Jump to ContentJump to Main Navigation
Who Controls the Internet?Illusions of a Borderless World$

Jack Goldsmith and Tim Wu

Print publication date: 2006

Print ISBN-13: 9780195152661

Published to Oxford Scholarship Online: November 2020

DOI: 10.1093/oso/9780195152661.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2022. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: null; date: 26 January 2022

Consequences of Borders

Consequences of Borders

(p.147) 9 Consequences of Borders
Who Controls the Internet?

Jack Goldsmith

Tim Wu

Oxford University Press

Abstract and Keywords

Australia’s Joseph Gutnick is a billionaire, a diamond and gold miner, a political player, a philanthropist, and a rabbi. On October 20, 2000, Gutnick awoke in Victoria to find himself accused of tax evasion and money laundering by the American business magazine Barron’s....

Australia’s Joseph Gutnick is a billionaire, a diamond and gold miner, a political player, a philanthropist, and a rabbi. On October 20, 2000, Gutnick awoke in Victoria to find himself accused of tax evasion and money laundering by the American business magazine Barron’s. The article, “UnHoly Gains,” suggested that Gutnick had engaged in shady dealings with Nachum Goldberg, a Melbourne money launderer jailed in 2000 for washing AU$42 million in used notes through a bogus Israeli charity.1 Gutnick read the story, not in the print version of Barron’s but on the online version of its sister publication, “wsj.com,” a website on a server physically located in New Jersey. Gutnick was not the only Australian to read the story. Approximately seventeen hundred Australians subscribed to wsj.com, including many Australian business and finance leaders. An enraged Gutnick vehemently denied the illicit association with Goldberg. To protect his reputation, he sued Dow Jones & Company—the parent company of both Barron’s and the Wall Street Journal— in an Australian court, taking advantage of tough Australian libel laws unleavened by the U.S. First Amendment.

The legal arguments in the Gutnick case mirrored those in the Yahoo litigation in France a few years earlier. Dow Jones argued that Australian courts were legally powerless (or “without jurisdiction”) to rule on the legality of information on a computer in the United States, even if it appeared in Australia.2 The Australian High Court, like the court in France, disagreed. For material published on the Internet, it (p.148) stated, the place where the person downloads the material “will be the place where the tort of defamation is committed.”3 Within two years of this decision, Dow Jones agreed to pay Gutnick AU$180,000 in damages and AU$400,000 in legal fees to settle the case.4 It also issued this retraction: “Barron’s has no reason to believe Mr. Gutnick was ever a customer of Mr. Goldberg, and has no reason to believe that Mr. Gutnick was a money laundering customer of, or had any criminal or other improper relationship with, Mr. Goldberg.”5

“The U.S. cannot impose their laws on this country,” rejoiced Gutnick in Australia. “They have to respect our law.”6 But the Australian decision also attracted the same predictably negative reaction as the French decision in the Yahoo case. Fierce libel laws are antiquated in the Internet age and are inconsistent with the Net’s First Amendment-inspired approach to free expression. Australia has no business getting involved when a newspaper writes an article in New Jersey meant for an American audience. The Australian rule will chill speech in the United States and elsewhere, forcing newspapers around the world to bow to the most restrictive laws in the world. As a gloomy New York Times editorial page said, “To subject distant providers of online content to sanctions in countries intent on curbing free speech—or even to 190 different libel laws—is to undermine the Internet’s viability.”7

The Dow Jones–Gutnick controversy is no different than thousands of other conflicts of laws that have arisen on the Internet during the last decade. These conflicts give the lie to Frances Cairncross’s 1999 prediction that the death of distance heralded by the Internet would be a “powerful force for peace” and mutual understanding among nations.8 These conflicts of laws have not, however, had the widely predicted devastating effects on the Internet itself. Publishing and commerce are flourishing on the Internet despite the dozens of “parochial” national laws to which e-businesses are supposedly subject. And individuals continue to send e-mails, create web pages, and write blogs despite the supposed prospect of having to figure out how to comply with every law in the world.

This chapter explains why the predicted doomsday scenarios have not materialized. It begins by summarizing and extending the normative case made throughout this book for the bordered Internet. It then (p.149) addresses the conflict of laws problem. The heated rhetoric about conflicts of laws masks two more salient operating principles: multinational firms want to minimize global operating costs, and libertarians want to extend the unusually tolerant values of the U.S. First Amendment across the globe. As we will see, national Internet laws are no more burdensome than the scores of conflicting national laws that multinational firms typically face, and they have no effect on the vast majority of individual Net users who, unlike global firms, lack a multinational presence. There is no denying that the bordered Internet replicates some of the familiar costs and pathologies that result when nations apply their laws to transnational communications and transactions. But like the international system itself, it also lets many different peoples coexist on the same planet while maintaining very different values and ideas of the good life. In this diversity lies a happier world than one governed by a single global law for all matters. When dreaming of a better society centered on the Internet, the many virtues of a bordered system must not be overlooked.

Borders 2.0

This book has described three reasons why what we once called a global network is becoming a collection of nation-state networks—networks still linked by the Internet protocol, but for many purposes separate. First, peoples in different nations tend to read and speak different languages and have different backgrounds, capacities, preferences, desires, and needs. These reflect local differences in history, culture, geography, and wealth. Internet users seek out, and content providers want to provide, congenial content that reflects these differences.

Technological developments are the second reason for Internet borders. The Chinas of the world are becoming remarkably sophisticated at firewalling their countries and creating closed national networks. Internet geo-ID technologies are becoming faster and cheaper and more prevalent. These technologies enable content to be tailored to Net users by geography and permit e-firms to avoid sending content to places where it is illegal. Even the deep structure of the Internet—bandwidth distribution, increasing Internet traffic within countries and (p.150) regions, and diminished traffic between countries—reinforces Internet borders.

The enforcement of national laws in cases like Gutnick and Yahoo are the third reason for Internet borders. One strong and important difference among peoples concerns their values, and people with different values disagree about the type of information they want to receive and the type of information they deem harmful. Some societies tolerate Nazi goods; others don’t. Some like privacy warning labels; others don’t. Some accept online gambling; others don’t. Some want strong protections for intellectual property; again, others differ. These differences are reflected in different national laws, and governmental officials charged with enforcing national values must enforce these laws, as cases like Gutnick and Yahoo make clear.

The bordered Internet is widely viewed to be a dreadful development that is antithetical to the Internet’s “true” purposes and undermines the Internet’s promise. The issue tends to be joined most fiercely in the context of speech regulation, as the storm over the Gutnick and Yahoo cases reveal. There are many reasons for this focus, but the main one is that the Internet is a revolutionary medium of communication, and communication is speech. In that sense, just about every debate about Internet governance is at bottom a debate about speech governance. The most basic question about the bordered Internet, therefore, is whether speech should be regulated globally or locally.

We think that there is very little to say in favor of a single global rule for Internet speech. “Every jurisdiction controls access to some speech . . . but what that speech is differs from jurisdiction to jurisdiction,” explain Lawrence Lessig and Paul Renick. “What constitutes ‘obscene’ speech in Tennessee is permitted in Holland; what constitutes porn in Japan is child porn in the United States; what is ‘harmful to minors’ in Bavaria is Disney in New York.”9 These dramatically different attitudes toward proper speech among mature democracies reflect important differences among the peoples that populate these countries—differences in culture, history, and tastes that are legitimately reflected in national and local laws.

In the United States, it is acceptable to join a political party that condones racism, and courts uphold the right of neo-Nazis to parade through predominantly Jewish towns wearing uniforms and swastikas. (p.151) Other democracies, influenced by very different histories and tradition, take a different view. Israel is a democracy that rose out of the Holocaust and that persists in a nearly constant state of emergency. It bans speech that is “offensive” or causes “emotional harm,” and it outlaws political parties that espouse racism or call for the destruction of the State of Israel.10 Germany bans Nazi speech for yet a different reason, the same reason that Japan’s Constitution outlaws aggressive war: it is a nation still coming to grips with the horrors it committed in the past, and it is terrified that they could happen again.11 As we saw in chapter 1, France too bans pro-Nazi speech, as well as speech that endorses or minimizes the Holocaust. French law reflects its occupation by Nazi Germany during World War II, and its related belief that a person’s right to be free from threatening and degrading speech trumps the right to voice one’s political ideas, however harmful.12

The Gutnick case reveals the same underlying tension. The case arises out of deep differences between the United States and Australia on the importance of free speech, reputation, and public order. Although both countries are former British colonies influenced by the English common law tradition, they have for forty years taken vastly different approaches to free speech protections for the press. In 1964, the U.S. Supreme Court in New York Times v. Sullivan broke with the British common law when it interpreted the First Amendment to make it much harder for public officials to recover in libel suits against newspapers. The Sullivan case embraced a new “commitment to the principle that debate on public issues should be uninhibited, robust, and wide-open.”13 After Sullivan, it has been very hard for public figures like Gutnick to win in libel cases in the United States. The Australian High Court, however, rejects the Sullivan rule, reasoning that it “tilts the balance unduly in favor of free speech” and “gives inadequate protection to reputation.”14 Whereas American libel law places a high burden of proof on the alleged victim of defamatory speech, Australia places the burden of proof on the publisher and requires the publisher to reasonably believe that its statement concerning the alleged victim is true, and to give the alleged victim a chance to reply.15 The dispute in Gutnick thus goes far beyond the dry rules of libel law. It reflects deeper disagreements between the United States and Australia about (p.152) the processes that best secure truth, and about the relative values of robust speech versus reputation and uninhibited debate versus order.

These examples show that deeply held differences in values, even among democracies, lie behind conflicts of laws. A bordered Internet is valuable precisely because it permits people of different value systems to coexist on the same planet.

A good way to understand the case for a bordered Internet is to consider its opposite. Imagine a global law in the form of a world government or a world treaty. Set aside the insurmountable problem of creating a legitimate and reliable global executive to enforce such global norms. A more fundamental problem is that the global norms would often be unattractive, even if they could be enforced. When you choose a single rule for six billion people, odds are that several billion, or more, will be unhappy with it. Should divorce and abortion and pornography be allowed, and if so to what degree? Ought economic and environmental policy reflect the interests of poorer or wealthier nations? Similar questions arise on the Internet: Is the American approach to Nazi speech right, or the French-German-Israeli variants? Should the competing interests of the free press and private reputation be balanced Australian style or U.S. style? To what degree should gambling and pornography on the web be allowed? Should data privacy be unregulated, modestly regulated, or heavily regulated? A single answer to these and thousands of other questions would leave the world divided and discontented.

The advantage of decentralized governance is that it can better reflect differences among peoples.16 Consider what would happen when three nations with equal populations of 100 people—A, B, and C—tried to decide whether web gambling should be allowed in their country.17 Assume that 75 percent of the people in A, 65 percent of the people in B, and 35 percent of the people in C want to ban web gambling, with the remainder of the population in all three countries opposed to the ban. If the countries decided on a “global” rule reflecting majority preferences among the 300 people in the three territories, the result would be a global ban on web gambling with 175 people pleased and 125 displeased. But if each country can decide whether to ban web gambling for itself, A and B will ban web gambling and C will not, and in the aggregate 205 people will be pleased and 95 displeased. (p.153) In this way, decentralized government can respond in a more fine-grained way to what people want and can best enhance overall welfare.

Of course, this and other arguments for a bordered Internet must confront the problem of China and other oppressive nations that do not purport to represent the interests or preferences of their people. But even the China example, as bad as it is, does not undermine the case for territorial control of the Internet. Governments did not create the technologies that China is using to keep unwanted content out. Rather, as we saw in chapter 4, the private sector created it in response to consumer demands that the Net’s content be better tailored to suit individual interests—interests that, as a brute fact, cluster by geography. And as the Yahoo case shows, governments in democratic states are starting to demand that this technology be used to respond to entirely appropriate constituent demands to protect them not only from Nazi goods but also from hate speech, credit card theft, invasions of privacy, sexual predators, spam, and much more. Technologies of control designed to serve legitimate and desired ends can rarely be limited to those ends, and will often be co-opted for illegitimate purposes. The Internet is no exception.

The question about the optimal form of Internet governance must always be “compared to what?” While it is easy to criticize traditional territorial government and bemoan its many failures, there is no reasonable prospect of any better system of governmental organization. Even acknowledging that in places like China the laws will often not reflect the wishes of people who live there, differences among laws in the many democratic governments in the world (such as the ones at issue in Yahoo and Gutnick) are presumptively legitimate. Many elements of China’s bordered Internet, moreover, do respond to legitimate Chinese preferences—for example, the language in which Net content is delivered, and the cheap digital goods that are helping the Chinese economy flourish.

In defending the system of decentralized national control, we are not arguing for the current number and size of territorial nation-states. Nations that are too small will lack the economic capacity to provide public goods like national defense and education. As a nation’s size increases it can address these deficiencies, but at the cost of increasingly diverse values, preferences, and commitments among citizens, (p.154) which makes it harder for the ever-more-distant government to promulgate and enforce rules that are viewed as legitimate by all.18 This is one reason why many large modern democratic nations—the United States, Australia, Germany—have federal systems that make important governmental decisions at the subunit level. The European Union is an emerging territorial nation-state created (like Italy and Germany in the nineteenth century) out of smaller ones. To work it will need to retain its federalist structure, and we may now be approaching the point where its increasingly heterogeneous peoples will stall further enlargement. By contrast, many nations (including some in the EU) face devolution pressures, as distinct groups within nations clamor for greater direct control over aspects of their lives. There is a natural limit here too. In the words of UN Secretary-General Boutros Boutros-Ghali, “If every ethnic, religious or linguistic group claimed state-hood, there would be no limit to fragmentation, and peace, security and well-being for all would become ever more difficult to achieve.”19

Nation-states have always faced these competing pressures for expansion and contraction, and the Internet will surely exacerbate them in many ways. But pressures to change the size of nation-states should not overshadow the many ways that the decentralized territorial system itself promotes diversity and self-determination, even with regard to Internet communications. There is, however, another objection to decentralized control. Even if differing national laws reflect what’s best for people in those countries, the argument goes, the global effects of national control of the Internet are ruinous for the Internet. It is to this argument that we now turn.


Australia’s Gutnick decision “puts at risk the ability of Americans to speak with each other and be protected by American law when they do so,” said First Amendment maven Floyd Abrams.20 Abrams was complaining about the extraterritorial effect of the Australian decision. The Australian court effectively applied Australian laws outside Australian borders to a publication in the United States that was intended primarily for an American audience. It applied Australian law to an American company, (p.155) Dow Jones, whose political interests were not formally represented in Australia. And in so doing, the court undoubtedly caused Dow Jones to become more cautious about what it published in America, thereby contravening American free-speech values and depriving Americans and others across the globe of information.

All of this is true—but it is also inevitable and commonplace. Seventy-five years ago an international arbitral panel ruled that Canada was responsible for preventing sulfur dioxide emissions from Canada that caused agricultural damage in the United States.21 U.S. officials, had they so desired, could have applied U.S. law to make the Canadian firm pay for the damage caused in the United States. The punishment in the U.S. of a Canadian polluter would have had the effect of raising the cost of smelting, and thus the price of metals, in Canada. But these “extraterritorial” effects of U.S. law do not call into question the United States’ right and duty to protect Americans in America. If the United States does not act against the Canadian polluter, then the permissive Canadian law would have resulted in the “extraterritorial” damage in the United States. In this sense, extraterritorial effects always run in both directions when two nations try to apply their different laws to the same transnational event. These inevitable cross-border effects do not undermine the legitimacy of a nation applying its laws to redress local harms.

Consider a more recent example. In the late 1990s, Boeing and McDonnell Douglas, two American aerospace giants that did business worldwide but had their productive resources in the United States, tried to merge. The U.S. Federal Trade Commission (FTC) approved the merger, but the European Commission threatened to stop the merger because it viewed Boeing’s exclusive contracts with other airlines to be harmful to European airline competition. Ultimately Boeing gave in to the commission’s demands and eliminated exclusive contracting. This meant that the commission’s threatened injunction raised the costs in the United States of a merger between two American companies and superceded the regulatory efforts of the FTC. But if the commission had not enforced the EU laws, the more permissive American laws would have caused harmful anticompetitive effects in Europe. Once again, whichever nation’s law ends up applying to transnational activity will inevitably have indirect effects in another state. But these (p.156) effects are perfectly legitimate by-products of the EU’s action to protect Europeans from what it deemed to be the harmful local effects of offshore activity.22

In international law, borders are fundamental. As a general matter, nations can exercise coercive powers within their borders but not beyond.23 But a nation can always take steps within its territory to stop and redress harms that come from abroad. Indeed, as we saw in chapter 5 control of local Internet intermediaries is the main way that governments control offshore Internet harms. The principle that gives a nation the right and duty to protect citizens from locally caused harms applies with equal if not greater force when the harm comes from abroad. Not surprisingly in our modern interconnected world, nations frequently apply local law to harms from abroad. In addition to the pollution and antitrust examples, nations have long applied local law to regulate unwanted television and radio broadcasts from abroad, the harmful local effects of offshore frauds, local crimes (like drug dealing) initiated elsewhere, and the like.

A government’s responsibility for redressing local harms caused by a foreign source does not change because the harms are caused by an Internet communication. Cross-border harms that occur via the Internet are not any different than those outside the Net. Both demand a response from governmental authorities charged with protecting public values. When Nepali scam-artists defraud Indian investors in India, the Indian government must act, regardless of whether the fraud occurred in a magazine from Nepal or an e-mail from there. The United States wants to stop the local consumption of child pornography produced in Russia regardless of the medium—World Wide Web, magazine, or video—in which the porn appears. The French view the sale of Nazi paraphernalia as repugnant, whether it is sold on Yahoo’s servers or by mail-order catalogue. In short, nations have a right and a duty to protect their citizens from harm, whatever the source and whatever the medium.

These points illuminate the Gutnick decision. The Australian decision had effects in the United States, to be sure. But if Australia had not applied its laws to redress the harm to Gutnick in Australia, U.S. First Amendment law and speech-protective U.S. libel laws would have produced harmful and unwanted effects in Australia. This point is invariably missed by the critics of government control over the Net, (p.157) who believe that the U.S. First Amendment reflects universal values and is somehow written into the architecture of the Internet. But the First Amendment does not reflect universal values; to the contrary, no other nation embraces these values, and they are certainly not written into the Internet’s architecture. Enforcing the outlier First Amendment in Gutnick would have meant eviscerating Australian laws that reflected Australian values and concerns. But there is no reason why Australia should yield local control over its territory in order to accommodate Internet users in the United States. Nor should it absorb the costs in Australia of U.S. Internet activity simply because the Australian law might produce costs in the United States. Australia can regulate the local harm of transnational Internet activity even if doing so harms Dow Jones in the United States.

This result is not unfair to Dow Jones. Dow Jones chose to publish in Australia, and, as the court in Gutnick noted, “there is nothing unique about multinational business” that makes it exempt from local law.24 Compliance with Australian libel laws—like compliance with Australian tax laws, Australian accounting laws, and Australian consumer protection laws—is a cost of doing business in Australia. As the Australian court noted, “If people wish to do business in, or indeed travel to, or live in, or utilize the infrastructure of different countries, they can hardly expect to be absolved from compliance with the laws of those countries.”25 Dow Jones reaps financial and other benefits from its presence in Australia. Without this presence Australian enforcement threats would be empty. Dow Jones need not remain in Australia; it can close its shop there if Australian laws become too burdensome. Its decision to continue operations in Australia after settling with Gutnick reflects the company’s judgment that the benefits of doing business in Australia outweigh its costs.

Nor is the Gutnick decision unfair to consumers of Dow Jones news in the United States and other countries. At first glance it seems unfair if the Australian decision causes Dow Jones to stop publishing news that might have been of interest to Dow Jones readers in the United States. But again, such a result would be a consequence of Dow Jones’s business decision to continue operating in Australia—a decision that weighed the financial benefits of doing business in Australia against the costs of not doing business there, including the cost of not publishing pieces globally that might run afoul of Australian (p.158) libel law. Since Barron’s chose to continue to do business in Australia, its consumers in the United States and Japan cannot legitimately expect to receive news from Barron’s that runs afoul of Australian law. If they do not like the reduced content that results from Dow Jones’s decision to remain in Australia, they can get the information from scores of other news sources that do not do business in Australia and thus have no fear of Australian libel law.26

The ultimate problem with criticisms of Gutnick is that they reject any legal outcome other than the American approach. The critics assume that wherever the Internet goes, it brings a single global cyberlaw with it, like a tortoise carrying its shell. The irony, of course, is that the tortoise shell is not a consensus global law, but rather the parochial U.S. First Amendment. Australia is a democracy that has a different conception of free speech, and tougher libel laws, than the United States. The outcome of the Gutnick case suggests that the “Unholy Gains” article was in fact a pack of lies that harmed Gutnick’s reputation. Australians need not forego redressing this harm to one of its citizens in Australia out of deference to the U.S. Constitution.

Multiple Laws

“It’s a bad thing, not a horrible thing,” said instapundit Glenn Reynolds of the Gutnick decision. “What moves you to a horrible thing is that because the Australian high court has done this, it will be acceptable for countries with systems of law far less congenial to free speech to do the same thing.”27 This is the result predicted by cyberscholars David Johnson and David Post, who argued in 1997 that if a territorial government could apply its laws to a Net communication, then all “Web-based activity . . . must be subject simultaneously to the laws of all territorial sovereigns.”28

Being subject to a patchwork of conflicting laws seems like a bad, unworkable idea. The idea seems to get worse when we contemplate its effect on the decision to publish. The Sydney Morning Herald warned that after the Gutnick decision, “publishers will be tempted to produce material that is innocuous enough not to fall foul of the most draconian legal regimes.”29 In other words, publishers will be chilled by the prospect of having to comply with dozens of different laws and, racing (p.159) to the bottom, will conform their content to the laws of the most restrictive nation. Glenn Reynolds warned of a “lowest common denominator approach in which Internet publishers strive not to be offensive according to anyone’s standards, which is likely to mean not publishing at all, or publishing only inoffensive pap.”30

A similar chorus of sky-is-falling rhetoric greets every judicial decision that applies local law to a Net transaction with an offshore source. And yet wsj.com and millions of other web content providers, both firms and individuals, continue to publish news and opinion online, and not only the “inoffensive pap” predicted by Reynolds.

To see why the specter of multiple laws is exaggerated, recall the main lesson from chapter 5: with few exceptions, governments can use their coercive powers only within their borders and can control offshore Internet communications only by controlling local intermediaries, local assets, and local persons. Australia can effectively coerce Dow Jones because Dow Jones is a multinational company with employees, facilities, contracts, and bank accounts in Australia. But the vast majority of Internet users—students, e-consumers, porn purveyors, chat room participants, web-page operators, bloggers, and over 99 percent of other Net users—have no connection to Australia or to any other country other than the one in which they live. Far from being subject to multiple laws, these persons are immune from every law but their own.

The implication of the Gutnick decision, then, is that small Internet content providers need not worry about complying with the laws of every nation, but large firms with a presence in many nations—content providers like CNN, Dow Jones, and The Economist; systems operations like Yahoo, Google, eBay, and AOL; and financial intermediaries like MasterCard, PayPal, and Citibank—must comply with local laws in the places where they do business. Australia can go after these large multinationals in Australia when the multinationals assist in violations of Australian law. But it can do nothing directly to control Internet users outside Australia who have no presence there and must instead focus on Internet intermediaries with a local presence (as we discussed in chapter 5).

This still leaves big Internet multinationals like Dow Jones to face a jumble of overlapping and contradictory laws. But there is nothing new here. McDonalds complies with different health regulations and tax laws everywhere in the world it does business. Microsoft abides by (p.160) varying consumer protection laws everywhere it sells its software. Honda builds cars to meet local emissions standards in different nations. The Red Cross must learn about and follow charitable registration requirements that differ among nations. And despite the hysteria over Gutnick, the Wall Street Journal employs lawyers to monitor and comply with the different libel laws in the various countries where it publishes. In each of these cases, multinational firms incur significant costs to keep abreast of laws in different nations and to take steps to comply with these laws. These are simply the costs of doing international business.

Why should the Internet be different? The conventional answer is that Internet multinationals are different from real-space multinationals because the Internet’s architecture precludes them from knowing where in the world their content goes, making it impossible to comply with all local laws or to keep prohibited content out of certain places. But the claim that companies like Dow Jones cannot reduce or eliminate risk on a geographical basis in particular states is false.

As noted above, Dow Jones can leave Australia altogether, eliminating its presence and assets there and with them any fear of Australian libel law. Having decided to stay, it could monitor or control the geographical flow of its news. Dow Jones knew it had approximately seventeen hundred Australian users, and it knew that Gutnick lived in Australia. It could have simply denied access to the Gutnick story to these seventeen hundred users. Or it could have employed one of the various geo-ID and blocking technologies that are increasingly accurate and inexpensive, and that e-firms around the world are beginning to use to avoid or manage legal risk in distant jurisdictions.31 As we learned in chapter 4, these technologies are not perfect. But no border control technology is, and it need not be to be effective. Moreover, neither Gutnick nor Yahoo nor any other decision has placed an absolute rule of exclusion on Internet companies. Rather, firms like Dow Jones are only responsible for content that they could, through best efforts, keep out of places where it is illegal. It is true that these measures are costly But compliance with the law has never been free, and these costs are no different from other legal compliance costs in transnational commerce.

In the late 1990s, the Internet appeared to be a corporation’s dream: a medium that facilitated unlimited and inexpensive access to consumers (p.161) without any regulatory restrictions. Yahoo and Gutnick mark the beginning of the end of that dream. When corporate activity causes cross-border harm, nations can, and will, assert their regulatory authority. The threat of multiple regulatory exposure will not, as many have histrionically claimed, destroy the Internet. Firms will have to filter content geographically to comply with local law for only a small fraction of their communications. This will impose costs on multinational Internet firms, which will have to adjust to this cost of business just as real-space multinationals do. In light of the Internet’s many efficiencies, this cost will be trivial in the long run.

The lesson of this chapter is that when communications on the Internet collide with sensitive local public policies like gambling, pornography, consumer protection, libel, and the like, there are strong reasons to prefer a decentralized approach. In these contexts, there is no legitimate basis for giving any single law a kind of global constitutional status. It does not follow from what we have said, however, that there is no place for global Internet rules. To the contrary, many aspects of the Internet need to be regulated on a global scale. As the next chapter shows, however, this is sometimes easier said than done, and even when global rules prevail, territorial governments, and especially powerful ones, have devised them to serve their interests.