well-functioning search platform. 

24 See Erick Schonfeld, Estimates Put Internet Advertising at $21 Billion in U.S., $45 Billion Globally, 

TECHCRUNCH, Feb. 26, 2008, http://techcrunch.com/2008/02/26/estimates-putinternet-advertising-at-21-billion-in-us-45-billion-globally/. 

25 See, e.g., Neil Walker, Google Organic Click Through Rate (CTR), UK SEO CONSULTANT, May 11, 

2010, http://www.seomad.com/SEOBlog/google-organic-click-through-ratectr.html. 

26 See Bill Slawski, A Look at Google Midpage Query Refinements, SEO BY THE SEA, Apr. 20, 2006, 

http://www.seobythesea.com/?p=174.


Google offers its organic search results and its other services as a solution to the 

two-sided platform problem mentioned above: In order to attract paying 

advertisers, Google also has to attract (and match up) the advertisers’ target 

audience. Google offers everything it does to its users in an effort to attract 

these users and to glean information from them that facilitates its all-important 

matching (relevance) function. In the process, Google generates revenue from 

advertisers eager to “sell” to this audience. For a host of reasons, Google (like 

all search engines) does not charge searchers to access its various services, but it 

does charge advertisers. Just because search is an ancillary business to Google’s 

true advertising business does not necessarily mean it is not a relevant market 

for purposes of antitrust analysis; nevertheless it is essential to avoid the pitfall 

of examining one side of a two-sided market in isolation. As David Evans 

notes, “[t]he analysis of either side of a two-sided platform in isolation yields a 

distorted picture of the business.” 27 Two-sided market definition is complex, 

and little understood—especially by non-experts throwing around various 

alleged markets in which companies like Google are said to be “dominant.” 

There is actually substantial reason to doubt the propriety of a narrow market 

definition limited to online search advertising. 28 Even where there are different 

purposes for different types of advertising—e.g. brand recognition for display 

ads and efforts to sell for search ads and other outlets like coupons—this is 

merely a difference in degree. Both are fundamentally forms of reducing the 

costs of a user’s search for a product, as we have understood since George 

Stigler’s seminal work on the subject in 1968, 29 and the relevant question is 

whether the difference is significant enough to render decisions in one market 

essentially unaffected by decisions or prices in the other. 

There is evidence that advertisers view online and offline advertising as 

substitutes, and this applies not only to traditional advertisers but also Internet 

companies. Thus, in 2009, Pepsi decided not to advertise during the 2010 Super 

Bowl, in order to focus instead on a particular type of online campaign. “This 

year for the first time in 23 years, Pepsi will not have ads in the Super Bowl 

telecast …. Instead it is redirecting the millions it has spent annually to the 

27 David S. Evans, Two-Sided Market Definition, ABA SECTION OF ANTITRUST LAW, MARKET 

DEFINITION IN ANTITRUST: THEORY AND CASE STUDIES (forthcoming), available at 

http://ssrn.com/abstract=1396751 at p. 9. 

28 Readers interested in a fuller treatment of the market definition question surrounding 

Google are directed toward Geoffrey A. Manne & Joshua D. Wright, Google and the Limits of 

Antitrust: The Case Against the Case Against Google, 34 HARV. J. L. & PUB. POL’Y 1 (2011) 

(forthcoming), from which much of the discussion of Google’s markets and economics in 

this essay is drawn. 

29 GEORGE JOSEPH STIGLER, THE ORGANIZATION OF INDUSTRY 201 (Univ. of Chi. Press 1983) 

(1968).


Internet.” 30 And even Google itself advertises offline. 31 Another study suggests 

that there is indeed a trade-off between online and more traditional types of 

advertising: Avid Goldfarb and Catherine Tucker have demonstrated that 

display advertising pricing is sensitive to the availability of offline alternatives. 32 

And of course companies have limited advertising budgets, distributed across a 

broad range of media and promotional efforts. As one commentator notes: “By 

2011 web advertising in the United States was expected to climb to sixty billion 

dollars, or 13 percent of all ad dollars. This meant more dollars siphoned from 

traditional media, with the largest slice probably going to Google.” 33 

Advertising revenue on the Internet is driven initially by the size of the 

audience, with a significant multiplier for the likelihood that those consumers 

will purchase the advertisers’ products 34 (based on a viewer’s propensity to 

“click through” to the advertiser’s site). Google’s competition in selling ads thus 

comes, in varying degrees, not only from other search sites, but also from any 

other site that offers a service, product, or experience that consumers might 

otherwise find in Google’s “organic” search results, for which Google is not 

paid. For Google’s competitors, this means seeking forced access to its users. 

But access to eyeballs can be had from a large range of access points around the 

Web. 

Social media sites like Twitter and Facebook are therefore significant access 

points, occupying, as they do, a considerable amount of Internet “eyeball” time. 

The Pepsi deviation of advertising revenue from the Super Bowl to the Internet 

is not likely to have inured much to Google’s benefit as the strategy was a 

“social media play,” building on the expressed brand loyalties and peer 

communications that propel social media. 35 In a world of scarce advertising 

dollars and effective marketing via social media sites, Google and all other 

advertisers, online and off, must compete with the growing threat to their 

revenue from these still-novel marketing outlets. “If Facebook’s community of 

30 Larry D. Woodard, Pepsi’s Big Gamble: Ditching Super Bowl for Social Media, ABC NEWS, Dec. 23, 

2009, http://abcnews.go.com/print?id=9402514. 

31 See Danny Sullivan, Google Pushes Chrome Browser Via Newspaper Ads, SEARCH ENGINE LAND, 

Nov. 21, 2010, http://searchengineland.com/google-pushes-chrome-browser-vianewspaper-ads-56600. 

32 Avi Goldfarb & Catherine Tucker, Search Engine Advertising: Pricing Ads to Context 96 (NET 

Institute Working Paper No. 07-23, 2007) available at 

http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1021451&rec=1&srcabs=10084 

33 KEN AULETTA, GOOGLED: THE END OF THE WORLD AS WE KNOW IT 16 (2009). 

34 David S. Evans, The Economics of the Online Advertising Industry, 7 REV. OF NETWORK ECON. 

359, 359-60 (2008). 

35 See Larry D. Woodard, Pepsi’s Big Gamble: Ditching Super Bowl for Social Media, ABC NEWS, Dec. 

23, 2009, http://abcnews.go.com/print?id=9402514.


users got more of their information through [the Facebook] network, their 

Internet search engine and navigator might become Facebook, not Google.” 36 

The upshot: To the extent that inclusion in Google search results is about 

“Stiglerian” search-cost reduction for websites (and it can hardly be anything 

else), the range of alternate facilities for this function is nearly limitless. 

Finally, Google competes not only with other general search engines (and 

possibly all other forms of advertising) but also with so-called vertical search 

engines. These are search engines and e-commerce websites with search 

functionality that specializes in specific content: Amazon in books, music, and 

other consumer goods; Kayak in travel services; eBay in consumer auctions; 

WebMD in medical information and products; SourceTool in business-tobusiness 

supplies; Yelp in local businesses, and many others. To the extent that 

Internet users bypass Google and begin their searches at one of these 

specialized sites (as is increasingly the case), the value to these heavily-trafficked 

websites from access to Google’s users decreases. 37 

Competition from vertical search engines is important because ad click-through 

rates likely are higher when consumers are actively searching for something to 

buy—just as search advertising targets consumers who express some interest in 

a particular search term, the effect is magnified if the searcher can be identified 

as an immediate consumer. Thus online retailers like CDnow that can establish 

their own brands and their own navigation channels 38 have a significant 

advantage in drawing searchers—and advertisers—away from Google: The fact 

that a consumer is performing a search on a retail site itself conveys important 

and valuable information to advertisers that is not otherwise available from 

most undifferentiated Google searches—it certainly increases the chance that 

the searcher is searching to buy a CD rather than learn something about the 

singer. Because this “ready-to-buy” traffic is the most valuable, there is a 

possibility of two separate search markets, with most high-value traffic 

bypassing general-purpose search engines for product search sites like eBay and 

Amazon.com, and with Google and other general-purpose search engines 

serving primarily non-targeted, lower-value traffic. The implication is that, while 

even relatively small-scale competition may present a potentially significant 

threat to Google’s search business, this threat does not depend on links to these 

sites from Google’s search results. And thus these competitors have a strong, 

36 KEN AULETTA, GOOGLED: THE END OF THE WORLD AS WE KNOW IT 172–73 (2009). 

37 For example, in the thirty days ending on February 23, 2010, less than ten percent of visits 

to eBay.com originated from a search engine. See ALEXA, eBay.com Site Info, 

http://www.alexa.com/siteinfo/ebay.com. 

38 See Donna L. Hoffman & Thomas P. Novak, How to Acquire Customers on the Web, HARV. BUS. 

REV., May–June 2000, at 3, 5, 7. (CDnow was acquired by Amazon.com in 2001.)


independent incentive to develop marketing programs outside of Google’s 

search pages—and there is good reason not to deputize Google in the process. 

Is Google an Essential Facility? 

Recall that the basic claim is that Google’s competitors are foreclosed from 

access to Google’s desirable (essential) marketing platform and thereby suffer 

significant harm. Of course from the outset, this has it backwards (and this is a 

core problem with the essential facilities doctrine as a whole). 

If there is a problem, it should be the problem of limited access by Google’s users 

to Google’s competitors. Sometimes the absence of access by competitors to 

consumers is the same thing as the absence of access by consumers to 

competitors, but it depends on how well the market has been defined. In the 

most fundamental sense Google has precisely zero control over access by 

consumers (meaning users who use Google to search the Internet) to 

competitors: Anyone with access to a browser can access any site on the 

Internet simply by typing its URL into the browser. Perhaps understanding 

this, proponents of the “Internet search is an essential facility” claim argue that 

mere access is insufficient, and that consumers are essentially ignorant about the 

valuable content on the web except by search engines, which are subject to the 

search engine’s editorial control over that access. To the typical Google user, 

according to this view, Google’s competitors are effectively non-existent unless 

they appear in the top few search results. 

Now we are dangerously close to the sort of arbitrary market definition exercise, 

devoid of the discipline imposed by economics, that identifies an 

anticompetitive problem by narrowing the market until every company is a 

monopolist over some small group of consumers. Indeed, one can always 

define a market by focusing on idiosyncratic preferences or product variations. 

Justice Fortas decried this type of analysis in his dissent in Grinnell (regarding 

home security systems), and it merits quoting at length: 

The trial court’s definition of the “product” market even more 

dramatically demonstrates that its action has been 

Procrustean—that it has tailored the market to the dimensions 

of the defendants. It recognizes that a person seeking 

protective services has many alternative sources. It lists 

“watchmen, watchdogs, automatic proprietary systems 

confined to one site, (often, but not always,) alarm systems 

connected with some local police or fire station, often 

unaccredited CSPS [central station protective services], and 

often accredited CSPS.” The court finds that even in the same 

city a single customer seeking protection for several premises 

may “exercise its option” differently for different locations. It 

may choose accredited CSPS for one of its locations and a 

different type of service for another.


But the court isolates from all of these alternatives only those 

services in which defendants engage. It eliminates all of the 

alternative sources despite its conscientious enumeration of 

them. Its definition of the “relevant market” is not merely 

confined to “central station” protective services, but to those 

central station protective services which are “accredited” by 

insurance companies. 

There is no pretense that these furnish peculiar services for 

which there is no alternative in the market place, on either a 

price or a functional basis. The court relies solely upon its finding that 

the services offered by accredited central stations are of better quality, and 

upon its conclusion that the insurance companies tend to give 

“noticeably larger” discounts to policyholders who use 

accredited central station protective services. This Court now 

approves this strange red-haired, bearded, one-eyed man-witha-limp 

classification. 39 

In Internet search as well, complainants imply a market based on the fact that 

Google offers “better quality” access to a larger set of Internet users than the 

myriad existing alternatives. But claiming essentiality based on a competitor’s 

relative high quality is deeply problematic. 

This point is of great importance in assessing the economics of the essential 

facilities doctrine generally and its application to Internet search in particular. It 

is clear, even under a fairly expansive reading of the essential facilities doctrine, 

that even a monopolist has no duty to subsidize the efforts of a less-effective 

rival. 40 Arguably the Aspen Skiing case should have been tossed out on this 

basis. As a practical matter, the Aspen Ski Company, by entering into a joint 

marketing agreement with its smaller rival, Aspen Highlands, allowed Highlands 

to take advantage of its markedly larger productivity (both in developing ski 

terrain and amenities, as well as marketing Aspen as a ski destination). Its 

subsequent decision to drop Highlands from its marketing program for failing 

to offer sufficient return on its investment should have been unobjectionable. 41 

Similarly, the explicit claim in cases brought against Google by its allegedlyforeclosed 

rivals is that these (relatively miniscule) sites should have access to 

Google’s effective and inexpensive marketing tool. But it is by no means clear 

that Google does or should have this duty to promote its rivals (without 

compensation to Google, as it happens). This is particularly true when, as 

discussed above, other modes of access exist for competitors’ activities, even if 

39 U.S. v. Grinnell Corp., 384 U.S. 563, 590-91 (1966) (emphasis added). 

40 See Olympia Equip. Leasing Co. v. Western Union Tel. Co., 797 F.2d 370 (7 th Cir. 1986). 

41 See KEITH N. HYLTON, ANTITRUST LAW: ECONOMIC THEORY AND COMMON LAW 

EVOLUTION 205-06 (2003).


these modes of access are of lower quality or higher cost. Particularly where, as 

here, the alleged bottleneck arises not out of a combination with another firm 

or firms but out of unilateral conduct (success in the marketplace), the claim 

that a superior access point among many (inferior) access points should be pried 

open for the benefit of its competitors is specious. 

It is worth noting that an alleged Google competitor, SourceTool, in the 

TradeComet complaint, 42 has made a version of this argument, alleging that 

Google once engaged in profitable commerce with SourceTool (by selling 

SourceTool ads next to Google search results) and then penalized SourceTool 

to its (Google’s) economic detriment. 43 The shape of this argument is a 

transparent effort to remain under what is left of the essential facilities doctrine 

following Trinko. But notice that even if it is true that Google intentionally 

ended a profitable arrangement with SourceTool (which is by no means clear), 

the claim still doesn’t pass muster. It is almost impossible that Google could be 

receiving less revenue from whatever site has replaced SourceTool in the paid 

search result spots SourceTool once paid for. As a result, even if Google were 

foregoing a previously-profitable relationship with SourceTool, it is not, in fact, 

suffering any economic harm because another advertiser has stepped into 

SourceTool’s shoes. 

Of course the argument that Google’s competitors are effectively absent 

without (guaranteed?) access to Google’s top few search results proves too 

much. There is a scarcity of “top few search results,” and any effective search 

engine must have the ability to ensure that those results are the most relevant 

possible, as well as that they do not violate various quality, safety, moral or 

other standards that the search engine chooses to promote. “Forcing [owners 

of essential facilities] to share access may not enhance consumer welfare.” 44 

Pure “neutrality” is neither possible nor desirable, and the exclusion of certain 

websites from these coveted positions should be deemed utterly unpersuasive in 

making out even a prima facie monopolization case against a search engine. 

And it is not even the case that SourceTool, Foundem, 45 and other competing 

websites are absent from Google; it is, however, sometimes the case that these 

42 Complaint , TradeComet.Com LLC v. Google, Inc., 693 F. Supp. 2d 370 (S.D.N.Y. 2009) (No. 

09Civ.1400(SHS)). 

43 Id. 8. 

44 KEITH N. HYLTON, ANTITRUST LAW: ECONOMIC THEORY AND COMMON LAW EVOLUTION 

208 (2003). 

45 Foundem is a “vertical search and price comparison” site in the UK. See 

www.foundem.co.uk. The company is at the heart of the “search neutrality” debate in 

Internet search. It has created a website to advocate its views on the neutrality issue at 

www.searchneutrality.org, and its claims are at the heart of the European Commission’s 

investigation of Google. See Foundem’s discussion of the EU action and its relationship to 

Foundem’s claims in European Commission Launches Antitrust Investigation of Google, SEARCH 

NEUTRALITY.ORG, Nov. 30, 2010, http://www.searchneutrality.org.


sites do not show up in the top few organic search results (and, often at the 

same time, Google’s own competing product search results do). But if access to 

the top few search results is required to ensure the requisite access sought by 

Google’s competitors, the relevant market has been narrowed considerably, 

creating a standard that can’t possibly be met, no matter how “neutral” a search 

engine’s results. 

Meanwhile, if Foundem were to disappear from the face of the Earth, who, 

other than its investors and employees (and perhaps their landlords), would be 

harmed? The implicit claim (if an antitrust case is to be made) is that websites 

like Foundem apply a constraint on Google’s ability to extract monopoly rents 

(presumably from advertisers). But this is a curious claim to make while 

simultaneously arguing that Google itself is made “better” (as in, searchers are 

indeed looking for Foundem in searches from which the site may be excluded) 

by the inclusion of Foundem in its search results (thus, presumably, increasing 

Google’s attractiveness to its users and thus its advertisers), while also claiming 

that Foundem would cease to exist without access to the top few Google search 

results. 

Google does not sell retail goods, and does not profit directly from its own 

product search offerings (which compete with Foundem), instead receiving 

benefit by increasing its customer base and the efficacy (presumably) of paid 

advertisements on its search pages that include a link to its own price 

comparison results. It is a remarkably tenuous claim to make that Google 

profits more by degrading its search results than by improving them. If the 

contrary claim is really true—if, that is, Google harms itself or its advertisers by 

intentionally penalizing competing sites like Foundem—then that argument and 

any evidence for it is absent from the current debate. And, of course, if Google 

is, as it claims, actually improving its product by applying qualitative decisions to 

demote sites like Foundem and others that, Google claims, merely re-publish 

information from elsewhere on the web with precious little original content, 

then Google’s efforts should be seen as a feature and not a bug. 

Moreover, the extension of the essential facilities logic to competition between 

Google and competitors like Foundem, MapQuest or Kayak is extremely 

problematic. To the extent that Google and Foundem, for example, are 

competitors, they are competitors not in the advertising space but rather in the 

“information dissemination and retail distribution channel” space. I’m not sure 

what else to call it. Foundem earns revenue by directing customers to retail 

sites to purchase goods. In this sense, Foundem acts like a shopping mall. 

Google does the same, only instead of receiving a cut from the sale, as 

Foundem does, Google sells advertisements. Thus, when Foundem complains 

about access to Google’s site, it is a competing channel of distribution, 

complaining that it needs access to its competitor’s distribution channel in order 

to compete. 

It’s a weird sort of complaint. It isn’t the same as the classic essential facilities 

sort of complaint where, to simplify, the owner of a vertically-integrated railroad


and rail transport company prevents access by other transport companies to its 

railroad line. Instead this would be like railroad company A arguing that 

railroad B must give A access to B’s tracks so A can sell access to those tracks 

to other rail transport companies. 

But even this doesn’t completely capture the audacity of the complaint, because 

for the analogy to hold, railroad A would actually be asking the court to force 

railroad B to put up a sign at the head of its tracks allowing railroad A to offer 

to trains already on B’s railroad the opportunity to jump off B’s railroad and 

start over again on A’s railroad that follows another route—but without 

knowing for sure if the route is better or worse until you jump onto A’s tracks. 

Something like that. Again, it’s weird. 

And note, of course, the problem that “at the head of the tracks” (as in 

something like “the first, second or third organic result”) is a problematic 

requirement as only three sites at any given time can occupy those spots—but 

there may be many more than three firms complaining of Google’s conduct 

and/or affected by the vagaries of its product design decisions. Or to keep with 

the shopping mall analogy, it’s like the owner of any of a number of small, new 

shopping malls requiring the owner of a large, established shopping mall to 

permit each of the new mall’s owners to set up a bus line to ferry shoppers to 

the new mall as they enter the established mall. Even where the established 

mall has a geographic, reputational and resource advantage, no one would argue 

that this access was essential to efficient commerce, and the cost to the 

successful incumbent would be manifestly too high. 

As discussed above, sites like Foundem do indeed have access to Google’s end 

users via any number of keywords on Google’s site. Type “UK price 

comparison site” into Google and a number of Google competitors come up 

including Foundem (and Google’s own price comparison site is seemingly 

absent). The claim thus becomes one that is either inappropriately aggregated 

(“for all search terms on average that may direct users to Foundem, Foundem is 

effectively denied access to the top search results”) or else overly narrow (“we 

prefer customers to find us by typing ‘Nikon camera’ into Google, not by typing 

‘price comparison Nikon camera’ into Google”). In any case, access is in fact 

available for these competitors, and “the indispensable requirement for 

invoking the [essential facilities] doctrine is the unavailability of access to the 

‘essential facilities’; where access exists, the doctrine serves no purpose.” 46 

Meanwhile, it is difficult to see how relevance (and thus efficiency) could be 

well-served by a neutrality principle that required a tool that reduces search costs 

to inherently increase those costs by directing searchers to a duplicate search on 

another site. If one is searching for a specific product and hoping to find price 

comparisons on Google, why on earth would that person be hoping to find not 

Google’s own efforts at price comparison, built right into its search engine, but 

46 Trinko, 540 U.S. at 411.


instead a link to another site that requires another several steps before finding 

the information? 

Seen this way, Google’s decision to promote its own price comparison results is 

a simple product pricing and design decision, protected by good sense and the 

Trinko decision (at least in the U.S.). Unlike the majority of its vertical search 

competitors and by design, Google makes no direct revenue from users clicking 

through to purchase anything from its shopping search results, and this allows it 

to offer a different (and, to many consumers, a significantly better) set of 

results. The page has paid search results only in small boxes at the top and 

bottom, the information is all algorithmically generated, and retailers do not pay 

to have their information on the page. For this product design—by definition 

of great value to users (in effect lowering the price to them of their product 

search)—to merit Google’s investment, it is necessary that its own, morerelevant 

and less-expensive results receive priority. If this is generating 

something of value for Google it is doing so only in the most salutary fashion: 

by offering additional resources for users to improve their “search experience” 

and thus induce them to use Google’s search engine. To require “neutrality” in 

this setting is to impair the site’s ability to design and price its own product. 

Even the Aspen Skiing decision didn’t go that far, requiring access to a joint 

marketing arrangement but not obligating Aspen Ski Company to alter its prices 

for skiers seeking to access only its own slopes. 

And the same analysis holds for assessments of Google’s other offerings (maps 

and videos, for example) that compete with other sites. Look for the nearest 

McDonalds in Google and a Google Map is bound to top the list (but not be 

the exclusive result, of course). But why should it be any other way? In effect, 

what Google does is give you the Web’s content in as accessible and appropriate 

a form as it can—design decisions that, Google must believe, increase quality 

and reduce effective price for its users. By offering not only a link to 

McDonalds’ web site, as well as various other links, but also a map showing the 

locations of the nearest restaurants, Google is offering up results in different 

forms, hoping that one is what the user is looking for. There is no economic 

justification for requiring a search engine in this setting to offer another site’s 

rather than its own simply because there happen to be other sites that do, 

indeed, offer such content (and would like cheaper access to consumers). 

Conclusion 

Search neutrality and forced access to Google’s results pages is based on the 

proposition that—Google’s users’ interests be damned—if Google is the easiest 

way competitors can get to potential users, Google must provide that access. 

The essential facilities doctrine, dealt a near-death blow by the Supreme Court 

in Trinko, has long been on the ropes. It should remain moribund here. On 

the one hand Google does not preclude, nor does it have the power to preclude, 

users from accessing competitors’ sites; all users need do is type 

“foundem.com” into their web browser—which works even if it’s Google’s


own Chrome browser! To the extent that Google can and does limit 

competitors’ access to its search results page, it is not controlling access to an 

“essential facility” in any sense other than Wal-Mart controls access to its own 

stores. “Google search results generated by its proprietary algorithm and found 

on its own web pages” do not constitute a market to which access should be 

forcibly granted by the courts or legislature. 

The set of claims that are adduced under the rubric of “search neutrality” or the 

“essential facilities doctrine” against Internet search engines in general and, as a 

practical matter, Google in particular, are deeply problematic. They risk 

encouraging courts and other decision makers to find antitrust violations where 

none actually exist, threatening to chill innovation and efficiency-enhancing 

conduct. In part for this reason, the essential facilities doctrine has been 

relegated by most antitrust experts to the dustbin of history. As Joshua Wright 

and I conclude elsewhere: 

Indeed, it is our view that in light of the antitrust claims arising 

out of innovative contractual and pricing conduct, and the 

apparent lack of any concrete evidence of anticompetitive 

effects or harm to competition, an enforcement action against 

Google on these grounds creates substantial risk for a “false 

positive” which would chill innovation and competition 

currently providing immense benefits to consumers. 47 

47 Geoffrey A. Manne & Joshua D. Wright, Google and the Limits of Antitrust: The Case Against the 

Case Against Google, 34 HARV. J. L. & PUB. POL’Y at 62.


Some Skepticism 

About Search Neutrality 

By James Grimmelmann * 

The perfect search engine would be like the mind of God. 1 

The God that holds you over the pit of hell, much as one holds a spider, or 

some loathsome insect, over the fire, abhors you, and is dreadfully 

provoked; his wrath towards you burns like fire; he looks upon you as 

worthy of nothing else, but to be cast into the fire … 2 

If God did not exist, it would be necessary to invent him. 3 

Search engines are attention lenses; they bring the online world into focus. 

They can redirect, reveal, magnify, and distort. They have immense power to 

help and to hide. We use them, to some extent, always at our own peril. And 

out of the many ways that search engines can cause harm, the thorniest 

problems of all stem from their ranking decisions. 4 

What makes ranking so problematic? Consider an example. The U.K. 

technology company Foundem offers “vertical search” 5—it helps users 

compare prices for electronics, books, and other goods. That makes it a Google 

competitor. 6 But in June 2006, Google applied a “penalty” to Foundem’s 

* Associate Professor of Law, New York Law School. I would like to thank Aislinn Black and 

Frank Pasquale for their comments. This essay is available for reuse under the Creative 

Commons Attribution 3.0 United States license, 

http://creativecommons.org/`licenses/by/3.0/us/. 

1 Charles Ferguson, What’s Next for Google, TECH. REV., Jan. 1, 2005, at 38, available at 

http://www.technologyreview.com/web/14065/ (quoting Sergey Brin, co-founder of 

Google). 

2 Jonathan Edwards, Sinners in the Hands of an Angry God (sermon delivered July 8, 1741 in 

Enfield, Connecticut), available in 22 WORKS OF JONATHAN EDWARDS 411 (Harry S. Stout & 

Nathan O. Hatch eds., Yale University Press 2003). 

3 Voltaire, Epître à l’auteur du livre des Trois imposteurs [Letter to the Author of The Three 

Impostors] (1768), available at http://www.whitman.edu/VSA/trois.imposteurs.html. 

4 See James Grimmelmann, The Structure of Search Engine Law, 93 IOWA L. REV. 1, 17–44 (2007) 

(identifying nine distinct types of harm search engines can cause to users, information 

providers, and third parties). 

5 See generally JOHN BATTELLE, THE SEARCH: HOW GOOGLE AND ITS RIVALS REWROTE THE 

RULES OF BUSINESS AND TRANSFORMED OUR CULTURE 274–76 (2005) (discussing “domainspecific 

search”). 

6 See Google Product Search Beta, GOOGLE, http://www.google.com/prdhp.


website, causing all of its pages to drop dramatically in Google’s rankings. 7 It 

took more than three years for Google to remove the penalty and restore 

Foundem to the first few pages of results for searches like “compare prices 

shoei xr-1000.” 8 Foundem’s traffic, and hence its business, dropped off 

dramatically as a result. The experience led Foundem’s co-founder, Adam Raff, 

to become an outspoken advocate: creating the site searchneutrality.org, 9 filing 

comments with the Federal Communications Commission (FCC), 10 and taking 

his story to the op-ed pages of The New York Times, 11 calling for legal protection 

for the Foundems of the world. 

Of course, the government doesn’t get involved every time a business is harmed 

by a bad ranking—or Consumer Reports would be out of business. 12 Instead, 

search-engine critics base their case for regulation on the immense power of 

search engines, which can “break the business of a Web site that is pushed 

down the rankings.” 13 They have the power to shape what millions of users, 

carrying out billions of searches a day, see. 14 At that scale, search engines are 

the new mass media 15—or perhaps the new meta media—capable of shaping 

public discourse itself. And while power itself may not be an evil, abuse of 

power is. 

Search-engine critics thus aim to keep search engines—although in the U.S. and 

much of the English-speaking world, it might be more accurate to say simply 

“Google” 16—from abusing their dominant position. The hard part comes in 

defining “abuse.” After a decade of various attempts, critics have hit on the 

7 See Foundem’s Google Story, SEARCHNEUTRALITY.ORG (Aug. 18, 2009), 

http://www.searchneutrality.org/foundem-google-story. 

8 The Shoei XR-1000 is a motorcycle helmet—according to Foundem, it’s £149.99 plus £10 

delivery from Helmet City. 

9 About, SEARCH NEUTRALITY.ORG, Oct. 9, 2009, http://www.searchneutrality.org/about. 

10 Reply Comments of Foundem, In the Matter of Preserving the Open Internet Broadband 

Industry Practices, GN Docket No. 09-191 (F..C.C). 

11 Adam Raff, Search, But You May Not Find, N.Y. TIMES, Dec. 27, 2009, at A27. 

12 Cf. Bose Corp. v. Consumers Union, 466 U.S. 485 (1984) (holding Consumer Reports not 

subject to product disparagement liability for negative review of Bose speaker). 

13 The Google Algorithm, N.Y. TIMES, July 14, 2010, at A30. 

14 See GRANT ESKELSEN ET AL, THE DIGITAL ECONOMY FACT BOOK 12–13 (10th ed. 2009), 

http://pff.org/issues-pubs/books/factbook_10th_Ed.pdf. 

15 See generally KEN AULETTA, GOOGLED: THE END OF THE WORLD AS WE KNOW IT (2009) 

(trying to understand Google by adopting the perspective of the media industry). Cf. Aaron 

Swartz, Googling for Sociopaths, RAW THOUGHT (Dec. 14, 2009), 

http://www.aaronsw.com/weblog/googled (describing Googled as “a history of [Google] 

as told by the incumbent sociopaths”). 

16 See ESKELSEN ET AL., FACT BOOK, supra note 14.


idea of “neutrality” as a governing principle. The idea is explicitly modeled on 

network neutrality, which would “forbid operators of broadband networks to 

discriminate against third-party applications, content or portals.” 17 Like 

broadband Internet service providers (ISPs), search engines “accumulate great 

power over the structure of online life.” 18 Thus, perhaps search engines should 

similarly be required not to discriminate among websites. 

For some academics, this idea is a thought experiment: a way to explore the 

implications of network neutrality ideas. 19 For others, it is a real proposal: a 

preliminary agenda for action. 20 Lawyers for ISPs fighting back against network 

neutrality have seized on it, either as a reductio ad absurdum or a way to kneecap 

their bitter rival Google. 21 Even the New York Times has gotten into the game, 

running an editorial calling for scrutiny of Google’s “editorial policy.” 22 Since 

New York Times editorials, as a rule, reflect no independent thought but only a 

kind of prevailing conventional wisdom, it is clear that search neutrality has 

truly arrived on the policy scene. 

Notwithstanding its sudden popularity, the case for search neutrality is a 

muddle. There is a fundamental misfit between its avowed policy goal of 

protecting users and most of the tests it proposes to protect them. Scratch 

beneath the surface of search neutrality and you will find that it would protect 

17 Barbara van Schewick, Towards an Economic Framework for Network Neutrality Regulation, 5 J. 

TELECOMM. & HIGH TECH. L. 329, 333 (2007), available at 


18 Frank Pasquale, Internet Nondiscrimination Principles: Commercial Ethics for Carries and Search 

Engines, 2008 U. CHI. LEGAL. FORUM 263, 298, available at 

http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1134159 [hereinafter Pasquale, 

Internet Nondiscrimination Principles]. 

19 See Mark R. Patterson, Non-Network Barriers to Network Neutrality, 78 FORDHAM L. REV. 2843 

(2010); Andrew Odlyzko, Network Neutrality, Search Neutrality, and the Never-ending Conflict 

Between Efficiency and Fairness in Markets, 8 REV. NETWORK ECON. 40 (2009). 

20 See DAWN C. NUNZIATO, VIRTUAL FREEDOM: NET NEUTRALITY AND FREE SPEECH IN THE 

INTERNET AGE (2009) [hereinafter NUNZIATO, VIRTUAL FREEDOM]; Pasquale, Internet 

Nondiscrimination Principles, supra note 18; Oren Bracha & Frank Pasquale, Federal Search 

Commission: Access, Fairness, and Accountability in the Law of Speech, 93 CORNELL L. REV. 1149 

(2008) [hereinafter Bracha & Pasquale, Federal Search Commission]; Jennifer A. Chandler, A 

Right to Reach an Audience: An Approach to Intermediary Bias on the Internet, 35 HOFSTRA L. REV. 

1095 (2007). 

21 Letter from Robert W. Quinn, Jr., Senior Vice President, AT&T, to Sharon Gillett, Chief, 

Wireline Competition Bureau Federal Communications Commission (Sept. 25, 2009), 

available at 

http://graphics8.nytimes.com/packages/pdf/technology/20090925_ATT- 

Letter.pdf. 

22 The Google Algorithm, supra note 13. But see Danny Sullivan, The New York Times Algorithm & 

Why It Needs Government Regulation, SEARCH ENGINE LAND (July 15, 2010) (parodying New 

York Times editorial on Google).


not search users, but websites. In the search space, however, websites are as 

often users’ enemies as not; the whole point of search is to help users avoid the 

sites they don’t want to see. 

In short, search neutrality’s ends and means don’t match. To explain why, I will 

deconstruct eight proposed search-neutrality principles: 

1. Equality: Search engines shouldn’t differentiate at all among websites. 

2. Objectivity: There are correct search results and incorrect ones, so search 

engines should return only the correct ones. 

3. Bias: Search engines should not distort the information landscape. 

4. Traffic: Websites that depend on a flow of visitors shouldn’t be cut off 

by search engines. 

5. Relevance: Search engines should maximize users’ satisfaction with 

search results. 

6. Self-interest: Search engines shouldn’t trade on their own account. 

7. Transparency: Search engines should disclose the algorithms they use to 

rank web pages. 

8. Manipulation: Search engines should rank sites only according to general 

rules, rather than promoting and demoting sites on an individual basis. 

As we shall see, all eight of these principles are unusable as bases for sound 

search regulation. 

I would like to be clear up front about the limits of my argument. Just because 

search neutrality is incoherent, it doesn’t follow that search engines deserve a 

free pass under antitrust, intellectual property, privacy, or other well-established 

bodies of law. 23 Nor is search-specific legal oversight out of the question. 

Search engines are capable of doing dastardly things: According to 

BusinessWeek, the Chinese search engine Baidu explicitly shakes down 

websites, demoting them in its rankings unless they buy ads. 24 It’s easy to tell 

horror stories about what search engines might do that are just plausible enough 

to be genuinely scary. 25 My argument is just that search neutrality, as currently 

proposed, is unlikely to be workable and quite likely to make things worse. It 

fails at its own goals, on its own definition of the problem. 

23 This essay is not the place for a full discussion of these issues (although we will meet 

antitrust and consumer protection law in passing). Grimmelmann, The Structure of Search 

Engine Law, supra note 4, provides a more detailed map. 

24 Chi-Chu Tschang, The Squeeze at China’s Baidu, BUSINESSWEEK, Dec. 31, 2008, 

http://www.businessweek.com/magazine/content/09_02/b4115021710265.htm 

(alleging that Baidu directly retaliates against sites that refuse to buy sponsored links by 

demoting them in its organic rankings). 

25 See, e.g., Cory Doctorow, Scroogled, http://craphound.com/scroogled.html; Tom Slee, Mr. 

Google’s Guidebook, WHIMSLEY (Mar. 7, 2008), 

http://whimsley.typepad.com/whimsley/2008/03/mr-googles-guid.html.


Theory 

Before delving into the specifics of search-neutrality proposals, it will help to 

understand the principles said to justify them. There are two broad types of 

arguments made to support search neutrality, one each focusing on users and 

on websites. A search engine that misuses its ranking power might be seen 

either as misleading users about what’s available online, or as blocking websites 

from reaching users. 26 Consider the arguments in turn. 

Users: Search helps people find the things they want and need. Good search 

results are better for them. And since search is both subjective and personal, 

users themselves are the ones who should define what makes search results 

good. The usual term for this goal is “relevance”: relevant results are the ones 

that users themselves are most satisfied with. 27 All else being equal, good search 

policy should try to maximize relevance. 

A libertarian might say that this goal is trivial. 28 Users are free to pick and 

choose among search engines and other informational tools. 29 They will 

naturally flock to the search engine that offers them the most relevant results; 

the market will provide just as much relevance as it is efficient to provide. 30 

There is no need for regulation; relevance, being demanded by users, will be 

26 Other arguments for search neutrality reduce to these two. Bracha and Pasquale, for 

example, are concerned about democracy. They want “an open and relatively equal chance 

to all members of society for participation in the cultural sphere.” Bracha & Pasquale, Federal 

Search Commission, supra note 20, at 1183–84. Search engines provide that chance if 

individuals can both find (as users) and be found (as websites) when they participate in 

politics and culture. Similarly, Bracha and Pasquale’s economic efficiency argument turns on 

users’ ability to find market information, id. at 1173–75. and their fairness concern speaks to 

websites’ losses of “audience or business,” id. at 1175–76. Whatever interest society has in 

search neutrality arises from users’ and websites’ interests in it—so we are justified in 

focusing our attention on users and websites. 

27 See BATTELLE, THE SEARCH, supra note 5, at 19–25. 

28 For a clear statement of a libertarian perspective on search neutrality, see Mike Masnick’s 

posts at Techdirt on the subject, collected at 

http:/www.techdirt.com/blog.php?tag=search+neutrality. Eric Goldman’s Search 

Engine Bias and the Demise of Search Engine Utopianism, 8 YALE J. L. & TECH. 188 (2006), makes a 

general case against the regulation of relevance on similar grounds. 

29 In Google’s words, “Competition is just one click away.” Adam Kovacevich, Google’s 

Approach to Competition, GOOGLE POLICY BLOG (May 8, 2009), 

http://googlepublicpolicy.blogspot.com/2009/05/googles-approach-tocompetition.html. 

30 See Eric Goldman, A Coasean Analysis of Marketing, 2006 WIS. L. REV. 1151.


supplied by search engines. And this is exactly what search engines themselves 

say: relevance is their principal, or only, goal. 31 

The response to this point of view—most carefully argued by Frank 

Pasquale 32—is best described as “liberal.” It focuses on maximizing the 

effective autonomy of search users, but questions whether market forces 

actually enable users to demand optimal relevance. For one thing, it questions 

whether users can actually detect deviations from relevance. 33 The user who 

turns to a search engine, by definition, doesn’t yet know what she’s looking for 

or where it is. Her own knowledge, therefore, doesn’t provide a fully reliable 

check on what the search engine shows her. The information she would need 

to know that the search engine is hiding something from her may be precisely 

the information it’s hiding from her—a relevant site that she didn’t know 

existed. 34 

Perhaps just as importantly, structural features of the search market can make it 

hard for users to discipline search engines by switching. Search-neutrality 

advocates have argued that search exhibits substantial barriers to entry. 35 The 

web is so big, and search algorithms so complex and refined, that there are 

substantial fixed costs to competing at all. 36 Moreover, the rise of personalized 

search both creates switching costs for individual users 37 and also makes it 

harder for them to share information about their experiences with multiple 

search engines. 38 

Websites: The case for protecting websites reaches back into free speech theory. 

Jerome Barron’s 1967 article, Access to the Press—A New First Amendment Right, 39 

31 See, e.g., Technology Overview, GOOGLE, www.google.com/corporate/tech.html; How Web 

Documents Are Ranked, YAHOO!, 

http://help.yahoo.com/l/us/yahoo/search/indexing/ranking-01.html; Ask Search 

Technology, ASK, http://sp.ask.com/en/docs/about/ask_technology.shtml. 

32 See Pasquale, Internet Nondiscrimination Principles, supra note 18; Bracha & Pasquale, Federal 

Search Commission, supra note 20; Frank Pasquale, Asterisk Revisited: Debating a Right of Reply on 

Search Results, 3 J. BUS. & TECH. L. 61 (2008); Frank Pasquale, Rankings, Reductionism, and 

Responsibility, 54 CLEV. ST. L. REV. 115 (2006). 

33 See Chandler, Right to Reach an Audience, supra note 20, at 1116; Patterson, Non-Network 

Barriers, supra note 19, at 2860-62. 

34 See Bracha & Pasquale, Federal Search Commission, supra note 20, at 1183–84. 

35 See id. at 1181–82. 

36 See id. at 1181. 

37 See Pasquale, Internet Nondiscrimination Principles, supra note 18, at 265. 

38 See Frank Pasquale, Could Personalized Search Ruin Your Life?, CONCURRING OPINIONS (Feb. 7, 

2008), 

http://www.concurringopinions.com/archives/2008/02/personalized_se.html. 

39 80 HARV. L. REV. 1641 (1967).


argued that freedom of speech is an empty right in a mass-media society unless 

one also has access to the mass media themselves. He thus argued that 

newspapers should be required to open their letters to the editor and their 

advertising to all points of view. 40 Although his proposed right of access is 

basically a dead letter as far as First Amendment doctrine goes, 41 it captured the 

imaginations of media-law scholars and media advocates. 42 

Scholars have begun to adapt Barron’s ideas to online intermediaries, including 

search engines. Dawn Nunziato’s book Virtual Freedom draws extensively on 

Barron to argue that Congress may need to “authorize the regulation of 

dominant search engines to require that they provide meaningful access to 

content.” 43 Jennifer Chandler applies Barron’s ideas to propose a “right to 

reach an audience” 44 that would give website owners various protections against 

exclusion 45 and demotion by search engines. 46 Similarly, Frank Pasquale 

suggests bringing “universal service” over into the search space, 47 perhaps 

through a government-provided search engine. 48 

The Barronian argument for access, however, needs to be qualified. The freespeech 

interest in access to search engine ranking placement is really audiences’ 

free speech interest; the real harm is that search users have been deprived of 

access to the speech of websites, not that websites have been deprived of access 

to users. Put another way, websites’ access interest is derivative of users’ 

interests. In the Supreme Court’s words, “The First Amendment protects the 

right of every citizen to ‘reach the minds of willing listeners.’” 49 Or, in Jerome 

40 Id. at 1667. 

41 See Miami Herald Pub’g Co. v. Tornillo, 418 U.S. 241 (1974) (striking down Florida law 

requiring newspapers to provide equal space for political responses). 

42 See, e.g., Reclaiming the First Amendment: Constitutional Theories of Media Reform, 35 HOFSTRA L. 

REV. 917–1582 (symposium issue collecting papers from conference honoring the 40 th 

anniversary of publication of Access to the Press). 

43 NUNZIATO, VIRTUAL FREEDOM, supra note 20, at 150. 

44 Chandler, Right to Reach an Audience, supra note 20, at 1103–17 (search engines), 1124-30 

(proposed right). 

45 Exclusion from a search index may sound like a bright-line category of abuse, but note that 

a demotion from, say, #1 to #58,610 will have the same effect. No one ever clicks through 

5861 pages of results. Thus, in practice, any rule against exclusion would also need to come 

with a—more problematic—rule against substantial demotions. 

46 Id. at 1117–18. 

47 Pasquale, Internet Nondiscrimination Principles, supra note 18, at 289–92. His example, which 

focuses on Google’s scans of books for its Book Search project, is interesting, but is 

“universal access” only in a loose, metaphorical sense. 

48 See Frank Pasquale, Dominant Search Engines: An Essential Cultural and Political Facility, infra 258. 

49 Hefron v. Int’l Soc. for Krishna Consciousness, Inc., 452 U.S. 640, 655 (1981) (quoting 

Kovacs v. Cooper, 336 U.S. 77, 87 (1949)) (emphasis added).


Barron’s, “[T]he point of ultimate interest is not the words of the speakers but 

the minds of the hearers.” 50 With these purposes in mind, let us turn to actual 

search-neutrality proposals. 

Equality 

Scott Cleland observes that Google’s “algorithm reportedly has over 1,000 

variables/discrimination biases which decide which content gets surfaced.” 51 

He concludes that “Google is not neutral” and thus should be subject to any 

FCC network-neutrality regulation. 52 On this view, a search engine does 

something wrong if it treats websites differently, “surfac[ing]” some, rather than 

others. This is a theory of neutrality as equality, it comes from the networkneutrality 

debates, and it is nonsensical as applied to search. 

Equality has a long pedigree in telecommunications. For years, common-carrier 

regulations required the AT&T system to offer its services on equal terms to 

anyone who wanted a phone. 53 This kind of equality is at the heart of proposed 

network neutrality regulations: treating all packets identically once they arrive at 

an ISP’s router, regardless of source or contents. 54 Whether or not equality in 

packet routing is a good idea as a technical matter, the rule itself is simple 

enough and relatively clear. One can, without difficulty, identify Comcast’s 

forging of packets to terminate BitTorrent connections as a violation of the 

principle. 55 As long as an ISP isn’t overloaded to the point of losing too many 

packets, equality does what it’s supposed to: ensures that every website enjoys 

access to the ISP’s network and customers. 

Try to apply this form of equality to search and the results are absurd. Of 

course Google differentiates among sites—that’s why we use it. Systematically 

favoring certain types of content over others isn’t a defect for a search engine— 

it’s the point. 56 If I search for “Machu Picchu pictures,” I want to see llamas in a 

50 Barron, Access to the Press, supra note 39, at 1653. 

51 Scott Cleland, Why Google Is Not Neutral, PRECURSOR BLOG (Nov. 4, 2009), 

http://precursorblog.com/content/why-google-is-not-neutral. 

52 Id. 

53 See generally JONATHAN E. NEUCHTERLEIN & PHILIP J. WEISER, DIGITAL CROSSROADS 45–68 

(2005). 

54 For an accessible introduction to the technical issues, see Edward W. Felten, The Nuts and 

Bolts of Network Neutrality (2006), http://itpolicy.princeton.edu/pub/neutrality.pdf. 

55 See In re Formal Compl. of Free Press & Public Knowledge Against Comcast Corp. for Secretly Degrading 

Peer-to-Peer Applications, WC Docket No. 07-52, Order, 23 F.C.C. Rcd. 13,028, 13,029–32 

(discussing blocking), 13,050–58 (finding that blocking violated federal policy) (2008), vacated, 

Comcast v. FCC, 600 F.3d 642 (D.C. Cir. 2010). 

56 See Karl Bode, Google Might Stop Violating “Search Neutrality”If Anybody Knew What That 

Actually Meant, TECHDIRT (May 7, 2010),


ruined city on a cloud-forest mountaintop, not horny housewives who whiten 

your teeth while you wait for them to refinance your mortgage. Search 

inevitably requires some form of editorial control. 57 A search engine cannot 

possibly treat all websites equally, not without turning into the phone book. But 

for that matter, even the phone book is not neutral in the sense of giving fully 

equal access to all comers, as the proliferation of AAA Locksmiths and Aabco 

Plumbers attests. Differentiating among websites, without something more, is 

not wrongful. 

Objectivity 

If search engines must make distinctions, perhaps we should insist that they 

make correct distinctions. Foundem, for example, argues that the Google 

penalty was unfair by pointing to positive write-ups of Foundem from “the 

UK’s leading technology television programme” and “the UK’s leading 

consumer body,” and to its high search ranks on Yahoo! and Bing. 58 The 

unvoiced assumption here is that search queries can have objectively right and 

wrong answers. A search on “James Grimmelmann blog” should come back 

with my weblog at http://laboratorium.net; anything else is a wrong answer. 

But this view of what search is and does is wrong. A search for “apple” could 

be looking for information about Fiji apples, Apple computers, or Fiona Apple. 

“bbs” could refer to airgun pellets, bulletin-board systems, or bed-andbreakfasts. 

Different people will have different intentions in mind; even the 

same person will have different intentions at different times. Sergey Brin’s 

theological comparison of perfect search to the “mind of God” 59 shows us why 

perfect search is impossible. Not even Google is—or ever could be— 

omniscient. The search query itself is necessarily an incomplete basis on which 

to guess at possible results. 60 

The objective view of search, then, fails for two related reasons. First, search 

users are profoundly diverse. They have highly personal, highly contextual 

goals. One size cannot fit all. And second, a search engine’s job always 

involves guesswork. 61 Some guesses are better than others, but the search 

http://www.techdirt.com/articles/20100504/1324279300.shtml (“[T]he entire purpose 

of search is to discriminate and point the user toward more pertinent results.”). 

57 See Goldman, Search Engine Bias, supra note 28, at 115–18. 

58 Foundem’s Google Story, supra note 7. 

59 Supra note 1. 

60 See generally ALEX HALAVAIS, THE SEARCH-ENGINE SOCIETY 32–55 (2009) (discussing 

difficulties of ascertaining meaning in search process). 

61 See Eric Goldman, Deregulating Relevancy in Internet Trademark Law, 54 EMORY L.J. 507, 521–28 

(2005).


engine will always have to guess. “James Grimmelmann blog” shouldn’t take 

users to Toyota’s corporate page—but perhaps they were interested in my 

guest-blogging at Concurring Opinions, or in blogs about me, or they have me 

mixed up with Eric Goldman and were actually looking for his blog. Time 

Warner Cable’s complaint that “significant components of [Google’s] Ad Rank 

scheme are subjective” 62 is beside the point. Search itself is subjective. 63 

Few scholars go so far as to advocate explicit re-ranking to correct search 

results. 64 But even those who acknowledge that search is subjective sometimes 

write as though it were not. Frank Pasquale gives a hypothetical in which 

“YouTube’s results always appear as the first thirty [Google] results in response 

to certain video queries for which [a rival video site] has demonstrably more 

relevant content.” 65 One might ask, “demonstrably more relevant” by what 

standard? Often the answer will be contentious. 

In Foundem’s case, what difference should it make that Yahoo! and others liked 

Foundem? So? That’s their opinion. Google had a different one. Who is to 

say that Yahoo! was right and Google was wrong? 66 One could equally well 

argue that Google’s low ranking was correct and Yahoo!’s high ranking was the 

mistake. “compare prices shoei xr-1000” is not the sort of question that admits 

62 Comments of Time Warner Cable Inc. 77, In the Matter of Preserving the Open Internet 

Broadband Industry Practices, GN Docket No. 09-191 (F.C.C. comments filed Jan. 14, 

2010). 

63 See Goldman, Search Engine Bias, supra note 28, at 112–13. This point should not be 

confused with a considered opinion on the question of how the First Amendment applies to 

search-ranking decisions. Search engines make editorial judgments about relevance, but they 

also present information that can only be described as factual (such as maps and addresses), 

extol their objectivity in marketing statements, and are perceived by users as having an aura 

of reliability. It is possible to make false statements even when speaking subjectively—for 

example, I would be lying to you if I said that I enjoy eating scallops. The fact that search 

engines’ judgments are expressed algorithmically, including in ways not contemplated by 

their programmers, complicates the analysis even further. The definitive First Amendment 

analysis of search-engine speech has yet to be written. Academic contributions to that 

conversation include Goldman, Search Engine Bias, supra note 28, at 112–15; Bracha & 

Pasquale, Federal Search Commission, supra note 20, at 1188–1201; Pasquale, Asterisk Revisited, 

supra note 32, at 68–85; NUNZIATO, VIRTUAL FREEDOM, supra note 20, passim (and 

particularly pages 149–51); Chandler, Right to Reach an Audience, supra note 20, at 1124–29; 

James Grimmelmann, The Google Dilemma, 53 N.Y.L.S. L. REV. 939, 946 (2009); 

Grimmelmann, The Structure of Search Engine Law, supra note 4, at 58–60. Some leading cases 

are listed in note 85, infra. 

64 But see Sandeep Pandey et al., Shuffling a Stacked Deck: The Case for Partially Randomized Search 

Results, PROC. 31ST VERY LARGE DATABASES CONF. 781 (2005) (arguing for randomization in 

search results to promote obscure websites). 

65 Pasquale, Internet Nondiscriminaion Principles, supra note 18, at 296. 

66 Cf. Rebecca Tushnet, It Depends on What the Meaning of “False” Is, Falsity and Misleadingness in 

Commercial Speech Doctrine, 41 LOYOLA L.A. L. REV. 101 (2008) (arguing that judgments about 

falsity frequently embody contested social policies).


of a right answer. This is why it doesn’t help to say that the Foundem vote is 

four-to-one against Google. If deviation from the majority opinion makes a 

search engine wrong, then so much for search engine innovation—and so much 

for unpopular views. 67 

Bias 

Ironically, it is the goal of protecting unpopular views that drives the concern 

with search engine “bias.” Lucas Introna and Helen Nissenbaum, for example, 

are concerned that search engines will direct users to sites that are already 

popular and away from obscure sites. 68 Alex Halavais calls for “resistance to 

the homogenizing process of major search engines,” 69 including governmental 

interventions. 70 These are structural concerns with popularity-based search. 

Others worry about more particular biases. AT&T complains that “Google’s 

algorithms unquestionably do favor some companies or sites.” 71 Scott Cleland 

objects that Google demotes content from other countries in its countryspecific 

search pages. 72 

The point that a technological system can display bias is one of those profound 

observations that is at once both startling and obvious. 73 It naturally leads to 

the question of whether, when, and how one could correct for the bias search 

engines introduce. 74 But to pull that off, one must have a working 

understanding of what constitutes search-engine bias. Batya Friedman and 

Helen Nissenbaum define a computer system to be “biased” if it “systematically 

and unfairly discriminates against certain individuals or groups of individuals in 

favor of others.” 75 Since search engines systematically discriminate by design, 

67 This last point should be especially troubling to Barron-inspired advocates of “access,” since 

the point of such a regime is to promote opinions that are not widely shared. 

68 Lucas D. Introna & Helen Nissenbaum, Shaping the Web: Why the Politics of Search Engines 

Matters, 16 INFO. SOC. 169, 175 (2000). 

69 HALAVAIS, SEARCH ENGINE SOCIETY, supra note 60, at 106. 

70 Id. at 132–38. 

71 Comments of AT&T Inc. 102, In the Matter of Preserving the Open Internet Broadband 

Industry Practices, GN Docket No. 09-191 (F.C.C. comments filed Jan. 14, 2010). 

72 Cleland, Why Google Is Not Neutral, supra note 51. 

73 In Landgon Winner’s phrase, “artifacts have politics.” LANGDON WINNER, THE WHALE 

AND THE REACTOR: A SEARCH FOR LIMITS IN AN AGE OF HIGH TECHNOLOGY19 (University 

of Chicago Press 1986). 

74 See, e.g., Pandey et al, Shuffling a Stacked Deck, supra note 64. 

75 Batya Friedman & Helen Nissenbaum, Bias in Computer Systems, 14 ACM TRANS. ON 

COMPUTER SYS. 330, 332 (1996). See also Alejandro M. Diaz, Through the Google Goggles: 

Sociopolitical Bias in Search Engine Design (May 23, 2005) (unpublished B.A. thesis, 

Stanford University), available at 

http://epl.scu.edu/~stsvalues/readings/Diaz_thesis_final.pdf.


all of the heavy lifting in the definition is done by the word “unfair.” But this 

just kicks the problem down the road. One still must explain when 

discrimination is “unfair” and when it is not. Friedman and Nissenbaum’s 

discussion is enlightening, but does not by itself help us identify which practices 

are abusive. 76 

The point that socio-technical systems have embedded biases also cuts against 

search neutrality. We should not assume that if only the search engine could be 

made properly neutral, the search results would be free of bias. Every search 

result requires both a user to contribute a search query, and websites to 

contribute the content to be ranked. Neither users nor websites are passive 

participants; both can be wildly, profoundly biased. 

On the website side, the web is anything but neutral. 77 Websites compete 

fiercely, and not always ethically, for readers. 78 It doesn’t matter what the search 

engine algorithm is; websites will try to game it. Search-engine optimization, or 

SEO, is as much a fixture of the Internet as spam. Link farms, 79 spam blog 

comments, hacked websites—you name it, and they’ll try it, all in the name of 

improving their search rankings. A fully invisible search engine, one that 

introduced no new values or biases of its own, would merely replicate the 

underlying biases of the web itself: 80 heavily commercial, and subject to a truly 

mindboggling quantity of spam. Raff says that search algorithms should be 

“comprehensive.” 81 But should users be subjected to a comprehensive 

presentation of discount Canadian pharmaceutical sites? 

On the user side, sometimes the bias is between the keyboard and the chair. 

Fully de-biasing search results would also require de-biasing search queries— 

and users’ ability to pick which results they click on. Take a search for “jew,” 

for example. Google has been criticized both for returning anti-Semitic sites (to 

76 If one fears, with Bracha and Pasquale, that “a handful of powerful gatekeepers” wield 

disproportionate influence, then the solution is simple: break up the bastards. If they 

reassemble or reacquire too much power, do it again. Neutrality will always be an imperfect 

half-measure if power itself is the problem. 

77 See Clay Shirky, Power Laws, Weblogs, and Inequality, SHIRKY.COM (Feb. 8, 2003), 

http://www.shirky.com/writings/powerlaw_weblog.html (discussing vast 

disproportion of prominence between famous and obscure weblogs). 

78 See IAN H. WITTEN ET AL., WEB DRAGONS: INSIDE THE MYTHS OF SEARCH TECHNOLOGY 

145–75 (Morgan Kaufmann Publishers 2007). 

79 A link farm is a group of automatically generated web sites that heavily link to each other. 

The point is to trick a popularity-based search engine into believing that all of the sites in the 

group are popular. See Grimmelmann, The Google Dilemma, supra note 63, at 946, 

80 See Patterson, Non-Network Barriers, supra note 19, at 2854–55. 

81 Raff, Search, But You May Not Find, supra note 11.


American users) and for not returning such sites (to German users). 82 The 

inescapable issue is that Google has users who want to read anti-Semitic web 

pages and users who don’t. One might call some of those users “biased,” but if 

they are, it’s not Google’s fault. 

Some bias is going to leak through as long as search engines help users find 

what they want. And helping users find what they want is such a profound 

social good that one should be skeptical of trying to inhibit it. 83 Telling users 

what they should see is a serious intrusion on personal autonomy, and thus 

deeply inconsistent with the liberal argument for search neutrality. If you want 

Google to steer users to websites with views that differ from their own, 84 your 

goal is not properly described as search neutrality. In effect, you have gone back 

to asserting the objective correctness of search results: Certain sites are good for 

users, like whole grains. 

Traffic 

The most common trope in the search debates is the website whose traffic 

vanishes overnight when it disappears from Google’s search results. 85 Because 

so much traffic flows through Google, it holds websites over the flames of 

website hell, ready at any instant to let them fall in the rankings. Chandler’s 

proposed right to reach an audience and Foundem’s proposed “effective, 

accessible, and transparent appeal process” 86 attempt to protect websites from 

82 See Grimmelmann, The Google Dilemma, supra note 63, at 943–45. 

83 See James Grimmelmann, Don’t Censor Search, 117 YALE L.J. POCKET PART 48 (2007). 

84 See generally CASS SUNSTEIN, REPUBLIC.COM 2.0 (Princeton University Press 2007). 

85 See, e.g., BATTELLE, THE SEARCH, supra note 5, at 153–59 (2bigfeet.com, main index); 

NUNZIATO, VIRTUAL FREEDOM, supra note 20, at 14–17 (various sites, AdWords and Google 

News); Chandler, Right to Reach an Audience, supra note 20, at 1110 (BMW Germany and 

Ricoh Germany, main index); Michael Y. Park, Journalist Who Exposes U.N. Corruption 

Disappears from Google, FOX NEWS, Feb. 18, 2008, 

http://www.foxnews.com/story/0,2933,331106,00.html (Inner City Press, Google 

News); Cleland, Why Google Is Not Neutral, supra note 51 (ExtremeTech.com and 

Fotolog.com, AdWords); Dan Mcsai, G-Railed: Why Did Google Bury the Web’s Oldest 

Entertainment Publication?, FASTCOMPANY.COM (Dec. 2, 2009), 

http://www.fastcompany.com/blog/dan-macsai/popwise/why-did-neutral-googlede-list-webs-oldest-entertainment-publication 

(Studio Briefing, AdWords and main 

index); Foundem’s Google Story, supra note 7 (Foundem, main index). Opinions in lawsuits 

challenging demotions or exclusions include Langdon v. Google Inc., 474 F. Supp. 622 (D. 

Del. 2007) (NCJusticeFraud.com and ChinaIsEvil.com, AdWords); Kinderstart.com LLC v. 

Google, Inc., No. C 06-2057 JF (RS), 2006 U.S. Dist. LEXIS 82481 (N.D. Cal. July 13, 2006) 

(Kinderstart.com, main index) and Search King, Inc. v. Google Tech., Inc., No. CIV-02- 

1457-M, 2003 U.S. Dist. LEXIS 27193 (W.D. Okla. 2003) (SearcchKing.com, main index). 

86 Foundem’s Google Story, supra note 7.


being dropped. Dawn Nunzatio, for her part, would require search engines to 

open their sponsored links to political candidates. 87 

A right to continued customer traffic would be a legal anomaly; offline 

businesses enjoy no such right. Some Manhattanites who take the free IKEA 

ferry to its store in Brooklyn eat at the nearby food trucks in the Red Hook Ball 

Fields. 88 The food truck owners would have no right to complain if IKEA 

discontinued the ferry or moved its store. Search neutrality advocates, however, 

would say that RedHookFoodTruck.com has a Jerome Barron-style free-speech 

interest in having access to the search engine’s result pages, and thus has more 

right to complain if the Google ferry no longer comes to its neighborhood. 89 

But, as we saw above, this is really an argument that users have a relevance interest 

in seeing the site. If no one actually wants to visit RedHookFoodTruck.com, 

then its owner shouldn’t be heard to complain about her poor search ranking. 

When push comes to shove, search neutrality advocates recognize that websites 

must plead their case in terms of users’ needs. Chandler’s modern right of 

access is a “right to reach a willing audience,” 90 which she describes as “the right 

to be free of the imposition of discriminatory filters that the listener would not 

otherwise have used.” 91 Even Foundem’s Adam Raff presents his actual searchneutrality 

principle in user-protective terms: “search engines should have no 

editorial policies other than that their results be comprehensive, impartial and 

based solely on relevance.” 92 Relevance is, of course, the touchstone of users’ 

interests, not websites’. 

Indeed, looking at the rankings from a website’s perspective, rather than from 

users’, can be counterproductive to free-speech values. If users really find other 

websites more relevant, then making them visit RedHookFoodTruck.com 

impinges on their autonomy and on their free speech interests as listeners. For 

any given search query, there may be dozens, hundreds, thousands of 

competing websites. The vast majority of them will thus have interests that 

diverge from users’—and every incentive to override users’ wishes. 

87 NUNZIATO, VIRTUAL FREEDOM, supra note 20, at 150–51. 

88 See Adam Kuban, Red Hook Vendors: A Quick Guide for the Uninitiated, SERIOUS EATS (July 18, 

2008), http://newyork.seriouseats.com/2008/07/red-hook-vendors-soccer-tacosguide-how-to-get-there-what-to-eat.html. 

89 See Chandler, Right to Reach an Audience, supra note 20; NUNZIATO, VIRTUAL FREEDOM, supra 

note 20. 

90 Chandler, Right to Reach an Audience, supra note 20, at 1099 (emphasis added). 

91 Id. at 1103 (emphasis added). 

92 Raff, Search, But You May Not Find, supra note 11 (emphasis added).


Even when users are genuinely indifferent among various websites, some search 

neutrality advocates think websites should be protected from “arbitrary” or 

“unaccountable” ranking changes as a matter of fairness. 93 We should call the 

websites that currently sit at the top of search engine rankings by their proper 

name—incumbents—and we should look as skeptically on their demands to 

remain in power as we would on any other incumbent’s. The search engine that 

ranks a site highly has conferred a benefit on it; turning that gratuitous benefit 

into a permanent entitlement gets the ethics of the situation exactly backwards. 

Indeed, giving highly-ranked websites what is in effect a property right in search 

rankings runs counter to everything we know about how to hand out property 

rights. Websites don’t create the rankings; search engines do. Similarly, search 

engines are in a better position to manage rankings and prevent waste. And if 

each individual search ranking came with a right to placement, every searchresults 

page would be an anti-commons in the making. 94 

Thus, it is irrelevant that Foundem had a prominent search placement on Google 

before it landed in the doghouse. Just as the subjectivity of search means that 

search engines will frequently disagree with each other, it also means that a 

search engine will disagree with itself over time. From the outside looking in, 

we have no basis to say whether the initial high ranking or the subsequent low 

ranking made more sense. To give Foundem—and every other website 

currently enjoying a good search ranking—the right to continue where it is 

would lock in search results for all time, obliterating search-engine 

experimentation and improvement. 

Relevance 

Given the importance of user autonomy to search-neutrality theory, relevance is 

a natural choice for a neutrality principle. In Foundem’s words, search results 

should be “based solely on relevance.” 95 Chandler proposes a rule against 

“discrimination that listeners would not have chosen.” 96 Bracha and Pasquale 

decry “search engines [that] highlight or suppress critical information” and 

thereby “shape and constrain [users’] choices”—that is, hide information that 

users would have found relevant. 97 

93 Bracha & Pasquale, Federal Search Commission, supra note 20, at 1175–76. 

94 See generally Michael Heller, The Tragedy of the Anticommons, 111 HARV. L. REV. 621 (1998) 

(arguing that when too many owners have exclusion rights over a resource, it is prone to 

underuse). 

95 Search Neutrality, SEARCH NEUTRALITY.ORG (Oct. 11, 2009), 

http://www.searchneutrality.org/. 

96 Chandler, Right to Reach an Audience, supra note 20, at 1098. 

97 Bracha & Pasquale, Federal Search Commission, supra note 20, at 1177.


Relevance, however, is such an obvious good that its virtue verges on the 

tautological. Search engines compete to give users relevant results; they exist at all 

only because they do. Telling a search engine to be more relevant is like telling 

a boxer to punch harder. Of course, sometimes boxers do throw fights, so it 

isn’t out of the question that a search engine might underplay its hand. How, 

though, could regulators tell? Regulators can’t declare a result “relevant” 

without expressing a view as to why other possibilities are “irrelevant,” and that 

is almost always going to be contested. 

Here’s an example: Foundem. Recall that Foundem is a “vertical search site” 

that specializes in consumer goods. Well, a great many vertical search sites are 

worthless. (If you don’t believe me, please try using a few for a bit.) Like other 

kinds of sites that simply roll up existing content and slap some of their own 

ads on it—Wikipedia clones and local business directories also come to mind— 

they superficially resemble legitimate sites that provide something of value to 

users. 98 But only superficially. The “penalties” that reduce vertical search sites’ 

Google ranks aren’t an attempt to reduce competition at the expense of 

relevance; they’re an attempt to implement relevance. 99 There are a few relatively 

good, usable product-search sites, but most of them are junk and good riddance 

to them. You’re welcome to disagree—search is subjective—but I’d rather have 

the anti-vertical penalty in place than not. Those who would argue that 

Google’s rankings don’t reflect relevance have a heavy burden of proof, in the 

face of ample, easily verified evidence to the contrary. 

In fact, behind almost every well-known story of search engine caprice, there is 

a more persuasive relevance-enhancing counter-story. For example, 

SourceTool, another vertical search engine, has sued Google under antitrust law 

for, in effect, demoting it in Google’s rankings for search ads. 100 SourceTool, 

though, is a “directory” with a taxonomic logic of dubious utility—the United 

Nations Standard Products and Services Code—and almost no content of its 

own. It’s the rare user indeed who will find SourceTool relevant. If you care 

about relevance and user autonomy, you should applaud Google’s decision to 

demote SourceTool. 

98 See Chris Lake, Foundem vs Google: A Case Study in SEO Fail, ECONSULTANCY (Aug. 18, 2009), 

http://econsultancy.com/blog/4456-foundem-vs-google-a-case-study-in-seo-fail; 

Little or No Original Content, GOOGLE WEBMASTER CENTRAL (updated June 10, 2009), 

www.google.com/support/webmasters/bin/answer.py?answer=66361. 

99 See John Lettice, When Algorithms Attack, Does Google Hear You Scream?, THE REGISTER (Nov. 

19, 2009), http://www.theregister.co.uk/2009/11/19/google_hand_of_god/. 

100 See TradeComet.com LLC v. Google Inc., No. 09–CIV-1400 (S.D.N.Y. complaint filed Feb. 

17, 2009). The District Court dismissed the case on the basis of the forum-selection clause 

in Google’s advertiser agreement, without reaching the merits of the case. See 

TradeComet.com LLC v. Google Inc., 693 F. Supp. 2d 370 (S.D.N.Y. 2010).


Self-Interest 

In practice, even as search-neutrality advocates claim “relevance” as their goal, 

they rely on proxies for it. The most common is self-interest. A Consumer 

Watchdog report accuses Google of “an abandonment of [its] pledge to provide 

neutral search capability” by “steering Internet searchers to its own services” to 

“muscle its way into new markets.” 101 Foundem alleges that Google demotes it 

and other vertical search sites to fend off competition, and alleges that Google’s 

links to itself give it “an unassailable competitive advantage.” 102 Bracha and 

Pasquale worry that search engines can change their rankings “in response to 

positive or negative inducements from other parties.” 103 

Bad motive may lead to bad relevance, but it’s also a bad proxy for it. The first 

problem is evidentiary. By definition, motivations are interior, personal. 104 Of 

course, the law has to guess at motives all the time, but the task is by its nature 

harder than looking to extrinsic evidence. People get it wrong all the time. In 

2009, an Amazon employee with a fat finger hit a wrong button and categorized 

tens of thousands of gay-themed books as “adult.” 105 An angry mob of 

Netizens assumed the company had deliberately pulled the books from its 

search engine out of anti-gay animus, and used the Twitter hashtag #amazonfail 

to express their very public outrage. 106 Amazon’s reclassification was a mistake 

(a quickly corrected one), and a vivid demonstration of the power of search 

algorithms—but not a case of bad motives. 107 

In all but the most blatant of cases, in fact, a search engine will be able to tell a 

plausible relevance story about its ranking decisions.Proving that a relevance 

story is pretextual will be extraordinarily difficult, in view of the complexity and 

subjectivity of search. But it would also be disastrous to adopt the opposite 

point of view and presume pretext. The absence of bad motive is a negative 

that it will often be impossible for the search engine to prove. How can it 

101 TRAFFIC REPORT: HOW GOOGLE IS SQUEEZING OUT COMPETITORS AND MUSCLING INTO 

NEW MARKETS (Consumer Watchdog 2010), 

http://www.consumerwatchdog.org/resources/TrafficStudy-Google.pdf. 

102 Reply Comments of Foundem, supra note 10, at 1. 

103 Bracha & Pasquale, Federal Search Commission, supra note 20, at 1170. 

104 As an artificial corporate entity, a search engine may not even have motives other than the 

ones the law attributes to it. 

105 See Nick Eaton, AmazonFail: An Inside Look at What Happened, AMAZON & THE ONLINE 

RETAIL BLOG (Apr. 13, 2009), http://blog.seattlepi.com/amazon/archives/166384.asp. 

106 See Clay Shirky, The Failure of #amazonfail, SHIRKY.COM (Apr. 15, 2009), 

http://www.shirky.com/weblog/2009/04/the-failure-of-amazonfail/. 

107 But see Mary Hodder, Why Amazon Didn’t Just Have a Glitch, TECHCRUNCH (Apr. 14, 2009), 

http://techcrunch.com/2009/04/14/guest-post-why-amazon-didnt-just-have-aglitch/.


establish, for example, that the engineer who added the anti-vertical penalty 

didn’t have a lunchroom conversation with an executive who played up the 

competition angle? This is not to say that serious cases of abuse are 

implausible, 108 just that investigation will be unusually hard and that false 

positives will be dangerously frequent. 

There is a nontrivial antitrust issue lurking here. In the United States, Google 

has a dominant market share in both search and search advertising, and one 

could argue that Google has started to leverage its position in anticompetitive 

ways. 109 Antitrust, however approaches such questions with a well-developed 

analytical toolkit: relevant markets, market power, pro-competitive and anticompetitive 

effects, and so on. 110 Antitrust rightly focuses on the effects of 

business practices on consumers; search neutrality should not short-circuit that 

consumer-centric analysis by overemphasizing the role of a search engine’s 

motives. Some things can be good for Google and good for its users. 

Thus, when Google links to its own products, not only can there be substantial 

technical benefits from integration, but often Google is helping users by 

pointing them to services that really are better than the competition. Consumer 

Watchdog, for example, cries foul that Google “put its own [map] service atop 

all others for generic address searches,” 111 and that Google Maps has taken half 

of the local search market at the expense of previously dominant MapQuest and 

Yahoo! Maps. 112 But perhaps MapQuest and Yahoo! Maps deserved to lose. 

Google Maps was groundbreaking when launched, and years later, it remains 

one of the best-implemented services on the Internet, with astonishingly clever 

scripting, flexible route-finding, and a powerful application programming 

interface (API). 113 

108 Baidu’s alleged shakedown (see supra note 24 and accompanying text), if true, would be an 

example. Willingness to buy Baidu search ads is not in itself a reliable indicator of relevance 

to Baidu searchers. But then again, even pay-for-placement was once considered a plausible 

model for main-column search results—and willingness to pay is not inherently a crazy 

proxy for relevance. See BATTELLE, THE SEARCH, supra note 5, at 104–14 (discussing GoTo’s 

pay-for-placement model). See also Goldman, Coasean Analysis, supra note 30 (envisioning a 

future in which advertisers and users negotiate over access to users’ attention). Indeed, 

search ads today are sold on an auction-based basis. They’re often as relevant as maincolumn 

search results, sometimes more so. It might be better to say that Baidu’s real 

problems are monopoly pricing and (compulsory) stealth marketing. 

109 See, e.g., Brad Stone, Sure, It’s Big, But Is That Bad?, N.Y. TIMES, May 21, 2010, at BU1. 

110 See generally Geoffrey A. Manne & Joshua D. Wright, Google and the Limits of Antitrust: The 

Case Against the Antitrust Case Against Google, HARV. J. L. & PUB. POL’Y. (forthcoming). 

111 TRAFFIC REPORT, supra note 101, at 5. 

112 Id. at 5–7. 

113 See, e.g., John Carroll, Google Maps and Innovation, A DEVELOPER’S VIEW (Oct. 12, 2005), 

http://www.zdnet.com/blog/carroll/google-maps-and-innovation/1499.


One form of self-interest that may be well-enough defined to justify regulatory 

scrutiny is the straightforward bribe: a payment from a website to change its 

ranking, or a competitor’s. Search-engine critics argue that search engines 

should disclose commercial relationships that bear on their ranking decisions. 114 

This is a standard, sensible policy response to the fear of stealth marketing. 115 

Indeed, the Federal Trade Commission (FTC) has specifically warned search 

engines not to mix their organic and paid search results. 116 More generally, the 

FTC endorsement guidelines provide that endorsements must “reflect the 

honest opinions, findings, beliefs, or experience of the endorser” 117 and that any 

connections between endorser and seller that “might materially affect the weight 

or credibility of the endorsement” 118 must be fully disclosed. These policies 

have a natural application to search engines. A search engine that factors 

payments from sponsors into its ranking decisions is lying to its users unless it 

discloses those relationships, and this sort of lie would trigger the FTC’s 

jurisdiction. 119 This isn’t a neutrality principle, or even unique to search; it’s just 

a natural application of a well-established legal norm. 

Transparency 

Search-engine critics generally go further and argue that search engines should 

also be required to disclose their algorithms in detail: 

� Introna and Nissenbaum: “As a first step we would demand full and 

truthful disclosure of the underlying rules (or algorithms) governing 

indexing, searching, and prioritizing, stated in a way that is meaningful 

to the majority of web users.” 120 

� Foundem: “Search Neutrality can be defined as the principle that 

search engines should be open and transparent about their editorial 

policies … .” 121 

114 See, e.g., Pasquale, Internet Nondiscrimination Principles, supra note 18, at 286. 

115 See generally Ellen P. Goodman, Stealth Marketing and Editorial Integrity, 85 TEX. L. REV. 

83 (2006). 

116 See Letter from Heather Hippsley, Acting Assoc. Dir., Div. of Adver. Practices, Fed. Trade 

Comm., to Gary Ruskin, Executive Dir. at Commercial Alert (June 27, 2007), available at 

http://www.ftc.gov/os/closings/staff/commercialalertletter.shtm. 

117 16 CFR § 255.1(a). 

118 16 CFR § 255.5. 

119 Disclosure in common cases need not be onerous. Where, for example, a search engine 

auctions off sponsored links on its results pages, telling users that those links are auctioned 

off should generally suffice. See generally Letter from Heather Hippsley, supra note 116. 

120 Introna and Nissenbaum, supra note 68, at 181. 

121 Search Neutrality, SEARCHNEUTRALITY.ORG (Oct. 11, 2009), 

http://www.searchneutrality.org/search-neutrality.


� Pasquale: “[Dominant search engines] should submit to regulation that 

bans stealth marketing and reliably verifies the absence of the practice.” 122 

These disclosures are meant to inform users about what they’re getting from a 

search engine (Introna and Nissenbaum), to inform websites about the 

standards they’re being judged by (Foundem), 123 or to inform regulators about 

what the search engine is actually doing (Pasquale). 124 

Algorithmic transparency is a delicate business. Full disclosure of the algorithm 

itself runs up against critical interests of the search engine. A fully public 

algorithm is one that the search engine’s competitors can copy wholesale. 125 

Worse, it is one that websites can use to create highly optimized search-engine 

spam. 126 Writing in 2000, long before the full extent of search-engine spam was 

as clear as it is today, Introna and Nissenbaum thought that the “impact of 

these unethical practices would be severely dampened if both seekers and those 

wishing to be found were aware of the particular biases inherent in any given 

search engine.” 127 That underestimates the scale of the problem. Imagine 

instead your inbox without a spam filter. You would doubtless be “aware of the 

particular biases” of the people trying to sell you fancy watches and penis 

pills—but that will do you little good if your inbox contains a thousand pieces 

of spam for every email you want to read. That is what will happen to search 

results if search algorithms are fully public; the spammers will win. 

For this reason, search-neutrality advocates now acknowledge the danger of 

SEO and thus propose only limited transparency. 128 Pasquale suggests, for 

example, that Google could respond to a question about its rankings with a list 

of a few factors that principally affected a particular result. 129 But search is 

immensely complicated—so complicated that it may not be possible to boil a 

ranking down to a simple explanation. When the law demands disclosure of 

complex matters in simple terms, we get pro forma statements and boilerplate. 

122 Pasquale, Internet Nondiscrimination Principles, supra note 18, at 299 (emphasis added). 

123 See also id. at 285 (arguing that search engines benefit from hidden algorithms because 

websites, lacking clear information about how to achieve high organic search rankings, must 

resort to buying paid search ads). 

124 See also The Google Algorithm, supra note 13 (recommending required disclosure). 

125 See Grimmelmann, Structure of Search Engine Law, supra note 4, at 49, 55. 

126 Id. at 44–46, 56. 

127 Introna and Nissenbaum, supra note 68, at 181. 

128 See Pasquale, Internet Nondiscrimination Principles, supra note 18, at 297; Bracha & Pasquale, 

Federal Search Commission, supra note 20, at 1201–02; Chandler, Right to Reach an Audience, supra 

note 20, at 1117. 

129 Pasquale, Internet Nondiscrimination Principles, supra note 18, at 296–97.


Consumer credit disclosures and securities prospectuses have brought 

important information into the open, but they haven’t done much to aid the 

understanding of their average recipient. 

Google’s algorithm depends on more than 200 different factors. 130 Google 

makes about 500 changes to it a year, 131 based on ten times as many 

experiments. 132 One sixth of the hundreds of millions of queries the algorithm 

handles daily are queries it has never seen before. 133 The PageRank of any 

webpage depends, in part, on every other page on the Internet. 134 And even 

with all the computational power Google can muster, a full PageRank 

recomputation takes weeks. 135 PageRank is, as algorithms go, elegantly 

simple—but I certainly wouldn’t want to have the job of making Markov chains 

and eigenvectors “meaningful to the majority of Web users.” 136 In practice, any 

simplified disclosure is likely to leave room for the search engine to bury plenty 

of bodies. 

Some scholars have suggested that concerns about transparency could be 

handled through regulatory opacity: The search engine discloses its algorithm to 

the government, which then keeps the details from the public. 137 This is a 

promising way of dealing with search engines’ operational needs for secrecy, but 

it sharpens the question of regulators’ technical competence. If the record is 

sealed, they won’t have third-party experts and interested amici to walk them 

through novel technical issues. Everything will hinge on their own ability to 

evaluate the implications of small details in search algorithms. The track record 

of agencies and courts in dealing with other digital technologies does not 

provide grounds for optimism on this score. 138 Pasquale makes an important 

130 See, e.g., Technology Overview, GOOGLE, http://www.google.com/corporate/tech.html. 

131 See Steven Levy, Inside the Box, WIRED, Mar. 2010, at 96. 

132 See Rob Hof, Google’s Udi Manber: Search Is About People, Not Just Data, THE TECH BEAT (Oct. 

1, 2009), 

http://www.businessweek.com/the_thread/techbeat/archives/2009/10/googles_ud 

i_manber_search_is_about_people_not_just_data.html. 

133 Id. 

134 See AMY N. LANGVILLE & CARL D. MEYER, GOOGLE’S PAGERANK AND BEYOND: THE 

SCIENCE OF SEARCH (Princeton University Press 2006). 

135 Id. 

136 In The Google Dilemma, supra note 63, I didn’t even try to explain the math to law professors. 

137 Pasquale, Internet Nondiscrimination Principles, supra note 18 at 297–98; Bracha & Pasquale, 

Federal Search Commission, supra note 20, at 1294–96. See generally Viva R. Moffat, Regulating 

Search, 22 HARV. J. L. & TECH. 475 (2009) (discussing institutional choice issues in search 

regulation). 

138 But see Frank Pasquale, Trusting (and Verifying) Online Intermediaries’ Policing, supra at 258 

(proposing “Internet Intermediary Regulatory Council” and arguing that it could develop


point that “it is essential that someone has the power to ‘look under the hood,’” 139 

but it is also important that algorithmic disclosure remain connected to a 

workable theory of what regulators are looking for and what they would do if 

they found it. 

Manipulation 

Perhaps the most interesting idea in the entire search neutrality debate is the 

“manipulation” of search results. It’s a slippery term, and used inconsistently in 

the search-engine debates—including by me. 140 In the dictionary sense of 

“process, organize, or operate on mentally or logically; to handle with mental or 

intellectual skill,” 141 all search results are manipulated and the more skillfully the 

better. But in the dictionary sense of “manage, control, or influence in a subtle, 

devious, or underhand manner,” 142 it’s a bad thing indeed: no one likes to be 

manipulated. 143 

In practice—although this is rarely made explicit—the concern is with what I 

have described elsewhere as “hand manipulation.” 144 This idea imagines the 

search engine as having both an automatic, general-purpose ranking algorithm 

and a human-created list of exceptions. Consumer Watchdog, for example, 

derides Google’s claim to rank results “automatically by algorithms,” saying, “It 

is hard to see how this can still be true, given the increasingly pronounced tilt 

toward its own services in Google’s search results.” 145 Foundem calls it 

“manual intervention,” “special treatment,” and “manual bias,” and documents 

how Google’s public statements have quietly backed away from claims that its 

rankings are “objective” and “automatic.” 146 

Put this way, the distinction between objective algorithm and subjective 

manipulation is incoherent. Both kinds of decisions come from the same 

sufficient technical expertise to “generate official and even public understanding of [search 

engines’] practices”). 

139 Pasquale, Internet Nondiscrimination Principles, supra note 18, at 286. 

140 Compare Grimmelmann, Structure of Search Engine Law, supra note 4, at 44 (“technical arms 

race between engines and manipulators”) with id., at 59–60 (“hand manipulation of results 

[by search engines]”). 

141 Oxford English Dictionary (June 2010 draft). 

142 Id. 

143 See Bracha & Pasquale, Federal Search Commission, supra note 20, at 1176–79 (discussing effects 

of manipulation on user autonomy). 

144 Grimmelmann, Structure of Search Engine Law, supra note 4, at 59. 

145 TRAFFIC REPORT, supra note 101, at 8. 

146 Foundem’s Google Story, supra note 7.


source: the search engine’s programmers. 147 Nor can the algorithm provide a 

stable baseline against which to measure manipulation, since each 

“manipulation” is a change to the algorithm itself. It’s not like Bing has rooms 

full of employees looking over search results pages and making last-minute 

tweaks before the pages are delivered to users. 

Academics, being more careful with concepts, have focused on intentionality: 

does the search engine intend the promotions and demotions that will result 

from an algorithmic change? Mark Patterson, for example, refers to 

“intentional manipulation of results.” 148 Bracha and Pasquale sharpen this idea 

to speak of “highly specific or local manipulations,” such as singling out 

websites for special treatment. 149 Chandler argues that “search engines should 

not manipulate individual search results except to address instances of suspected 

abuse.” 150 Google itself is remarkably coy about whether and when it changes 

rankings on an individual basis. 151 

Surprisingly, no one has explained why special-casing in and of itself is a 

problem. One possibility is that it captures the distinction between individual 

adjudication and general rulemaking: changes that only affect a few websites 

trigger a kind of due process interest in individualized procedural protections. 152 

There is also a kind of Rawlsian argument 153 here, that algorithmic decisions 

should be made from behind a veil of ignorance, not knowing which websites 

they will favor. For whatever reason, local manipulations make people nervous, 

nervous enough that most of the stories told to instill fear of search engines 

involve what is or looks like manipulation. 154 

Local manipulation, however, is a distraction. The real goal is relevance. From 

that point of view, most local manipulations aren’t wrongful at all. Foundem 

should know; it benefited from a local manipulation. The penalty that afflicted 

147 See Goldman, Search Engine Bias, supra note 28, at 112–15; Grimmelmann, Structure of Search 

Engine Law, supra note 4, at 59–60. 

148 Patterson, Non-Network Barriers, supra note 19, at 2854. 

149 Bracha & Pasquale, Federal Search Commission, supra note 20, at 1168. 

150 Chandler, Right to Reach an Audience, supra note 20, at 1117 (emphasis added). 

151 See, e.g., Lettice, When Algorithms Attack, supra note 99; James Grimmelmann, Google Replies to 

SearchKing Lawsuit, LAWMEME (Jan. 9, 2003), 

http://lawmeme.research.yale.edu/modules.php?name=News&file=article&sid=80 

7. 

152 Compare Londoner v. Denver, 210 U.S. 373, 386 (1908) (hearing required when tax 

assessment affects only a few people) with Bi-Metallic Inv. Co. v. Colorado, 239 U.S. 441, 

445–46 (hearing not required when tax assessment affects all citizens equally). 

153 See JOHN RAWLS, A THEORY OF JUSTICE (1971). 

154 See, e.g., Lettice, When Algorithms Attack, supra note 99.


it for three years appears to have been a relatively general change to Google’s 

algorithm, one designed to affect a great many low-value vertical search sites. 155 

When Foundem was promoted back to prominent search placement, that was 

actually the manipulation, since it affected Foundem and Foundem alone. 

Google thus “manipulated” its search results to exempt Foundem from what 

would otherwise have been a generally applicable rule. To condemn 

manipulation on the basis of its specificity is to say that Google acted more 

rightfully when it demoted Foundem in 2006 than when it promoted it back in 

2009. 156 

The point is that local manipulations, being quick and easy to implement, are 

often a useful part of a search engine’s toolkit for delivering relevance. Searchengine-optimization 

is an endless game of loopholing. Regulators who attempt 

to prohibit unfair manipulations will have to wade quite far into the swamp of 

white-hat and black-hat SEO. 157 Prohibiting local manipulation altogether 

would keep the search engine from closing loopholes quickly and punishing the 

loopholers—giving them a substantial leg up in the SEO wars. Search results 

pages would fill up with spam, and users would be the real losers. 

Conclusion 

Search neutrality gets one thing very right: Search is about user autonomy. A 

good search engine is more exquisitely sensitive to a user’s interests than any 

other communications technology. 158 Search helps her find whatever she wants, 

whatever she needs to live a self-directed life. It turns passive media recipients 

into active seekers and participants. If search did not exist, then for the sake of 

human freedom it would be necessary to invent it. Search neutrality properly 

seeks to make sure that search is living up to its liberating potential. 

Having asked the right question—are structural forces thwarting search’s ability to 

promote user autonomy?—search neutrality advocates give answers concerned with 

protecting websites rather than users. With disturbing frequency, though, 

websites are not users’ friends. Sometimes they are, but often, the websites 

want visitors, and will be willing to do what it takes to grab them. 

155 Id. 

156 If you are bothered more by demotions than promotions, remember that search rankings are 

zero-sum. Foundem’s 50-place rise is balanced out by 50 one-place falls for other websites. 

157 On the distinction between ethical, permitted “white-hat” SEO and unethical, forbidden 

“black-hat” SEO, see Frank Pasquale, Trusting (and Verifying) Online Intermediaries’ Policing, supra 

at 258. I believe that what Pasquale calls the intermediate “grey-hat” zone between the two 

is generally less grey than he and his sources perceive it to be. 

158 Except, perhaps, the library reference desk. Unfortunately, librarians don’t scale.


If Flowers by Irene sells a bouquet for $30 that Bob’s Flowers sells for $50, 

then Bob’s interest in being found is in direct conflict with users’ interest in 

being directed to Irene. The last thing that Bob wants is for the search engine 

to maximize relevance. Search-neutrality advocates fear that Bob will pay off 

the search engine to point users at his site. But that’s not the only way the story 

can play out. Bob could also engage in self-help SEO to try to boost his 

ranking. In that case, the search engine may respond by demoting his site. And 

if that happens, then Bob has another card to play: search-neutrality itself. 

Regulators bearing search neutrality can inadvertently prevent search engines 

from helping users find the websites they want. The typical model assumed by 

search neutrality is of a website and a search engine corruptly conspiring to put 

one over on users. But much, indeed most, of the time, the real alliance is 

between search engines and users, together trying to sort through the clamor of 

millions of websites’ sales pitches. Giving websites search-neutrality rights gives 

them a powerful weapon in their wars with each other—one that need not be 

wielded with users’ interests in mind. 159 Search neutrality will be born with one 

foot already in the grave of regulatory capture. 

There is a profound irony at the heart of the liberal case for search neutrality. 

Requiring search engines to behave “neutrally” will not produce the desired goal 

of neutral search results. The web is a place where site owners compete fiercely, 

sometimes viciously, for viewers and users turn to intermediaries to defend 

them from the sometimes-abusive tactics of information providers. Taking the 

search engine out of the equation leaves users vulnerable to precisely the sorts 

of manipulation search neutrality aims to protect them from. Whether it ranks 

sites by popularity, by personalization, or even by the idiosyncratic whims of its 

operator, a search engine provides an alternative to the Hobbesian world of the 

unmediated Internet, in which the richest voices are the loudest, and the 

greatest authority on any subject is the spammer with the fastest server. Search 

neutrality is cynical about the Internet—but perhaps not cynical enough. 

159 This has already happened in trademark law, which is supposed to prevent consumer 

confusion, but just as often is a form of offensive warfare among companies, consumer 

interests be damned. See Mark A. Lemley & Mark P. McKenna, Owning Mark(et)s (Stanford 

Law and Economics Olin Working Paper No. 395, May 2010), available at 

http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1604845. For an exploration of 

the competitive dynamics of trademark in the search-engine context, see Goldman, 

Deregulating Relevancy, supra note 61.



Search Engine Bias & the 

Demise of Search Engine 

Utopianism 

By Eric Goldman * 

In the past few years, search engines have emerged as a major force in our 

information economy, helping searchers perform hundreds of millions (or even 

billions) of searches per day. 1 With this broad reach, search engines have 

significant power to shape searcher behavior and perceptions. In turn, the 

choices that search engines make about how to collect and present data can 

have significant social implications. 

Typically, search engines automate their core operations, including the 

processes that search engines use to aggregate their databases and then 

sort/rank the data for presentation to searchers. This automation gives search 

engines a veneer of objectivity and credibility. 2 Machines, not humans, appear 

to make the crucial judgments, creating the impression that search engines 

bypass the structural biases and skewed data presentations inherent in any 

human-edited media. 3 Search engines’ marketing disclosures typically reinforce 

this perception of objectivity. 

Unfortunately, this romanticized view of search engines does not match reality. 

Search engines are media companies. Like other media companies, search 

* Associate Professor, Santa Clara University School of Law and Director, High Tech Law 

Institute. Home page: http://www.ericgoldman.org. Email: egoldman@gmail.com. I 

appreciate the comments of Nico Brooks, Soumen Chakrabarti, Ben Edelman, Elizabeth 

Van Couvering and the participants at the Yale Law School Regulating Search Symposium 

and the 2005 Association of Internet Researchers (AoIR) Annual Meeting. This essay 

focuses principally on American law and consumer behavior. Consumer behavior and 

marketplace offerings vary by country, so this discussion may not be readily generalizable to 

other jurisdictions. 

1 In 2003, search engines performed over a half-billion searches a day. See Danny Sullivan, 

Searches Per Day, SEARCH ENGINE WATCH, Feb. 25, 2003, 

http://searchenginewatch.com/reports/article.php/2156461. 

2 See Jason Lee Miller, Left, Right, or Center? Can a Search Engine Be Biased?, WEBPRONEWS.COM, 

May 10, 2005, http://www.webpronews.com/insidesearch/insidesearch/wpn-56- 

20050510LeftRightorCenterCanaSearchEngineBeBiased.html. 

3 There is a broad perception that search engines present search results passively and neutrally. 

See Leslie Marable, False Oracles: Consumer Reaction to Learning the Truth About How Search 

Engines Work, CONSUMER REPORTS WEBWATCH, June 30, 2003, 

http://www.consumerwebwatch.org/dynamic/search-report-false-oraclesabstract.cfm; 

Maureen O’Rourke, Defining the Limits of Free-Riding in Cyberspace: Trademark 

Liability for Metatagging, 33 GONZ. L. REV. 277 (1998).


engines make editorial choices designed to satisfy their audience. 4 These 

choices systematically favor certain types of content over others, producing a 

phenomenon called “search engine bias.” 

Search engine bias sounds scary, but this essay explains why such bias is both 

necessary and desirable. The essay also explains how emerging personalization 

technology will soon ameliorate many concerns about search engine bias. 

Search Engines Make Editorial Choices 

Search engines frequently claim that their core operations are completely 

automated and free from human intervention, 5 but this characterization is false. 

Instead, humans make numerous editorial judgments about what data to collect 

and how to present that data. 6 

Indexing. Search engines do not index every scrap of data available on the 

Internet. Search engines omit (deliberately or accidentally) some web pages 

entirely 7 or may incorporate only part of a web page. 8 

4 See, e.g., C. EDWIN BAKER, ADVERTISING AND A DEMOCRATIC PRESS (1994). 

5 See, e.g., Does Google Ever Manipulate Its Search Results?, GOOGLE.COM, 

http://www.google.com/support/bin/answer.py?answer=4115&topic=368 (“The 

order and contents of Google search results are completely automated. No one hand picks a 

particular result for a given search query, nor does Google ever insert jokes or send messages 

by changing the order of results.”); Does Google Censor Search Results?, GOOGLE.COM, 

http://www.google.com/support/bin/answer.py?answer=17795&topic=368 (“Google 

does not censor results for any search terms. The order and content of our results are 

completely automated; we do not manipulate our search results by hand.”); Technology 

Overview, GOOGLE.COM, http://www.google.com/corporate/tech.html (“There is no 

human involvement or manipulation of results….”); How Can I Improve My Site’s Ranking?, 

GOOGLE.COM, http://www.google.com/support/webmasters/bin/answer.py?answer 

=34432&topic=8524 (“Sites’ positions in our search results are determined automatically 

based on a number of factors, which are explained in more detail at 

http://www.google.com/technology/index.html. We don’t manually assign keywords to 

sites, nor do we manipulate the ranking of any site in our search results.”) see also Complaint 

at 37-38, 52-56, KinderStart.com LLC v. Google, Inc., Case No. C 06-2057 RS (N.D. Cal. 

Mar. 17, 2006) (giving other examples of Google’s claims to be passive). Note that Google 

has subsequently revised some of these cited pages after its censorship controversy in China. 

6 See generally Abbe Mowshowitz & Akira Kawaguchi, Bias on the Web, COMM. ACM, Sept. 2002, 

at 56 (distinguishing “indexical bias” and “content bias”). 

7 See Judit Bar-Ilan, Expectations Versus Reality – Search Engine Features Needed for Web Research at 

Mid-2005, 9 CYBERMETRICS 2 (2005), 

http://www.cindoc.csic.es/cybermetrics/articles/v9i1p2.html. 

8 For example, many search engines ignore metatags. See Eric Goldman, Deregulating Relevancy 

in Internet Trademark Law, 54 EMORY L.J. 507, 567-68 (2005). Search engines also incorporate 

only portions of very large files. See Bar-Ilan, supra note 7; Why Doesn’t My Site Have a Cached 

Copy or a Description?, GOOGLE.COM, http://www.google.com/support/bin/ 

answer.py?answer=515&topic=365 (describing how some pages are “partially indexed”);


During indexing, search engines are designed to associate third party 

“metadata” (data about data) with the indexed web page. For example, search 

engines may use and display third party descriptions of the website in the search 

results. 9 Search engines may also index “anchor text” (the text that third parties 

use in hyperlinking to a website), 10 which can cause a website to appear in 

search results for a term the website never used (and may object to). 11 

Finally, once indexed, search engines may choose to exclude web pages from 

their indexes for a variety of reasons, ranging from violations of quasi-objective 

search engine technical requirements 12 to simple capriciousness. 13 

Ranking. To determine the order of search results, search engines use complex 

proprietary “ranking algorithms.” Ranking algorithms obviate the need for 

humans to make individualized ranking decisions for the millions of search 

Has Google Dropped Their 101K Cache Limit?, RESEARCHBUZZ!, Jan. 31, 2005, 

http://www.researchbuzz.org/2005/01/has_google_dropped_their_101k.shtml 

(discussing how historically Google indexed only the first 101k of a document). 

9 See My Site’s Listing Is Incorrect and I Need it Changed, GOOGLE.COM, 

http://www.google.com/webmasters/3.html. Google’s automated descriptions have 

spawned at least one lawsuit by a web publisher who believed the compilation created a false 

characterization. See Seth Fineberg, Calif. CPA Sues Google Over “Misleading” Search Results, 

ACCT. TODAY, Apr. 19, 2004, at 5, available at 

http://www.webcpa.com/article.cfm?articleid=193&pg-acctoday&print=yes. 

10 See Jagdeep S. Pannu, Anchor Text Optimization, WEBPRONEWS.COM, Apr. 8, 2004, 

http://www.webpronews.com/ebusiness/seo/wpn-4- 

20040408AnchorTextOptimization.html. 

11 For example, the first search result in Google and Yahoo! for the keyword “miserable 

failure” is President George W. Bush’s home page because so many websites have linked to 

the biography using the term “miserable failure.” See Tom McNichol, Your Message Here, N.Y. 

TIMES, Jan. 22, 2004, at G1. This algorithmic vulnerability has spawned a phenomenon 

called “Google bombing,” where websites coordinate an anchor text attack to intentionally 

distort search results. See John Hiler, Google Time Bomb, MICROCONTENT NEWS, Mar. 3, 2002, 

http://www.microcontentnews.com/articles/googlebombs.htm. 

12 See, e.g., Stefanie Olsen, Search Engines Delete Adware Company, CNET NEWS.COM, May 13, 

2004, http://news.com.com/2102-1024_3-5212479.html?tag=st.util.print (Google and 

Yahoo kicked WhenU.com out of their indexes for allegedly displaying different web pages 

to searchers and search engine robots, a process called “cloaking”). 

13 This is the heart of KinderStart’s allegations against Google. See Complaint, 

KinderStart.com LLC v. Google, Inc., Case No. C 06-2057 (N.D. Cal. Mar. 17, 2006). 

Although the complaint’s allegations about Google’s core algorithmic search may not be 

proven, Google does liberally excise sources from Google News. For example, Google 

claims that “news sources are selected without regard to political viewpoint or ideology,” see 

Google News (Beta), GOOGLE.COM, 

http://news.google.com/intl/en_us/about_google_news.html#25, but Google 

dropped a white supremacist news source from Google News because it allegedly 

promulgated “hate content.” See Susan Kuchinskas, Google Axes Hate News, 

INTERNETNEWS.COM, Mar. 23, 2005, 

http://www.internetnews.com/xSP/article.php/3492361.


terms used by searchers, but they do not lessen the role of human editorial 

judgment in the process. Instead, the choice of which factors to include in the 

ranking algorithm (and how to weight them) reflects the search engine 

operator’s editorial judgments about what makes content valuable. Indeed, to 

ensure that these judgments are producing the desired results, search engines 

manually inspect search results 14 and make adjustments accordingly. 

Additionally, search engines claim they do not modify algorithmically-generated 

search results, but there is some evidence to the contrary. Search engines 

allegedly make manual adjustments to a web publisher’s overall ranking. 15 Also, 

search engines occasionally modify search results presented in response to 

particular keyword searches. Consider the following: 

� Some search engines blocked certain search terms containing the 

keyword “phpBB.” 16 

� In response to the search term “Jew,” for a period of time (including, at 

minimum November 2005 when the author observed the 

phenomenon), Google displayed a special result in the sponsored link, 

saying “Offensive Search Results: We’re disturbed about these results 

as well. Please read our note here.” The link led to a page explaining 

the results. 17 

� Reportedly, Ask.com blocked search results for certain terms like 

“pedophile,” “bestiality,” “sex with children” and “child sex.” 18 

� Google removed some websites from its index in response to a Digital 

Millenium Copyright Act (DMCA) take-down demand from the 

Church of Scientology. However, Google displayed the following 

legend at the bottom of affected search results pages (such as search 

results for “scientology site:xenu.net”): “In response to a complaint we 

received under the US Digital Millennium Copyright Act, we have 

14 See Posting of Eric Goldman to Technology & Marketing Law Blog, Google’s Human 

Algorithm, http://blog.ericgoldman.org/archives/2005/06/googles_human_a.htm 

(June 5, 2005, 14:11 EST) (Google hires students to manually review search results for 

quality purposes). 

15 See Search King, Inc. v. Google Tech., Inc., No. CIV-02-1457-M, at 4 (W.D. Okla. Jan. 13, 

2003) (“Google knowingly and intentionally decreased the PageRanks assigned to both 

SearchKing and PRAN.”). This manual adjustment has also been alleged in the recent 

KinderStart lawsuit. See Complaint, KinderStart.com L.L.C. v. Google, Inc., Case No. C 06- 

2057 RS (N.D. Cal. Mar. 17, 2006). 

16 See MSN Blockades phpBB Searchers, TRIMMAIL’S EMAIL BATTLES, Jan. 18, 2006, 

http://www.emailbattles.com/archive/battles/vuln_aacgfbgdcb_jd/. 

17 See http://www.google.com/explanation.html. 

18 See Jennifer Laycock, Ask.com Actively Censoring Some Search Phrases, SEARCH ENGINE GUIDE, 

June 25, 2006, http://www.searchengineguide.com/searchbrief/senews/007837.html. 

On Aug. 1, 2006, I was unable to replicate these results.


removed 2 result(s) from this page. If you wish, you may read the 

DMCA complaint that caused the removal(s) at ChillingEffects.org.” 19 

Conclusion. Search engines have some duality in their self-perceptions, and this 

duality creates much confusion. 20 Search engines perceive themselves as 

objective and neutral because they let automated technology do most of the 

hard work. However, in practice, search engines make editorial judgments just 

like any other media company. Principally, these editorial judgments are 

instantiated in the parameters set for the automated operations, but search 

engines also make individualized judgments about what data to collect and how 

to present it. These manual interventions may be the exception and not the 

rule, but these exceptions only reinforce that search engines play an active role 

in shaping their users’ experiences when necessary to accomplish their editorial 

goals. 

Search Engine Editorial Choices 

Create Biases 

Search results ordering has a significant effect on searchers and web publishers. 

Searchers usually consider only the top few search results; the top-ranked search 

result gets a high percentage of searcher clicks, and click-through rates quickly 

decline from there. 21 Therefore, even if a search engine delivers hundreds or 

even thousands of search results in response to a searcher’s query, searchers 

19 See Chris Sherman, Google Makes Scientology Infringement Demand Public, SEARCH ENGINE 

WATCH, Apr. 15, 2002, 

http://searchenginewatch.com/searchday/article.php/2159691. 

20 See Danny Sullivan, KinderStart Becomes KinderStopped In Ranking Lawsuit Against Google, SEARCH 

ENGINE WATCH, July 14, 2006, http://blog.searchenginewatch.com/blog/060714- 

084842. This duality, if it ends up leading to the dissemination of false information, could 

also create some legal liability. See KinderStart v. Google, No. 5:06-cv-02057-JF (N.D. Cal. 

motion to dismiss granted July 13, 2006) (pointing out the potential inconsistency of 

Google’s position that PageRank is both Google’s subjective opinion but an objective 

reflection of its algorithmic determinations). 

21 See iProspect Search Engine User Behavior Study, IPROSPECT, Apr. 2006, 

http://www.iprospect.com/premiumPDFs/WhitePaper_2006_SearchEngineUserBe 

havior.pdf (62% of searchers click on a search result on the first results page); Jakob 

Nielsen, The Power of Defaults, JAKOB NIELSEN’S ALERTBOX, Sept. 26, 2005, 

http://www.useit.com/alertbox/defaults.html (citing a study by Cornell professor 

Thorsten Joachims that the first search result gets 42% of clicks and the second search result 

gets 8%; further, when the first two search results are switched, the first search result gets 

34%—meaning that positioning dictated searcher behavior); Nico Brooks, The Atlas Rank 

Report: How Search Engine Rank Impacts Traffic, ATLAS INSTITUTE DIGITAL MARKETING 

INSIGHTS, June 2004, http://app.atlasonepoint.com/pdf/AtlasRankReport.pdf (the 

first-ranked search result may get ten times the quantity of clicks as the tenth-ranked search 

result).


effectively ignore the vast majority of those search results. Accordingly, web 

publishers desperately want to be listed among the top few search results. 22 

For search engines, results placement determines how the searcher perceives the 

search experience. If the top few search results do not satisfy the searcher’s 

objectives, the searcher may deem the search a failure. Therefore, to maximize 

searcher perceptions of search success, search engines generally tune their 

ranking algorithms to support majority interests. 23 In turn, minority interests 

(and the websites catering to them) often receive marginal exposure in search 

results. 

To gauge majority interests, search engines frequently include a popularity 

metric in their ranking algorithm. Google’s popularity metric, PageRank, treats 

inbound links to a website as popularity votes, but votes are not counted 

equally; links from more popular websites count more than links from lesserknown 

websites. 24 

Beyond promoting search results designed to satisfy majority interests, 

PageRank’s non-egalitarian voting structure causes search results to be biased 

towards websites with economic power 25 because these websites get more links 

due to their marketing expenditures and general prominence. 

Indeed, popularity-based ranking algorithms may reinforce and perpetuate 

existing power structures. 26 Websites that are part of the current power elite get 

better search result placement, which leads to greater consideration of their 

messages and views. Furthermore, the increased exposure attributable to better 

placement means that these websites are likely to get more votes in the future, 

22 See Michael Totty & Mylene Mangalindan, Web Sites Try Everything To Climb Google Rankings, 

WALL ST. J. ONLINE, Feb. 26, 2003, 

http://online.wsj.com/article/SB1046226160884963943.html?emailf=yes. 

23 See Lucas D. Introna & Helen Nissenbaum, Shaping the Web: Why the Politics of Search Engines 

Matters, INFO. SOC’Y, July-Sept. 2000, at 169. 

24 See Our Search: Google Technology, GOOGLE.COM, http://www.google.com/technology/. 

25 See Niva Elkin-Koren, Let the Crawlers Crawl: On Virtual Gatekeepers and the Right to Exclude 

Indexing, 26 U. DAYTON L. REV. 179, 188 (2001); Frank Pasquale, Rankings, Reductionism, and 

Responsibility, SETON HALL PUBLIC LAW RESEARCH PAPER NO. 888327, at 25, Feb. 25, 2006, 

http://papers.ssrn.com/sol3/papers.cfm?abstract_id=888327; Trystan Upstill et al., 

Predicting Fame and Fortune: PageRank or Indegree?, PROC. OF THE 8TH AUSTRALASIAN 

DOCUMENT COMPUTING SYMP., Dec. 15, 2003, 

http://research.microsoft.com/users/nickcr/pubs/upstill_adcs03.pdf (showing that 

BusinessWeek Top Brand, Fortune 500 and Fortune Most Admired companies get 

disproportionately high PageRank). 

26 See Introna & Nissenbaum, supra note 23; Matthew Hindman et al., “Googlearchy”: How a Few 

Heavily-Linked Sites Dominate Politics on the Web, Mar. 31, 2003, 

http://www.princeton.edu/~mhindman/googlearchy--hindman.pdf.


leading to a self-reinforcing process. 27 In contrast, minority-interest and 

disenfranchised websites may have a difficult time cracking through the 

popularity contest, potentially leaving them perpetually relegated to the search 

results hinterlands. 28 

A number of commentators have lamented these effects and offered some 

proposals in response: 

� Improve Search Engine Transparency. Search engines keep their ranking 

algorithms secret. 29 This secrecy hinders search engine spammers from 

gaining more prominence than search engines want them to have, but 

the secrecy also prevents searchers and commentators from accurately 

assessing any bias. To enlighten searchers, search engines could be 

required to disclose more about their practices and their algorithms. 30 

This additional information has two putative benefits. First, it may 

improve market mechanisms by helping searchers make informed 

choices among search engine competitors. Second, it may help 

searchers determine the appropriate level of cognitive authority to 

assign to their search results. 

� Publicly Fund Search Engines. Arguably, search engines have “public 

good”-like attributes, such as reducing the social costs of search 

behavior. If so, private actors will not incorporate these social benefits 

into their decision-making. In that case, public funding of search 

engines may be required to produce socially-optimal search results. 31 

27 See Egalitarian Engines, ECONOMIST, Nov. 17, 2005 (“there is a widespread belief among 

computer, social and political scientists that search engines create a vicious circle that 

amplifies the dominance of established and already popular websites”); see also Junghoo Cho 

& Sourashis Roy, Impact of Search Engines on Page Popularity, WWW 2004, May 2004, 

http://oak.cs.ucla.edu/~cho/papers/cho-bias.pdf; Upstill, supra note 25. But see Santo 

Fortunato et al., The Egalitarian Effect of Search Engines, Nov. 2005, 

http://arxiv.org/pdf/cs.CY/0511005 (questioning the consequences of the “rich-getsricher” 

effect). 

28 See Cho & Roy, supra note 27; but see Filippo Menczer et al., Googlearchy or Googlocracy?, IEEE 

SPECTRUM, Feb. 2006 (providing empirical evidence suggesting that “search engines direct 

more traffic than expected to less popular sites”). 

29 See Search King Inc. v. Google Tech., Inc., No. CIV-02-1457-M, at 3 n.2 (W.D. Okla. Jan. 13, 

2003) (“Google’s mathematical algorithm is a trade secret, and it has been characterized by 

the company as ‘one of Google’s most valuable assets.’”); Stefanie Olsen, Project Searches for 

Open-Source Niche, CNET NEWS.COM, Aug. 18, 2003, http://news.com.com/2102-1032_3- 

5064913.html?tag=st_util_print. 

30 See Introna & Nissenbaum, supra note 23. 

31 See id.; Eszter Hargittai, Open Portals or Closed Gates? Channeling Content on the World Wide Web, 

27 POETICS 233 (2000); cf. CASS SUNSTEIN, REPUBLIC.COM 170-72 (2001) (advocating 

publicly funded “deliberative domains”).


Indeed, there have been several proposals to create government-funded 

search engines. 32 

� Mandate Changes to Ranking/Sorting Practices. Search engines could be 

forced to increase the exposure of otherwise-marginalized websites. At 

least five lawsuits 33 have requested judges to force search engines to 

reorder search results to increase the plaintiff’s visibility. 34 

In addition to plaintiffs, some academics have supported mandatory reordering 

of search results. For example, Pandey et al. advocate a “randomized rank 

promotion” scheme where obscure websites randomly should get extra credit in 

ranking algorithms, appearing higher in the search results on occasion and 

getting additional exposure to searchers accordingly. 35 In another essay in this 

collection, Frank Pasquale proposes that, when people think the search engines 

are providing false or misleading information, search engines should be forced 

to include a link to corrective information. 36 

Search Engine Bias Is 

Necessary and Desirable 

Before trying to solve the problem of search engine bias, we should be clear 

how search engine bias creates a problem that requires correction. From my 

perspective, search engine bias is the unavoidable consequence of search 

32 See Kevin J. O’Brien, Europeans Weigh Plan on Google Challenge, INT’L HERALD TRIB., Jan. 18, 

2006 (discussing a European initiative called Quaero, which is intended to break the 

American hegemony implicit in Google’s dominant market position); Graeme Wearden, 

Japan May Create Its Own Search Engine, CNET NEWS.COM, Dec. 21, 2005, 

http://news.com.com/Japan+may+create+its+own+search+engine/2100-1025_3- 

004037.html. 

33 See Search King, Inc. v. Google Tech., Inc., No. CIV-02-1457-M (W.D. Okla. Jan. 13, 2003); 

KinderStart.com LLC v. Google, Inc., No. C 06-2057 RS (N.D. Cal. dismissed July 13, 2006); 

Langdon v. Google, Inc., No. 1:06-cv-00319-JJF (D. Del. complaint filed May 17, 2006); 

Roberts v. Google, No. 1-06-CV-063047 (Cal. Superior Ct. complaint filed May 5, 2006); 

Datner v. Yahoo! Inc, Case No. BC355217 (Cal. Superior Ct. complaint filed July 11, 2006) 

[note: this list updated as of July 24, 2006]. 

34 As Google said in its response to the KinderStart lawsuit, “Plaintiff KinderStart contends 

that the judiciary should have the final say over [search engines’] editorial process. It has 

brought this litigation in the hopes that the Court will second-guess Google’s search rankings 

and order Google to view KinderStart’s site more favorably.” Motion to Dismiss at 1, 

KinderStart.com LLC v. Google, Inc., No. C 06-2057 RS (N.D. Cal. May 2, 2006). 

35 See Sandeep Pandey et al., Shuffling a Stacked Deck: the Case for Partially Randomized Ranking of 

Search Engine Results, 

http://www.cs.cmu.edu/~olston/publications/randomRanking.pdf; cf. SUNSTEIN, 

supra note 31 (explaining that websites should be forced to link to contrary views as a way of 

increasing exposure to alternative viewpoints). 

36 See Pasquale, supra at 401; see also Pasquale, supra note 25, at 28-30 (proposing that the link be 

displayed as an asterisk to the search results).


engines exercising editorial control over their databases. Like any other media 

company, search engines simply cannot passively and neutrally redistribute third 

party content (in this case, web publisher content). If a search engine does not 

attempt to organize web content, its system quickly and inevitably will be 

overtaken by spammers, fraudsters and malcontents. 37 At that point, the search 

engine becomes worthless to searchers. 

Instead, searchers (like other media consumers) expect search engines to create 

order from the information glut. To prevent anarchy and preserve credibility, 

search engines must exercise some editorial control over their systems. In turn, 

this editorial control necessarily will create some bias. 

Fortunately, market forces limit the scope of search engine bias. 38 Searchers 

have high expectations for search engines: they expect search engines to read 

their minds 39 and infer their intent based solely on a small number of search 

keywords. 40 Search engines that disappoint (either by failing to deliver relevant 

results, or by burying relevant results under too many unhelpful results) are held 

37 Every Internet venue accepting user-submitted content inevitably gets attacked by unwanted 

content. If left untended, the venue inexorably degrades into anarchy. See, e.g., Step-by-Step: 

How to Get BILLIONS of Pages Indexed by Google, MONETIZE BLOG, June 17, 2006, 

http://merged.ca/monetize/flat/how-to-get-billions-of-pages-indexed-by- 

Google.html (Google indexed over five billion “spam” pages from a single spammer before 

manually de-indexing the sites); Alorie Gilbert, Google Fixes Glitch That Unleashed Flood of Porn, 

CNET NEWS.COM, Nov. 28, 2005, http://news.com.com/2102-1025_3- 

5969799.html?tag=st.util.print (describing how Google Base, a venue for user-submitted 

content, was overtaken by pornographers: “the amount of adult content on Google Base 

was staggering considering Google only launched the tool a week ago.”); Josh Quittner, The 

War Between alt.tasteless and rec.pets.cats, WIRED, May 1994, at 46 (describing how a group of 

anarchists, for fun, took over a USENET newsgroup about pets). 

38 See Mowshowitz & Kawaguchi, supra note 6, at 60 (market forces are the best way to counter 

adverse effects of search engine bias). 

39 See Our Philosophy, GOOGLE.COM, http://www.google.com/corporate/tenthings.html 

(“The perfect search engine … would understand exactly what you mean and give back 

exactly what you want.”); Chris Sherman, If Search Engines Could Read Your Mind, SEARCH 

ENGINE WATCH, May 11, 2005, 

http://searchenginewatch.com/searchday/article.php/3503931. 

40 Searchers routinely use a very small number of keywords to express their search interests. 

See iProspect.com, Inc., iProspect Natural SEO Keyword Length Study, Nov. 2004, 

http://www.iprospect.com/premiumPDFs/keyword_length_study.pdf (eighty-eight 

percent of search engine referrals are based on only one or two keywords); see also Declan 

Butler, Souped-Up Search Engines, NATURE, May 11, 2000, at 112, 115 (citing an NEC Research 

Institute study showing that up to 70% of searchers use only a single keyword as a search 

term); Bernard J. Jansen et al., Real Life Information Retrieval: A Study of User Queries on the Web, 

32 SIGIR FORUM 5, 15 (1998) (stating that the average keyword length was 2.35 words; onethird 

of searches used one keyword and 80% used three keywords or fewer); Jakob Nielsen, 

JAKOB NIELSEN’S ALERTBOX, Search: Visible and Simple, May 13, 2001, 

http://www.useit.com/alertbox/20010513.html (stating that the average keyword length 

was 2.0 words).


accountable by fickle searchers. 41 There are multiple search engines available to 

searchers, 42 and few barriers to switching between them. 43 

As a result, searchers will shop around if they do not get the results they want, 44 

and this competitive pressure constrains search engine bias. If a search engine’s 

bias degrades the relevancy of search results, searchers will explore alternatives 

even if searchers do not realize that the results are biased. Meanwhile, search 

engine proliferation means that niche search engines can segment the market 

and cater to underserved minority interests. 45 Admittedly, these market forces 

41 See Kim Peterson, Microsoft Learns to Crawl, SEATTLE TIMES, May 2, 2005 (MSN Search 

“learned that the arcane searches were the make-or-break moments for Web searchers. 

People weren’t just happy when a search engine could find answers to their most bizarre, 

obscure and difficult queries. They would switch loyalties.”); Bob Tedeschi, Every Click You 

Make, They’ll Be Watching You, N.Y. TIMES, Apr. 3, 2006, 

http://www.nytimes.com/2006/04/03/business/03ecom.html?ei=5090&en=9e55ae6 

4f692433a&ex=1301716800&partner=rssuserland&emc=rss&pagewanted=print. 

42 In addition to the recent launch of major new search engines by providers like MSN, the 

open-source software community is developing Nutch to allow anyone to build and 

customize his or her own web search engine. http://nutch.apache.org/; see also Olsen, 

Open-Source Niche, supra note 29.While there are multiple major search engines, the market 

may still resemble an oligopoly; a few major players (Google, Yahoo, MSN, Ask Jeeves) have 

the lion’s share of the search engine market. However, this may construe the search engine 

market too narrowly. Many types of search providers compete with the big mass-market 

search engines, ranging from specialty search engines (e.g., Technorati) to alternative types of 

search technology (e.g., adware) to non-search information retrieval processes (e.g., link 

navigation). Ultimately, every search engine competes against other search engines and these 

other search/retrieval options. 

43 See Rahul Telang et al., An Empirical Analysis of Internet Search Engine Choice, Aug. 2002 (on file 

with author). For example, search engines use the same basic interface (a white search box), 

and searchers rarely use advanced search features that might require additional learning time 

at other search engines. 

44 See Grant Crowell, Understanding Searcher Behavior, SEARCH ENGINE WATCH, June 14, 2006, 

http://searchenginewatch.com/showPage.html?page=3613291 (citing a Kelsey 

Research study that 63% of searchers used two or more search engines); Press Release, 

Vividence, Inc., Google Wins Users’ Hearts, But Not Their Ad Clicks (May 25, 2004), 

http://www.vividence.com/public/company/news+and+events/press+releases/20 

04-05-25+ce+rankings+search.htm (stating that up to 47% of searchers try another 

search engine when their search expectations are not met). 

45 See Rahul Telang et al., The Market Structure for Internet Search Engines, 21 J. MGMT. INFO. SYS. 

137 (2004), available at http://www.heinz.cmu.edu/~rtelang/engine_jmis_final.pdf 

(describing how searchers sample heterogeneous ranking algorithms, which support a 

diversity of search engines); Mário J. Silva, The Case for a Portuguese Web Search Engine, 

http://xldb.fc.ul.pt/data/Publications_attach/tumba-icwi2003-final.pdf (describing 

the value of a Portuguese-oriented search engine); Jeffrey McMurray, Social Search Promises 

Better Intelligence, ASSOCIATED PRESS, July 9, 2006 (discussing niche search engines that draw 

on social networking); cf. Jakob Nielsen, Diversity is Power for Specialized Sites, JAKOB NIELSEN’S 

ALERTBOX, June 16, 2003, http://www.useit.com/alertbox/20030616.html (describing 

how specialized sites will flourish on the Internet).


are incomplete—searchers may never consider what results they are not 

seeing—but they are powerful nonetheless. 

In contrast, it is hard to imagine how regulatory intervention will improve the 

situation. First, regulatory solutions become a vehicle for normative views 

about what searchers should see—or should want to see. 46 How should we 

select among these normative views? What makes one bias better than another? 

Second, regulatory intervention that promotes some search results over others 

does not ensure that searchers will find the promoted search results useful. 

Determining relevancy based on very limited data (such as decontextualized 

keywords) is a challenging process, and search engines struggle with this 

challenge daily. Due to the complexity of the relevancy matching process, 

government regulation rarely can do better than market forces at delivering 

results that searchers find relevant. As a result, searchers likely will find some of 

the promoted results irrelevant. 

The clutter of unhelpful results may hinder searchers’ ability to satisfy their 

search objectives, undermining searchers’ confidence in search engines’ mindreading 

abilities. 47 In this case, regulatory intervention could counterproductively 

degrade search engines’ value to searchers. Whatever the adverse 

consequences of search engine bias, the consequences of regulatory correction 

are probably worse. 48 

Technological Evolution Will 

Moot Search Engine Bias 

Currently, search engines principally use “one-size-fits-all” ranking algorithms to 

deliver homogeneous search results to searchers with heterogeneous search 

objectives. 49 One-size-fits-all algorithms exacerbate the consequences of search 

engine bias in two ways: (1) they create winners (websites listed high in the 

46 See, e.g., Susan L. Gerhart, Do Web Search Engines Suppress Controversy?, FIRST MONDAY, Jan. 

2004, http://www.firstmonday.org/issues/issue9_1/gerhart/. Gerhart argues that 

search engines do not adequately prioritize search results that expose controversies about the 

search topic. However, her argument assumes that controversy-related information has 

value to consumers, an assumption that deserves careful evaluation. 

47 See Eric Goldman, A Coasean Analysis of Marketing, 2006 WIS. L. REV. 1151, available at 


48 See Susan P. Crawford, Shortness of Vision: Regulatory Ambition in the Digital Age, 74 FORDHAM 

L. REV. 695 (2005) (discussing the shortcomings of regulatory intervention in organic 

information systems). 

49 See James Pitkow et al., Personalized Search, COMM. ACM, Vol. 45:9 (Sept. 2002) at 50-1.


search results) and losers (those with marginal placement), and (2) they deliver 

suboptimal results for searchers with minority interests. 50 

These consequences will abate when search engines migrate away from onesize-fits-all 

algorithms towards “personalized” ranking algorithms. 51 

Personalized algorithms produce search results that are custom-tailored to each 

searcher’s interests, so searchers will see different results in response to the 

same search query. For example, Google offers searchers an option that “orders 

your search results based on your past searches, as well as the search results and 

news headlines you’ve clicked on.” 52 

Personalized ranking algorithms represent the next major advance in search 

relevancy. One-size-fits-all ranking algorithms have inherent limits on their 

maximum relevancy potential, and further improvements in one-size-fits 

algorithms will yield progressively smaller relevancy benefits. Personalized 

algorithms transcend those limits, optimizing relevancy for each searcher and 

thus implicitly doing a better job of searcher mind-reading. 53 

Personalized ranking algorithms also reduce the effects of search engine bias. 

Personalized algorithms mean that there are multiple “top” search results for a 

particular search term instead of a single “winner,” 54 so web publishers will not 

compete against each other in a zero-sum game. In turn, searchers will get 

results more influenced by their idiosyncratic preferences and less influenced by 

the embedded preferences of the algorithm-writers. Also, personalized 

algorithms necessarily will diminish the weight given to popularity-based metrics 

(to give more weight for searcher-specific factors), reducing the structural biases 

due to popularity. 

50 See Michael Kanellos, Microsoft Aims for Search on Its Own Terms, CNET NEWS.COM, Nov. 24, 

2003, http://news.com.com/2102-1008_3-5110910.html?tag=st.util.print (quoting a 

Microsoft researcher as saying “If the two of us type a query [into a search engine], we get 

the same thing back, and that is just brain dead. There is no way an intelligent human being 

would tell us the same thing about the same topic.”); David H. Freedman, Why Privacy Won’t 

Matter, NEWSWEEK, Apr. 3, 2006; Personalization of Placed Content Ordering in Search 

Results, U.S. Patent App. 0050240580 (filed July 13, 2004). 

51 See Pitkow, supra note 49, at 50. 

52 What’s Personalized Search?, GOOGLE.COM, 

http://www.google.com/support/bin/answer.py?answer=26651&topic=1593. 

53 See Jaime Teevan et al., Personalizing Search via Automated Analysis of Interests and Activities, 

SIGIR ‘05, http://haystack.lcs.mit.edu/papers/teevan.sigir05.pdf; Terry McCarthy, On 

the Frontier of Search, TIME, Aug. 28, 2005 (“Search will ultimately be as good as having 1,000 

human experts who know your tastes scanning billions of documents within a split 

second.”) (quoting Gary Flake, Microsoft Distinguished Engineer). 

54 See Kevin Lee, Search Personalization and PPC Search Marketing, CLICKZ NEWS, July 15, 2005, 

http://www.clickz.com/experts/search/strat/print.php/3519876.


Personalized ranking algorithms are not a panacea—any process where humans 

select and weight algorithmic factors will produce some bias 55—but 

personalized algorithms will eliminate many of the current concerns about 

search engine bias. 

Conclusion 

Complaints about search engine bias implicitly reflect some disappointed 

expectations. In theory, search engines can transcend the deficiencies of 

predecessor media to produce a type of utopian media. In practice, search 

engines are just like every other medium—heavily reliant on editorial control 

and susceptible to human biases. This fact shatters any illusions of search 

engine utopianism. 

Fortunately, search engine bias may be largely temporal. In this respect, I see 

strong parallels between search engine bias and the late 1990s keyword metatag 

“problem.” 56 Web publishers used keyword metatags to distort search results, 

but these techniques worked only so long as search engines considered keyword 

metatags in their ranking algorithms. When search engines recognized the 

distortive effects of keyword metatags, they changed their algorithms to ignore 

keyword metatags. 57 Search result relevancy improved, and the problem was 

solved without regulatory intervention. 

Similarly, search engines naturally will continue to evolve their ranking 

algorithms and improve search result relevancy—a process that, organically, will 

cause the most problematic aspects of search engine bias to largely disappear. 

To avoid undercutting search engines’ quest for relevance, this effort should 

proceed without regulatory distortion. 

55 Personalized algorithms have other potentially adverse consequences, such as creating selfreinforcing 

information flows. See SUNSTEIN, supra note 31. For a critique of these 

consequences, see Goldman, Coasean Analysis, supra note 47. 

56 See generally Goldman, Deregulating Relevancy, supra note 8. 

57 See Danny Sullivan, Death of a Meta Tag, SEARCH ENGINE WATCH, Oct. 1, 2002, 

http://www.searchenginewatch.com/sereport/print.php/34721_2165061.


CHAPTER 8 

WHAT FUTURE FOR PRIVACY? 

475 

Privacy Protection in the Next Digital Decade: 

“Trading Up” or a “Race to the Bottom”? 477 


The Privacy Problem: What’s Wrong with Privacy? 483 

Stewart Baker 

A Market Approach to Privacy Policy 509 

Larry Downes

476 CHAPTER 8: WHAT FUTURE FOR PRIVACY?


Privacy Protection in the Next 

Digital Decade: “Trading Up” or 

a “Race to the Bottom”? 

By Michael Zimmer * 

Apparent to most citizens of contemporary, industrialized society, people no 

longer exist and live in fixed locations and spaces. Instead people are on the 

move in their personal, professional, intellectual, and social spheres. Within and 

across these spheres, mobility, rather than permanence, is likely to be the norm. 

Manuel Castells captures this feature of modern life in his theory of the space of 

flows, arguing that “our society is constructed around flows: flows of capital, 

flows of information, flows of technology, flows of organizational interaction, 

flows of images, sounds, and symbols.” 1 These flows—particularly information 

flows—constitute what Castells describes as the “network society,” where 

“networks constitute the new social morphology of our societies, and the 

diffusion of networking logic substantially modifies the operation and outcomes 

in processes of production, experience, power and culture.” 2 

Nowhere is Castells “network society” more apparent than in our contemporary 

global digital information network, with the Internet as its backbone. 

Originating from a handful of universities and research laboratories in the 

1960s, the Internet began to take shape as a ubiquitous information network 

with the emergence of the “dot-com” economy in the 1990s. Dot-com business 

models varied—and met varied levels of success—but most relied on the rapid 

delivery of services and exchange of information. While much of the dot-com 

economy burst with the dot-com bubble in 2000, the Internet remained a 

powerful network enabling robust flows of information, continually modifying 

“experience, power and culture,” just as Castells described. 

In the past digital decade, the Internet has provided new linkages and spaces for 

information flows, and has particularly emerged as a potent infrastructure for 

the flow and capture of personal information. These flows take many forms and 

stem from various motivations. Large-scale web advertising platforms and 

search engines utilize robust infrastructures to collect data about web browsing 

and search activities in order to provide relevant advertising. Users’ 

consumption habits are captured by online service providers like Amazon and 

Netflix, fueling powerful recommendation systems meant to improve user 

* School of Information Studies, University of Wisconsin-Milwaukee 

1 MANUEL CASTELLS, THE RISE OF THE NETWORK SOCIETY 412 (1996). 

2 Id. at 469.

478 CHAPTER 8: WHAT FUTURE FOR PRIVACY? 

satisfaction. Individuals openly share personal information with friends and 

colleagues on social networking services such as Facebook and LinkedIn, and 

their thoughts with the world on platforms like Blogger and Twitter. Looking 

back at the past decade, the Internet has become a platform for the open flow 

of personal information—flows that are largely voluntarily provided by users— 

and as such, appear to have validated Scott McNealy’s (in)famous 1999 remark 

that “You have zero privacy anyway … get over it.” 3 

Notwithstanding McNealy’s view, privacy has remained a central concern amid 

the open information flows in our contemporary network society, including 

worries about the growing size and role of networked databases, 4 the possibility 

of tracking and surveillance by Internet service providers 5 and Web search 

engines, 6 privacy threats from digital rights management technologies, 7 and 

growing concerns about protecting the privacy of users of social networking 

sites and related Web 2.0 services. 8 

While scholars continue to detail possible threats to privacy spawned by the last 

decade of innovations on the Internet, governments have struggled with 

whether—and how—to regulate information flows across these global networks 

3 Polly Sprenger, Sun on Privacy: ‘Get Over It,’ WIRED, March 31, 2007, 

http://www.wired.com/politics/law/news/1999/01/17538. 

4 SIMSON GARFINKEL, DATABASE NATION: THE DEATH OF PRIVACY IN THE 21ST CENTURY 

(2000). 

5 Colin J. Bennett, Cookies, Web Bugs, Webcams and Cue Cats: Patterns of Surveillance on the World 

Wide Web, 3(3) ETHICS AND INFORMATION TECHNOLOGY 195, 197-210 (2001); Paul Ohm, 

The Rise and Fall of Invasive ISP Surveillance, 2009 UNIVERSITY OF ILLINOIS LAW REVIEW 1417- 

1496. 

6 M. Goldberg, The Googling of Online Privacy: Gmail, Search-Engine Histories, and the New Frontier 

of Protecting Private Information on the Web, 9 Lewis & Clark Law Review 249-272 (2005); 

Michael Zimmer, The Gaze of the Perfect Search Engine: Google as an Infrastructure of Dataveillance 

in WEB SEARCHING: MULTIDISCIPLINARY PERSPECTIVES 77-99 (Amanda Spink & Michael 

Zimmer, eds., 2008). 

7 Julie E. Cohen, A Right to Read Anonymously: A Closer Look at ‘Copyright Management’ in 

Cyberspace, 28(4) CONNECTICUT LAW REVIEW 981-1039 (1996); Julie E. Cohen, DRM and 

privacy, 18 BERKELEY TECHNOLOGY LAW JOURNAL 575-617 (2003). 

8 Ralph Gross & Alessandro Acquisti, Information Revelation and Privacy in Online Social Networks 

(ACM Workshop on Privacy in the Electronic Society, Alexandria, VA, 2005); Michael 

Zimmer, The Externalities of Search 2.0: The Emerging Privacy Threats When the Drive for the Perfect 

Search Engine Meets Web 2.0, FIRST MONDAY, Mar. 3, 2010, 

http://firstmonday.org/htbin/cgiwrap/bin/ojs/index.php/fm/article/view/2136/1 

944; Joseph Bonneau & Sören Preibusch, The Privacy Jungle: On the Market for Data Protection in 

Social Networks (The Eighth Workshop on the Economics of Information Security (WEIS 

2009)); James Grimmelmann, Facebook and the Social Dynamics of Privacy, 95(4) IOWA LAW 

REVIEW 1137 (2009); Marc Parry, Library of Congress, Facing Privacy Concerns, Clarifies Twitter 

Archive Plan, THE CHRONICLE OF HIGHER EDUCATION, June 1, 2010, 

http://chronicle.com/blogPost/Library-of-Congress-Facing/23818/.


to protect the privacy of their citizens. Given the diversity of interests, histories, 

and cultural contexts, a complicated terrain of trans-national laws and policies 

for the protection of privacy and personal data flows across networks has 

emerged across the globe. Some jurisdictions have opted for broad, and 

relatively strict, laws regulating the collection, use and disclosure of personal 

information, such as Canada’s Personal Information Protection and Electronic 

Documents Act (PIPEDA) 9 or the European Union’s Data Protection 

Directive. 10 The United States, however, maintains a more sectoral approach to 

privacy legislation, with laws addressing only specific types of personal 

information. For example, the Health Insurance Portability and Accountability 

Act (HIPAA) 11 offers protection of personal medical information, the Fair 

Credit Reporting Act 12 regulates the collection and flow of personal financial 

data, and the Video Privacy Protection Act 13 makes the wrongful disclosure of 

video rental records illegal. 

The differences between Canadian/EU approaches to privacy and that of the 

United States have been well documented and analyzed. 14 Put bluntly, the 

Canadian/EU regulators can be described as embracing a more paternalist 

approach to data protection policy, aiming to preserve a fundamental human 

right of its citizens through preemptive governmental action. In contrast, the 

governance of privacy in the U.S. typically emerges only after some 

informational harm has occurred, often taking the form of industry selfregulation 

or very targeted legislation, with the responsibility of initiating 

enforcement resting on the harmed data subject herself. As Dorothee 

Heisenberg summarizes, “In practical terms, the EU and the US reached very 

different conclusions about the rights of businesses and individuals related to 

personal data.” 15 While the EU and Canada focus on direct and preemptive 

9 R.S., 1985, c. P-21, http://laws.justice.gc.ca/en/P-21/index.html. 

10 Directive 95/46/EC of the European Parliament and of the Council of 24 October 1995 

on the protection of individuals with regard to the processing of personal data and on the 

free movement of such data, http://eurlex.europa.eu/LexUriServ/LexUriServ.do?uri=CELEX:31995L0046:EN:HTML. 

11 Health Insurance Portability and Accountability Act of 1996, H. Rept. 104-736, 

http://www.gpo.gov/fdsys/search/pagedetails.action?granuleId=CRPT- 

104hrpt736&packageId=CRPT-104hrpt736. 

12 Fair Credit Reporting Act (FCRA), 15 U.S.C. § 1681 et seq., 

http://www.ftc.gov/os/statutes/031224fcra.pdf. 

13 Video Privacy Protection Act of 1988, Pub. L. 100-618 (codified at 18 U.S.C. § 2710), 

http://www.law.cornell.edu/uscode/18/2710.html. 

14 See, e.g., DOROTHEE HEISENBERG, NEGOTIATING PRIVACY: THE EUROPEAN UNION, THE 

UNITED STATES, AND PERSONAL DATA PROTECTION (2005); COLIN J. BENNETT & CHARLES 

D. RAAB, THE GOVERNANCE OF PRIVACY: POLICY INSTRUMENTS IN GLOBAL PERSPECTIVE 

(2003). 

15 Heisenberg, supra note 14 at 2.


regulation of the collection and use of personal data, prohibiting “excess” data 

collection and restricting use to the original and stated purposes of the 

collection, the U.S, framework begins with the assumption that most data 

collection and use is both acceptable and beneficial, that guidelines should be 

primarily voluntary and non-invasive, and that regulation should only address 

documented instances of abuse. 

This difference in regulatory approaches to privacy—and the underpinning 

tensions between different jurisdictions’ views towards the rights of data 

subjects—becomes complicated further given the increasing flows of personal 

information across and between transnational networks, and thus, across 

jurisdictions. Internet companies like Google have customers accessing their 

products and services from across the globe, with data processing and storage 

facilities equally scattered. A Canadian citizen, for example, might be accessing a 

Google product in the United States, while the record of the particular 

information exchange might be stored on a server in Ireland. Each jurisdiction 

has its own complex set of regulations and rights assigned to the treatment of 

any personal information shared and stored. 

These kinds of scenarios have prompted growing concerns about whether the 

global diversity of privacy governance will result in a “race to the bottom” 

where corporate interests in processing personal data will migrate to 

jurisdictions where there is little or no control over the circulation and capture 

of personal information flows, or a “race to the top” where the fashioning of 

privacy policy to the highest possible standards in order to be perceived as the 

“best” protector of personal information flows. After considering the available 

evidence, political scientists Colin Bennett and Charles Raab have suggested that 

privacy protection is actually improving globally—a “trading up” of the 

governance of privacy. 16 Companies are, on the whole, not moving around in 

order to avoid strict privacy regulations, such as those developed in the EU; 

instead, there has been a gradual increase in awareness and action on the issue 

of privacy. Examples of this “trading up” include Facebook’s strengthening of 

its privacy policies and practices in reaction to an investigation by the Office of 

the Privacy Commissioner of Canada, or Google’s modifying its Web cookie 

and partially anonymizing search logs in response to Norwegian privacy 

regulators. 17 In each case, large multi-national Internet companies reacted to 

strong regional privacy laws in ways that benefited all users across the globe. 

16 Bennett & Raab, supra note 14. 

17 Office of the Privacy Commissioner of Canada, Facebook Agrees to Address Privacy 

Commissioner’s Concerns, Aug. 20, 2010, http://www.priv.gc.ca/media/nr-c/2009/nrc_090827_e.cfm; 

Nate Anderson, Google To Anonymize Logs In A Nod To Privacy Advocates, 

ARS TECHNICA, Aug. 20, 2010, http://arstechnica.com/business/news/2007/03/ 

google-to-anonymize-logs-in-a-nod-to-privacy-advocates.ars.


Offsetting this positive note, however, is the realization that privacy protection 

may not be “trading up” as rapidly as other global factors, such as the extensive, 

intensive processing of personal data across borders and platforms; the 

increased focus on economic growth through the use of electronic 

communications and information infrastructures; and the harmonization of law 

enforcement and security objectives. Bennett and Raab go to some length to 

expose the limitations of relying solely on individual countries to impose 

isolated privacy policies in the face of a globally-networked computer system 

permitting—indeed encouraging—transnational information flows. 18 While 

state-specific data protection governance might have been sufficient in the past, 

they argue, today’s digitally networked society demands that any country’s 

efforts to protect its citizens will inescapably be linked with (as well as 

dependent on) the actions and laws of other, often disparate, jurisdictions. 

This leads to obvious problems, when, for example, a legal approach like that of 

the United States, with an emphasis on self-regulation and public-sector 

enforcement, meets a different philosophy, such as the more top-down, 

paternalistic approach to data protection held by Canada and the European 

Union. This clash between U.S. and non-U.S. standards for governing personal 

information flows has prompted large, multi-national companies dependent on 

the relatively unfettered flow of information across global digital networks to 

lobby for some middle ground to be reached. In the case of the U.S. and the 

European Union, the result was the 2000 Safe Harbor agreement 19 between the 

two global economic powers to avoid the most egregious misuse of Europeans’ 

private data, while at the same time creating a semi-permanent “cease fire” that 

would allow transatlantic data (and hence commerce) to flow, despite failing to 

meet the letter, and perhaps not even the intent, of the E.U. Data Protection 

Directive. In the end, while U.S. based companies are forced to provide more 

privacy protections than U.S. law demands, the Safe Harbor provisions are 

weaker than the full European Directive on Data Protection. As Heisenberg 

explains, “the evolution … of the [European Union] Commission’s stance on 

data protection seems to have been one of softening a bit” during the Safe 

Harbor negotiations, as the “Commission began to accommodate the US as 

privacy legislation clashed with first commercial, and then security concerns.” 20 

So, while there has been no clear “race to the bottom” in global privacy 

protections, the “trading up” to an increased level of protection of personal 

18 Bennett & Raab, supra note 14. 

19 2000/520/EC: Commission Decision of 26 July 2000 pursuant to Directive 95/46/EC of 

the European Parliament and of the Council on the adequacy of the protection provided by 

the safe harbour privacy principles and related frequently asked questions issued by the US 

Department of Commerce (notified under document number C(2000) 2441), http://eurlex.europa.eu/LexUriServ/LexUriServ.do?uri=CELEX:32000D0520:EN:HTML. 

20 Heisenberg, supra note 14 at 136.


information flows on our transnational digital networks has not materialized as 

quickly or clearly as one might expect. Heisenberg correctly notes that with the 

Safe Harbor Agreement, the EU was able to force the U.S. to deal with the 

privacy issues that might have otherwise been ignored, force some minor 

concessions, and show that the EU’s privacy standard was significant, granting 

the EU something like a “first-mover advantage” in future trans-border privacy 

disputes. 21 Yet, beyond isolated examples of Internet companies’ hesitant 

acquiescence to non-U.S. regulatory bodies—like the Facebook and Google 

examples provided above—new norms of personal data protection are unlikely 

to emerge in the next digital decade, as data protection officials in Europe have 

begun to publicly question the appropriateness of the current levels of 

protections. 22 

Recalling Castells’ warning that “networks constitute the new social morphology 

of our societies, and the diffusion of networking logic substantially modifies the 

operation and outcomes in processes of production, experience, power and 

culture,” we are left to consider the status of privacy protections in the next 

digital decade. Our network society will continue to grow in size and density, as 

well as in its global importance and interconnectedness. Without concerted 

efforts to ensure a “trading up” in global privacy protections—a renewed 

commitment to the rights of data subjects embodied in the Canadian and 

European Union approach to data protection—those caught within the 

inescapable “diffusion of networking logic” may have little control over how 

the increased flows of their personal information will modify “experience, 

power and culture” over the next digital decade. 

21 Id. at 170. 

22 W. Scott Blackmer, The Information Law Group, European Reservations?, Aug. 26, 2010, 

http://www.infolawgroup.com/2010/08/articles/eu-1/european-reservations/.


The Privacy Problem: 

What’s Wrong with Privacy? 

By Stewart Baker * 

Why are privacy groups so viscerally opposed to government action that could 

reduce the risks posed by exponential technologies? The cost of their stance was 

made clear on September 11, 2001. That tragedy might not have occurred if not 

for the aggressive privacy and civil liberties protection imposed by the Foreign 

Intelligence Surveillance Court and the Department of Justice’s Office of 

Intelligence; and it might have been avoided if border authorities had been able 

to use airline reservation data to screen the hijackers as they entered the United 

States. 

But even after 9/11, privacy campaigners tried to rebuild the wall and to keep 

the Department of Homeland Security (DHS) from using airline reservation 

data effectively. They failed; too much blood had been spilled. 

But in the fields where disaster has not yet struck—computer security and 

biotechnology—privacy groups have blocked the government from taking even 

modest steps to head off danger. 

I like to think that I care about privacy, too. But I had no sympathy for privacy 

crusaders’ ferocious objection to any new government use of technology and 

data. Where, I wondered, did their objection come from? 

So I looked into the history of privacy crusading. And that’s where I found the 

answer. 

The Birth of the Right of Privacy 

In the 1880s, Samuel Dennis Warren was near the top of the Boston 

aristocracy. He had finished second in his class at Harvard Law School. He 

founded a law firm with the man who finished just ahead of him, Louis 

Brandeis, and they prospered mightily. Brandeis was a brilliant, creative lawyer 

and social reformer who would eventually become a great Supreme Court 

justice. 

But Samuel Dennis Warren was haunted. There was a canker in the rose of his 

life. His wife was a great hostess, and her parties were carefully planned. When 

* Stewart A. Baker is a partner in the Washington office of Steptoe & Johnson LLP. He 

returned to the firm following 3½ years at the Department of Homeland Security as its first 

Assistant Secretary for Policy.


Warren’s cousin married, Mabel Warren held a wedding breakfast and filled her 

house with flowers for the event. The papers described her home as a “veritable 

floral bower.” 

No one should have to put up with this. Surely you see the problem. No? Well, 

Brandeis did. 

He and Warren both thought that, by covering a private social event, the 

newspapers had reached new heights of impertinence and intrusiveness. The 

parties and guest lists of a Boston Brahmin and his wife were no one’s business 

but their own, he thought. And so was born the right to privacy. 

Angered by the press coverage of these private events, Brandeis and Warren 

wrote one of the most frequently cited law review articles ever published. In 

fact, “The Right to Privacy,” which appeared in the 1890 Harvard Law Review, 

is more often cited than read—for good reason, as we’ll see. 1 But a close 

reading of the article actually tells us a lot about the modern concept of privacy. 

Brandeis, 2 also the father of the policy-oriented legal brief, begins the article 

with a candid exposition of the policy reasons why courts should recognize a 

new right to privacy. His argument is uncompromising: 

The press is overstepping in every direction the obvious 

bounds of propriety and of decency. Gossip is no longer the 

resource of the idle and of the vicious, but has become a trade, 

which is pursued with industry as well as effrontery … To 

occupy the indolent, column upon column is filled with idle 

gossip, which can only be procured by intrusion upon the 

domestic circle. The intensity and complexity of life, attendant 

upon advancing civilization, have rendered necessary some 

retreat from the world, and man, under the refining influence 

of culture, has become more sensitive to publicity, so that 

solitude and privacy have become more essential to the 

individual; but modern enterprise and invention have, through 

invasions upon his privacy, subjected him to mental pain and 

distress, far greater than could be inflicted by mere bodily 

injury … Even gossip apparently harmless, when widely and 

persistently circulated, is potent for evil … When personal 

gossip attains the dignity of print, and crowds the space 

available for matters of real interest to the community, what 

1 Samuel Warren & Louis D. Brandeis, The Right to Privacy, 4 HARVARD L. REV. 193 (1890). 

2 Because the article owes much of its current fame to Brandeis’s later career, I will from this 

point on discuss only his views without each time laboriously giving credit, if that is the right 

word, to his coauthor.


wonder that the ignorant and thoughtless mistake its relative 

importance … Triviality destroys at once robustness of 

thought and delicacy of feeling. 3 

What does Brandeis mean by this? To be brief, he thinks it should be illegal for 

the newspapers to publish harmless information about himself and his family. 

That, he says, is idle gossip, and it distracts “ignorant and thoughtless” 

newspaper readers from more high-minded subjects. It also afflicts the refined 

and cultured members of society—like, say, Samuel Dennis Warren and his 

wife—who need solitude but who are instead harassed by the fruits of “modern 

enterprise and invention.” 

What’s remarkable about “The Right to Privacy” is that the article’s title still 

invokes reverence, even though its substance is, well, laughable. 

Is there anyone alive who thinks it should be illegal for the media to reveal the 

guest-list at a prominent socialite’s dinner party or to describe how elaborate the 

floral arrangements were? Today, it’s more likely that the hostess of a prominent 

dinner party will blog it in advance, and that the guests will send Twitter updates 

while it’s under way. For most socialites, what would really hurt is a lack of 

media coverage. To be blunt, when he complains so bitterly about media 

interest in a dinner party, Brandeis sounds to modern ears like a wuss. 

Equally peculiar is the suggestion that we should keep such information from 

the inferior classes lest they abandon self-improvement and wallow instead in 

gossip about their betters. That makes Brandeis sound like a wuss and a snob. 

He does sound quite up-to-date when he complains that “modern enterprise 

and invention” are invading our solitude. That is a familiar complaint. It’s what 

privacy advocates are saying today about Google, not to mention the National 

Security Agency (NSA). Until you realize that he’s complaining about the 

scourge of “instantaneous photographs and newspaper enterprise.” 4 Huh? 

Brandeis evidently thinks that publishing a private citizen’s photo in the 

newspaper causes “mental pain and distress, far greater than could be inflicted 

by mere bodily injury.” 5 

If we agreed today, of course, we probably wouldn’t have posted 5 billion 

photographs of ourselves and our friends on Flickr. 6 

3 The Right to Privacy, at 196 (1890). 

4 Id. at 195. 

5 Id. at 196. 

6 Zack Sheppard, 5,000,000,000, Flickr Blog, Sept. 19, 2010, 

http://blog.flickr.net/en/2010/09/19/5000000000/.


Spirit of the Privacy Movement Today 

Anachronistic as it seems, the spirit of Brandeis’s article is still the spirit of the 

privacy movement. The right to privacy was born as a reactionary defense of the 

status quo, and so it remains. Then, as now, new technology suddenly made it 

possible to spread information more cheaply and more easily. This was new, 

and uncomfortable. But apart from a howl of pain—pain “far greater than … 

mere bodily injury”—Brandeis doesn’t tell us why it’s so bad. I guess you had to 

be there—literally. Unless you were an adult when photography came to 

newspapers, you’ll probably never really understand what the fuss was about. 

We’ve all been photographed, and most of us aren’t happy with the results, at 

least not all the time. But that’s life, and we’ve learned to live with it. Most of us 

can’t imagine suing to prevent the distribution of our photographs—which was 

the tort Brandeis wanted the courts to create. 

We should not mock Brandeis too harshly. His article clearly conveys a heartfelt 

sense of invasion. But it is a sense of invasion we can never share. The 

sensitivity about being photographed or mentioned in the newspapers, a raw 

spot that rubbed Brandeis so painfully, has calloused over. So thick is the 

callous that most of us would be tickled, not appalled, to have our dinner parties 

make the local paper, and especially so if it included our photos. 

And that’s the second thing that Brandeis’s article can tell us about more 

contemporary privacy flaps. His brand of resistance to change is still alive and 

well in privacy circles, even if the targets have been updated. Each new privacy 

kerfuffle inspires strong feelings precisely because we are reacting against the 

effects of a new technology. Yet as time goes on, the new technology becomes 

commonplace. Our reaction dwindles away. The raw spot grows a callous. And 

once the initial reaction has passed, so does the sense that our privacy has been 

invaded. In short, we get used to it. 

At the beginning, of course, we don’t want to get used to it. We want to keep 

on living the way we did before, except with a few more amenities. And so, like 

Brandeis, we are tempted to ask the law to stop the changes we see coming. 

There’s nothing more natural, or more reactionary, than that. 

Most privacy advocates don’t see themselves as reactionaries or advocates for 

the status quo, of course. Right and left, they cast themselves as underdogs 

battling for change against the entrenched forces of big government. But 

virtually all of their activism is actually devoted to stopping change—keeping 

the government (and sometimes industry) from taking advantage of new 

technology to process and use information. 

But simply opposing change, especially technological change, is a losing battle. 

At heart, the privacy groups know it, which may explain some of their shrillness 

and lack of perspective. Information really does “want to be free”—or at least


cheap. And the spread of cheap information about all of us will change our 

relationship to the world. We will have fewer secrets. Crippling government by 

preventing it from using information that everyone else can get will not give us 

back our secrets. 

In the 1970s, well before the personal computer and the Internet, privacy 

campaigners persuaded the country that the FBI’s newspaper clipping files 

about U.S. citizens were a threat to privacy. Sure, the information was public, 

they acknowledged, but gathering it all in one file was viewed as vaguely sinister. 

The attorney general banned the practice in the absence of some legal reason 

for doing so, usually called an investigative “predicate.” 

So, in 2001, when Google had made it possible for anyone to assemble a clips 

file about anyone in seconds, the one institution in the country that could not 

print out the results of its Google searches about Americans was the FBI. This 

was bad for our security, and it didn’t protect anyone’s privacy either. 

The privacy campaigners are fighting the inevitable. The “permanent record” 

our high school principals threatened us with is already here—in Facebook. 

Anonymity, its thrills and its freedom, has been characteristic of big cities for 

centuries. But anonymity will also grow scarce as data becomes easier and easier 

to gather and correlate. We will lose something as a result, no question about it. 

The privacy groups’ response is profoundly conservative in the William F. 

Buckley sense—standing athwart history yelling, “Stop!” 7 

I’m all for conservatism, even in unlikely quarters. But using laws to fight the 

inevitable looks a lot like Prohibition. Prohibition was put in place by an Anglo- 

Saxon Protestant majority that was sure of its moral superiority but not of its 

future. What the privacy community wants is a kind of data Prohibition for 

government, while the rest of us get to spend more and more time in the corner 

bar. 

That might work if governments didn’t need the data for important goals such 

as preventing terrorists from entering the country. After September 11, though, 

we can no longer afford the forced inefficiency of denying modern information 

technology to government. In the long run, any effective method of ensuring 

privacy is going to have to focus on using technology in a smart way, not just 

trying to make government slow and stupid. 

7 See William F. Buckley Jr., Publisher’s Statement, NATIONAL REVIEW, Nov. 19, 1955, at 5, 

available at www.nationalreview.com/articles/223549/our-mission-statement/williamf-buckley-jr.


The Evolution of Technology 

& the “Zone of Privacy” 

That doesn’t mean we have to give up all privacy protection. It just means that 

we have to look for protections that work with technology instead of against it. 

We can’t stop technology from making information cheap and reducing 

anonymity, but we can deploy that same technology to make sure that 

government officials can’t misuse data and hide their tracks. This new privacy 

model is partially procedural—greater oversight and transparency—and partly 

substantive—protecting individuals from actual adverse consequences rather 

than hypothetical informational injuries. 

Under this approach, the first people who should lose their privacy are the 

government workers with access to personal data. They should be subject to 

audit, to challenge, and to punishment if they use the data for improper 

purposes. That’s an approach that works with emerging technology to build the 

world we want to live in. In contrast, it is simple Luddism to keep government 

from doing with information technology what every other part of society can 

do. 

The problem is that Luddism always has appeal. “Change is bad” is a slogan 

that has never lacked for adherents, and privacy advocates sounded alarm after 

alarm with that slogan as the backdrop when we tried to put in place a databased 

border screening system. 

But would we really thank our ancestors if they’d taken the substance of 

Brandeis’s article as seriously as its title? If, without a legislature ever 

considering the question, judges had declared that no one could publish true 

facts about a man’s nonpolitical life, or even his photograph, without his 

permission? 

I don’t think so. Things change. Americans grow less private about their sex 

lives but more private about financial matters. Today, few of us are willing to 

have strangers living in our homes, listening to our family conversations, and 

then gossiping about us over the back fence with the strangers who live in our 

friends’ homes. Yet I’ll bet that both Brandeis and Warren tolerated without a 

second thought the limits that having servants put on their privacy. 

Why does our concept of privacy vary from time to time? Here’s one theory: 

Privacy is allied with shame. We are all ashamed of something about ourselves, 

something we would prefer that no one, or just a few people, know about. We 

want to keep it private. Sometimes, of course, we should be ashamed. Criminals 

always want privacy for their acts. But we’re also ashamed—or at least feel 

embarrassment, the first cousin of shame—about a lot of things that aren’t 

crimes.


We may be ashamed of our bodies, at least until we’re sure we won’t be mocked 

for our physical shortcomings. Privacy is similar; we are often quite willing to 

share information about ourselves, including what we look like without our 

clothes, when we trust our audience, or when the context makes us believe that 

our shortcomings will go unnoticed. Most of us would rather be naked with our 

spouse than a random stranger. And we would not appear at the office in our 

underwear, even if it covers more than the bathing suit we wore at the beach on 

the weekend. 

For that reason, enforced nudity often feels like a profound invasion of our 

privacy. At least at first. In fact, though, we can get used to it pretty quickly, as 

anyone who has played high school sports or served in the army can attest. 

That’s because the fear of mockery is usually worse than the experience. So 

when we discover that being naked in a crowd of other naked people doesn’t 

lead to mockery and shame, we begin to adapt. We develop a callous where we 

once were tender. 

The things that Brandeis considered privacy invasions are similar. Very few of 

us are happy the first time we see our photograph or an interview in the 

newspaper. But pretty soon we realize it’s just not that big a deal. Our nose and 

our style of speech are things that the people we know have already accepted, 

and no one else cares enough to embarrass us about them. The same is true 

when we Google ourselves and see that a bad review of our dinner-theater 

performance is number three on the list. Our first reaction is embarrassment 

and unhappiness, but the reaction is oddly evanescent. 

If this is so, then the “zone of privacy” is going to vary from time to time and 

place to place—just as our concept of physical modesty does. The zone of 

privacy has boundaries on two sides. We don’t care about some information 

that might be revealed about us, probably because the revelation causes us no 

harm—or we’ve gotten used to it. If the information is still embarrassing, we 

want to keep it private, and society may agree. But we can’t expect privacy for 

information that society views as truly shameful or criminal. 

Over time, information will move into and out of the zone of privacy on both 

sides. Some information will simply become so unthreatening that we’ll laugh at 

the idea that it is part of the privacy zone. Photographs long ago entered that 

category, despite Brandeis’s campaigning. Some information will move from 

criminal evidence into the zone of privacy, as sexual preference has. Conversely, 

it may move in the other direction: information that a man beats his wife is no 

longer protected by a zone of familial privacy, as it once was; now it’s viewed as 

evidence of a crime. 

The biggest privacy battles will often be in circumstances where the rules are 

changing. The subtext of many Internet privacy fights, for example, is whether 

some new measure will expose the identities of people who download


pornography or copyrighted music and movies. Society is divided about how 

shameful it is to download these items, and it displaces that moral and legal 

debate into a fight about privacy. 

Divorce litigation, for instance, is brutal in part because information shared in a 

context of love and confidence ends up being disclosed to the world in a 

deliberately harmful way. Often the activity in question (like making a telephone 

call or a credit card purchase) is something that the individual does freely, with 

clear knowledge that some other people (his bank or his phone company) know 

what he is doing. Sometimes the activities are proudly public in nature— 

protests against government policy, for example. 

In those cases, the privacy concern is not that the bank or the phone company 

(or our spouse) actually has the information, but rather what they will do with 

the information they have—whether they will use the data in ways we didn’t 

expect or give the data to someone who can harm us. We want to make sure the 

data will not be used to harm us in unexpected ways. 

And that helps explain why privacy advocates are so often Luddite in 

inclination. Modern technology keeps changing the ways in which information 

is used. Once, we could count on practical obscurity—the difficulty of finding 

bits of data from our past—to protect us from unexpected disclosures. Now, 

storage costs are virtually nil, and processing power is increasing exponentially. 

It is no longer possible to assume that your data, even though technically public, 

will never actually be used. It is dirt cheap for data processors to compile 

dossiers on individuals, and to use the data in ways we didn’t expect. 

Some would argue that this isn’t really “privacy” so much as a concern about 

abuse of information. However it’s defined, though, the real question is what 

kind of protection is it reasonable for us to expect. Can we really write a 

detailed legislative or contractual pre-nup for each disclosure, setting forth 

exactly how our data will be used before we hand it over? I doubt it. Maybe we 

can forbid obvious misuses, but the more detailed we try to get, the more we 

run into the problem that our notions of what is private, and indeed of what is 

embarrassing, are certain to change over time. If so, does it make sense to freeze 

today’s privacy preferences into law? 

In fact, that’s the mistake that Brandeis made—and the last lesson we can learn 

from the odd mix of veneration and guffawing that his article provokes. 

Brandeis wanted to extend common law copyright until it covered everything 

that can be recorded about an individual. The purpose was to protect the 

individual from all the new technologies and businesses that had suddenly made 

it easy to gather and disseminate personal information: “the too enterprising


press, the photographer, or the possessor of any other modern device for 

rewording or reproducing scenes or sounds.” 8 

This proposal is wacky in two ways. First it tries to freeze in 1890 our sense of 

what is private and what is not. Second, it tries to defy the gravitational force of 

technology. 

Every year, information gets cheaper to store and to duplicate. Computers, 

iPods, and the Internet are all “modern devices” for “reproducing scenes or 

sounds,” which means that any effort to control reproduction of pictures, 

sounds, and scenes becomes extraordinarily difficult if not impossible. In fact, it 

can’t be done. 

There is a deep irony here. Brandeis thought that the way to ensure the strength 

of his new right to privacy was to enforce it just like state copyright law. If you 

don’t like the way “your” private information is distributed, you can sue 

everyone who publishes it. One hundred years later, the owners of federal 

statutory copyrights in popular music and movies followed this prescription to a 

T. They began to use litigation to protect their data rights against “the 

possessor[s] of any other modern device for … reproducing scenes or 

sounds,” 9 a class that now included many of their customers. The Recording 

Industry Association of America (RIAA) sued consumers by the tens of 

thousands for using their devices to copy and distribute songs. 

Unwittingly, the RIAA gave a thorough test to Brandeis’s notion that the law 

could simply stand in front of new technology and bring it to a halt through 

litigation. There aren’t a lot of people who think that that has worked out well 

for the RIAA’s members, or for their rights. 

Brandeis wanted to protect privacy by outlawing the use of a common new 

technology to distribute “private” facts. His approach has fared no better than 

the RIAA’s. Information that is easy to gather, copy and distribute will be 

gathered, copied, and distributed, no matter what the law says. 

It may seem a little bit odd for me to criticize Brandeis and other privacy 

campaigners for resisting the spread of technology. After all, we can’t simply 

accept the world that technology and commerce serve up. 

It’s one thing to redirect the path of technological change by a few degrees. It’s 

another to insist that it take a right angle. Brandeis wanted it to take a right 

8 Warren & Brandeis, supra note 1 at 206. 

9 Id.


angle; he wanted to defy the changes that technology was pressing upon him. So 

did the RIAA. 

Both were embracing a kind of Luddism—a reactionary spasm in the face of 

technological change. They were doomed to fail. The new technologies, after all, 

empowered ordinary citizens and consumers in ways that could not be resisted. 

If the law tries to keep people from enjoying the new technologies, in the end it 

is the law that will suffer. 

But just because technologies are irresistible does not mean that they cannot be 

guided, or cannot have their worst effects offset by other technologies. The 

solutions I’m advocating will only work if they allow the world to keep 

practically all the benefits of the exponential empowerment that new technology 

makes possible. 

Privacy for the Real World: 

Proposed Solutions 

So what’s my solution to the tension between information technology and our 

current sense of privacy? The short answer is that we should protect privacy, 

but not by defying the course of technology or by crippling government when it 

investigates crimes. We can do it by working with technology, not against it. In 

particular, we can use information technology to make sure that government 

officials lose their privacy when they misuse data that has been gathered for 

legitimate reasons. Information technology now makes it easier to track every 

database search made by every user, and then to follow any distribution of that 

data outside the system. In other words, it can make misuse of the data in 

government files much more difficult and much more dangerous. 

But before talking about what might work, let’s take a closer look at some of the 

ideas that don’t. 

Ownership of Personal Data 

The first privacy solution is one we’ve already seen. It’s the Brandeisian notion 

that we should all “own” our personal data. That has some appeal, of course. If 

I have a secret, it feels a lot like property. I can choose to keep it to myself, or I 

can share it with a few people whom I trust. And I would like to believe that 

sharing a secret with a few trusted friends doesn’t turn it into public property. 

It’s like my home. Just because I’ve invited one guest home doesn’t mean the 

public is welcome. 

But in the end, information is not really like property. Property can only be held 

by one person at a time, or at most by a few people. But information can be 

shared and kept at the same time. And those with whom it is shared can pass it 

on to others at little or no cost. If you ever told a friend about your secret crush


in junior high, you’ve already learned that information cannot be controlled like 

property. As Ben Franklin is credited with saying, “Three may keep a secret if 

two of them are dead.” 10 The redistribution of information cannot be easily 

controlled in the best of times, and Moore’s Law is making the control of 

information nearly impossible. 11 

The recording and movie industries discovered the same thing. If these 

industries with their enormous lobbying and litigation budgets cannot control 

information that they own as a matter of law, the rest of us are unlikely to be 

able to control information about ourselves. Gossip is not going to become 

illegal simply because technology amplifies it. 

That’s why Brandeis’s proposal never really got off the ground, at least not as he 

envisioned it. Buoyed by Brandeis’s prestige, the idea that private facts are 

private property lingered on in the courts for years, but what survived of his 

proposal is scarcely recognizable today. 

In fact, so transformed is Brandeis’s privacy doctrine that it is now described, 

accurately, as a “right of publicity,” which surely would have him turning in his 

grave. Currently, most states honor Brandeis by allowing lawsuits for 

unauthorized commercial use of a person’s likeness, either by statute or judgemade 

law. 

Over time, courts lost sight of Brandeis’s purpose. They began to take the 

analogy to property literally. Brandeis wanted to treat private information like 

property because that was the only way to give a remedy for the “mental pain 

and distress, far greater than could be inflicted by mere bodily injury,” that he 

thought a man suffered when his photo was published without permission. But 

as people got used to having their pictures taken, the mental pain and distress 

slowly drained out of the experience. 

All that was left was the property analogy. And so judges began shrinking the 

right until it only had bite in the one set of circumstances where the right to 

control one’s image actually feels like a property right—when the image is 

worth real bucks. Thus, the courts require disgorgement of profits made when a 

celebrity’s name, face, voice, or even personal style is used without permission 

to sell or endorse products. As a result, the right to exploit a celebrity’s image 

really is property today; it can be sold, transferred, and even inherited. 

10 Benjamin Franklin, POOR RICHARD’S ALMANAC, July 1735. 

11 Moore’s Law describes the long-term trend that the number of transistors that can be placed 

inexpensively on an integrated circuit has doubled approximately every two years. It is named 

after Intel’s co-founder Gordon E. Moore, who described the trend in the essay Cramming 

More Components Onto Integrated Circuits, ELECTRONICS MAGAZINE 4, 1965.


There’s only one problem with this effort to turn privacy into property: it hasn’t 

done much for privacy. It simply protects the right of celebrities to make money 

off their fame. In fact, by monetizing things like celebrity images, it rewards 

those who have most relentlessly sacrificed their privacy to gain fame. 

The right of publicity is well named. It is the right to put your privacy up for 

sale. Not surprisingly, a lot of people have been inspired to do just that. 

Ironically, Brandeis’s doctrine has helped to destroy the essence of what he 

hoped to preserve. 

Oh, and in the process, Brandeis’s approach has stifled creativity and restricted 

free speech—muzzling artists, social commentators, and businesspeople who 

want to make creative use of images that are an essential part of our cultural 

environment. It’s a disaster. Slowly, courts are waking up to the irony and 

limiting the right of publicity. 

The same “private information as property” approach has also made a modest 

appearance in some consumer privacy laws, and it’s worked out just as badly. At 

bottom, consumer privacy protection laws like the Right to Financial Privacy 

Act 12treat a consumer’s data like a consumer’s money: You can give your data 

(or your money) to a company in exchange for some benefit, but only if you’ve 

been told the terms of the transaction and have consented. Similarly, the Cable 

Communications Policy Act of 1984 prevents cable providers from using or 

releasing personal information in most cases unless the providers get the 

customer’s consent. The fruit of this approach is clear to anyone with a bank 

account or an Internet connection. Everywhere you turn, you’re confronted 

with “informed consent” and “terms of service” disclosures; these are uniformly 

impenetrable and non-negotiable. No one reads them before clicking the box, 

so the “consent” is more fiction than reality; certainly it does little to protect 

privacy. Indeed, it’s turning out a lot like the right of publicity. By treating 

privacy as property, consumer privacy protection law invites all of us to sell our 

privacy. 

And we do. Only for most of us, the going price turns out to be disconcertingly 

cheap. 

Mandatory Predicates for Information Access 

The second way of protecting privacy is to require what’s called a “predicate” 

for access to information. That’s a name only a lawyer could love. In fact, the 

whole concept is one that only lawyers love. 

12 The Right to Financial Privacy Act of 1978, Pub. L. No. 95-630, 92 Stat. 3695 (1978) 

(codified as amended at 12 U.S.C. § 3401 et seq.).


Simply put, the notion is that government shouldn’t get certain private 

information unless it satisfies a threshold requirement—a “predicate” for access 

to the data. Lawyers have played a huge role in shaping American thinking 

about privacy, and the predicate approach has been widely adopted as a privacy 

protection. But its value for that purpose is quite doubtful. 

The predicate approach to privacy can be traced to the Fourth Amendment, 

which guarantees that “no Warrants shall issue, but upon probable cause.” 

Translated from legalese, this means that the government may not search your 

home unless it has a good reason to do so. When the government asks for a 

search warrant, it must show the judge “probable cause”—evidence that the 

search will likely turn up criminal evidence or contraband. Probable cause is the 

predicate for the search. 

When a flap arose in the 1970s over the FBI practice of assembling domestic 

security dossiers on Americans who had not broken the law, the attorney 

general stepped in to protect their privacy. He issued new guidelines for the 

FBI. He was a lawyer, so he declared that the FBI could not do domestic 

security investigations of Americans without a predicate. 

The predicate wasn’t probable cause; that was too high a standard. Instead, the 

attorney general allowed the launching of a domestic security investigation only 

if the bureau presented “specific and articulable facts giving reason to believe” 

that the subject of the investigation may be involved in violence. 13 

Actually, the story of the FBI guidelines shows why the predicate approach 

often fails. The dossiers being assembled by the FBI were often just clippings 

and other public information. They usually weren’t the product of a search in 

the classic sense; no federal agents had entered private property to obtain the 

information. Nonetheless, the FBI guidelines treated the gathering of the 

information itself as though it were a kind of search. 

In so doing, the guidelines were following in Brandeis’s footsteps—treating 

information as though it were physical property. The collection of the 

information was equated to a physical intrusion into the home or office of the 

individual. Implicitly, it assumes that data can be locked up like property. 

But that analogy has already failed. It failed for Brandeis and it failed for the 

RIAA. It failed for the FBI guidelines, too. As clippings became easier to 

13 The Right to Financial Privacy Act of 1978, Pub. L. No. 95-630, 92 Stat. 3695 (1978) 

(codified at 12 U.S.C. § 3414(a)(5)(A)), amended by the Uniting and Strengthening America 

by Providing Appropriate Tools Required to Intercept and Obstruct Terrorism (USA 

PATRIOT) Act of 2001, Pub. L. 107–56, 115 STAT. 272, § 505(b), 

www.law.cornell.edu/uscode/pdf/uscode12/lii_usc_TI_12_CH_35_SE_3401.pdf.


retrieve, clippings files became easier to assemble. Then Google made it 

possible for anyone to assemble an electronic clips file on anyone. There was 

nothing secret about the clippings then. They were about as private as a bus 

terminal. 

But the law was stuck in another era. Under the guidelines, only the FBI and 

CIA needed a predicate to do Google searches. You have to be a pretty resilient 

society to decide that you want to deny to your law enforcement agencies a tool 

that is freely available to nine-year-old girls and terrorist gangs. Resilient, but 

stupid. (Not surprisingly, the guidelines were revised after 9/11.) 

That’s one reason we shouldn’t treat the assembling of data as though it were a 

search of physical property. As technology makes it easier and easier to collect 

data, the analogy between doing that and conducting a search of a truly private 

space will become less and less persuasive. No one thinks government agencies 

should have a predicate to use the White Pages. Soon, predicates that keep law 

enforcement from collecting information in other ways will become equally 

anachronistic, leaving law enforcement stuck in the 1950s while everyone else 

gets to live in the twenty-first century. 

I saw this lawyerly affinity for predicates up close at DHS. The issue was laptop 

searches at the border. The government has always had the right to search 

anything crossing the border without probable cause. Smugglers are smart and 

highly motivated; they would find a way to exploit any limitations on the 

authority to conduct searches. The first Congress knew that quite well, and in 

1789, two months before it sent the Fourth Amendment to the states for 

approval, Congress gave the customs service “full power and authority” to 

search “any ship or vessel, in which they shall have reason to suspect any goods, 

wares or merchandise subject to duty shall be concealed.” 14 

Obviously, DHS and its border predecessors didn’t search laptops in 1789. But 

they did search books, papers, correspondence, and anything else that could 

store information. That was the law for two hundred years, with one exception. 

The Supreme Court has ruled that a few extraordinarily intrusive techniques— 

body cavity searches and forced x-rays—require a “reasonable suspicion.” 15 

Laptops are treated like books and papers. They are searched whenever border 

officials think that such a search is likely to be productive. Even the famously 

14 An Act to Regulate the Collection of Duties, 1st Cong. 1st Sess., Stat.1, Ch. V, Sec. 24 at 43 

(July 31, 1789). 

15 U.S. v. Flores-Montano, 541 U.S. 149 (2004).


liberal Ninth Circuit, the court of appeal that includes California, has had no 

trouble approving that practice, 16 and for good reason—laptop searches pay off. 

In 2006, for example, border officials at the Minneapolis-St. Paul airport 

referred a suspect traveler to secondary inspection. There they found that his 

computer contained video clips of IEDs being used to kill soldiers and destroy 

vehicles and a video on martyrdom. He was also carrying a manual on how to 

make improvised explosive devices, or IEDs—a weapon of choice for terrorists 

in Afghanistan and Iraq. 

Despite two hundred years of history and precedent, as well as the proven value 

of searching electronic media, privacy groups launched a campaign against 

laptop searches toward the end of the Bush administration. This was a strange 

and unhappy era in the debate over privacy. By 2005, privacy advocates had 

found a growing audience for claims that the Bush administration had 

abandoned all limits in pursuing terrorism—that it had swung the pendulum 

violently away from privacy and in favor of government authority. 

The privacy advocates’ solution to the laptop issue was the lawyer’s favorite—a 

predicate requirement. Laptops should not be searched at the border, they 

argued, unless the border official could articulate some specific reason for 

conducting the search. That argument was rejected by both the Bush and the 

Obama administrations after careful consideration. 

We rejected it for two reasons. It wouldn’t have protected privacy in any 

meaningful way, and it would have helped criminals like pedophiles and 

terrorists defeat our border defenses. Other than that, it was jim-dandy. 

Why wouldn’t it help protect privacy? Because, as a practical matter, no border 

official today searches a laptop without some reasonable suspicion about the 

traveler. The exponential increase in commercial jet travel and the unforgiving 

thirty-second rule mean that only one traveler in two hundred is sent to 

secondary inspection for a closer look. Once there, many travelers quickly 

satisfy the officials that they don’t deserve more detailed inspection. 

Everyone at the border is busy; border officers don’t have the luxury of 

hooking up the laptops of random travelers for inspection without a good 

reason. Officers who waste their time and DHS’s resources that way are going 

to hear from their supervisors long before they hear from the travelers’ lawyers. 

If border officials only search laptops today when they have a good reason to 

do so, why not make that a requirement? What harm can it do to make 

reasonable suspicion a predicate for laptop searches at the border? Plenty. 

16 U.S. v. Arnold, 523 F.3d 941 (9th Cir. 2008).


Requiring reasonable suspicion before a laptop search will open every border 

search to litigation. And in court, it may be hard to justify even some very 

reasonable judgments. 

Inevitably, enforcement of a predicate requirement for border searches will 

produce litigation. The litigation will focus on the motives of the border 

officials. The courts will tell those officials that some reasons are not good 

enough. Defense lawyers will want to see the personnel records of border 

officials, hoping to show that they’ve inspected a disproportionate number of 

laptops belonging to minorities, or to Saudis, or to men, or any other pattern 

that might get the case thrown out. Border officials will have to start keeping 

detailed records justifying each laptop search. New paperwork and new 

procedures will clog the inspection process, backing up travelers and penalizing 

any inspector who searches a laptop. 

Wait a minute, you might ask, what if those officials are racists or sexists? 

Let’s assume that this concern is legitimate, at least sometimes, and that there 

are biased officials at work on the border. Surely there’s a better way to find 

them and get them off the job than to count on criminal defense lawyers 

exposing them on the witness stand years after the event. 

By now, notice, we’re not even talking about privacy anymore. The “predicate” 

solution has, in effect, changed the subject. We’re talking about the motives of 

border officials, or ethnic profiling, or something—but it isn’t privacy. We’re 

also moving the whole discussion into territory that lawyers find comfortable 

but that ordinary people might question. 

The Fourth Amendment approach to privacy assumes that privacy is best 

protected by letting criminals challenge the search that produced the evidence 

against them, but before adopting that solution, we ought to be pretty sure that 

we’re going to get benefits that match the cost of letting guilty defendants go 

free, something that isn’t obvious here. 

Limits on Information Use 

That leaves the third approach to privacy, one we’ve already seen in action. If 

requiring a predicate is the lawyer’s solution; this third approach is the 

bureaucrat’s solution. It is at heart the approach adopted by the European 

Union: Instead of putting limits on when information may be collected, it sets 

limits on how the information is used. 

The European Union’s data protection principles cover a lot of ground, but 

their unifying theme is imposing limits on how private data is used. Under those 

principles, personal data may only be used in ways that are consistent with the


purposes for which the data were gathered. Any data that is retained must be 

relevant to the original purposes and must be stored securely to prevent misuse. 

The EU’s negotiating position in the passenger name records conflict was 

largely derived from this set of principles. The principles also explain Europe’s 

enthusiasm for a wall between law enforcement and intelligence. If DHS 

gathered reservation data for the purpose of screening travelers when they cross 

the border, why should any other agency be given access to the data? This also 

explains the EU’s insistence on short deadlines for the destruction of PNR data. 

Once it had been used to screen passengers, it had served the purpose for which 

it was gathered and should be promptly discarded. 

There is a core of sense in this solution. It focuses mainly on the consequences 

of collecting information, and not on the act of collection. It doesn’t try to insist 

that information is property. It recognizes that when we give information to 

others, we usually have an expectation about how it will be used, and as long as 

the use fits our expectations, we aren’t too fussy about who exactly gets to see 

it. By concentrating on how personal information is used, this solution may get 

closer to the core of privacy than one that focuses on how personal information 

is collected. 

It has another advantage, too. In the case of government databases, focusing on 

use also allows us to acknowledge the overriding importance of some 

government data systems while still protecting against petty uses of highly 

personal information. 

Call it the deadbeat-dad problem, or call it mission creep, but there’s an 

uncomfortable pattern to the use of data by governments. Often, personal data 

must be gathered for a pressing reason—the prevention of crime or terrorism, 

perhaps, or the administration of a social security system. Then, as time goes on, 

it becomes attractive to use the data for other, less pressing purposes— 

collecting child support, perhaps, or enforcing parking tickets. No one would 

support the gathering of a large personal database simply to collect unpaid 

parking fines; but “mission creep” can easily carry the database well beyond its 

original purpose. A limitation on use prevents mission creep, or at least forces a 

debate about each step in the expansion. 

That’s all fine. But in the end, this solution is also flawed. 

It, too, is fighting technology, though less obviously than the predicate and 

property approaches. Data that has already been gathered is easier to use for 

other purposes. It’s foolish to pretend otherwise. Indeed, developments in 

information technology in recent years have produced real strides in searching 

unstructured data or in finding relationships in data without knowing for sure 

that the data will actually produce anything useful. In short, there are now good


reasons to collate data gathered for widely differing purposes, just to see the 

patterns that emerge. 

This new technical capability is hard to square with use limitations or with early 

destruction of data. For if collating data in the government’s hands could have 

prevented a successful terrorist attack, no one will congratulate the agency that 

refused to allow the collation because the data was collected for tax or 

regulatory purposes, say, and not to catch terrorists. 

What’s more, use limitations have caused great harm when applied too 

aggressively. The notorious “wall” between law enforcement and intelligence 

was at heart a use limitation. It assumed that law enforcement agencies would 

gather information using their authority, and then would use the information 

only for law enforcement purposes. Intelligence agencies would do the same. 

Or so the theory went. But strict enforcement of this use limitation was 

unimaginably costly. In August 2001, two terrorists were known to have 

entered the United States. As the search for them began, the government’s top 

priority was enforcing the wall -- keeping intelligence about the terrorists from 

being used by the “wrong” part of the FBI. Government lawyers insisted that 

law enforcement resources could not be used to pursue intelligence that two 

known al Qaeda agents were in the United States in August 2001. 

This was a fatal blunder. The criminal investigators were well-resourced and 

eager. They might have found the men. The intelligence investigators, in 

contrast, had few resources and did not locate the terrorists, at least not until 

September 11, when the terrorists’ names were discovered on the manifests of 

the hijacked planes. It was a high price to pay for the modest comfort of “use” 

limitations. 

Like all use limitations, the “wall” between law enforcement sounded reasonable 

enough in the abstract. While no one could point to a real privacy abuse arising 

from cooperation between the intelligence and law enforcement agencies in the 

United States, it was easy to point to the Gestapo and other totalitarian 

organizations where there had been too much cooperation among agencies. 

What was the harm in a little organizational insurance against misuse of 

personal data, the argument ran. The rules allowed cooperation where that was 

strictly necessary, and we could count on the agencies to crowd right up to the 

line in doing their jobs. Or so we thought. In fact, we couldn’t. As the pressure 

and the risk ratcheted up, agents were discouraged from pushing for greater 

communication and cooperation across the wall. All the Washington-wise knew 

that the way to bureaucratic glory and a good press lay in defending privacy. 

Actually, more to the point, they knew that bad press and bureaucratic disgrace 

were the likely result if your actions could be characterized as hurting privacy. 

Congress would hold hearings; appropriators would zero out your office; the 

second-guessing arms of the Justice Department, from the inspectors general to


the Office of Professional Responsibility, would feast on every detail of your 

misstep. So, what might have been a sensible, modest use restriction preventing 

the dissemination of information without a good reason became an 

impermeable barrier. 

That’s why the bureaucratic system for protecting privacy so often fails. The use 

restrictions and related limits are abstract. They make a kind of modest sense, 

but if they are enforced too strictly, they prevent new uses of information that 

may be critically important. 

And often they are enforced too strictly. You don’t have to tell a bureaucrat 

twice to withhold information from a rival agency. Lawsuits, bad press, and 

Congressional investigations all seem to push against a flexible reading of the 

rules. If a use for information is not identified at the outset, it can be nearly 

impossible to add the use later, no matter how sensible the change may seem. 

This leads agencies to try to draft broad uses for the data they collect, which 

defeats the original point of setting use restrictions. 

It’s like wearing someone else’s dress. Over time, use restrictions end up tight 

where they should be roomy—and loose where they should be tight. No one is 

left satisfied. 

The Audit Approach: Enforced Accountability 

So what will work? Simple: accountability, especially electronically-enforced 

accountability. 

The best way to understand this solution is to begin with Barack Obama’s 

passport records—and with “Joe the Plumber.” These were two minor flaps 

that punctuated the 2008 presidential campaign. But both tell us something 

about how privacy is really protected these days. 

In March of 2008, Barack Obama and Hillary Clinton were dueling across the 

country in weekly primary showdowns. Suddenly, the campaign took an odd 

turn. The Bush administration’s State Department announced that it had fired 

or disciplined several contractors for examining Obama’s passport records. 

Democrats erupted. It wasn’t hard to jump to the conclusion that the 

candidate’s files had been searched for partisan purposes. 17After an 

investigation, the flap slowly deflated. It soon emerged that all three of the main 

presidential candidates’ passport files had been improperly accessed. 

Investigators reported that the State Department was able to quickly identify 

17 Karen Tumulty, Snooping Into Obama’s Passport, TIME, Mar. 21, 2008, 

http://www.time.com/time/politics/article/0,8599,1724520,00.html.


who had examined the files by using its computer audit system. This system 

flagged any unusual requests for access to the files of prominent Americans. 

The fired contractors did not deny the computer record. Several of them were 

charged with crimes and pleaded guilty. All, it turned out, had acted purely out 

of “curiosity.” 

Six months later, it was the Republicans’ turn to howl about privacy violations 

in the campaign. Samuel “Joe” Wurzelbacher, a plumber, became an overnight 

hero to Republicans in October 2008 after he was practically the only person 

who laid a glove on Barack Obama during the campaign. The candidate made 

an impromptu stop in Wurzelbacher’s Ohio neighborhood and was surprised 

when the plumber forced him into a detailed on-camera defense of his tax plan. 

Three days later, “Joe the Plumber” and his taxes were invoked dozens of times 

in the presidential debates. 

The price of fame was high. A media frenzy quickly stripped Wurzelbacher of 

anonymity. Scouring the public record, reporters found that the plumber had 

been hit with a tax lien; they also found government data that raised doubts 

about the status of his plumbing license. 

Reporters weren’t the only ones digging. Ohio state employees also queried 

confidential state records about Wurzelbacher. In all, they conducted eighteen 

state records checks on Wurzelbacher. They asked whether the plumber owed 

child support, whether he’d ever received welfare or unemployment benefits, 

and whether he was in any Ohio law enforcement databases. Some of these 

searches were proper responses to media requests under Ohio open records 

laws; others looked more like an effort to dig dirt on the man. 

Ohio’s inspector general launched an investigation and in less than a month was 

able to classify all but one of the eighteen records searches as either legitimate 

or improper. 18 Thirteen searches were traced and deemed proper, but three 

particularly intrusive searches were found improper; they had been carried out 

at the request of a high-ranking state employee who was also a strong Obama 

supporter. She was suspended from her job and soon stepped down. A fourth 

search was traced to a former information technology contractor who had not 

been authorized to search the system he accessed; he was placed under criminal 

investigation. 

What do these two flaps have in common? They were investigated within weeks 

of the improper access, and practically everyone involved was immediately 

caught. That’s vitally important. Information technology isn’t just taking away 

your privacy or mine. It’s taking away the privacy of government workers even 

18 See State of Ohio, Office of Inspector General, Report of Investigation, File ID Number 

2008299, Nov. 20, 2008, www.judicialwatch.org/documents/2009/IGReport.pdf.


faster. Data is cheap to gather and cheap to store. It’s even getting cheap to 

analyze. 

So it isn’t hard to identify every official who accessed a particular file on a 

particular day. That’s what happened here. And the consequences for privacy 

are profound. 

If the lawyer’s solution is to put a predicate between government and the data 

and the bureaucrat’s solution is to put use restrictions on the data, then this is 

the auditor’s solution. Government access to personal data need not be 

restricted by speed bumps or walls. Instead, it can be protected by rules, so long 

as the rules are enforced. 

What’s new is that network security and audit tools now make it easy to enforce 

the rules. That’s important because it takes the profit motive out of misuse of 

government data. No profit-motivated official is going to take the risk of 

stealing personal data if it’s obvious that he’ll be caught as soon as people start 

to complain about identity theft. Systematic misuse of government databases is 

a lot harder and more dangerous if good auditing is in place. 

Take another look at why government officials accessed these files. It wasn’t to 

steal identities. The reason most of these people accessed the data was simple 

curiosity. Even the one access that may have been for more reprehensible 

reasons—the woman who checked confidential child support and welfare 

records for Joe the Plumber—was quickly caught and the data never leaked. 

The speed and nearly complete effectiveness of the audit process in these cases 

tells us that network auditing tools can transform the way we enforce the rules 

for handling data in government. For example, if we catch every error, we can 

improve compliance and at the same time reduce the penalties for mistakes. 

Harsh penalties are not the most effective way to enforce rules. In fact, they’re 

usually a confession of failure. 

When we catch every offender, we can afford to lower the penalty. Lighter, 

more certain penalties for privacy violations serve another purpose, too. We’ve 

talked a lot about the oddly protean nature of privacy. Not causing harm in 

unexpected ways is at the core of the concept, but it’s nearly impossible to write 

detailed rules spelling out what is and is not a violation of privacy. Indeed, the 

effort to write such rules and stick to them is what gave us the wall, and 

thousands of American dead. So something must be left to discretion. 

Government employees must use good sense in handling personal data. If they 

don’t, they should be punished. But if we are confident that we can identify any 

questionable use of personal data and correct it quickly, the punishments can be 

smaller. They can be learning experiences rather than penological experiences.


So why did we criminally prosecute the poor schlubs whose hobby was looking 

at the passport pictures of famous people? The election happened. Everything 

that touched on the election was put under a microscope. Evil motives were 

always ascribed to the other side. The State Department had to make a blood 

sacrifice to show that accessing the data was not part of an evil plot by one 

party against the other. Opening a criminal investigation was a way of 

condemning the access in the clearest possible fashion. That the poor schlubs 

probably only deserved demotions counted for little in the super-heated 

atmosphere of a presidential campaign. 

That shows one of the problems with the audit approach. It is too easily turned 

into a phony privacy scandal. In both the Wurzelbacher and Obama cases, the 

audits did their job. With one possible exception, they caught the government 

staff that broke the rules. They prevented any harm to either Wurzelbacher or 

Obama. And they made sure that the officials who were responsible would 

never repeat their errors again. 

The system worked. Privacy was protected. But that’s certainly not the 

impression that was left by coverage of the affairs. Indeed, the chairman of the 

Senate Judiciary Committee, Senator Leahy, used the passport flap to tout new 

legislation strengthening privacy protections on government databases. From a 

political point of view, then, the system failed. There were no thanks for the 

government officials who put the system in place, who checked the audit logs, 

who confronted and disciplined the wrongdoers, and who brought the solved 

problem to public attention. To the contrary, they were pilloried for allowing 

the access in the first place—even though preventing such access is an 

impossible task unless we intend to re-erect walls all across government. 

How’s that for irony? Audits work. But they work too well. Every time they 

catch someone and put a stop to misuse of personal data they also provide an 

opening for political grandstanding. In the end, the finger pointing will 

discourage audits. And that will mean less privacy enforcement. So, the more we 

turn every successful audit into a privacy scandal, the less real privacy we’re 

likely to have. 

That would be a shame, because the auditor’s solution to the problem is the 

only privacy solution that will get more effective as technology advances. And 

we’re going to need more solutions that allow flexible, easy access to sensitive 

databases while still protecting privacy. 

If the plight of government investigators trying to prevent terrorist attacks 

doesn’t move you, think about the plight of medical technicians trying to keep 

you alive after a bad traffic accident. 

The Obama administration has launched a long-overdue effort to bring 

electronic medical records into common use. But the privacy problem in this


area is severe. Few of us want our medical records to be available to casual 

browsers. At the same time, we can’t personally verify the bona fides of the 

people accessing our records, especially if we’re lying by the side of the road 

suffering from what looks like brain or spine damage. 

But the electronic record system won’t work if it can’t tell the first responders 

that you have unusual allergies or a pacemaker. It has to do that quickly and 

without a lot of formalities. Auditing access after the fact is likely to be our best 

answer to this problem, as it is to the very similar problem of how to let law 

enforcement and intelligence agencies share information smoothly and quickly 

in response to changing and urgent circumstances. The Markle Foundation has 

done pioneering work in this area, and its path-breaking 2003 report on privacy 

and security in the war on terror recommends embracing technologies that 

watch the watchers. 19 A unique mix of security, privacy, and technology experts 

managed to reach agreement in that report; they found that one key to 

protecting privacy without sacrificing security was a network that included 

“access control, authentication, and full auditing capability.” 20 

The Markle report urges that large databases with personal information use 

emerging technologies that can identify all users of the system with certainty 

and then give them access that depends on their roles at any particular time. 

This includes “the ability to restrict access privileges so that data can be used 

only for a particular purpose, for a finite period of time, and by people with the 

necessary permissions.” 21 The technologies they cited are not pie in the sky. 

They exist today: “smart cards with embedded chips, tokens, biometrics, and 

security circuits” as well as “[i]nformation rights management technologies.” 22 

The Markle task force later did a thoughtful paper on one of those technologies, 

which would preserve audit logs even if high-ranking officials seek to destroy or 

modify them later. 23 

These technologies can be very flexible. This makes them especially suitable for 

cases where outright denial of data access could have fatal results. The tools can 

be set to give some people immediate access, or to open the databases in certain 

situations, with an audit to follow. They can monitor each person with access to 

19 Markle Foundation Task Force, Creating a Trusted Network for Homeland Security, Dec. 2003, 

http://www.markle.org/downloadable_assets/nstf_report2_full_report.pdf. 

20 Id. at 15. 

21 Id. 

22 Id. 

23 MARKLE FOUNDATION TASK FORCE, IMPLEMENTING A TRUSTED INFORMATION SHARING 

ENVIRONMENT: USING IMMUTABLE AUDIT LOGS TO INCREASE SECURITY, TRUST, AND 

ACCOUNTABILITY (2006), available at 

http://www.markle.org/downloadable_assets/nstf_IAL_020906.pdf.


the data and learn that person’s access patterns—what kinds of data, at what 

time, for how long, with or without copying, and the like. Deviations from the 

established pattern can have many consequences. Perhaps access will be granted 

but the person will be alerted that an explanation must be offered within 

twenty-four hours. Or access could be granted while a silent alarm sounds, 

allowing systems administrators to begin a real-time investigation. 

There’s a kind of paradox at the heart of this solution. We can protect people 

from misuse of their data, but only by stripping network users of any privacy or 

anonymity when they look at the data. The privacy campaigners aren’t likely to 

complain, though. In our experience, their interest in preserving the privacy of 

intelligence and law enforcement officers is pretty limited. 

When I was general counsel of the National Security Agency, a well-known 

privacy group headed by Marc Rotenberg filed a Freedom of Information Act 

request asking the NSA to assemble all documents and emails sent “to or from 

Stewart Baker.” Then as now, the NSA was forbidden to assemble files on 

American citizens who were not agents of a foreign power. Even so, Rotenberg 

was asking NSA to assemble a dossier on me. Since NSA and I were locked in a 

battle with Rotenberg over encryption policy at the time, the purpose of the 

dossier was almost certainly to look for embarrassing information that might 

help Rotenberg in his political fight. Indeed, Rotenberg claimed when I 

confronted him that he was planning to scrutinize my dossier for evidence of 

misconduct. 

Had the FBI or NSA assembled a dossier on their political adversaries, it would 

have been a violation of law. In fact, it would have caused a privacy scandal. But 

Rotenberg saw no irony in his request. It wasn’t a privacy problem, in his view, 

because government officials deserve no privacy. 

I still think Rotenberg’s tactics were reprehensible: He had singled me out for a 

selective loss of privacy because he didn’t like my views. But I’ve come to 

appreciate that there’s a core of truth to his view of government. Anyone who 

has access to government files containing personal data has special 

responsibilities. He should not expect the same privacy when he searches that 

data as he has while he’s surfing the net at home. And now that technology 

makes it easy to authenticate and track every person, every device, and every 

action on a network, perhaps it’s time to use that technology to preserve 

everyone else’s privacy. 

In the end, that’s the difference between a privacy policy that makes sense and 

one that doesn’t. We can’t lock up data that is getting cheaper every day. 

Pretending that it’s property won’t work. Putting “predicates” between 

government and the data it needs won’t work, and neither will insisting that they 

may only be used for purposes foreseen when it was collected.


What we can do is use new information technology tools to deter government 

officials from misusing their access to that data. 

As you know by now, I think that some technology poses extraordinary risks. 

But we can avoid the worst risks if we take action early. We shouldn’t try to 

stop the trajectory of new technology. But we can bend it just a little. Call it a 

course correction on an exponential curve. 

That’s also true for privacy. The future is coming—like it or not. Our data will 

be everywhere. But we can bend the curve of technology to make those who 

hold the data more accountable. Bending the exponential curve a bit—that’s a 

privacy policy that could work. And a technology policy that makes sense.



A Market Approach 

to Privacy Policy 

By Larry Downes * 

Privacy: The Problem 

What happens when the cost of deleting information is higher than the cost of 

retaining it? 

The answer is that nothing gets deleted. In the age of cloud computing, mobile 

devices, and social networking, what that really means is that more and more 

data—some of it enterprise data, some of it personal information, and more and 

more of it something that merges the two—is being saved. 

Soon, perhaps already, much of it will be consolidated, aggregated, reorganized, 

and mined for valuable patterns, behaviors, and insights. Privacy has become 

an unintended casualty of Moore’s Law—collateral damage from friendly fire. 

That, at least, is one way of thinking about privacy in the digital age, one that 

has been on my mind for the last several months. I wrote about the privacy 

problem in my recent book, The Laws of Disruption, in which I argued that the 

real solution to concerns about privacy in the digital age would be the 

emergence of robust markets for private information, where consumers would 

be able to trade personal information with other individuals and enterprises 

when doing so generated mutual benefit. 1 

The privacy problem has morphed since then into the latest terror of the digital 

age, surpassing earlier shibboleths, such as copyright piracy, identity theft, cyber 

war and net neutrality. 2 Daily media coverage of the latest privacy policy 

* Larry Downes is an Internet analyst and consultant, helping clients develop business 

strategies in an age of constant disruption caused by information technology. He is the 

author of UNLEASHING THE KILLER APP: DIGITAL STRATEGIES FOR MARKET 

DOMINANCE (Harvard Business School Press 1998) and, most recently, of THE LAWS OF 

DISRUPTION: HARNESSING THE NEW FORCES THAT GOVERN LIE AND BUSINESS IN THE 

DIGITAL AGE (Basic Books 2009) [hereinafter THE LAWS OF DISRUPTION]. 

1 See LARRY DOWNES, THE LAWS OF DISRUPTION, Law Two: Privacy. 

2 As I’ve written elsewhere, all of these problems share a common core. Each raises the 

fundamental question about the nature of digital life and by whom and how its basic 

infrastructure is to be governed. In some sense, each is another view of the same regulatory 

problem, seen through lenses that are equally unfocused, but in different ways. See Larry 

Downes, After the Deluge, More Deluge, THE TECHNOLOGY LIBERATION FRONT, July 22, 2010, 

http://techliberation.com/2010/07/22/after-the-deluge-more-deluge/.


change, hacking incident, stolen government laptop or inadvertent disclosure 

has raised the stakes and the tension in a problem that, it seems, people react to 

with such strong emotions that rational discussion of any solution is now 

impossible. 3 

The privacy crisis is very much on the mind of regulators around the world, 

who see the emergence of privacy fears among consumers as the latest and 

perhaps the best opportunity to gain a toehold in regulating (and perhaps 

taxing) content on the web. Nearly all of the earlier efforts—including outright 

censorship, imposition of protectionist laws on global e-commerce, and 

enforcement of strict copyright, trademark and patent regimes onto the 

evolving collaborative ethos of digital life—have failed utterly. 4 By aligning 

themselves with consumer interests (and perhaps helping to stoke the fires of 

anxiety), regulators may have at last found their point of entry into the market 

for Internet regulation. 

That certainly seems to be the attitude adopted by the once-moribund U.S. 

Federal Trade Commission (FTC), which began a series of workshops in late 

2009 aimed at exploring “the privacy challenges posed by the vast array of 21st 

century technology and business practices that collect and use consumer data.” 5 

On January 28th, 2010, which was also dubbed Data Privacy Day by the nonprofit 

group The Privacy Projects, 6 the second workshop in the FTC’s threepart 

7 series took place at the University of California, Berkeley campus. 

Attendees heard from government, business, and public interest speakers on 

3 Recent examples include the Google Maps drive-by, see Robert Graham, Technical Details of 

the Street View WiFi Payload Controversy, ERRATA SECURITY, May 19, 2010, 

http://erratasec.blogspot.com/2010/05/technical-details-of-street-view-wifi.html, 

Facebook’s on-going changes to its privacy policy and user options, Twitter’s FTC 

settlement, see Press Release, Federal Trade Commission, Twitter Settles Charges that it 

Failed to Protect Consumers’ Personal Information; Company Will Establish Independently 

Audited Information Security Program, June 24, 2010, 

http://www.ftc.gov/opa/2010/06/twitter.shtm, the botched launch of Google Buzz, see 

Danny Goodwin, Google to Pay $8.5 Million in Buzz Privacy Class Action Settlement, 

SEARCHENGINEWATCH.COM, Nov. 3, 2010, http://blog.searchenginewatch.com/101103- 

081738, shocking behavior on Chatroulette, the conviction of Google executives in an Italian 

court over a video showing the bullying of a minor with disabilities by a Google Video user, 

see Reuters, Italy Convicts Google Execs for Down Syndrome Video, Feb. 24, 2010, 

http://www.wired.com/epicenter/2010/02/google-executive-convicted-in-italy-fordowns-video/, 

and the launch of “social shopping” websites such as Blippy and Swipely. 

4 See generally, THE LAWS OF DISRUPTION, supra note 1. 

5 See Exploring Privacy: A Roundtable Series, Fed. Trade Comm’n, 

www.ftc.gov/bcp/workshops/privacyroundtables/ [hereinafter FTC Roundtable Series]. 

6 For more information on Data Privacy Day, visit http://dataprivacyday2010.org. 

7 See FTC Roundtable Series, supra note 5.


whether and how the FTC should regulate the private collection and use of data 

to protect consumer privacy interests. The conversation that day, characterized 

by histrionic rhetoric, self-congratulatory moralizing, and an utter lack of focus, 

reflects well the current state of the so-called “privacy problem.” 

Why is the FTC holding such hearings in the first place? The agency’s charter, 

which has evolved over its long history, includes policing anticompetitive 

behavior 8 and enforcing a Congressional ban on “unfair and deceptive acts or 

practices.” So far, the FTC’s main contribution to the debate about digital 

privacy has been the drafting of non-binding guidelines for consumer notice of 

online services’ privacy policies, the so-called “Fair Information Practice 

Principles (FIPs).” 9 In the United States, the adoption of the FIPs is voluntary, 

but failure to abide by them can lead to FTC enforcement. 10 

The limits of the so-called “notice” regime are pretty obvious. Consumers 

don’t read privacy policies. Even if they did, they would find them to be 

absurdly long, most of them written in some of the worst legalese I’ve ever 

seen. 

Notice is also difficult to achieve in practice. During the last few years, 

Facebook has repeatedly landed in trouble for its mostly-admirable efforts to 

craft a working privacy regime for its now 500 million users. The generally poor 

response to these efforts, I think, stems from a growing privacy paranoia fueled 

by the media and governments, kindled by the growing pains of a company that 

by its nature deals with very personal, even intimate, information and whose 

growth rate challenges pretty much everything. 

At the same time, the company’s founder, Mark Zuckerberg, has demonstrated 

remarkably poor timing and nearly perfect political tone deafness. Even as the 

company dug itself out of criticism of a new set of privacy tools in the fall, 

Zuckerberg told an audience in January 2010 that “the social norm” for sharing 

private information had “evolved.” 11 Well, he is only 25 years old, and many of 

8 The FTC recently reached a settlement with Intel on a broad regulatory action brought 

against the company. See Press Release, Fed. Trade Comm., FTC Settles Charges of 

Anticompetitive Conduct Against Intel, Aug. 4, 2010, 

http://www.ftc.gov/opa/2010/08/intel.shtm. 

9 For more information on the Fair Information Practice Principles, visit 

http://www.ftc.gov/reports/privacy3/fairinfo.shtm. 

10 The FTC can take action if a website violates its own policy on the basis of its jurisdiction 

over deceptive practices. But there is no requirement for a website to have a privacy policy, 

and no explicit privacy protection in U.S. law for most categories of personally-identifiable 

information. 

11 See Marshall Kirkpatrick, Facebook’s Zuckerberg Says Age of Privacy is Over, READWRITEWEB, 

Jan. 9, 2010, http://www.readwriteweb.com/archives/ 

facebooks_zuckerberg_says_the_age_of_privacy_is_ov.php.


his comments, including these, have been unfairly taken out of context by 

mainstream, business, and technology media companies. 12 

The limits of notice become acute when the data collection device is not a 

computer. Cell phones, in addition to taking on more and more data functions, 

collect a great deal of information about the location and movements of their 

users—they have to in order to function. But many applications such as 

Loopt 13 and Foursquare 14 take advantage of GPS data to offer services not 

possible on a fixed computing device, including locating friends or providing 

location-specific ads. On the smaller screen of a cell phone, reading any 

document is difficult enough. What consumer in any case is going to read a 

separate privacy policy for every application they download? 

Even as it continues to refine and promote FIPs, the FTC has held hearings, 

workshops and other information-gathering sessions regarding emerging 

technologies that seem to raise new and worrisome privacy concerns. These 

have included Radio Frequency ID tags, targeted or “behavioral” 

advertisements, cookies and now Flash-based “super cookies.” 15 

There was plenty of evidence at the Berkeley session of these and more of what 

Microsoft’s Peter Cullen called “anxiety-based conversations.” 16 This year’s 

themes include the dangers of mobile computing, social networking, and cloudbased 

computing, as well as continuing hand-wringing over targeted or 

contextual advertising. 

The structure of these conversations doesn’t change much over time. The new 

technology is discussed by law professors, company representatives, and FTC 

12 Caroline McCarthy, Facebook Follies: A Brief History, CNET NEWS.COM, May 13, 2010, 

http://news.cnet.com/8301-13577_3-20004853-36.html; Guilbert Gates, Facebook Privacy: 

A Bewildering Tangle of Options, N.Y. TIMES, May 12, 2010, 

http://www.nytimes.com/interactive/2010/05/12/business/facebook-privacy.html. 

13 Loopt is a mobile mapping service that allows users to find local information using the 

geographic location of their mobile phones. For more information, visit 

http://www.loopt.com. 

14 Foursquare is a location-based social networking application that allows users to connect 

with friends in their geographic location. For more information, visit 

http://foursquare.com. 

15 The super cookie “problem,” piled on by many speakers at the Berkeley workshop, turned 

out to be a red herring, easily controlled by engineered fixes to major browsers. It’s hard not 

to read too much into that example. See Berin Szoka, Privacy Innovation: Adobe Flash Supports 

Private Browsing & Deletes Flash Cookies, THE TECH. LIBERATION FRONT, Feb. 17, 2010, 

http://techliberation.com/2010/02/17/privacy-innovation-adobe-flash-supportsprivate-browsing-deletes-flash-cookies/. 

16 Private Conversation with Peter Cullen, Jan. 28, 2010 (notes on file with author).


staff (everybody but the engineers who know how the technology works), who 

try to point out the benefits and mostly dangers of the technology. 

Extremists on either end of the spectrum call for the FTC to either ban or 

ignore the development. In typical fashion, satisfied that all sides have been 

heard, the agency takes the problem under advisement, and then waits for the 

next crisis du jour to emerge. No legislation or regulations are enacted. 

Well, that’s probably just as well—assuming there is no real privacy crisis that 

needs to be addressed, or rather that needs to be addressed by an organization 

with the FTC’s institutional limitations. I mean no disrespect to the FTC’s hardworking 

staff. By institutional limits, I am thinking of the inherent constraints 

on a U.S. regulatory agency. These include the mismatch of a national regulator 

supervising behavior that is natively global, problems in revising rules and 

jurisdictions for an environment that is evolving at accelerating speeds using a 

process designed to be slow and deliberative, and the dangers of solving what 

are largely engineering problems with staff whose expertise is policy—and 

offline policy at that. 

It’s probably clear that I don’t think there is a crisis. But I admit that it’s 

difficult to know. Both sides in this non-debate have an unfortunate habit of 

relying on unscientific survey data and anecdotal evidence, the latter of which, 

on closer inspection, turns out to be highly incomplete if not urban myth. Pam 

Dixon of the World Privacy Forum, for example, told a story about grocer 

Whole Foods using facial recognition software in stores to collect data from the 

tomato aisle for what Dixon called “direct marketing purposes.” But according 

to the Forum’s own report, The One Way Mirror Society, (whose investigation of 

the Whole Foods story was limited to parsing company press releases) the 

software could at best distinguish gender, not specifically identify customers. 17 

Direct marketing is targeted to an individual (or their likely interests) rather than 

demographic characteristics such as gender. So whatever Whole Foods was 

doing, it wasn’t direct marketing. 

The surveys that purportedly show a privacy panic are, for the most part, poorly 

constructed and unscientifically executed. For some reason law professors and 

their students, who have at best a casual acquaintance with the methods and 

rigors of any social science, are the ones called on by private and public actors 

to conduct these studies. 

Consider, for example, two questions about the same technical feature of cell 

phones. If you ask for an emotional response to the statement, “My cell phone 

17 Pam Dixon, The One-Way-Mirror Society: Privacy Implications of the New Digital Signage Networks, 

WORLD PRIVACY FORUM, Jan. 27, 2010, 

http://www.worldprivacyforum.org/pdf/onewaymirrorsocietyfs.pdf.


tracks where I go” you’ll get one answer, but if you phrase it, “My cell phone 

can tell me where I am,” you’ll get a very different result. It’s the same feature. 

The “findings” are useless. The choice and wording of the questions are the 

only valuable information, in that it reveals a great deal about the authors of the 

survey. But the surveys tell us nothing about the respondents or the choices 

they would make when faced with real-world tradeoffs between restricting data 

and the benefits that flow from it—such as getting more relevant ads or a 

greater quality and quantity of “free” online content and services supported by 

advertising revenue. 

Privacy: Defining the Problem 

There’s a bigger problem here, and that is with the terms of the debate. In most 

conversations, no one knows what anyone else means by “privacy,” or what 

information is included in the term “personally-identifiable information,” which 

drives much of the privacy regulations in the European Union. The discussion 

at the FTC’s Berkeley roundtable, as with all privacy discussions, conflated 

several different information problems into one, freely mixing and matching 

issues and regulatory solutions that don’t actually go together. Until we separate 

the problems and solve them individually, the chances for meaningful policy 

solutions are nil. 

To start with, it’s essential to understand the unique properties of information 

as an economic good. Information has very different properties from 

traditional commodities such as farm products, timber, and oil. Information 

can be used simultaneously by everyone, for one thing, and when we’re done 

using it, it’s still there, potentially more valuable because of the use. These 

remarkable features make information what economists call a non-rivalrous or 

“public” good, and they are the main reason that information now drives 

economic activity in much of the developed world. (The other is the continued 

decline in the cost of computing power.) 

So rather than talking about who “owns” private information or who is 

“stealing” data (words that make more sense when talking about traditional 

commodities), I find it much more constructive to talk about whether any 

particular use of information is “productive” or “destructive” A productive 

use of information is one that makes it more valuable, including collaboration, 

remixing, and validation. Destructive uses leave the information less valuable, 

and include misrepresentation, misidentification, and dilution. 18 

18 Some information uses include both productive and destructive elements. Arguably, 

remixing and other information sampling adds value to information protected by copyright 

and trademark law while potentially diluting markets the law protects on behalf of the 

information producer. See Larry Downes, Viacom v. YouTube: The Principle of Least Cost 

Avoidance, The Tech. Liberation Front, June 26, 2010,


As I argue in my book, The Laws of Disruption, privacy laws are best understood 

as legal protections against destructive uses of information by different 

categories of users. Seen that way, there is not one overarching and 

overwhelming privacy problem, but several very different privacy problems, 

each deserving of particular analysis and, one hopes, its own resolution. Here 

are the main categories of destructive information uses: 

Information User Destructive Uses 

Criminals Identity theft, phishing, malware and other forms of 

fraud 

Commercial 

enterprises 

Other consumers, 

friends, family 

Government and 

other state actors 

Surreptitious collection of consumer information for sale 

or use in marketing, often without adequate 

compensation or revenue sharing with the consumer 

Stalking, bullying, accidental or intentional disclosure of 

embarrassing or secret information 

Unlawful search and seizure, accidental disclosure 

News media Publication of defamatory or erroneous information that 

damages reputation 

Employers and 

business associates 

Insurers and health 

care professionals 

Eavesdropping and other monitoring to identify poor 

performance, violation of employer rules, or business 

secrets 

Collection and use of known and potential risks to 

determine coverage or the danger of accidental 

disclosure 

Though many of these destructive uses were discussed at the FTC hearing, it 

should be noted at the outset that the agency’s charter only extends to the first 

and second category. 19 (To be fair, FTC staff frequently reminded the speakers 

to limit their discussion to topics over which the agency had jurisdiction.) Why, 

then, did the speakers repeatedly bring up all the others? For one thing, some 

http://techliberation.com/2010/06/26/viacom-v-youtube-the-principle-of-leastcost-avoidance/. 

For now, I’ll stick to the “easier” problems of uses that are almost purely 

destructive. 

19 See Fed. Trade Comm’n, Privacy Initiatives, Introduction, http://www.ftc.gov/privacy. As the 

agency explains the scope of its privacy initiatives, “The Federal Trade Commission is 

educating consumers and businesses about the importance of personal information privacy, 

including the security of personal information. Under the FTC Act, the Commission guards 

against unfairness and deception by enforcing companies’ privacy promises about how 

they collect, use and secure consumers’ personal information. Under the Gramm-Leach- 

Bliley Act, the Commission has implemented rules concerning financial privacy notices 

and the administrative, technical and physical safeguarding of personal information, and it 

aggressively enforces against pretexting. The Commission also protects consumer privacy 

under the Fair Credit Reporting Act and the Children’s Online Privacy Protection Act.”


of the most lurid stories suggesting a privacy crisis come from the other 

categories, making them irresistible to those arguing for a crisis response. 

Unfortunately, most parties in the privacy “debate” so far have shown little 

interest in cabining the discussion to manageable and discrete problems when 

an emotional point is there to be scored. 

Indeed, one important reason to evaluate the categories of destructive 

information use separately is to help us see that some goals of the privacy 

movement are mutually exclusive. In the abstract, for example, most people are 

uncomfortable with the proliferation of surveillance cameras in urban locations. 

But listen to the indignation that erupts the minute a serious crime or terrorist 

act occurs and the police turn out not to have caught it on film. 

Or take the often-repeated example of the victim of domestic violence, used as 

a stalking horse for the proposition that search engines, cell phone carriers, and 

other service providers, who collect bits and pieces of information that might be 

used to identify and locate an individual, should immediately purge their 

databases, lest they fall into the wrong hands. 

Turn the problem around, however, and you can make the exact opposite case. 

For the victim, it’s important to erase all traces of their online activity. But for 

the perpetrator, effective law enforcement requires as much information as 

possible. 20 Optimally, we’d like to tell information collectors to purge data 

about victims but retain it for criminals, but of course, we don’t know who is 

who until after the fact. What’s a data collector to do? 21 

This isn’t a hypothetical problem. Lawmakers in the United States and the 

European Union are simultaneously putting pressure on phone companies, 

search engines and social networking sites to both purge and retain the same 

data, a kind of whipsaw that has led these providers uncharacteristically to call 

for new laws—laws that would give them a straight answer on what is expected 

of them. 

It’s also worth noting that in the United States in particular, most of the existing 

legal protections for private information are squarely aimed at deterring 

destructive uses by criminals and by governments themselves. Every year I 

have to convince another batch of students that the right to privacy recognized 

by U.S. courts and grounded in the U.S. Constitution does not apply to conflicts 

20 Declan McCullagh, Web Searches Lead to Murder Conviction, CNET NEWS.COM, Feb. 12, 2010, 


21 See Miguel Helft, For Data, Tug Grows Over Privacy vs. Security, N. Y. TIMES, Aug. 2, 2010, 

http://www.nytimes.com/2010/08/03/technology/03blackberry.html. See also Lance 

Whitney, German Court Rules Against Data Retention Policy, CNET NEWS.COM, March 2, 2010, 

http://news.cnet.com/8301-13578_3-10462117-38.html.


with commercial enterprises, parents, friends, or the news media. (Indeed, the 

latter are strongly protected by the First Amendment against such regulation.) 

With at least a whiff of First Amendment rationale, even the common law torts 

that deal with conflicts between individuals over information use—including 

rights of publicity, defamation, and “false light” claims—have fallen into 

disrepute during the last fifty years. 22 

Here there are also economic forces at work: information technology has 

erased the temporary mask of anonymity created by 19 th century urban life, 

reviving a social goal of transparency as old as Hawthorne’s The Scarlet Letter. 23 

American life, built of equal parts frontier necessity and Puritan aspiration, calls 

for complete and accurate information about each other. That goal increasingly 

weighs more heavily in private disputes over what Warren and Brandeis in 1890 

famously termed “the right to be left alone.” 24 More about that in a moment. 

Why the focus on state actors? The principal fear of the drafters of the 

Constitution, obviously informed by the experience of the colonies, was with 

potential tyranny from government. The government, after all, practically holds 

a monopoly on the coercive power of the military and the ability to incarcerate. 

For historical reasons, the focus is very different in Europe and many parts of 

Asia, which have enacted strong privacy laws that align citizens and democratic 

governments against everybody else. The difference between U.S. and 

European law, in particular here, is perhaps the broadest effort to apply 

terrestrial laws to digital life generally. 25 

In the United States, distrust of government evolved to include a fear that 

private information could be misused to achieve the same ends as more overt 

repression. These fears were underscored by the late 20 th century scandals, 

22 See Haynes v. Alfred A. Knopf, Inc., 8 F.3d 1222 (7th Cir. 1993) (opinion by Posner). 

23 NATHANIEL HAWTHORNE, THE SCARLET LETTER (Ticknor, Reed & Fields 1850). 

24 Samuel Warren & Louis D. Brandeis, The Right to Privacy, 4 HARV. L. REV. 193 (1890), 

(discussed below). See also Peggy Noonan, The Eyes Have It, WALL ST. J., May 21, 2010, 

http://online.wsj.com/article/NA_WSJ_PUB:SB10001424052748703559004575256732 

042885638.html. 

25 See Adam Liptak, When American and European Ideas of Privacy Collide, N.Y. TIMES, Feb. 26, 

2010, http://www.nytimes.com/2010/02/28/weekinreview/28liptak.html; Kevin J. 

O’Brien, Europe Says Search Firms Are Violating Data Rules, N.Y. TIMES, May 26, 2010, 

http://www.nytimes.com/2010/05/27/technology/27data.html; Jessica E. Vascellaro, 

Ten Countries Ask Google to do More to Protect Privacy, WALL ST. J., April 20, 2010, 


879579682.html.


including The Pentagon Papers and later Watergate, which led to the first 

comprehensive anti-wiretapping law at the federal level. 26 

So, with notable exceptions, all existing U.S. privacy laws afford limited—and 

perhaps inadequate—protection to citizens against their governments. That fact is 

often whitewashed in the furor of the privacy debate. The ACLU of Northern 

California, for example, recently published a white paper called Cloud Computing: 

Storm Warnings for Privacy? 27 The paper points out the mismatch between 

existing privacy law and the reality of cloud computing, where personal 

information is turned over for storage and processing to a variety of third 

parties and often to their unnamed and changing business partners. 

The report never quite says so directly, but all the proposed reforms are aimed 

at curbing the ability of state actors, not private parties, to gain access and make 

use of data in the cloud. It is not “consumers” as the report characterizes them, 

who need to be worried about their “privacy protections” in the cloud. It is 

citizens. I think the ACLU is right to be worried about government access to 

cloud data sources, but I wish it wouldn’t pretend to have a broader agenda 

than it does. 

Though the right to privacy against government is now firmly established, it’s 

also worth remembering that this is a relatively new right. Despite all the talk of 

one’s right to privacy, you will scour the Bill of Rights in vain for any reference 

to privacy even against state actors. 

It wasn’t until the 1960s that the Supreme Court began to interpret the First 

Amendment’s free speech provisions, along with the Fourth Amendment’s 

prohibition against unreasonable searches and seizures of people and their 

property, as implying a more general right to privacy enforceable against state 

actors. 28 In key cases, including Griswold v. Connecticut (birth control), Roe v. Wade 

(abortion), and more recently Lawrence v. Texas (homosexuality), the Court 

struggled to reign in government intrusions into the private lives of citizens, 

intrusions the Founders would never have imagined possible. Lacking a 

26 See The Omnibus Crime Control and Safe Streets Act of 1968, 18 U.S.C. §§ 2510 et. seq. 

27 Am. Civil Liberties Union of N. Cal., Cloud Computing: Storm Warnings for Privacy?, Jan. 2010, 

http://www.dotrights.org/sites/default/files/Cloud%20Computing%20Issue%20Pa 

per.pdf. 

28 See, e.g., Griswold v. Connecticut, 381 U.S. 479, 483 (1965) (“In other words, the First 

Amendment has a penumbra where privacy is protected from governmental intrusion.”); 

Roe v. Wade, 410 U.S. 113, 152 (1973) (“The Constitution does not explicitly mention any 

right of privacy. In a line of decisions, however, going back perhaps as far as [1891], the 

Court has recognized that a right of personal privacy, or a guarantee of certain areas or 

zones of privacy, does exist under the Constitution.”).


Constitutional provision that guaranteed the right to privacy for basic human 

activities, the Court more-or-less invented one. 

At the same time, as 20 th century society came to recognize that information 

was a kind of property (something with value, in any case), the perception of a 

right to privacy emerged. It was found, to use Justice Douglas’s famous but 

unfortunate phrase, in “the penumbras and emanations” of the Bill of Rights. 29 

But to reiterate, the right to privacy, as it is currently understood, is a right to be 

free of unreasonable interference from government, not from each other or the 

businesses with which we interact. 

Some of today’s most vocal privacy advocates are calling for a broader right of 

privacy, one that could be asserted against any or all of the destructive 

information uses and perhaps against many of the productive uses as well. 

An earlier effort to create such a right, it is worth noting, failed. In 1890, 

Samuel Warren and Louis Brandeis wrote a famous article for the Harvard Law 

Review in which they called for the creation of a general right of privacy, defined 

as the “right to be left alone.” 30 The article was inspired by Warren’s personal 

experience. Warren’s wife was the daughter of a U.S. Senator, and Warren was 

appalled to discover that their daughter’s wedding was reported, with 

photographs, in The Washington Post. 31 

Warren and Brandeis’s proposal led to some experimentation, mostly at the 

state level and mostly through judge-made law. Common law courts invented 

new tort injuries for damage to reputation, rights of publicity, and more 

expanded forms of defamation, including the portrayal of someone in a “false 

light.” Many of these rights are no longer recognized, or have become nearly 

impossible to enforce. 

Under the Supreme Court’s 1964 decision in New York Times v. Sullivan, 32 for 

example, public officials must prove actual malice to recover damages for 

defamation. Extending that decision, the Court held in the Florida Star 33 case 

that a rape victim could not prevent a newspaper from publishing information 

from a police report about the crime committed against her. More recently, 

29 Griswold, 381 U.S. at 484. Justice Thomas, a strict constructionist, has a sign on his desk 

asking visitors to kindly keep their penumbras off his emanations. 

30 Warren & Brandeis, supra note 24. 

31 What would Warren think of today’s celebrity media, or reality TV, in which non-celebrities 

volunteer to give themselves the celebrity treatment? 

32 376 U.S. 254 (1964). 

33 Florida Star v. B.J.F., 491 U.S. 524 (1989).


homeowners have been rebuffed in efforts to prohibit Google maps from 

photographing their homes. 34 

The Warren and Brandeis experiment failed for a good reason, at least from an 

economic standpoint. Many of the uses of private information that Warren and 

Brandeis sought to outlaw are in fact productive uses, generating substantial 

social value that outweighs the costs to the individuals of keeping such 

information public. 

It may be very important for you to keep secret the fact that you have a criminal 

record, a communicable disease, a Swiss bank account or a secret liaison with an 

employee. But it is more valuable to the rest of us to know these things about 

you, if only to know how many others have the same attribute so we can take 

appropriate actions—quarantine the sick, pass stricter banking laws, etc. The 

benefits of disclosure, the courts have determined, generally outweigh the costs 

of secrecy. 

Warren and Brandeis weren’t entirely wrong, however. It’s important to note 

that the same “private” information can also be used destructively. I might 

overestimate the risks of hiring a former felon, for example, or even exclude 

potential tenants for my apartments based on an irrational reliance on personal 

traits and associated stereotypes. That’s why we have anti-discrimination laws, 

one of the notable exceptions where protections for misuse of personallyidentifiable 

information extend to commercial and other non-governmental 

users. But it is an exception, narrowly focused on a particular abuse. 

Generally speaking, in fact, privacy legislation in the U.S. has only been enacted 

when legislators find particular and persistent market failures—failures, that is, 

to use personally-identifiable information in rational ways. Along with antidiscrimination 

laws, we have laws that control private information use in credit 

reporting, medical records, and identity theft and other forms of financial 

fraud. 35 In each of these cases, the law is focused on a particular destructive use 

in a particular context, with remedies for consumers narrowly-tailored to leave 

as much information unprotected as possible. 

Privacy: The Solution 

Particular solutions to particular information use failures are, I believe, a model 

lawmakers ought to be encouraged to continue following. 

34 See Steven Musil, Google Wins Street View Privacy Suit, CNET NEWS.COM, Feb. 18, 2009, 


35 See Larry Downes, If Feds Fail, What Can Stop Identity Theft? CIO INSIGHT, July 2005, 

http://www.cioinsight.com/c/a/Past-Opinions/If-Feds-Fail-What-can-Stop- 

Identity-Theft/.


But maybe it’s time to revisit Warren and Brandeis’s call for broader privacy 

protections against non-state actors. The argument in favor would go 

something like this: Now that the cost of information retention is less than the 

cost of information deletion, commercial enterprises may soon wield the same 

kind of coercive power that until now has usually been the domain of 

governments. 

Perhaps the same kinds of risks to society that led the courts to recognize a 

“zone of privacy” for information collection and use now justify extending that 

zone to the complicated web of enterprises that collect, consolidate, and resell 

information about a wide range of consumer behaviors—as well as to 

companies that by design collect intimate information, including social 

networking sites and anything mobile. 

But note that even if this argument carries the day—that is, if the benefits of 

reducing destructive uses of private information in some of the other categories 

exceed the costs (including inadvertent limits on some productive uses, such as 

the advertising that supports so much “free” media), it doesn’t necessarily 

follow that the FTC or any other government entity is the right institution to 

define, enact, and enforce those new legal rights. 

Again, the FTC’s authority is limited to investigating anticompetitive behavior 

and “unfair” or “deceptive” trade practices—terms that have clear meanings 

under the agency’s statute, its previous decisions and policy statements, and 

court cases interpreting the law. 

Moreover, there is the agency’s tendency simply to react with dismay to new 

technologies and then to move on when the technology soon after resolves its 

own problems. This habit—not limited to the FTC by any means—leads me to 

doubt its institutional capacity to define what kind of new privacy rights 

consumers should have, independent of always-changing technological 

capacities. The FTC’s staff invokes the phrase “privacy by design” like a mantra 

in conversation with policy representatives and consumer advocates. The hope 

behind that phrase is that future technologies can be engineered to protect 

privacy interests. But the phrase is meaningless without first defining what 

interests are to be protected. 

There are bigger reasons to question whether federal or state government is 

well-suited to the role of privacy cop. For one thing, competing government 

interests in effective law enforcement create a kind of regulatory schizophrenia 

over the use of privacy-enhancing technologies. As an example, consider 

encryption. Given that one common theme of destructive information use in 

several categories, as noted in the chart above, involves accidental disclosure 

(including access gained illegally by hackers or by fraud), it might be thought 

that more widespread use of encryption technology would reduce those risks. 

But the federal government has been ambiguous at best about encouraging


enterprises to adopt encryption. In the 1990s, basic encryption methods were 

classified as a “munition” by the Department of State (and later, the 

Department of Commerce). Phil Zimmerman, the inventor of an open source 

encryption protocol known as Pretty Good Privacy, was the subject of a lengthy 

criminal investigation for exporting “weapons” without a license. Only under 

intense pressure from the courts in a series of First Amendment decisions did 

the Commerce Department finally liberalize these export controls. 

There was also the memorable fight in the early 1990s over the Clipper Chip, a 

government-developed encryption technology for use in cell phones. 36 Had the 

chip been made mandatory or even widely deployed (opponents, including the 

Electronic Frontier Foundation (EFF), successfully prevented that outcome), it 

would have given law enforcement a back door into mobile phone calls and 

might have led to the prohibition of other encryption technologies. 37 

Federal and state governments also receive failing grades at adopting and using 

the kinds of safe data handling practices some think these governments should 

enforce against everyone else. That failure is evidenced by numerous 

embarrassing breaches and stolen unsecured laptops containing millions of 

records of sensitive citizen data. Government in this sense offers a good 

example (several, actually) of how not to manage privacy and the worst, not best, 

safe handling practices. Finally, a general problem of state, as opposed to 

federal, legislation is the potential for fifty different definitions of privacy, all 

imposed on what is a global information economy. 

More broadly, there’s a disturbing irony to handing privacy enforcement over to 

governments, an argument I have also made with regard to enforcement of net 

neutrality principles. Enforcement powers would require the agency to examine 

the information and how it was disclosed. For agencies to investigate and 

punish banned uses of private information, they would necessarily need to have 

access to the data sources in question. 

If I complain, to the FTC, for example, that a social networking enterprise is 

selling my information in violation of some future privacy protection law, how 

else will the agency evaluate my claim without looking deeply into the 

company’s information practices? Like any good audit, they will need to follow 

the flow of information from beginning to end to determine what, if any, 

violations of law occurred. 

36 For more information on the Clipper Chip, including government documents and public 

response, visit http://epic.org/crypto/clipper. 

37 The Clipper Chip fight was instrumental in the formation of EFF, proving once again the 

law of unintended consequences. The government lost the battle, but more importantly, it 

inadvertently helped to organize a permanent and effective opposition.


Yet state actors, as noted above, are currently the most restricted from seeing 

this information. In a bitter irony, enforcement of privacy rights against 

enterprises would have the effect of exposing data to the government that it 

otherwise would be forbidden to see. As a result, there is the potential for 

abuse of that privilege in the name of law enforcement or other government 

priorities (e.g., terrorism or tax fraud). 38 

Who else can define, let alone enforce, new privacy protections against 

enterprises and other non-state actors? One obvious alternative is selfregulation 

by the enterprises themselves, perhaps encouraged by the threat of 

government intervention if the market fails. 

There’s already a good deal of self-regulation, including voluntary adoption of 

the FIPs and certification from third parties such as TRUSTe 39 and the Better 

Business Bureau. 40 Much of this self-regulation, however, presumes the 

usefulness of the notice regime which, as noted, no one presumes any more. 

The potential is there, however, for more effective self-regulation. Even 

companies who make money from information collection and its use are 

worried about the privacy problem, or at least the perception of one. Some are 

even calling for new legislation, in part to solve the whipsaw problem described 

above. 

Microsoft and other companies who are counting on cloud computing as the 

next major computer architecture are eager for regulatory frameworks that will 

both allay consumer anxiety and give cloud-based service providers safe harbors 

in which to develop their offerings. 

An emerging consensus among a wide range of technology companies and 

information service providers would move the discussion from guidelines about 

private information notice to rules about information use. The “use” approach 

would develop acceptable principles for how collected and retained data (private 

or otherwise) can be used, by whom, and with what controls reserved to 

consumers to limit or block those uses. 

As part of the Business Forum for Consumer Privacy (BFCP), Microsoft and 

others are calling for a legislated framework for use-based rules. 41 What does 

38 See Declan McCullaugh, Amazon Fights Demand for Customer Records, CNET NEWS.COM, April 

19, 2010, http://news.cnet.com/8301-13578_3-20002870-38.html. 

39 See http://www.truste.com. 

40 See http://www.bbb.org/us/Business-Accreditation. 

41 The Business Forum for Consumer Privacy, A Use and Obligations Approach to Protecting Privacy: 

A Discussion Document, Dec. 7, 2009,


that mean? According to a 2009 BFCP white paper, a “use‐and‐obligations” 

model requires all organizations to be transparent, offer and honor appropriate 

choice, ensure that risks to consumers related to use are assessed and managed, 

and implement security for the information they maintain. 42 The white paper 

describes a taxonomy of use types (for example, marketing, internal operations) 

within the enterprise category, and tries to define an appropriate set of default 

rules that should apply for each. 

The BFCP approach is certainly more productive than the anxiety-based 

conversations. Focusing on use, for one thing, moves the conversation away 

from the emotional subject of privacy to the more rational subject of propriety, 

by which I mean the recognition that both enterprises and consumers 

participate in the creation of valuable new sources of data and both should have 

rights to monetize that value. Consumers ought to be able to buy out 

enterprises for productive uses they don’t want, and vice-versa. 

In this scenario, privacy policy evolves from self-regulation by the market to 

actually becoming a market for private and other data. Assuming this market 

works, data will be put to the most highly-valued productive uses, and those 

uses that are not valued or are destructive will not be implemented. 43 As with 

any market, the government will stand in the background to ensure the rules are 

obeyed and to intervene when market failures occur. 

One immediate concern about the creation of a privacy market is that 

consumers will have no voice in its creation or operation. Consumers, after all, 

are individuals, easily overwhelmed by the economic might of large 

corporations, who can be expected to develop rules that give consumers an 

unfair disadvantage in the information marketplace. 

That would have been an entirely reasonable concern in the pre-digital age, and 

one of the principal justifications for the creation of private and public 

consumer watchdog groups such as the FTC in the first place (the agency is 

nearly 100 years old). Such groups, at least in theory, lobby and sue on behalf 

http://www.huntonfiles.com/files/webupload/CIPL_Use_and_Obligations_White_ 

Paper.pdf. 

42 See Hunton & Williams LLP, Business Forum for Consumer Privacy Introduces New Data Protection 

Model, Dec. 21, 2009, 

http://www.huntonprivacyblog.com/2009/12/articles/events/business-forum-forconsumer-privacy-introduces-new-data-protection-model/. 

43 This emerging privacy market, operating with minimal transaction costs or “friction,” could 

present a wonderful opportunity to test Ronald Coase’s theory that absent transaction costs, 

stakeholders will necessarily bargain to the most productive use of a given resource. See 

Ronald Coase, “The Problem of Social Cost,” 3 Journal of Law & Economics 1 (1960) (the 

so-called “Coase Theorem”).


of large groups of consumers to ensure their interests are fairly represented in a 

wide variety of market activities. (The class-action mechanism is another 

example.) To use the economic terminology, these legal constructs overcome 

the collective action problem of individual users whose individual losses may be 

too small to justify the expense of enforcing or even negotiating rights. 

But the same technologies that create the privacy problem are also proving to 

be the source of its solution. Even without government intervention, 

consumers increasingly have the ability to organize, identify their common 

demands, and enforce their will on enterprises. 44 

Facebook’s journey to a fair privacy policy continues to be instructive here. 

The company’s ongoing privacy crisis actually started in early 2009, when 

Facebook announced what was in fact a modest change to its terms of service. 

Some users who read the modification objected to it, and used the very tools 

Facebook provides for group formation to create a “People Against the New 

Terms of Service” page, which quickly signed up 100,000 fans. 45 (Still a 

relatively small number given Facebook’s size: now over 500 million users 

worldwide.) 

The revolt and (perhaps more influential) the ensuing bad publicity led 

Facebook to withdraw the changes—no lawsuit or legislation necessary. Even 

more, the company soon announced that it was changing its entire approach to 

governance. In the future, the company said, it would rewrite its user 

agreement to be understandable to lay readers, and circulate future changes as 

proposals to the entire population of users. 

If enough users objected to a proposed change, the changes would be put to a 

vote. (The changes to privacy settings that set off the firestorm in the fall of 

2009 were circulated ahead of time, but didn’t qualify for a vote, perhaps 

suggesting that Facebook’s “constitution” still needs some tweaking.) 

So assuming that buyers and sellers have equal access and power, how would a 

market for private data work in practice? 46 There are already some good 

44 Steve Lohr, You Want My Personal Data? Reward Me for It, N.Y. TIMES, July 16, 2010, 

http://www.nytimes.com/2010/07/18/business/18unboxed.html; L. Gordon Crovitz, 

Privacy Isn’t Everything on the Web,” WALL ST. J., May 24, 2010, 


054326304.html. 

45 To view the Facebook page, visit 

http://www.facebook.com/group.php?gid=77069107432. 

46 Venture capitalists have begun to recognize the potential of privacy markets and are 

investing accordingly. See Pui-Wing Tam & Ben Worthen, Funds Invest in Privacy Start-ups, 

WALL ST. J., June 20, 2010,


prototypes in place in the form of consumer loyalty and affinity programs that 

have been around for decades. The grocery store scanner data can already 

determine what items have been bought at the same time, but in order to tie 

that collection to particular demographic characteristics (gender and zip code 

are the most important for marketing purposes), it requires the cooperation of 

the customer. 

So, in exchange for using a club card and giving the store permission to connect 

purchases to a particular customer, the consumer gets special discounts. In 

effect, the store recognizes the value of the identifying information and pays the 

consumer for it. 

That approach is attractive for several reasons. For one thing, it is entirely 

voluntary—consumers make the decision to sign up for the programs or not, or 

even whether or not to participate in any individual transaction. This protomarket 

also operates under very low transaction costs. Instead of negotiating 

for the purchase of each individual data element with each individual consumer, 

a framework is established that allows both parties to opt out at will. 

Discounted prices become a new form of currency requiring no oversight. No 

lawyers and no regulators are involved. 

The loyalty program also makes explicit what I think is the most important 

feature of the privacy debate, one that is usually lost in the extreme rhetoric of 

both sides. Information only attains its true monetizable value when every 

participant in the transaction has an incentive to provide it, authenticate it and 

protect it. Retail data is of limited use to the supply chain without demographic 

linkages. Individual data is not marketable unless someone is able to collect and 

analyze large volumes of it. As with all information exchanges, the whole is 

greater than the sum of its parts. 

Today, even fifty years into the computer revolution, most aspects of retail 

engineering, including production planning, pricing and promotions, product 

design, forecasting, and marketing, operate with precious little real information. 

Most are more black arts than science. With the potential capture of complete 

life-cycle transaction data, there is at last the hope of discipline, and with it great 

increases in efficiency. 

And in the next digital decade, we now have the potential for information to be 

collected post-sale: How did you like the product? Did it perform as expected? 

What other products did you use it with? It becomes even clearer that without 


025721578.html.


cooperation from the consumer, no real value will come from vast data 

warehouses that enterprises have built but rarely use effectively. 47 

The privacy marketplace is already here, and there is every indication that as 

technology continues to evolve, an increasingly robust and valuable set of 

institutions will develop alongside of it. As new forms of data enter the digital 

domain, new tools and techniques will emerge to harness their value and 

allocate it among those who develop it. 

In order for the technology of information processing to reach its potential, 

however, the histrionics of the privacy debate must stop. Instead, consumers 

must be encouraged and educated to think about information use in terms of 

productive and destructive costs and benefits. In short, we must learn to think 

rationally about information value. If the FTC and other regulators are looking 

for a role in solving the problems of online privacy, a good starting point would 

be to contribute constructively to the emergence of this organic, elegant 

solution. 

47 See LARRY DOWNES, THE STRATEGY MACHINE: BUILDING YOUR BUSINESS ONE IDEA AT A 

TIME, (Harper Business 2002), Chapter 3, The Information Supply Chain. See also Jules 

Polontesky & Christopher Wolf, Solving the Privacy Dilemma, THE HUFFINGTON POST, July 27, 

2010, http://www.huffingtonpost.com/jules-polonetsky/solving-the-privacydilem_b_660689.html.


CHAPTER 9 

CAN SPEECH BE POLICED 

IN A BORDERLESS WORLD? 

529 

The Global Problem of State Censorship 

& the Need to Confront It 531 

John G. Palfrey, Jr. 

The Role of the Internet Community 

in Combating Hate Speech 547 

Christopher Wolf

530 CHAPTER 9: CAN SPEECH BE POLICED IN A BORDERLESS WORLD?


The Global Problem of State 

Censorship & the Need 

to Confront It 

By John G. Palfrey, Jr. * 

Speech is policed through technical Internet filtering in more than three dozen 

states around the world. This practice is increasingly widespread. States 

including China, Iran, Syria, Tunisia, and Uzbekistan have extensive Internet 

filtering regimes in place. Censorship using technological filters is often 

combined with restrictive laws related to what the press can publish, opaque 

surveillance practices, and severe penalties for people who break the state’s rules 

of using the Internet. This trend has been emerging, and documented with 

precision, for nearly a decade. 1 

An empirical study of technical Internet filtering tells only part of the story, 

however. Speech is policed actively in parts of the world with regimes that are 

substantially more democratic than China or Iran. Through mechanisms that 

include surveillance, “encouraging” self-censorship, intellectual property 

restrictions, and defamation laws, virtually every state in the world polices 

online speech through multiple means. 2 As more and more of everyday life 

moves onto the Internet, so has regulation of that activity. These forms of 

regulation are driven by the same types of concerns that animate the regulation 

of speech in traditional environments. 

* John Palfrey is the Henry N. Ess III Professor of Law at Harvard Law School and Faculty 

Co-director of the Berkman Center for Internet & Society at Harvard University. This 

chapter draws upon research by the OpenNet Initiative, which is a collaboration that joins 

researchers at the Citizen Lab at the Munk Centre, University of Toronto (Prof. Ron 

Deibert, principal investigator), the SecDev Group (formerly the University of Cambridge 

where Rafal Rohozinski is principal investigator), and the Berkman Center (where the author 

and Jonathan Zittrain are co-principal investigators). The author is grateful to the large 

number of researchers who have participated in gathering, over nearly a decade, the data on 

which this chapter draws. Parts, though not all, of this argument have been published in 

other volumes. 

1 See www.opennet.net for the results of the OpenNet Initiative’s research since 2002. See 

also Ronald Deibert, John Palfrey, Rafal Rohozinski, & Jonathan Zittrain, eds., ACCESS 

DENIED: THE PRACTICE AND POLICY OF GLOBAL INTERNET FILTERING (MIT Press, 2008) 

and Ronald Deibert, John Palfrey, Rafal Rohozinski, & Jonathan Zittrain, eds., ACCESS 

CONTROLLED: THE SHAPING OF POWER, RIGHTS, AND RULE IN CYBERSPACE (MIT Press, 

2010), in which variations on these themes appear throughout. 

2 Like technical Internet filtering, this is not a new phenomenon. See Adam D. Thierer, 190 

Internet Censors? Rising Global Threats to Online Speech, 38 TechKnowledge, July 26, 2002, 

www.cato.org/pub_display.php?pub_id=11535.

532 CHAPTER 9: CAN SPEECH BE POLICED IN A BORDERLESS WORLD? 

The social, political, and cultural issues that give rise to online speech 

restrictions are familiar. Child safety is one of the primary drivers of online 

speech regulation, most commonly related to pornography. Police use 

surveillance to track the misdeeds of wrong-doers, in turn causing a chilling 

effect on online activities. Business people, authors, musicians, and others want 

their intellectual property rights vindicated to the fullest extent, prompting 

extensive intellectual property restrictions online. Those who perceive harm to 

their reputations seek the extension of defamation laws to the digital world. In 

each of these instances, speech restrictions cross traditional and digital 

environments more or less seamlessly. 

Is It True that “All Politics Is Local” 

in the Digitally-Mediated World? 

The problem of policing speech on the Internet arises at every level of 

government, from local political conversations about community norms to the 

international debate over Internet governance. The principal difference at these 

multiple levels of governance is the willingness to address the difficult problems 

of when to regulate speech online and the ramifications that flow from doing 

so. State legislatures fight over whether to constrain speech in schools when 

young people engage in cyber-bullying. 3 In the United States, Congress takes up 

the same issue from the perspective of federal law enforcement and funding for 

schools and education. At the same time, the topic of speech-based controls is 

an important aspect of global Internet regulation. However, it rarely makes it 

into the agenda at the international level as a genuine, openly-discussed matter. 

At the local level, there is often substantial appetite to regulate speech online, 

primarily as a response to fears about child safety. The clearest example at the 

local level in the United States is the effort to curb cyber-bullying through 

school environments, which states, regional, and municipal governments seek to 

regulate. 4 Speech, largely online in social network environments, is policed by 

local regulation that bars young people from expression that may harm their 

peers psychologically. While the debate rages as to the appropriateness of these 

regulations, dozens of states in the United States have enacted, or have 

considered enacting, these types of laws at the local level. Courts have split as 

to the constitutionality of such provisions. In short, intense local concerns, 

3 The First Amendment Center published a fine series of essays and research materials related 

to the speech restrictions associated with the anti-bullying efforts, with special attention paid 

to state legislative activities. See First Amendment Ctr., Online Symposium: Cyberbullying & 

Public Schools, 

www.firstamendmentcenter.org/collection.aspx?item=cyberbullying_public_schools. 

4 Sameer Hinduja, & Justin W. Patchin, Cyberbullying Research Center, State Cyberbullying Laws: 

A Brief Review of State Cyberbullying Laws and Policies, July 2010, 

www.cyberbullying.us/Bullying_and_Cyberbullying_Laws_20100701.pdf.


such as about how young people treat one another, are driving legislative 

attempts to regulate online speech through school-based enforcement. 5 

At the federal level, online speech restrictions have arisen repeatedly as 

proposals and as enacted law, in the United States and around the world. 

Twenty-five years after the creation of the .COM top-level domain, it is 

apparent that national governments can, and often do, assert sovereignty over 

the acts of their citizens in the online environment, including limitations on 

their speech. At a basic level, in the United States and in many other 

jurisdictions, speech that is deemed to be harmful in the offline world is 

considered equally unlawful in the online environment. For the speaker, there is 

no free pass simply because the utterance in question appears online. 

The primary difference between the policing of speech online and offline lies in 

terms of how intermediaries are treated. Under United States federal law, and 

the national law of many other jurisdictions, intermediaries that enable people 

to publish speech online are exempt from liability in most cases, for example, 

from claims of defamation under Section 230 of the Communications Decency 

Act. 6 There are exceptions to this rule, even in the United States: Criminal 

matters and copyright violations fall outside of the statute’s safe harbor 

provisions. 7 In other jurisdictions, regulation of intermediaries may soon be the 

law. For instance, in Sweden, the Data Inspection Board issued a report in July 

2010 asserting that companies offering social media services, such as blogs, 

Facebook, or Twitter, have a legal obligation to monitor personal data posted to 

the pages on their site. 8 

At the international level, we observe extensive policing of speech online, but 

discussion of the issues involved is largely invisible—or else not happening at 

all. While there is extensive and healthy debate about many aspects of the 

problem of Internet governance, the discussion does not reach the hard 

problems of when online speech regulation should be permitted. There are 

many issues worthy of the attention of the many capable minds focused on 

Internet governance, but the topic of speech regulation rarely makes the list of 

what is, in fact, publicly discussed and vetted seriously. 9 The primary focus for 

5 See Ronald Collins, A Look at “Cyber-bullying and Public Schools,” March 31, 2009, 

www.firstamendmentcenter.org/analysis.aspx?id=21410. 

6 47 U.S.C. § 230. 

7 See id. 

8 See Swedish Data Inspection Board Report, July 5, 2010, www.datainspektionen.se/inenglish/ 

(on intermediary liability); see also Companies Responsible for Social Media Content, THE 

LOCAL: SWEDEN’S NEWS IN ENGLISH, July 5, 2010, www.thelocal.se/27606/20100705/. 

9 The International Telecommunication Union (ITU), official host of the World Summit on 

the Information Society (WSIS) in Geneva, has held several events designed to refine the 

debate further. Through these events, the ITU has convinced dozens of observers to


Internet governance discussions continues to be issues related to the 

management of Internet resources, including the domain name system and 

related policy issues. Discussion of the non-profit Internet Corporation for 

Assigned Names and Numbers (ICANN) continues to play a central role. 

ICANN occupies an arcane bit of turf—essentially, the port allocation business. 

That is important in some respects but does not appear to concern most users 

of the Internet, particularly in a world in which most people find Internet 

resources through search engines and, increasingly, mobile devices and 

applications. 10 As an example, within the context of the Internet Governance 

Forum 2009 meeting in Egypt, the first substantive panel of the event was 

devoted to traditional ICANN-related matters such as the transition from 

Internet Protocol version 4 (IPv4) to IPv6 and the addition of new top-level 

domains (TLDs). 11 Possible topics for consideration, other than ICANN 

reform and these highly specific technical issues, each more important to the 

end-users of the Internet and their sovereigns, have included a fund for 

developing countries to build Internet infrastructure, the quandary of what to 

do about spam, and a cluster of intellectual property concerns. 

Internet speech restrictions should serve as the focal point for the world’s heads 

of state and their designees when Internet governance is on the table. While 

online speech restrictions raise a wide array of issues, a discussion of Internet 

filtering would hone in on whether states actually want their citizens to have full 

access to the Internet or not. It would help guide a public conversation about 

what is truly most important about having access to the Internet and the extent 

to which states place a premium, if at all, on the global flow of information. 

Without collective action, the Internet will likely continue to become balkanized 

into a series of local networks, each governed by local laws, technologies, 

markets, and norms. As Jonathan Zittrain argued in Who Rule the Net?, the 

predecessor of this collection, we may be headed toward a localized version of 

the Internet, governed in each instance by local laws. 12 If such a version of the 

Internet is inevitably part of our future, there ought to be open and transparent 

publish what comprises an extensive body of work on this topic on the ITU website. In 

addition, long-time experts in this field, such as Prof. Milton Mueller of Syracuse, and 

others, have constructed helpful models to structure the conversation. For suggestions on 

further information of this general nature, please see www.netdialogue.org, a joint project 

of Harvard Law School and Stanford Law School. 

10 Witness the abysmal turnout for ICANN’s election of 2000, in which a free and open 

election for five ICANN directors attracted fewer than 100,000 votes globally. 

11 Internet Governance Forum of 2009, Managing Critical Internet Resources, Transcript, Nov. 

16, 2009, 

www.intgovforum.org/cms/2009/sharm_el_Sheikh/Transcripts/Sharm%20El%20S 

heikh%2016%20November%202009%20Managing%20Critical%20Internet%20Resour 

ces.pdf. 

12 Jonathan Zittrain, Be Careful What You Ask For, in WHO RULES THE NET? INTERNET 

GOVERNANCE AND JURISDICTION 13-30 (Adam Thierer et al. eds., Cato Inst. 2003).


consideration of ways to embrace it that can preserve elements of the network 

that are the most important. 

The Internet Filtering Problem 

The world may appear borderless when viewed from cyberspace, but 

geopolitical lines are, in fact, well-established online. The fact that extensive 

Internet filtering occurs at a national level around the world is clearly 

documented. Through a collaborative research effort called the OpenNet 

Initiative, 13 the Citizen Lab at the University of Toronto, the Berkman Center 

for Internet and Society at Harvard University, and the SecDev Group are 

working together to compare Internet filtering practices of states in a 

systematic, methodologically rigorous fashion. In the past several years, 

OpenNet Initiative has sought to reach substantive conclusions about the 

nature and extent of Internet filtering in roughly 70 states and to compare 

practices across regions of the world. The OpenNet Initiative has released 

extensive reports that document and provide context for Internet filtering, 

previously reported anecdotally, in each of the states that it has studied closely. 

Reports released to date have focused on states in the Middle East and North 

Africa, Asia, and Central Asia, where the world’s most extensive filtering takes 

place. OpenNet Initiative’s research also covers states in every region of the 

world, including North America and Western Europe, where forms of speech 

regulation other than technical Internet filtering at the state level are the norm. 

Filtering implementations (and their respective scopes and levels of 

effectiveness) vary widely among the countries OpenNet Initiative has studied. 

China continues to institute by far the most intricate filtering regime in the 

world, with blocking occurring at multiple levels of the network and covering 

content that spans a wide range of topic areas. Though its filtering program is 

widely discussed, Singapore, by contrast, blocks access to only a handful of 

sites, each pornographic in nature. Most other states that OpenNet Initiative is 

studying implement filtering regimes that fall between the poles of China and 

Singapore, each with significant variation from one to the next. These filtering 

regimes are properly understood only in the political, legal, religious and social 

context in which they arise. 

Internet filtering occurs in different ways in different parts of the world. Some 

states implement a software application developed by one of a small handful of 

U.S.-based technology providers. Burma, in the first incarnation of its filtering 

regime, used an open source product for filtering, called DansGuardian. 14 

Others rely less heavily on technology solutions and more extensively on “soft 

controls.” Sometimes the filtering regime is supported explicitly by the state’s 

13 For more information, see www.opennetinitiative.net/. 

14 For more information on the DansGuardian filtering product, see dansguardian.org/.


legal code; in other cases, the filtering regime is carried out through a national 

security authority. In yet other instances, the regulation is simply presumed to 

be permissible. The content blocked spans a wide range of social, religious, and 

political information. Studies by OpenNet Initiative have reviewed whether 

individual citizens could access sites in a “global basket” of bellwether sites, 

testing every jurisdiction across a variety of sensitive areas—akin to a stock 

index sorted by sector—as well as a list of websites likely to be sensitive in some 

categories, but only in some countries. 

Extent, Character & Locus of Filtering 

More than three dozen states around the world practice technical Internet 

filtering of various sorts. 15 That number has grown over time. Those states 

that do filter the Internet have established a network of laws and technical 

measures to carry out substantial amounts of filtering that could allow the 

practice to become further embedded in their political and cultural 

environments. Web content is constantly changing, which poses a problem for 

the censors. Mobile devices and social networks have further complicated the 

task of speech regulation online. No state yet studied, even China, seems able 

to carry out its Web filtering in a comprehensive manner, i.e., consistently 

blocking access to a range of sites meeting specified criteria. China appears to 

be the most nimble at responding to the shifting Web, likely reflecting a 

devotion of the most resources, not to mention political will, to the filtering 

enterprise. 

A state wishing to filter its citizens’ access to the Internet has several initial 

options: Domain Name System (DNS) filtering, Internet Protocol (IP) filtering, 

or Uniform Resource Locator (URL) filtering. 16 Most states with advanced 

filtering regimes implement URL filtering, as it can avoid even more drastic 

over-filtering or under-filtering situations presented by the other choices, 

discussed below. 17 To implement URL filtering, a state must first identify 

where to place the filters; if the state directly controls the Internet service 

provider(s) (ISP), the answer is clear. Otherwise, the state may require private 

or semi-private ISPs to implement the blocking as part of their service. The 

technical complexities presented by URL filtering become non-trivial as the 

15 See Deibert et al., ACCESS DENIED, supra note 1. 

16 Nart Villeneuve, Why Block IP Addresses?, NART VILLENEUVE: INTERNET CENSORSHIP 

EXPLORER, Feb. 14, 2005, www.nartv.org/2005/02/14/why-block-by-ip-address/. 

17 For instance, IP filtering forces the choice of blocking all sites sharing an IP address. A 

recent OpenNet Initiative bulletin found more than 3,000 web sites blocked in an attempt to 

prevent access to only 31 sites. See Collateral Blocking: Filtering by South Korean 

Government of Pro-North Korean Websites, OpenNet Initiative Bulletin 009, Jan. 31, 2005, 

www.opennetinitiative.net/bulletins/009/. DNS blocking requires an entire domain and all 

subdomains to be either wholly blocked or wholly unblocked. See Villeneuve, supra note 16.


number of users grows to millions rather than tens of thousands. Some states 

appear to have limited overall access to the Internet in order to keep URL 

filtering manageable. The government of Saudi Arabia, for example, made 

filtering a pre-requisite for public Internet access, delaying any such access for a 

period of several years until the resources to filter were set in place. 

Citizens with technical knowledge can generally circumvent filters that a state 

has put in place. Some states acknowledge as much: The overseer of Saudi 

Arabia’s filtering program, via the state-run Internet Services Unit, admits that 

technically savvy users can simply not be stopped from accessing blocked 

content. Expatriates in China, as well as those citizens who resist the state’s 

control, frequently find up-to-date proxy servers through which they can 

connect to the Internet while evading filters. While no state will ultimately win 

a game of cat-and-mouse with those citizens who are resourceful and dedicated 

enough to employ circumvention measures, many users will never do so— 

rendering filtering regimes at least partially effective despite the obvious 

workarounds. 

Some of the earliest theorizing about control in the online environment, as 

discussed in Who Rules the Net?, 18 suggested that such state-run control of 

Internet activity would not work. It is important to note that states such as 

China have proven that an ambitious state can, by devoting substantial 

technical, financial, and human resources, exert a large measure of control over 

what their citizens do online. States, if they want, can erect certain forms of 

gates at their borders, even in cyberspace, and can render them effective 

through a wide variety of modes of control. 19 These controls have proven the 

claims of Jack L. Goldsmith and others who have emphasized the extent to 

which the online environment can be regulated and the ways in which 

traditional international relations theory will govern in cyberspace the same as 

they do in real-space. 20 

That does not mean that the issue is simple. For starters, states ordinarily need 

a great deal of help in carrying out filtering and surveillance regimes. Enter 

ISPs, many of which require a license from the government to lawfully provide 

Internet access to citizens. Much Internet filtering is effected by these private 

ISPs under respective states’ jurisdictions, though some governments partially 

centralize the filtering operation at private Internet Exchange 

Points�topological crossroads for network traffic�or through explicit state- 

18 Supra note 12. 

19 See Jack L. Goldsmith and Tim Wu, WHO CONTROLS THE INTERNET: ILLUSIONS OF A 

BORDERLESS WORLD 65-86 (Oxford University Press 2006). 

20 See Jack L. Goldsmith, Against Cyberanarchy, in WHO RULES THE NET?: INTERNET 

GOVERNANCE AND JURISDICTION (Thierer et al., eds., Cato Inst. 2003).


run clearing points established to serve as gatekeepers for Internet traffic. Some 

governments implement filtering at public Internet access points such as the 

computers found within cybercafés. Such filtering can take the form of 

software used in many American libraries and schools for filtering purposes, or 

“normative” filtering—government-encouraged interventions by shop-owners 

and others as citizens surf the Internet in a public place. 

Sometimes technical controls are not enough to constrain speech in the manner 

that the censors want. The exercise of more traditional state powers can have a 

meaningful impact on Internet usage that does not require the complete 

technical inaccessibility of particular categories of content. China, Vietnam, 

Syria, and Iran have each jailed “cyber-dissidents.” 21 Against this backdrop, the 

blocking of Web pages may be intended to deliver a message to users that state 

officials monitor Internet usage—in other words, making it clear to citizens that 

“someone is watching what you do online.” This message is reinforced by 

methods to gather what sites a particular user has visited after the fact, such as 

the requirement of passports to set up accounts with ISPs and tighter controls 

of users at cybercafés. 

As we learn more and more about how Internet filtering takes place, the 

problems of “governing” the Internet come more sharply into relief—about 

how control is exerted, about how citizens in one state can or cannot connect to 

others in another state, about the relationship between each state and its 

citizens, and about the relationships between states. 

Types of Content Filtered 

Around the world, states are blocking access to information online based upon 

its content—or what applications hosted at certain sites can do—for political, 

religious, cultural, security, and social reasons. Sensitivities within these 

categories vary greatly from country to country. These sensitivities often track, 

to a large extent, local conflicts. The Internet content blocked for social 

reasons—commonly child safety, pornography, information about gay and 

lesbian issues, and sex education—is more likely to be the same across countries 

than the political and religious information to which access is blocked. 

21 Reporters Sans Frontières, Internet Enemies: China, en.rsf.org/internet-enemiechina,36677.html 

(last accessed Aug. 25, 2010) (“Thirty journalists and seventy-two 

netizens are now behind bars for freely expressing their views.”); Reporters Sans Frontières, 

Internet Enemies: Viet Nam, en.rsf.org/internet-enemie-viet-nam,36694.html (last accessed 

Aug. 25, 2010) (“Vietnam is the world’s second biggest prison for netizens: it now has 

seventeen of them behind bars.”); Reporters Sans Frontières, Internet Enemies: Syria, 

en.rsf.org/internet-enemie-syria,36689.html (last accessed Aug. 25, 2010) (“At least four 

netizens are currently behind bars.”); Reporters Sans Frontières, Internet Enemies:Iran, 

http://en.rsf.org/internet-enemie-iran,36684.html (last accessed Aug. 25, 2010) (“Some 

thirty netizens have been arrested since June 2009, and a dozen are still being detained.”).


Several states carry out extensive filtering on certain topics. OpenNet Initiative 

testing has shown that 50% or more of the sites tested on a given topic (like sex 

education) or application (such as anonymization tools) are inaccessible. Very 

rarely does any state manage to achieve complete filtering on any 

topic/application. The only areas in which 100% filtering is approached are 

pornography and anonymizers (sites that, if left unfiltered, would defeat filtering 

of other sites by allowing a user to access any Internet destination through the 

anonymizers’ gateways). States like Burma, which reportedly monitors e-mail 

traffic, also block a high percentage of free e-mail service providers. Such 

complete, or near-complete, filtering is additionally only found in countries that 

have outsourced the task of identifying pornographic sites to one of several forprofit 

American companies, and is inevitably accompanied by over-blocking. 

Outside of these three areas, OpenNet Initiative testers are consistently able to 

access some material of a similar nature to the sites that were blocked. 

Filtering & Over-breadth 

Internet filtering is almost impossible to accomplish with any degree of 

precision. There is no way to stem the global flow of information in a 

consistently accurate fashion. A country that is deciding to filter the Internet 

must make an “over-broad” or “under-broad” decision at the outset. The 

filtering regime will either block access to too much or too little Internet 

content. Very often, this decision is tied to whether to use a home-grown 

system or whether to adopt a commercial software product, such as SmartFilter, 

WebSense, or an offering from security provider Fortinet, each of which are 

products made in the United States and are believed to be licensed to countries 

that filter the Internet. Bahrain, for instance, has opted for an “under-broad” 

solution for pornography; its ISPs appear to block access to a small and 

essentially fixed number of “black-listed” sites. Bahrain may seek to block 

access to pornographic material online, while actually blocking only token 

access to such material. The United Arab Emirates, by contrast, seem to have 

made the opposite decision by attempting to block much more extensively in 

similar categories, thereby sweeping into its filtering basket a number of sites 

that appear to have innocuous content by any metric. And Yemen was rebuked 

by the United States-based WebSense for allegedly using the company’s filtering 

system to block access to material that was not pornographic in nature, contrary 

to the company’s policies. 22 

Most of the time, states make blocking determinations to cover a wide range of 

Web content, commonly grouped around a second-level domain name or the IP 

address of a Web service (such as www.twitter.com or 66.102.15.100), rather 

22 See Jillian C. York, WebSense Bars Yemen’s Government from Further Software Updates, OpenNet 

Initiative, Aug. 12, 2009, opennet.net/blog/2009/08/websense-bars-yemensgovernment-further-software-updates.


than based on the precise URL of a given Web page (such as 

www.twitter.com/username), or a subset of content found on that page (such as 

a particular image or string of text). Iran, for instance, has used such an 

approach to block a cluster of weblogs that the state prefers not to reach its 

citizens. This approach means that the filtering process will often not 

distinguish between permissible and impermissible content so long as any 

impermissible content is deemed “nearby” from a network standpoint. 

Because of this wholesale acceptance or rejection of a particular speaker or site, 

it is difficult to know exactly what speech was deemed unacceptable for citizens 

to access. It’s even harder to ascertain why, exactly, the speech is blocked. 

Bahrain, a country in which we only found a handful of blocked sites at the 

outset of our first round of testing, blocked access to a discussion board at 

www.bahrainonline.org. The message board likely contains a combination of 

messages that would be tolerated independently as well as some that would 

appear to meet the state’s criteria for filtering. Likewise, we found minimal 

blocking for internal political purposes in the United Arab Emirates, but the 

state did block a site that essentially acted as a catalog of criticism of the state. 

Our tests cannot determine whether it was the material covering human rights 

abuses or discussion of historical border disputes with Iran, but in as much as 

the discussion of these topics is taking place within a broad dissention-based 

site, the calculation we project onto the censor in the United Arab Emirates 

looks significantly different than that for a site with a different ratio of 

“offensive” to “approved” content. 

For those states using commercial filtering software and update services to 

maintain a current list of blocked sites matching particular criteria, OpenNet 

Initiative has noted multiple instances where such software has mistaken sites 

containing gay and lesbian content for pornography. For instance, the site for 

the Log Cabin Republicans of Texas was blocked by the U.S.-based SmartFilter 

as pornography, and therefore, the apparent basis for its blocking by the United 

Arab Emirates. (OpenNet Initiative research shows that gay and lesbian 

content is itself often targeted for filtering, and even when it is not explicitly 

targeted, states may not be overly concerned with its unavailability.) 23 

As content changes increasingly faster on the Web and generalizations become 

more difficult to make by URLs or domains,—thanks in part to the rise of 

simpler, faster, and aggregated publishing tools, like those found on weblog 

sites and via other social networking applications—accurate filtering is getting 

trickier for filtering regimes to address unless they want to ban nearly 

everything. Mobile devices have further added to the complexity of the 

problem from the censor’s viewpoint. 

23 OpenNet Initiative, INTERNET FILTERING IN THE UNITED ARAB EMIRATES IN 2004-2005: A 

COUNTRY STUDY, Feb. 2005, opennet.net/studies/uae.


For example, free web hosting domains tend to group an enormous array of 

changing content and thus provoke very different responses from state 

governments. In 2004, Saudi Arabia blocked every page on freespace.virgin.net 

and www.erols.com. 24 However, research indicated the www.erols.com sites 

had been only minimally blocked in 2002, and the freespace.virgin.net sites had 

been blocked in 2002, but were accessible in 2003 before being re-blocked in 

2004. In all three tests, Saudi Arabia imposed URL blocking on 

www.geocities.com (possibly through SmartFilter categorization), but only 

blocked 3% of more than a thousand sites tested in 2004. Vietnam blocked all 

sites tested on the www.geocities.com and members.tripod.org domains. In 

OpenNet Initiative’s recent testing, it has found that Turkey and Syria have 

been blocking all blogs hosted on the free Blogspot service. 25 

China’s response to the same problem provides an instructive contrast. When 

China became worried about bloggers, they shut down the main blogging 

domains for weeks in the summer of 2004. When the domains came back 

online, the blogging systems contained filters that would reject posts containing 

particular keywords. 26 Even Microsoft’s MSN Spaces blogging software 

prevented writers from publishing terms like “democracy” from China. In 

effect, China moved to a content-based filtering system, but determined that the 

best place for such content evaluation was not the point of Web page access, 

but the point of publication, and it possessed the authority to force these filters 

on the downstream application provider. This approach is similar to that taken 

with Google in response to the accessibility of disfavored content via Google’s 

caching function. Google was blocked in China until a mechanism was 

implemented to prevent cache access. 27 These examples clearly demonstrate the 

length to which regimes will go to preserve “good” access instead of simply 

blocking an entire service. 

These examples also demonstrate the increasing reliance by states on “just-intime” 

filtering, rather than filtering that occurs in the same, consistent way over 

time. While the paradigmatic case of Internet filtering was initially the state that 

wished to block its citizens from viewing any pornography online at any time 

24 Saudi Arabia blocked every page on www.erols.com except for the root page at 

www.erols.com itself, potentially indicating a desire to manage perceptions as to the extent 

of the blocking. 

25 All data from OpenNet Initiative testing can be found in the country-by-country summaries 

at www.opennet.net/. 

26 Filtering by Domestic Blog Providers in China, OpenNet Initiative Bulletin 008, Jan. 14, 

2005, www.opennetinitiative.net/bulletins/008/. 

27 This mechanism turned out to be extremely rudimentary, as outlined in a previous OpenNet 

Initiative bulletin. See Google Search & Cache Filtering Behind China’s Great Firewall, 

OpenNet Initiative Bulletin 008, Sept. 3, 2004, 

www.opennetinitiative.net/bulletins/006/.


(for instance, Saudi Arabia), the phenomenon of a state blocking particular 

speech or types of speech at a sensitive moment has become commonplace. 

For instance, China blocked applications such as Twitter and YouTube at the 

time of the 20 th anniversary of the Tiananmen Square demonstrations in June 

2009. A few weeks later, Iran blocked similar applications, including Facebook, 

at the time of demonstrations in the streets of Tehran. These blocks are often 

lifted once the trouble has past. One means of tracking these changes in the 

availability of applications and websites is a project called Herdict.org, which 

enables people from around the world to submit reports on what they can and 

cannot access in real-time. 28 

Alternate approaches that demand a finer-grained means of filtering, such as the 

use of automated keywords to identify and expunge sensitive information on 

the fly, or greater manual involvement in choosing individual Web pages to be 

filtered, are possible, so long as a state is willing to invest in them. China in 

particular appears prepared to make such an investment, one mirrored by 

choices demonstrated through more traditional media. For example, China 

allows CNN to be broadcast within the country with a form of time delay, so 

the feed can be temporarily turned off as when, in one case, stories about the 

death of Zhao Ziyang were broadcast. 29 Online policing of speech, even in 

what appears to be a “borderless world,” can be carried out through technical 

controls at many layers. 

Law, Surveillance & Soft Controls 

Just as dozens of states use technical means to block citizens from accessing 

content on the Internet, most also employ legal and other soft controls. 

Surveillance practices are most commonly coupled with outright technical 

censorship. Many states that filter use a combination of media, 

telecommunications, national security, and Internet-specific laws and regulatory 

schemes to restrict the publication of and access to information on the Internet. 

States often require ISPs to obtain licenses before providing Internet access to 

citizens. Some states—China and Turkey, for instance, which have each 

enacted special regulations to this effect—apply pressure on cybercafés and 

ISPs to monitor Internet usage by their customers. With the exceptions of 

Saudi Arabia and Qatar, no country seems to explicitly communicate to the 

public about its process for blocking and unblocking content on the Internet. 

Most countries, instead, have a series of broad laws that cover content issues 

online, both empowering states that need these laws to carry out filtering 

28 See www.herdict.org. The histories of reports of these just-in-time blocking patterns can 

be viewed from this website. 

29 See Eric Priest, Reactions to the Internet & Society 2004 Session on “Business” (December 11, 10:45- 

12:15), Spring 2005, 

cyber.law.harvard.edu/blogs/gems/tka/EPriestReactionPaper2.pdf.


regimes, and putting citizens on general notice not to publish or to access 

content online that violates certain norms. 

Often these soft controls are exercised through social norms or through control 

at the far edges of the network. Sometimes the state requires non-governmental 

organizations and religious leaders to register before using the Internet to 

communicate about their work. In China and in parts of the former Soviet 

Union, very often the most fearsome enforcer of the state’s will is the old 

woman on one’s block, who may or may not be on the state’s payroll. The 

control might be exercised, as in Singapore, largely through family dynamics. 

The call by the local police force to the Malaysian blogger to come and talk 

about his web publishing might have as much of an effect on expression as any 

law on the books or technical blocking system. 

Whether through advanced information technology, legal mechanisms, or soft 

controls, a growing number of states around the world are seeking to control 

the global flow of information. Ordinarily, this control takes the form of 

blocking, through technical means, state’s citizens from accessing certain 

information online. In other instances, the blocking stops the state’s citizens 

from publishing information online, in effect disallowing people outside the 

state from hearing the voices of the state’s citizens. As a result, most filtering 

regimes cause a chilling effect on the use of information technologies as a 

means of free expression, whether for political, religious, or cultural purposes. 

From “How to Police Speech Online” 

to “How to Limit Speech 

Restrictions Online” 

It is commonplace to argue that states have generally regulated the Internet 

lightly, but it is increasingly untrue. The author of a chapter in an important 

recent book wrote, “governments exercise relatively little control over the 

Internet, even though it has a tremendous impact on society.” 30 This statement 

misleads readers into thinking that the Internet might somehow be a freer, more 

open environment than traditional spaces are. Such a statement might have 

been true in the United States fifteen years ago. But as of today, it is inaccurate. 

From a global perspective, both the importance of digitally-mediated 

communications and the extent of regulation of speech continue to grow over 

time. 

The policing of speech in a borderless world brings with it a series of problems 

that merit public discussion. The types of controls that take place at a local 

30 Harold Kwalwasser, Internet Governance, in CYBERPOWER AND NATIONAL SECURITY, 491 

(Franklin D. Kramer et al., eds., Nat’l Defense Univ. Press 2009).


level, about bullying or hate speech or any other issue that may arise, can be 

vetted in a public meeting of a school district or in a state legislature. The 

online speech issues that are consistently on federal agendas in the United 

States—intellectual property restrictions, child safety, network neutrality, 

defamation, and so forth—tend to play out in public forums like the legislatures 

or in the court system. Of course, there are instances where the debate about 

speech controls occurs behind closed doors at the local and national levels as 

well. Corporate-level filtering, which can affect millions of employees, is one 

example where the issues are rarely vetted in a meaningful way. But the place 

where the debate is most consistently and conspicuously absent is on the 

international stage. It ought to merit meaningful consideration in the Internet 

governance debate. 

The practice of state-mandated Internet filtering, and related regulations like 

surveillance, is now a widely-known fact, but the hard problems that stem from 

this practice are infrequently discussed as a matter of public policy outside of 

human rights and academic circles and the occasional national-level hearing. 

The blocking and surveillance of citizens’ activity on the Internet—by virtue of 

the network’s architecture, an issue of international dimensions—calls for 

discussion at a multilateral level. Rather than fretting over the finer points of 

the domain name system, time would be better spent in Internet governance 

discussions on issues like transparency in Internet filtering or broad issues of 

interconnection of the global network. The Internet filtering problem offers 

much more to be gained—through frank discussion, if not action—and 

provides an exercise worthy of an extraordinary gathering of world leaders who 

want to talk about the global “Information Society.” 

On one level, Internet filtering is a private matter between a state and its 

citizens as to what information citizens may access online. 31 States that censor 

the Internet assert the right to sovereignty. From the state’s perspective, the 

public interest, as defined in one state, such as Saudi Arabia, is different from 

the public interest as defined by the state in Uzbekistan, China, or the United 

Kingdom. States can and do exercise their sovereignty through control of the 

information environment. 

But even if one accepts the state sovereignty argument, that viewpoint should 

not end the conversation about Internet filtering. The state-based censorship 

31 Some states make an effort to suggest that their citizens (in Saudi Arabia and the United 

Arab Emirates specifically) are largely in support of the filtering regime, particularly when it 

comes to blocking access to pornographic material. For instance, the agency responsible for 

both internet access and filtering in Saudi Arabia conducted a user study in 1999, and 

reported that 45% of respondents thought “too much” was blocked, 41% thought it 

“reasonable,” and 14% found it “not enough.” These studies stand for the proposition, in 

the context of our report, that some states that filter seek to make the case that their filtering 

regime enjoys popular support, not that such support necessarily exists.


and surveillance practices of any state affect the citizens and businesses of other 

states in the context of the interconnected global communications network. 

Increasingly, the censorship and surveillance practices of states reach past the 

web-browsing habits of their own citizens. High-profile debates between 

Canadian company Research in Motion, maker of BlackBerry smartphones, and 

the United Arab Emirates and Saudi Arabia about the ability of the state to 

intercept BlackBerry communications make this point clear. 32 The Internet 

blocking that takes place in one state can also affect the network at a technical 

level in other jurisdictions, as Pakistan found out when it brought down 

YouTube globally for two hours in early 2009. 33 

A global discussion about the relationship between these filtering and 

surveillance practices and human rights is necessary and could be extremely 

fruitful. Specifically, states might consider rules that relate to common 

standards for transparency in Internet filtering and surveillance practices as they 

relate to individuals and those corporations drawn into the process. On a 

broader level, the issue raised here is about interconnection between states and 

the citizens of those states—and ultimately about what sort of an Internet we 

want to be building and whether the global flow of information is a sustainable 

vision. 

For instance, we have yet to join the ethical interests at play in filtering. States 

vary greatly in terms of how explicitly the filtering regime is discussed and the 

amount that citizens can come to know about it. No state OpenNet Initiative 

studied makes its block list generally available. 34 The world leaders who gather 

32 See Adam Shreck, UAE BlackBerry Crackdown Affects Visitors Too, ASSOCIATED PRESS, August 

3, 2010, 

www.google.com/hostednews/ap/article/ALeqM5iJ1MLhAMIeRDhT4heu4LKwxgH3QD9HBKDJO0. 

See also Anthony DiPaola & Vivian Salama, UAE to Suspend BlackBerry 

Service Citing Security, BLOOMBERG, August 1, 2010, www.businessweek.com/news/2010- 

08-01/u-a-e-to-suspend-blackberry-services-citing-security.html. 

33 See Declan McCullagh, How Pakistan Knocked YouTube Offline (and how to make sure it never 

happens again),” CNET, February 25, 2008, news.cnet.com/8301-10784_3-9878655-7.html. 

34 Saudi Arabia publishes its rationale and its blocking practices on an easily accessible website, 

at www.isu.net.sa/saudi-internet/contenet-filtring/filtring.htm (“The Internet Services 

Unit oversees and implements the filtration of web pages in order to block those pages of 

an offensive or harmful nature to the society, and which violate the tenants of the Islamic 

religion or societal norms. This service is offered in fulfillment of the directions of the 

government of Saudi Arabia and under the direction of the Permanent Security Committee 

chaired by the Ministry of the Interior.”). In Saudi Arabia, citizens may suggest sites for 

blocking or for unblocking, in either Arabic or English, via a public website. Most sites 

include a block-page, indicating to those seeking to access a website that they have reached a 

disallowed site. Most states have enacted laws that support the filtering regime and provide 

citizens with some context for why and how it is occurring, though rarely with any degree of 

precision. As among the states we have studied, China is one of the states that obscures the 

nature and extent of its filtering regime to the greatest extent through a long-running series 

of conflicting public statements about its practices in this respect.


periodically at United Nations-sponsored meetings and at the Internet 

Governance Forums could make the most of their leadership by seeking to 

establish a set of best practices related to Internet filtering and the transparency 

related to filtering regimes. They might also focus profitably on the difficult 

problems facing those multinational companies that do business in regimes that 

require filtering and surveillance of the network in ways that would not be 

legally permissible in the company’s home jurisdiction, as the Global Network 

Initiative has in its private capacity. 35 The issue of speech regulation in a 

borderless world is too important to leave only to those handful of companies, 

right-minded though they may be, which are seeking to do the right thing in a 

geopolitically complex regulatory environment. 

The critical question in the next digital decade is not whether speech can be 

policed in a borderless world, but whether we should set new limits on the 

extent and manner in which it is policed today. We should ask, too, whether it 

is today easier, in fact, than in the past to police speech in a digitally-mediated 

world, and what the ramifications are for civil liberties if so. The Internet is 

becoming larger and more fractured each day. We should not pretend the 

Internet is a “lightly regulated” medium as though the calendar on our kitchen 

wall reads “1985” at the top. Trends that support more speech from more 

people in more places around the globe—using mobile applications generally, 

such as blogs, wikis, Twitter, SMS, and so forth—are countered by the 

increasing sophistication and reach of Internet filtering and surveillance 

practices. A richer understanding of the complexities at play in Internet filtering 

and other speech restrictions would help develop a foundation that does not yet 

exist for building a sustainable, and truly global, network that will continue to 

bring with it the innovation, jobs, services, and other social benefits that it 

promises. 

35 See www.globalnetworkinitiative.org.


The Role of the Internet 

Community in Combating 

Hate Speech 

By Christopher Wolf * 

In less than twenty years, the Internet has evolved to become an 

unprecedented tool for information, communication, entertainment and 

commerce. Much of the progress was made possible by the protections 

of the First Amendment and the absence of legal constraints. But there 

is a growing dark side to the Internet, which raises the question of how 

to police harmful content. Online child predators are a well-known 

blight, a frequent focus of headlines and television news. Cyber-bullying 

also has received significant attention recently, especially in the wake of 

teen suicides. 

Less reported but equally troubling is the fact that the Internet has 

become a technology embraced by racists, anti-Semites, homophobes 

and bigots of all kinds to spread their messages of hate. The online 

haters use all of the tools of the Internet, from static websites, to 

streaming audio and video, to social networking sites like Facebook. 

No longer relegated to meeting in dark alleys and the basements of 

abandoned buildings, or to mailing their propaganda in plain brown 

wrappers, hate groups have a platform to reach millions around the 

world. They seek to victimize minorities, to embolden and mobilize likeminded 

haters, and to recruit followers. And in their wake, an online 

culture has developed—aided by the mask of anonymity—in which 

people who would never consider themselves members of hate groups 

employ racial, religious and other epithets as part of their vocabulary in 

posting comments to news stories on mainstream sites and in other 

aspects of online life. In turn, the common appearance of such epithets 

desensitizes readers, making hate speech and the denigration of 

minorities “normal.” 

* Christopher Wolf is a partner in the law firm Hogan Lovells LLP where he leads the Privacy 

and Information Management practice. He founded and chairs the Internet Task Force of 

the Anti-Defamation League, and is Past Chair of the International Network Against Cyber- 

Hate (INACH).


One recent example of how haters are using the Internet occurred on the 

Fourth of July in 2010. As Americans were celebrating that event, a new 

“Event” was announced on Facebook, entitled “Kill a Jew Day.” 1 The 

Facebook “host” for the Event wrote, “You know the drill guys,” and he 

urged followers to engage in violence “anywhere you see a Jew” between 

July 4 and July 22. A Nazi swastika adorned the Event page. 

The posting of that sickening Event prompted a wave of anti-Semitic 

rants on Facebook in support of the targeting of Jewish people. But it 

also prompted a counter-event on Facebook entitled “One Million 

Strong Against Kill a Jew Day” (whose supporters actually numbered, 

more modestly, in the thousands). 2 And, pursuant to the Facebook 

Terms of Service, complaints about the “Kill a Jew Day” event to 

Facebook administrators resulted in the company disabling the Event 

page. 

The outrage over the Facebook Event site was justified, not just because 

of the vile anti-Semitism underneath it or the glorified display of a 

swastika. People also objected to the site because they know that 

Internet messages can and do inspire violence. Online anti-Semitic hate 

speech has been implicated in real-world acts of violence, such as an 

attack on Nobel Laureate and Holocaust survivor Elie Wiesel by a 

Holocaust denier in 2007, 3 and the 2009 murder of a guard at the 

Holocaust Memorial Museum in Washington, D.C., by a white 

supremacist who maintained his own online hate site and who was 

egged-on by fellow haters. 4 Words have consequences, and indeed 

inspire acts of hate and violence. 

This recent example of online hate provides another opportunity to 

examine what society’s response to online hate speech should be. What 

1 See Yaakov Lappin, “Kill a Jew” Page on Facebook Sparks Furor, THE JERUSALEM POST, July 5, 

2010, http://www.jpost.com/JewishWorld/JewishNews/Article.aspx?id=180456. 

2 See Amada Schwartz, Anti-Semitism v. Facebook, JEWISH JOURNAL, July 13, 2010, 

http://www.jewishjournal.com/community/article/antisemitism_vs_facebook_20100713/. 

3 Suzanne Herel, Holocaust Survivor, Nobel Peace Prize Winner Elie Wiesel Attacked in S.F. Hotel, 

S.F. CHRONICLE, Feb. 9, 2007, http://www.sfgate.com/cgibin/blogs/nwzchik/detail?blogid=32&entry_id=13385. 

4 Michael E. Ruane, Paul Duggan & Clarence Williams, At a Moment of Sorrow, A Burst of 

Deadly Violence, THE WASHINGTON POST, June 11, 2009, 

http://www.washingtonpost.com/wpdyn/content/article/2009/06/10/AR2009061001768.html.


is the best way to police sites like the “Kill a Jew Day” Event page on 

Facebook, and the thousands of hate-filled videos uploaded to YouTube, 

and the white supremacist websites designed to recruit young people, 

glorifying violence against minorities? 

One visceral response to the proliferation of online hate is: “There 

ought to be a law.” Legal rules are the way a society decrees what is right 

and what is wrong, and since hate speech is wrong, it seems logical that 

the law would be employed to police it. A legal ban on hate speech and 

the criminalization of its publication is indeed an alternative in some 

jurisdictions. But, of course, it is not an option in the United States where 

the First Amendment gives broad latitude to virtually all speech, even the 

most repugnant. (Only direct threats against identifiable targets are 

criminalized.) 

Legislatures around the world have heeded the call for laws 

encompassing Internet hate speech. The hate speech protocol to the 

Cybercrime Treaty is a prime example of a heralded legal solution to the 

problem. 5 It was designed to eliminate racist sites from the Internet 

through criminal penalties. 

From Brazil to Canada, and from South Africa to Great Britain, there are 

legal restrictions on hate speech, online and offline. In much of Europe, 

denial of the Holocaust (online or offline) is forbidden. In Germany, 

even displaying the swastika is a crime. The enforcement of laws against 

Holocaust deniers—given the bitterly sad history of those countries— 

serves as a message to all citizens (especially impressionable children) that 

it is literally unspeakable to deny the Holocaust given the horrors of 

genocide inflicted in those countries. 

Still, there are many who believe that prosecutions, such as that of 

Holocaust denier David Irving in Austria, 6 promoted his visibility and 

stirred up his benighted supporters, rather than quelling future hate 

speech and enlightening the public. 

Moreover, laws against hate speech have not demonstrably reduced hate 

speech or deterred haters. The hate speech protocol to the Cybercrime 

5 The Convention on Cybercrime Nov. 23, 2001, Europ. T.S. No. 185, 

http://conventions.coe.int/Treaty/en/Treaties/Html/185.htm. 

6 See Austria Holds “Holocaust Denier,” BBC NEWS, Nov. 17, 2005, 

http://news.bbc.co.uk/2/hi/europe/4446646.stm.


Treaty, for example, has not reduced online hate. The shield of Internet 

anonymity and the viral nature of online hate make legal policing an 

unrealistic challenge, except in cases where authorities want to “set an 

example.” And since the U.S., with its First Amendment is essentially a 

safe-haven for virtually all web content, shutting down a website in 

Europe or Canada through legal channels is far from a guarantee that the 

contents have been censored for all time. The borderless nature of the 

Internet means that, like chasing cockroaches, squashing one does not 

solve the problem when there are many more waiting behind the walls— 

or across the border. 

Many see prosecution of Internet speech in one country as a futile 

gesture when the speech can re-appear on the Internet almost 

instantaneously, hosted by an Internet service provider (ISP) or social 

networking site in the United States. Moreover, in the social networking 

era, the ability of people to upload far outpaces the ability of the police to 

track and pursue offending speech. 

Like the prosecution in Austria of David Irving, the prosecutions in 

Germany of notorious Holocaust deniers and hate site publishers Ernst 

Zundel 7 and Frederick Töben 8 sent messages of deterrence to people that 

make it their life’s work to spread hate around the world that they may 

well go to jail as well. And, again, such prosecutions expressed society’s 

outrage at the messages. But all one need do is insert the names of those 

criminals in a search engine, and you will find websites of supporters 

paying homage to them as martyrs and republishing their messages. 

Even some free speech advocates around the world applaud the use of 

the law to censor speech when it is hate speech because of the pernicious 

effects of hate speech on minorities and children, and because of its 

potential to incite violence. But many of those same people object to the 

use of the law by repressive regimes like China to censor speech it deems 

to be objectionable as hate directed towards the Chinese government. It 

is not easy to draw the line between good and bad state use of censorship 

because defining what is hate speech can be quite subjective. Giving the 

state the power to censor is problematic, especially given the potential for 

abuse. 

7 See Zundel Gets Five Years from German Count, WINNIPEG FREE PRESS, Feb. 16, 2007, 

http://www.winnipegfreepress.com/historic/32115694.html. 

8 See Steve Kettmann, German Hate Law: No Denying It, WIRED, Dec. 15, 2000, 

http://www.wired.com/politics/law/news/2000/12/40669#ixzz0jTf5lnLh.


This is not to say that law has no role to play in fighting online hate 

speech—far from it. But countries with speech codes intended to 

protect minorities should make sure that the proper discretion is 

employed to use those laws against Internet hate speech, lest the 

enforcement be seen as ineffectual and result in a diminished respect for 

the law. And, again, the realities of the Internet are such that shutting 

down a website through legal means in one country is far from a 

guarantee that the website is shuttered for all time. 

Thus, the law is but one, albeit limited, tool in the fight against online 

hate. 

Counter-speech—exposing hate speech for its deceitful and false 

content, setting the record straight, and promoting the values of 

tolerance and diversity, has an important role to play in the policing of 

online hate speech. That is the thrust of the First Amendment. To 

paraphrase U.S. Supreme Court Justice Brandeis, sunlight is still the best 

disinfectant—it is always better to expose hate to the light of day than to 

let it fester in the darkness. One answer to hate speech is more speech. 

The Facebook event “One Million Strong Against Kill a Jew Day,” even 

if far short of a million members, is a vivid example of the power of 

counter-speech as a vehicle for society to stand up against hate speech. 

And, of course, education from an early age on Internet civility and 

tolerance would go far to stem the next generation of online haters. 

An equally important and powerful tool against hate speech is the 

voluntary cooperation of the Internet community, including Internet 

Service Providers, social networking companies and others. When 

Facebook enforced its Terms of Service (which requires users not to 

“post content that: is hateful, threatening, or … incites violence…” 9 ) and 

disabled the “Kill Jew Day” event site, 10 that was a powerful example of 

an Internet company exercising its own First Amendment rights to 

ensure that it remained an online service with standards of decency. That 

voluntary act was quick and effective. A legal action against Facebook for 

hosting the site—impossible in the U.S. but viable elsewhere around the 

world—would have been expensive, time-consuming and no more 

9 Statement of Rights and Responsibilities, Facebook, 

http://www.facebook.com/terms.php. 

10 Lappin, supra note 1.


effective. The chilling effect of a legal action against Facebook may have 

resulted in undue restrictions by Facebook on future user postings. 

Voluntary enforcement by Internet companies of self-established 

standards against hate speech is effective. If more Internet companies in 

the U.S. block content that violate their Terms of Service, it will at least 

be more difficult for haters to gain access through respectable hosts. The 

challenge, of course, is with social media sites where postings occur 

constantly and rapidly. Social media companies normally wait for a user 

complaint before they investigate hate speech posted on their service, but 

the proliferation of hate-filled postings outpaces the effectiveness of such 

a “notice and take down” arrangement. New monitoring techniques to 

identify hate speech as it is posted may be in order. 

In the era of search engines as the primary portals for Internet users, 11 

cooperation from the Googles of the world is an increasingly important 

goal. The example of the Anti-Defamation League 12 and Google with 

the site “Jew Watch” is a good one. 13 The high ranking of Jew Watch in 

response to a search inquiry was not due to a conscious choice by 

Google, but was solely a result of an automated system of ranking. 

Google placed text on its site that explained the ranking, and gave users a 

clear explanation of how search results are obtained, to refute the 

impression that Jew Watch was a reliable source of information, and 

linked to the ADL site for counter-speech. 14 

In short, vigilance and voluntary standards are more effective than the 

law in dealing with the increasing scourge of online hate speech. Hate 

speech can be “policed” in a borderless world, but not principally by the 

traditional police of law enforcement. The Internet community must 

continue to serve as a “neighborhood watch” against hate speech online, 

11 A study by Nielsen found that 37 percent of respondents used search engines when looking 

for information, as compared to 34 percent using portals and only 18 percent using social 

media. See Jon Gibs, Social Media: The Next Great Gateway for Content Discovery?, NIELSEN, Oct. 

5, 2009, http://blog.nielsen.com/nielsenwire/online_mobile/social-media-the-nextgreat-gateway-for-content-discovery/. 

12 The Anti-Defamation League was founded to combat anti-Semitism. For more information, 

see http://www.adl.org/. 

13 See Google Responds to Jew Watch Controversy, WEBPRONEWS, April 15, 2004, 

http://www.webpronews.com/topnews/2004/04/15/google-responds-to-jew-watchcontroversy. 

14 Google, An Explanation of Our Search Results, http://www.google.com/explanation.html.


“saying something when it sees something,” and working with online 

providers to enforce community standards.

554 CHAPTER 9: CAN SPEECH BE POLICED IN A BORDERLESS WORLD?

CHAPTER 10 

WILL THE NET LIBERATE THE WORLD? 

Can the Internet Liberate the World? 557 


Internet Freedom: Beyond Circumvention 565 


555

556 CHAPTER 10: WILL THE NET LIBERATE THE WORLD?


Can the Internet 

Liberate the World? 

By Evgeny Morozov * 

It may be useful to start by laying out the basics. Anyone pondering the 

question posed in the title of this essay most likely assumes that there exists 

some powerful forces of oppression from which the world could and should be 

liberated. A list of such problems is as infinite as it is intuitive: From poverty to 

racism and from pollution to obesity, “oppressive forces” seem to be all around 

us. 

Yet, for some reason, these are rarely the kind of problems that one wants to 

fight with the help of technology, let alone the Internet. It’s in solving political 

rather than socio-economic problems that the Internet is presumed to hold the 

greatest promise. Most specifically, it is its ability to undermine repressive 

governments that is widely discussed and admired, even more so as Internet 

companies like Google find themselves struggling with the likes of the Chinese 

government. 

Two features of the Internet are often praised in particular: 1) its ability to 

quickly disseminate any kind of information—including the information that 

authoritarian governments may not like—and 2) its ability to allow like-minded 

individuals to find each other, to mobilize supporters and to collectively pursue 

future goals—including democratization. The hype surrounding Iran’s Twitter 

Revolution of 2009 was probably the strongest public manifestation of high 

hopes for the transformative potential of the Internet; only a rare pundit did not 

predict the eventual collapse of the Iranian regime under the barrage of angry 

tweets from its citizens. 

Still, such praise is not without merit. Even the hardest skeptics would grant 

these two features to the Internet; to deny that it does enhance the citizens’ 

ability to inform (and get informed) and to mobilize—what the Internet theorist 

Clay Shirky calls “ridiculously easy group forming” 1—would be to deny the 

obvious. The skeptics would also have no trouble acknowledging that both of 

these features are constantly under threat, as governments keep implementing 

new systems of censorship and control. 

* Evgeny Morozov is the author of THE NET DELUSION: THE DARK SIDE OF INTERNET 

FREEDOM (Public Affairs, 2011). He’s also a visiting scholar at Stanford University, a fellow 

at the New America Foundation and a contributing editor to Foreign Policy magazine. 

1 CLAY SHIRKY, HERE COMES EVERYBODY: THE POWER OF ORGANIZING WITHOUT 

ORGANIZATIONS 54 (Penguin Press 2008), quoting social scientist Sébastien Paquet.

558 CHAPTER 10: WILL THE NET LIBERATE THE WORLD? 

Nor does anyone really contest the fact that the Internet has proved quite 

resilient against such attacks, giving rise to numerous tools to circumvent 

government censorship. For many, the fact that an institution as powerful as the 

U.S. government has trouble reining in a fluid and mostly virtual organization 

like WikiLeaks is a testament to the power of the Internet, even though the 

morality of WikiLeaks’ actions (in publishing leaked information about the U.S. 

military occupation of Afghanistan) is still widely disputed. 

However, conceding that the Internet helps to disseminate information and 

mobilize campaigners around certain causes is not quite the same as conceding 

that authoritarian regimes are doomed or that democracy is inevitable. There 

may be good independent reasons to campaign for greater freedom of 

expression on the Internet—but one shouldn’t presume that such freedoms 

would necessarily translate into democratization. 

For the Net—and its two powerful features discussed above—to be able to 

“liberate the world” from authoritarianism, one needs to make a few further 

assumptions. First, one needs to assume that modern authoritarian regimes 

derive their power primarily by suppressing the activities that the Internet helps 

to amplify: i.e., dissemination of information and popular mobilization around 

specific causes. Second, one also needs to assume that the Internet won’t have 

any other political and social effects that may—if only indirectly—create new 

modes of “oppression,” entrenching authoritarianism as a result. In other 

words, the Internet can only deliver on its liberating promise as long as the 

things it offers are also the things that the fight against authoritarianism 

requires—and as long as it doesn’t produce any other regime-strengthening 

effects that may inadvertently undermine that fight. 

On initial examination, the first assumption seems to hold. There is no shortage 

of suppression of both information and mobilization opportunities in modern 

authoritarian states. Their rulers have not lost the desire to guard their secrets or 

regulate how their citizens participate in public life. The fact that some forms of 

censorship persist even in democratic societies suggests that governments do 

not really aspire to lift all the digital gates and let information flow freely. The 

urge, then, is to find ways to break through those gates—and the Internet seems 

to excel at the job. If there is one thing that techies and hackers know how to 

do well, it’s to build tools to pierce firewalls. 

But suppose that such tools can be found and can even achieve the kind of 

scale where all Internet users in China or Iran have access to them. What would 

the effect be on their populations and their governments? I’d like to propose 

that one’s answer to this question depends mostly on one’s views about the 

sources of legitimacy of modern authoritarianism. 

Those who believe that such legitimacy is derived primarily through the 

brainwashing of their citizens are justifiably very excited about the Internet.


Moreover, they are usually very quick to predict the inevitable fall of 

authoritarian governments. After all, their theoretical conception of 

authoritarianism posits that once the information gates are open, brainwashing 

loses much of its effectiveness; people realize they have been lied to all along— 

and they rebel as a result. 

At first sight, the contemporary global situation may seem to vindicate such 

views. Many modern authoritarian governments—and here cases like Belarus, 

China, Russia come to mind—enjoy strong levels of support from vast swathes 

of their populations. One may quibble about the ways by which the Kremlin has 

solidified its power in the last decade—many of those ways are far from 

democratic—but virtually all opinion polls reveal that the Kremlin’s policies are 

genuinely popular. Ditto China, where the government is one of the most 

trusted institutions in the country, enjoying a level of trust that the U.S. 

Congress could only dream of. 

Is it all because of brainwashing? If the answer is “yes,” then there are, indeed, 

good reasons to be optimistic about the power of the Internet. The moment the 

authoritarian governments’ monopoly over information disappears, any 

manipulations of truth that were possible in an age of information scarcity 

would be impossible. 

This, I’d like to propose, may be a very simplistic reading of the situation—and 

a reading that is also extremely insensitive to historical and social forces. There 

is much more to the legitimacy of modern authoritarian states than just their 

skillful manipulation of information. Many authoritarian regimes—Belarus, 

China and Russia are again excellent examples but one could also add 

Azerbaijan, Kazakhstan, and Vietnam to the list—have made genuine advances 

in economic development, many of them thanks to their embrace of technology 

in general and the Internet in particular. Furthermore, in many of these 

countries, the once extremely contentious political life has stabilized as well, 

allowing their populations to enjoy a rare period of peace and prosperity—even 

it came at the hefty price of having their governments tighten the valves on 

freedom of expression or freedom of assembly. 

It seems disingenuous to argue that modern Russians or Chinese do not 

appreciate the fact that today they purchase considerably more commodities— 

including luxurious ones—than they could 20 years ago; or that they can travel 

the world much more freely; or that they—at least online—can consume any 

kind of entertainment they want, regardless of its origins. Placed in the historical 

context and compared against other possible scenarios of where these two 

countries may have been had their rulers not embarked on a series of reforms, 

such achievements look even more impressive. 

Should it turn out that large segments of the populations of authoritarian states 

are well aware of the kind of human rights violations that are needed in order to


sustain the impressive rates of economic progress—or, even more shockingly, 

that they are actually supportive of such violations—a strategy of “unbrainwashing” 

simply would not work. If modern authoritarianism is as much a 

product of a social contract as modern democracy is, then changing the 

attitudes of those who have long given up the fight for freedom would take 

much more than just exposing them to facts. 

The big question that Western do-gooders should be asking themselves here is 

not whether the Internet can liberate the world but whether the world actually 

wants to be liberated. Above all, this is a question of whether capitalism 

unburdened by democratic norms and ideals is sustainable in the long term—a 

possibility that goes directly against the theory that the logic of capitalism 

inevitably leads to democracy, a view that was extremely popular in the early 

1990s. 2 

If capitalism can get by without democracy, the ability to spread subversive 

information that might reveal the horrors of the regime looks considerably less 

impressive, for the populations that the West seeks to liberate are already wellaware 

of what’s going on and many of them may have simply chosen to look 

the other way in expectation of a better life for their children. 

Granted, it’s not just facts that may help change their attitudes. One may use the 

Net to distribute subversive poetry and fiction that would reawaken (or, in most 

cases, create from scratch), the political consciousness of those living under 

authoritarianism. Technology would certainly be of great help here, both in 

terms of helping to distribute such materials but also in terms of protecting 

those who access them. But the ability of such materials to incite people to 

democratic change is not just a function of how many people read them; rather, 

it’s a function of how well-argued such materials are. 

It seems that even if the West succeeded in distributing 1984, Darkness at Noon, 

or Brave New World to every single citizen of an authoritarian state, this might 

not lead to a revolution, simply because those books offer a poor critique of the 

actually existing modern-day authoritarianism, which has come to terms with 

both Western popular culture and globalization. 

2 See, e.g., Commission on Security & Cooperation in Europe (U.S. Helsinki Commission), 

Briefing on Twitter Against Tyrants: New Media In Authoritarian Regimes, Oct. 22, 2009, available 

at http://csce.gov/index.cfm?FuseAction=ContentRecords.ViewDetail& 

ContentRecord_id=462&Region_id=0&Issue_id=0&ContentType=H,B&ContentRe 

cordType=B&CFID=32177263&CFTOKEN=96274551; Nicholas Kristof, Tear Down 

This Cyberwall!, N.Y. TIMES, June 17, 2009, available at http://www.nytimes.com/2009/06/ 

18/opinion/18kristof.html?_r=1; L. Gordon Crovitz, Mrs. Clinton, Tear Down this Cyberwall, 

WALL ST. JOURNAL, May 3, 2010, available at http://online.wsj.com/article/ 

SB10001424052748704608104575219022492475364.html; FRANCIS FUKUYAMA, THE END 

OF HISTORY AND THE LAST MAN (1992).


If support for modern authoritarianism is not the product of ignorance and 

brainwashing but rather of a rational calculation that, under the present 

conditions, authoritarianism is the best way to generate and preserve economic 

growth, the Internet’s ability to disseminate and mobilize, while very impressive 

in itself, may not deliver the kind of benefits that so many in the West expect. 

To put it simply, citizens of authoritarian states may not be uninformed—so 

getting them informed is going to be of only limited value. As such, the Net’s 

ability to liberate the world is severely constrained by the absence of a strong 

intellectual vision for how a liberated Russia or China would look (and work) 

like. 

Now, if the Soviet experience is anything to judge by, revolutions don’t need 

the absolute majority of the population to be successful. In other words, it may 

be possible that a small group of politically active citizens could take advantage 

of political openings at the right moment and push for significant reform—or 

the overthrow of a government altogether. In situations like this, the Internet’s 

ability to mobilize may indeed come very handy. 

Several caveats are in order here. First, obviously, this doesn’t mean that the 

Internet can help create such political openings—those are usually created by 

structural factors. As much as it is tempting to believe that it was fax machines 

and photocopiers that brought down communism in Eastern European 

countries, one would probably be better off studying their dismal economic 

record in the late 1980s. So far, it seems that the information revolution—which 

many have taken to mean the end of authoritarianism everywhere—has, overall, 

had a positive impact on the rates of economic growth in modern authoritarian 

regimes—and, to this extent, that revolution may actually have strengthened these 

regimes. 

The second caveat is that there is little certainty that the group with the best 

ability to mobilize will also be the group with the most impressive democratic 

agenda. Once again, the Soviet example—with a tiny group of Bolsheviks 

gaining control of a country as massive as Russia—is quite instructive. Not all 

revolutions are democratic in character, and more than one of them ended up 

with the least democratic groups gaining power. Al-Qaeda is far better at using 

technology to mobilize the masses than the liberal voices of the Middle East; 

the Russian nationalists, likewise, are far more creative online than the 

democratic and pro-Western opposition. 

Third—and, perhaps, most important—what happens in between political 

openings matters a great deal as well. It’s simply not true that, from the 

perspective of an authoritarian state, all social mobilization is harmful. Take the 

case of China. Thanks to the extremely vibrant nationalist sector of the 

country’s blogosphere, the Chinese government is often pushed to adopt a 

much more aggressive posture—towards Taiwan, Japan, South Korea—than


they might otherwise have done. Such hawkishness in their foreign policy may 

or may not bolster their legitimacy; the short answer is that we have to look at 

the context. For our purposes, it seems clear that we won’t be able to 

understand the role of the Internet—even if one concedes that it is, indeed, 

conducive to more mobilization and contestation—if we study that role outside 

of the socio-political environment in which it is embedded. The assumption that 

“greater opportunities for social mobilization equals greater odds that 

democracy will prevail in the long run” simply is not true. 

This last point highlights the problem with the second grand assumption that 

we still need to examine—namely, that the Net won’t have any other primary or 

secondary effects on the quality and sustainability of authoritarianism. The 

account that prioritizes the role of information dissemination and mobilization 

usually rests on a very simplistic, even reductionist, theory of authoritarianism. 

It presumes the existence of an authoritarian chimera—the government—which 

controls its citizens through a combination of surveillance and coercion. 

Citizens, the theory goes, would take immediate advantage of the Internet and 

use it to push against the government. 

But why wouldn’t the government do the same to push against the citizens? In 

order to understand the overall impact of the Net on the “struggle for 

liberation,” one must study how it may have also facilitated government 

monitoring and control of what their citizens do. Even the most optimistic 

observers of the Net would concede that social networking, fun as it is, may not 

necessarily be the best way to protect one’s data—both because no social 

networking site is secure from occasional data leaks and because secret police 

around the world have now, inadvertently, obtained the ability to map the 

connections between different activists, see how they are related to foreign 

funders, and so on. 

Furthermore, there is a vibrant and rapidly-expanding global market in activities 

like face recognition, which makes the identification of those who participate in 

anti-government protests much easier—often this actually happens by 

comparing party photos they themselves upload to social networking sites with 

the photos taken at the protest rallies. Seasoned activists may, of course, be 

smart enough to steer away from social networking sites, but this hardly applies 

to the rest of the population. 

But conceding that it’s not just anti-government activists who have been 

empowered by the Net is only part of the story. The truth is that we simply 

can’t easily classify all social forces into “pro-” and “anti-government” simply 

based on the location of their offices (e.g., the secret police are in; the unions are 

out). In reality, modern authoritarian regimes derive their power from a much 

more diverse pool of resources than sheer brute force or surveillance. To 

understand what makes modern authoritarian regimes tick, one thus needs to


look at a whole range of other political, social, and cultural factors: religion, 

history, nationalism, geography (e.g., the relations between the federal center and 

the periphery), rates of economic growth, corruption, government efficiency, 

fear of a foreign invasion and so forth. Many of these factors have successfully 

been co-opted by modern authoritarian rulers to justify and prolong their rule. 

It’s easy to imagine an authoritarian regime that would become stronger as a 

result of (a) an increase in religious sentiment among its population, (b) the 

promotion of a particular interpretation of recent history that would justify the 

current political regime as inevitable and an unambiguous improvement over its 

predecessors, and (c) impressive rates of economic growth, with little 

corruption or government bureaucracy. Likewise, it’s easy to imagine how all of 

these developments would be amplified if (d) the Internet ends up providing 

more access to more religious materials to more believers (e.g., through mobile 

phones), (e) governments find a way to hire and compensate loyal bloggers for 

touting a particular reading of history, and (f) governments set up websites that 

allow citizens to report on corrupt officials, problems with existing 

infrastructure, or government waste. 

That last development may seem like a good thing—until one realizes that an 

authoritarian government with less government waste is not necessarily a 

weaker authoritarian government. It may actually be more effective and the 

country may enjoy faster rates of economic growth—but that, alas, still does not 

always translate into a more democratic government. 

All of this is to say that the only way to understand how the Internet influences 

authoritarianism is to first define a theory of authoritarianism itself—preferably, 

a theory that goes beyond Manichean theories of “the totalitarian state” versus 

“the dissidents”—and then use it to closely investigate how the Internet affects 

each of its components. 

As such, our ability to harvest the potential of the Net to “liberate the world” 

depends not so much on our ability to understand the Net but on our ability to 

understand the world itself. It’s much easier to understand how the Internet 

affects government efficiency than to understand how government efficiency 

affects government legitimacy under conditions of capitalism-friendly 

authoritarianism. 

Political scientists, unfortunately, don’t have much to boast of on this front: 

their understanding of this completely new breed of authoritarianism is at best 

rudimentary—and their understanding of how such a fluid and complex 

technology as the Internet can affect it is even worse. Given the immense 

poverty of our current conceptual apparatus, even if the Net does end up 

liberating the world, most likely we won’t know it for quite some time.

564 CHAPTER 10: WILL THE NET LIBERATE THE WORLD?


Internet Freedom: 

Beyond Circumvention 

By Ethan Zuckerman * 

U.S. Secretary of State Hillary Clinton’s January 2010 speech on Internet 

Freedom signaled a strong interest from the State Department in encouraging 

the use of the Internet to promote political reforms in closed societies. 1 It 

makes sense that the State Department would look to support existing projects 

to circumvent Internet censorship. The New York Times reports that a group of 

senators subsequently urged the Secretary to apply existing funding to support 

the development and expansion of censorship circumvention programs. 2 

My colleagues Hal Roberts, John Palfrey and I have studied the development of 

Internet circumvention systems over the past five years, and released a study last 

year that compared the strengths and weaknesses of different circumvention 

tools. 3 Some of my work at The Berkman Center for Internet & Society at 

Harvard University is funded by a U.S. State Department grant that focuses on 

the continuing study and evaluation of these sorts of tools. As a result, I spend 

a lot of time coordinating efforts between tool developers and people who need 

access to circumvention tools to publish sensitive content. 

I strongly believe that we need strong, anonymized and useable censorship 

circumvention tools. But I also believe that we need lots more than censorship 

circumvention tools, and I fear that both funders and technologists may overfocus 

on this one particular aspect of Internet freedom at the expense of other 

avenues. I wonder whether we’re looking closely enough at the fundamental 

limitations of circumvention as a strategy and asking ourselves what we’re 

hoping Internet freedom will do for users in closed societies. 

* Senior Researcher at the Berkman Center for Internet and Society at Harvard University. 

Thanks to Hal Roberts, Janet Haven and Rebecca MacKinnon for help editing and 

improving this essay. They’re responsible for the good parts. You can blame the rest on 

me. 

1 Hillary Rodham Clinton, Secretary of State, Remarks on Internet Freedom at the Newseum 

(Jan. 21, 2010), http://www.state.gov/secretary/rm/2010/01/135519.htm. 

2 Brad Stone, Aid Urged for Groups Fighting Internet Censors, N.Y. TIMES, Jan. 20, 2010, 

http://www.nytimes.com/2010/01/21/technology/21censor.html?_r=1. For more 

information on Tor, see http://www.torproject.org. For more information on Psiphon, 

see http://psiphon.ca. For more information on Freegate, see http://www.ditinc.us/freegate. 

3 HAL ROBERTS, ETHAN ZUCKERMAN & JOHN PALFREY, 2007 CIRCUMVENTION LANDSCAPE 

REPORT: METHODS, USES, AND TOOLS (March 2009), http://dash.harvard.edu/ 

bitstream/handle/1/2794933/2007_Circumvention_Landscape.pdf?sequence=2.


So here’s a provocation: We can’t circumvent our way around Internet 

censorship. 

I don’t mean that Internet censorship circumvention systems don’t work. They 

do—our research tested several popular circumvention tools in censored 

nations and discovered that most can retrieve blocked content from behind the 

Chinese firewall or a similar system. 4 There are problems with privacy, data 

leakage, the rendering of certain types of content, and particularly with usability 

and performance, but the systems we tested can indeed circumvent censorship. 

What I mean is this: We couldn’t afford to scale today’s existing circumvention 

tools to “liberate” all of China’s Internet users even if they all wanted to be 

liberated. 

Circumvention systems share a basic mode of operation—they act as proxies to 

let users retrieve blocked content. A user is blocked from accessing a website 

by her Internet Service Provider (ISP) or that ISP’s ISP. She may want to read a 

page from Human Rights Watch’s (HRW) website, which is accessible at IP 

address 70.32.76.212. But that IP address is on a national blacklist, and she’s 

prevented from receiving any content from it. So, she points her browser to a 

proxy server at another address—say, 123.45.67.89—and asks a program on 

that server to retrieve a page from the HRW website. Assuming that 

123.45.67.89 isn’t on the national blacklist, she should be able to receive the 

HRW page via the proxy. 

During the transaction, the proxy is acting like an Internet service provider. Its 

ability to provide reliable service to its users is constrained by bandwidth— 

bandwidth to access the destination site and to deliver the content to the proxy 

user. Bandwidth is costly in aggregate, and it costs real money to run a proxy 

that’s heavily used. 

Some systems have tried to reduce these costs by asking volunteers to share 

them—the first release of Citizen Lab’s Psiphon used home computers hosted 

by volunteers around the world as proxies, and then used their consumer 

bandwidth to access the public Internet. Unfortunately, in many countries, 

consumer Internet connections are optimized to download content and are 

much slower when they are uploading content. These proxies could access the 

Human Rights Watch website pretty quickly, but they took a very long time to 

deliver the page to the user behind the firewall. As a result, Psiphon is no 

longer primarily focused on trying to make proxies hosted by volunteers work. 

Tor, on the other hand, is, but Tor nodes are frequently hosted by universities 

and companies that have access to large pools of bandwidth. Still, available 

bandwidth is a major constraint of the Tor system. The most usable 

circumvention systems today—virtual private network (VPN) tools like 

4 See, generally, ROBERTS, ZUCKERMAN & PALFREY, supra note 3.


Relakks 5 or WiTopia 6—charge users between $3 and $6 per month to defray 

bandwidth costs. 

Assume that systems like Tor, Psiphon and Freegate receive additional funding 

from the U.S. State Department. How much would it cost to provide proxy 

Internet access for … well, China? China reports 384 million Internet users, 7 

meaning we’re talking about running an ISP capable of serving more than 25 

times as many users as the largest U.S. ISP. 8 According to the China Internet 

Network Information Center (CNNIC), China consumes 998,217 Mbps of 

international Internet bandwidth. 9 It’s hard to get estimates for what ISPs pay 

for bandwidth, though conventional wisdom suggests prices between $0.05 and 

$0.10 per gigabyte. Using $0.05 as a cost per gigabyte, the cost to provide the 

uncensored Internet to China would be $13,608,000 per month, or $163.3 

million a year in pure bandwidth charges, not including the costs of proxy 

servers, routers, system administrators and customer service. Faced with a bill 

of that magnitude, the $45 million U.S. senators are asking Secretary Clinton to 

spend quickly looks pretty paltry. 10 

There’s an additional complication—we’re not just talking about running an 

ISP—we’re talking about running an ISP that’s very likely to be abused by bad 

actors. Spammers, fraudsters and other Internet criminals use proxy servers to 

conduct their activities, both to protect their identities and to avoid systems on 

free webmail providers, for instance, which prevent users from signing up for 

dozens of accounts by limiting an IP address to a certain number of signups in a 

limited time period. For example, Wikipedia found that many users used open 

proxies to deface their system and now reserve the right to block proxy users 

from editing pages. 11 Proxy operators have a tough balancing act—for their 

proxies to be useful, people need to be able to use them to access sites like 

Wikipedia or YouTube, but if people use those proxies to abuse those sites, the 

proxy will be blocked. As such, proxy operators can find themselves at war 

with their own users, trying to ban bad actors to keep the tool useful for the rest 

of the users. 

5 For more information on Relakks, see http://www.relakks.com. 

6 For more information on WiTopia, see http://www.witopia.net. 

7 Chris Buckley, China Internet Population Hits 384 million, REUTERS, Jan. 15, 2010, 

http://www.reuters.com/article/idUSTOE60E06S20100115. 

8 See Top 23 U.S. ISPs by Subscriber: Q3 2008, ISP Planet, http://www.ispplanet.com/research/rankings/usa.html. 

9 See China Internet Network Info. Ctr., Internet Fundamental Data, 

http://www.cnnic.net.cn/en/index/0O/index.htm (last visited July 29, 2010). 

10 Brad Stone, Aid Urged for Groups Fighting Internet Censors, N.Y. TIMES, Jan. 20, 2010, 

http://www.nytimes.com/2010/01/21/technology/21censor.html. 

11 See Open Proxies, WIKIPEDIA, http://en.wikipedia.org/wiki/Wikipedia:Open_proxies.


I’m skeptical that the U.S. State Department can or wants to build or fund a 

free ISP that can be used by millions of simultaneous users, many of whom may 

be using it to commit click fraud or send spam. 12 I know—because I’ve talked 

with many of them—that the people who fund blocking-resistant Internet 

proxies don’t think of what they’re doing in these terms. Instead, they assume 

that proxies are used by users only in special circumstances, to access blocked 

content. 

Here’s the problem: A government like China is blocking a lot of content. As 

Donnie Dong notes in a recent blog post, five of the ten most popular websites 

worldwide are blocked in China. 13 Those sites include YouTube and Facebook, 

sites that eat bandwidth through large downloads and long sessions. Perhaps it 

would be realistic to act as an ISP to China if we were just providing access to 

Human Rights Watch—but it’s not realistic if we’re providing access to 

YouTube, too. 

Proxy operators have dealt with this question by putting constraints on the use 

of their tools. Some proxy operators block access to YouTube because it’s such 

a bandwidth hog. Others block access to pornography, both because it uses 

bandwidth and to protect the sensibilities of their sponsors. Others constrain 

who can use their tools, limiting access to people coming from Iranian or 

Chinese IP addresses, trying to reduce bandwidth use by American high school 

kids whose schools have blocked YouTube. In deciding who or what to block, 

proxy operators are offering their personal answers to a complicated question: 

What parts of the Internet are we trying to open up to people in closed societies? As we’ll 

address in a moment, that’s not such an easy question to answer. 

Imagine for a moment that we could afford to proxy China, Iran, Myanmar and 

others’ international traffic. We figure out how to keep these proxies unblocked 

and accessible (it’s not easy—the operators of heavily used proxy systems are 

engaged in a fast-moving cat and mouse game) and determine how to mitigate 

the abuse challenges presented by open proxies. We still have problems. 

Most Internet traffic is domestic. In China, we estimate that, at minimum, 95% 

of total traffic is within the country. Domestic censorship matters a great deal, 

and perhaps a great deal more than censorship at national borders. As Rebecca 

12 Matthew Broersma, Researchers Eye Open Proxy Attacks, TECHWORLD, Nov. 15, 2007, 

http://news.techworld.com/security/10663/researchers-eye-open-proxy-attacks. 

13 Donnie Dong, Google’s Angry, Sacrifice and the Accelerated Splitting Internet, BLAWGDOG, Jan. 13, 

2010, http://english.blawgdog.com/2010/01/googles-angry-sacrifice-andaccelerated.html.


MacKinnon documented in China’s Censorship 2.0, 14 Chinese companies censor 

user-generated content in a complex, decentralized way. As a result, a good deal 

of controversial material is never published in the first place, either because it’s 

blocked from publication or because authors decline to publish it for fear of 

having their blog account locked or cancelled. We might assume that if Chinese 

users had unfettered access to Blogger, they’d publish there. Perhaps not— 

people use the tools that are easiest to use and that their friends use. A 

seasoned Chinese dissident might use Blogger, knowing she’s likely to be 

censored—an average user, posting photos of his cat, would more likely use a 

domestic platform and not consider the possibility of censorship until he found 

himself posting controversial content. 

In promoting Internet freedom, we need to consider strategies to overcome 

censorship inside closed societies. We also need to address “soft censorship”: 

the co-opting of online public spaces by authoritarian regimes, which sponsor 

pro-government bloggers, seed sympathetic message board threads, and pay for 

sympathetic comments. Evgeny Morozov offers a thoroughly dark view of 

authoritarian use of social media in “How Dictators Watch Us on the Web.” 15 

We also need to address a growing menace to online speech—attacks on sites 

that host controversial speech. When Turkey blocks YouTube 16 to prevent 

Turkish citizens from seeing videos that defame Ataturk, they prevent 20 

million Turkish Internet users from seeing everything on YouTube. When 

someone—the Myanmar government, patriotic Burmese, mischievous 

hackers—mount a distributed denial of service attack on The Irrawaddy, 17 an 

online newspaper highly critical of the Myanmar government, this temporarily 

prevents everyone everywhere from seeing it. 

Circumvention tools help Turks who want to see YouTube get around a 

government block, but they don’t help Americans, Chinese or Burmese see The 

Irrawaddy if the site has been taken down by a Distributed Denial of Service 

(DDoS) 18 or hacking attack. Publishers of controversial online content have 

begun to realize that they’re not just going to face censorship by national 

14 Rebecca MacKinnon, China’s Censorship 2.0: How Companies Censor Bloggers, 14 FIRST MONDAY 

(Feb. 2009), http://firstmonday.org/htbin/cgiwrap/bin/ojs/index.php/fm/article/ 

view/2378/2089. 

15 Evgeny Morozov, How Dictators Watch Us on the Web, PROSPECT, Nov. 18, 2009, 

http://www.prospectmagazine.co.uk/2009/11/how-dictators-watch-us-on-the-web. 

16 Nico Hines, YouTube Banned in Turkey After Video Insults, THE TIMES, Mar. 7, 2007, 

http://www.timesonline.co.uk/tol/news/world/europe/article1483840.ece. 

17 Aung Zaw, The Burmese Regime’s Cyber Offensive, THE IRRAWADDY, Sept. 18, 2008, 

http://www.irrawaddy.org/opinion_story.php?art_id=14280 . 

18 A DDoS attack uses multiple computer systems to target and attack a single system, or 

website, thus preventing users from accessing the targeted system.


filtering systems—they’re going to face a variety of technical and legal attacks 

that seek to make their servers inaccessible. 

There’s quite a bit publishers can do to increase the resilience of their sites to 

DDoS attack and to make their sites more difficult to filter. To avoid blockage 

in Turkey, YouTube could increase the number of IP addresses that lead to the 

web server and use a technique called “fast-flux DNS” 19 to give the Turkish 

government more IP addresses to block. They could maintain a mailing list to 

alert users to unblocked IP addresses where they could access YouTube, or 

create a custom application that disseminates unblocked IPs to YouTube users 

who download the application. These are all techniques employed by content 

sites that are frequently blocked in closed societies. 

YouTube doesn’t utilize these anti-blocking measures for two reasons. One, it 

has historically preferred to negotiate with nations who filter the Internet to 

make YouTube sites accessible again, rather than to work against these nations 

by fighting filtering. (This may be changing, now that Google has decided to 

disengage from China due to censorship and hacking issues.) Second, YouTube 

doesn’t really have an economic incentive to be unblocked in Turkey. If 

anything, being blocked in Turkey, and perhaps even in China, may even be to 

its economic advantage, since serving these countries is likely to be unprofitable. 

Sites that enable distribution of user-created content are supported by 

advertising traffic. Advertisers are generally more excited about reaching users 

in the U.S. who have credit cards, more disposable income and are inclined to 

buy online than users in China or Turkey. Some suspect that the introduction 

of “lite” versions of services like Facebook is designed to serve users in the 

developing world at lower cost, since those users rarely create income for the 

sites. 20 In economic terms, it may be hard to convince Facebook, YouTube and 

others to continue providing services to closed societies, where they have a 

tough time selling ads. We also may need to ask more of them to take steps to 

ensure that they remain accessible and useful in censorious countries. 

In short: 

� Internet circumvention is difficult and expensive. It can make it easier 

for people to send spam and steal identities. 

� Circumventing censorship through proxies gives people access to 

international content, but doesn’t address domestic censorship, which 

likely affects the majority of people’s Internet behavior. 

19 Fast-flux DNS prevents the identification of a host server’s IP address. 

20 Brad Stone & Miguel Helft, In Developing Countries, Web Grows Without Profit, N.Y. TIMES, 

April 26, 2009, http://www.nytimes.com/2009/04/27/technology/startups/27global.html?_r=1.


� Circumventing censorship doesn’t offer a defense against DDoS or 

other attacks that target publishers. 

To figure out how to promote Internet freedom, we need to start addressing the 

question: “How do we think the Internet changes closed societies?” In other 

words, do we have a “theory of change” 21 behind our desire to ensure people in 

Iran, Burma, China, etc., can access the Internet? Why do we believe this is a 

priority for the U.S. State Department or for public diplomacy as a whole? 

Much work on Internet censorship isn’t motivated by a theory of change—it’s 

motivated by a deeply-held conviction—one that I share—that the ability to 

share information is a basic human right. Article 19 of the Universal 

Declaration of Human Rights states that “Everyone has the right to freedom of 

opinion and expression; this right includes freedom to hold opinions without 

interference and to seek, receive and impart information and ideas through any 

media and regardless of frontiers.” 22 The Internet is the most efficient system 

we’ve ever built to allow people to seek, receive and impart information and 

ideas, and therefore, we need to ensure everyone has unfettered Internet access. 

The problem with the Article 19 approach to censorship circumvention is that it 

doesn’t help us prioritize. It simply makes it imperative that we solve what may 

be an unsolvable problem. 

If we believe that access to the Internet will change closed societies in a 

particular way, we can prioritize access to those aspects of the Internet. Our 

theory of change helps us figure out what we must provide access to. The four 

theories I list below are rarely explicitly stated, but I believe they underlie much 

of the work behind censorship circumvention. 

The Suppressed Information Theory: If we can provide certain 

suppressed information to people in closed societies, they’ll rise up, 

challenge their leaders and usher in a different government. We might 

choose to call this the “Hungary ‘56 theory” 23—reports of struggles against 

communist governments around the world, reported into Hungary via 

Radio Free Europe, encouraged Hungarians to rebel against their leaders. 

(Unfortunately, the U.S. didn’t support the revolutionaries militarily—as 

many in Hungary had expected—and the revolution was brutally quashed 

by a Soviet invasion.) 

21 Mark Schmitt, The “Theory of Change” Primary, THE AMERICAN PROSPECT, Dec. 21, 2007, 

http://www.prospect.org/cs/articles?article=the_theory_of_change_primary. 

22 The Universal Declaration of Human Rights, art. 19, G.A. Res. 217A(III), U.N. GAOR, 3d 

Sess., U.N. Doc. A/810 (Dec. 10, 1948), 

http://www.un.org/en/documents/udhr/index.shtml. 

23 For more information on the Hungarian Revolution of 1965, see Hungarian Revolution of 1956, 

WIKIPEDIA, http://en.wikipedia.org/wiki/Hungarian_Revolution_of_1956.


Or we could term this the “North Korea theory,” because a state as closed 

as North Korea might be a place where unsuppressed information—about 

the fiscal success of South Korea, for instance—could provoke revolution. 

Barbara Demick’s beautiful piece in the New Yorker, “The Good Cook,” 

gives a sense of how little information most North Koreans have about the 

outside world and how different the world looks from Seoul. 24 

Nonetheless, even North Korea is less informationally isolated than we 

think—The Dong-A Ilbo, a South Korean newspaper, reports an 

“information belt” along the North Korea/China border where calls on 

smuggled mobile phones are possible between North and South Korea. 25 

Other nations are far more open—the Chinese tend to be extremely well 

informed about both domestic and international politics, both through 

using circumvention tools and because Chinese media reports a great deal 

of domestic and international news. 

It’s possible that access to information is a necessary, though not sufficient, 

condition for political revolution. It’s also possible that we overestimate 

the power and potency of suppressed information, especially as information 

is so difficult to suppress in a connected age. 

The Twitter Revolution Theory: If citizens in closed societies can use the 

powerful communications tools made possible by the Internet, they can 

unite and overthrow their oppressors. This is the theory that led the U.S. 

State Department to urge Twitter to postpone a period of scheduled 

downtime during the Iran election protests. 26 While it’s hard to make the 

case that technologies of connection are going to bring down the Iranian 

government, 27 good examples exist, like the role of the mobile phone in 

helping to topple President Estrada in the Philippines. 28 

There’s been a great deal of enthusiasm in the popular press for the Twitter 

revolution theory, but careful analysis reveals some limitations. The 

communications channels opened online tend to be compromised quickly, 

24 Barbara Demick, The Good Cook, THE NEW YORKER, Nov. 2, 2009, at 58, 

http://www.newyorker.com/reporting/2009/11/02/091102fa_fact_demick. 

25 North Koreans Directly Connect with South Koreans via Chinese Cell Phones, ASK A KOREAN!, Jan. 17, 

2010, http://askakorean.blogspot.com/2010/01/excellent-article-on-dong-ilboabout.html. 

26 Sue Pleming, U.S. State Department Speaks to Twitter Over Iran, REUTERS, June 16, 2009, 

http://www.reuters.com/article/idUSWBT01137420090616. 

27 See Cameron Abadi, Iran, Facebook, and the Limits of Online Activism, FOREIGN POLICY, Feb. 12, 

2010, http://www.foreignpolicy.com/articles/2010/02/12/irans_failed_facebook_ 

revolution. 

28 See Joseph Estrada Controversies, WIKIPEDIA, 

http://en.wikipedia.org/wiki/Joseph_Estrada#Controversies.


used for disinformation and monitoring activists. And when protests get 

out of hand, governments of closed societies don’t hesitate to pull the plug 

on networks—China has blocked Internet access in Xinjiang for months, 

and Ethiopia turned off SMS on mobile phone networks for years after 

they were used to organize street protests. And it’s worth noting that 

prophesied “twitter revolutions” in Moldova and Iran both failed in the 

face of authoritarian governments. 

The Public Sphere Theory: Communication tools may not lead to 

revolution immediately, but they provide a new rhetorical space where a 

new generation of leaders can think and speak freely. In the long run, this 

ability to create a new public sphere, parallel to the one controlled by the 

state, will empower a new generation of social actors, though perhaps not 

for many years. 

Marc Lynch made a pretty persuasive case for this theory in a talk last year 

about online activism in the Middle East. 29 In the former Soviet Union, 

samizdat (self-published, clandestine media) was probably more important 

as a space for free expression than it was as a channel for disseminating 

suppressed information. 30 The emergence of leaders, like Vaclav Havel, 

whose authority was rooted in cultural expression as well as political power, 

makes the case that simply speaking out is powerful. But the long timescale 

of this theory makes it hard to test. 

The theory we accept shapes our policy decisions. If we believe that 

disseminating suppressed information is critical—either to the public at large or 

to a small group of influencers—we might focus our efforts on spreading 

content from Voice of America or Radio Free Europe. Indeed, this is how 

many government forays into censorship circumvention began—national news 

services began supporting circumvention tools so their content, painstakingly 

created in languages like Burmese or Farsi, would be accessible in closed 

societies. This is a very efficient approach to anti-censorship—we can ignore 

many of the problems associated with abusing proxies and focus on prioritizing 

news over other less-important bandwidth-hogging uses, like the video of the 

cat flushing the toilet. Unfortunately, we’ve got a long track record that shows 

that this form of anti-censorship doesn’t magically open closed regimes, which 

suggests that increasing our reliance on this strategy might be a poor idea. 

29 Ethan Zuckerman, Marc Lynch Asks Us to be Realistic About Digital Activism in the Middle East, 

April 27, 2009, http://www.ethanzuckerman.com/blog/2009/04/27/marc-lynchasks-us-to-be-realistic-about-digital-activism-in-the-middle-east. 

30 See generally, Peter Steiner, Introduction: On Samizdat, Tamizdat, Magnitizdat, and Other Strange 

Words, 29 POETICS TODAY 613 (2008) 

http://poeticstoday.dukejournals.org/cgi/reprint/29/4/613.pdf.


If we adopt the Twitter Revolution theory, we should focus on systems that 

allow for rapid communication within trusted networks. This might mean tools 

like Twitter or Facebook, but it probably means tools like LiveJournal and 

Yahoo! Groups, which gain their utility through exclusivity, allowing small 

groups to organize outside the gaze of the authorities. If we adopt the public 

sphere approach, we want to open any technologies that allow public 

communication and debate—blogs, Twitter, YouTube, and virtually anything 

else that fits under the banner of Web 2.0. This, unfortunately, presents 

technical challenges that are proving extremely difficult to solve. 

What does all this mean in terms of how the U.S. State Department should 

allocate their money to promote Internet Freedom? My goal was primarily to 

outline the questions they should be considering, rather than offering specific 

prescriptions. But here are some possible implications of these questions: 

If we believe the U.S. government should be exporting “Internet freedom”— 

and there are good reasons to argue that a government, and particularly the US 

government, shouldn’t take on this task—we need to continue supporting 

circumvention efforts, at least in the short term. But we need to disabuse 

ourselves of the idea that we can “solve” censorship through circumvention. 

We should support circumvention until we find better technical and policy 

solutions to censorship, not because we can tear down the Great Firewall by 

spending more on proxies, etc. 

Second, if we want more people using circumvention tools, we need to find 

ways to make these systems fiscally sustainable. Sustainable circumvention is 

becoming an attractive business for some companies. 31 It needs to be part of a 

comprehensive Internet freedom strategy, and we need to develop strategies 

that are sustainable and provide low- to zero-cost access to users in closed 

societies. 

Third, as we continue to fund circumvention, we need to address usage of these 

tools to send spam, commit fraud and steal personal data. We might do this by 

relying less on IP addresses as an extensive, fundamental means of regulating 

bad behavior, but we have to find a solution that protects networks against 

abuse while maintaining the possibility of anonymity, a difficult balancing act. 

Additionally, we need to shift our thinking from helping users in closed 

societies access blocked content to helping publishers reach all audiences. In 

doing so, we may gain those publishers as a valuable new set of allies as well as 

opening a new class of technical solutions. 

31 Lara Farrar, Cashing in on Internet Censorship, CNN, Feb. 19, 2010, 

http://www.cnn.com/2010/TECH/02/18/internet.censorship.business/?hpt=Sbin.


Furthermore, if our goal is to allow people in closed societies to access an 

online public sphere or to use online tools to organize protests, we need to 

bring the administrators of these tools into the dialog. Secretary Clinton 

suggests that we make free speech part of the American brand identity—let’s 

find ways to challenge companies to build blocking resistance into their 

platforms and to consider Internet freedom as a central part of their business 

mission. We need to address the fact that making platforms unblockable has a 

cost for content hosts and that their business models currently don’t reward 

companies for providing services to blocked users. 

The U.S. government should treat Internet filtering—and more aggressive 

hacking and DDoS attacks—as a barrier to trade. The U.S, should strongly 

pressure governments in open societies like Australia and France to resist the 

temptation to restrict Internet access, as this behavior helps China and Iran 

make the case that their censorship is in line with international norms. And we 

need to fix U.S. treasury regulations that make it difficult and legally ambiguous 

for companies like Microsoft and projects like SourceForge 32 to operate in 

closed societies. If we believe in Internet Freedom, the first step is rethinking 

these policies so they don’t hurt ordinary Internet users. 

Finally, if attempts to export Internet freedom are to be met with something 

other than cynicism or skepticism, the U.S. government needs to do a better job 

of protecting free speech domestically. The pressure exerted by individual 

Senators and by the State Department on companies like Amazon and PayPal to 

terminate services to WikiLeaks calls into question the U.S. government’s 

commitment to online free speech. If the U.S. wants countries like China to 

consider a more free and open Internet, control of the Internet in the U.S. must 

also follow the rule of law, and not fall victim to political expediency. 

If we take seriously Secretary Clinton’s call, the danger is that we increase our 

speed marching in the wrong direction. As we embrace the goal of Internet 

Freedom, now is the time to ask what we’re hoping to accomplish and to shape 

our strategy accordingly. 

32 SourceForge is an open source code depository from which software can be developed and 

downloaded. For more information, see http://sourceforge.net/.

The BesT Thinking aBouT 

The FuTure oF DigiTal Policy 

This unique collection brings together 26 thought leaders on Internet 

law, philosophy, policy and economics to consider what the next 

digital decade might bring. Has the Internet been good for our culture? 

Is the Internet at risk from the drive to build more secure, but less 

“open” systems and devices? Is the Internet really so “exceptional?” 

Has it fundamentally changed economics? Who—and what ideas— 

will govern the Net in 2020? Should online intermediaries like access 

providers, hosting providers, search engines and social networks do 

more to “police” their networks, increase transparency, or operate 

“neutrally?” What future is there for privacy online? Can online free 

speech be regulated? Can it really unseat tyrants? These 31 thoughtprovoking 

essays tackle these questions and more. This book is 

essential reading for anyone gazing toward the digital future. 

conTriBuTors 

Rob Atkinson 

Stewart Baker 

Ann Bartow 


Larry Downes 

Josh Goldfoot 

Eric Goldman 

James Grimmelman 


TechFreedom 

techfreedom.org 

1899 l sT nW, 12th Floor 

Washington, D.c. 20036 

David Johnson 

Andrew Keen 

Hon. Alex Kozinski 


Geoff Manne 



John Palfrey 


Paul Szynol 

Adam Thierer 

Hal Varian 

Christopher Wolf 

Tim Wu 




NextDigitalDecade.com

downloaded as a PDF - Next Digital Decade

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?