National Governments and Control of the Internet

A Digital Challenge

Giampiero Giacomello

• Publisher: Taylor & Francis
• Year: 2004
• eBook ISBN: 9781134323920
• Edition: 1
• Topic: Politics & International Relations
• Subtopic: Politics
• Index: Politics & International Relations

In recent years, Internet control has become one of the major indicators to assess the balance between freedom and security in democracies. This book explores and compares why, and to what extent, national governments decide to control the Internet and how this impacts on crucial socio-economic activities and fundamental civil rights. The author provides detailed studies on the US, Germany, Italy and further case studies on Brazil, Canada, India, the Netherlands, South Africa and Switzerland, to address topics such national security, freedom of expression and privacy.

1 Introduction

Everyone has the right to freedom of opinion and expression.

(Universal Declaration of Human Rights, Article 19, 1948)

The Internet is for everyone, but it won’t be if governments restrict access to it.

(Vint Cerf, co-founder, Internet Society, and co-inventor of TCP/IP, April 7, 1999)

“On the Net, nobody knows you’re a dog”

In 1993, a popular cartoon in the The New Yorker pictured a dog in front of a computer, describing the Internet (or, more familiarly, “the Net”) to a fellow dog with this wry but appropriate punch-line.¹ It was an early attempt to lighten up a stodgy, though brand-new, piece of technology: the “network of computer networks.” The joke summarized well one of the features of the Net that many users already treasured at that early date: on the Internet, one’s identity could be concealed, or altered, or falsified at will. This important feature, however, like many aspects of the Internet, had not been an explicit part of the network’s original concept.

The birth and expansion of the Internet have been riddled with contradictions and surprises.² Originally a communication network designed to connect the many different machines of the US Department of Defense (DOD), the Internet had its roots “in the darkness of the Cold War” (Rosenzweig, 1998: 1533). DOD planners and engineers thought it might further be useful in a nuclear war. Otherwise “the Defense Department would never have committed funds to projects like ARPANET without the beliefs that they would ultimately serve specific military objectives and larger Cold War goals” (Rosenzweig, 1998: 1533). Indeed, it was the Soviet success with the Sputnik satellite that prompted the DOD Advanced Research Projects Agency (ARPA) to start pouring money into what would become, in the mid-1990s, the “accidental superhighway” as The Economist correctly labeled it (Anderson, 1995). Concepts and ideas such as packet-switching, distributed networks, and TCP/IP (the “Internet” protocol) slowly but steadily built on each other thanks to the dedication and ingenuity of several engineers and technicians.³ Ironically, “anarchy” (that is, the lack of effective control by a superior authority, despite the origins in the US government ARPANET) is the one feature that the early Internet shared with the “international system.”

Despite those murky origins, almost by accident the Internet then developed into a “network of networks” to disseminate knowledge among universities. There was no need to build the information highways because the Internet was already there. It now carries news, data, encrypted financial transactions, and impressive sex-related material. It has, ultimately (again by coincidence), revolutionized the way private companies do business. As it expanded in both size and scope, the computer network was seen simultaneously as a threat to, a tool for, or the object of statutory control (Mulgan, 1991: 5). National governments grew “schizophrenic” about it, because they found its “libertarian culture and contempt for national borders subversive and frankly terrifying” (The Economist, 1998: 18). The growing importance of economic liberalization has also led to substantial changes in the way that state control is exercised. By the late 1980s and early 1990s, Mulgan (1991: 137) noted, “deregulations coincided with an ever more active role for governments and state agencies in creating what they believed to be the best climate for the communications economy.” What happened was not the end of state control, but rather a change in its forms.

These remarks trigger some puzzling questions. Why do national governments want to control what is, apparently, so difficult to control?⁴ Why bother to put resources into monitoring a network that may route around control points the way it was designed to do around interruptions after a nuclear exchange? And if governments press on with their control programs, what levels of control would be reasonable to attain, since countries and their governments tend to be socially, culturally, and economically very different from one another? Answering these dilemmas is the central goal of this book.

Control, with its many variants, is a much studied theme in the social sciences. The form of control examined in this work is statutory control, that is, control exercised by the “state” and, in its place, by its material counterpart, the government. In the classification of international relations, this means that this work focuses on the unit level of analysis. Control often goes hand in hand with “power,” although the two concepts do not coincide (Lukes, 1974; Horowitz, 1990; McMahon, 2002). Indeed, power is often identified with a property, an ability or mechanical energy, that can be directed or stored. But control carries with it “the sense of information environment” (Mulgan, 1991: 52). This is the reason why so many authors have analyzed information (and communications) and control, whether social or political, together (Pool, 1983, 1990; Beniger, 1986; Buchner, 1988; Horowitz, 1990; Chomsky, 1997; Pound, 1997; Shapiro, 1999; Wright, 2000; McMahon, 2002; Mueller, 2002; Zittrain, 2004). Information, communications (conveyed information), and control are all functions in the same equation (Mowlana, 1997).⁵ Given these conditions, one would hardly fail in summoning the broad class of “media” to explain Internet control.

Radio, television, and one-to-many broadcasting in general in twentieth-century Europe were born under state structures with centralized control. This centralized control was what made (and makes, according to Chomsky, 1997) propaganda possible. The identification of the Internet as the newest communication medium would therefore explain why it is controlled. After all, “governments, fearful of a loss of control over sovereignty and culture, would continue to resist opening new communication channels” (Noam, 1990: vii), despite the century-old, self-defeating experience of that restrictiveness.⁶ The answer “the Internet is controlled because it is a new communication medium” is an incomplete and very partial answer.

A most striking feature of the Internet is its complexity. Engineers and computer scientists have long lost track of how the “whole” works. They know and understand the single parts, but nobody can comprehend the whole structure. This complexity is the outcome of three crucial sets of factors. First, there is dual nature of the Internet that makes it, at the same time, an infrastructure (that is, the actual computer network) and a communication medium. The Internet “as infrastructure” allows the functioning of other distribution networks such as water, gas, or energy.⁷ In the case of telecommunications (telecom) the Internet is so embedded into the overall structure that it is simply impossible to make any distinction at all. Despite using the same wires and software, the Internet communication medium is a substantially different phenomenon from the Internet infrastructure. Unlike radio and television, the former is a many-to-many broadcasting system in which users are, at the same time, audience and producers of information, data, knowledge, outright lies, and propaganda.⁸

The crucial point here is that information and data about the utilities travel on the Internet as packets that are exactly the same as the packets that carry the bytes of an e-mail to a friend, a newspaper article, my travel plans, or the request for a Web server to open the weather Web page. The bytes of financial transactions are usually encrypted, thus the information cannot be easily read. But even those packets are not any different for the router computer that is sending them to their destination. Hence, the disruption of that router would also affect the encrypted packets of a multimillion-dollar financial transaction. Information and data on the utilities are hardly ever encrypted. All the packets travel together along the same wires and routes and are indistinguishable, but the information they carry does not have equal value. The loss of an e-mail to a friend, the reservation for a car rental, or a bank transfer have different consequences and meaning for their senders and receivers.

Second, there is the paradoxical multiple jurisdiction of the Internet. What authorities, domestic as well as international, are institutionally competent for regulating, monitoring, and maintaining the “two Internets?” The answer to this question is simple but with complicated implications: multiple authorities that have different goals and different tools at their disposal. Therefore, if there is a threat, domestically originated, to the Internet as infrastructure, then law enforcement officers should have jurisdiction. If the origin of the threat is not domestic but international, then a mix of law enforcement agencies and the national security agencies could be the solution. A further distinction is whether the cause of an international menace is a state or a non-state actor, which would influence the proportion of military/police personnel necessary. On the other hand, if the threat is not to the Internet as infrastructure per se, but it uses the Internet as communication medium, then it becomes a crime. In these circumstances, regardless of whether the source is domestic or international, the proper answer would be domestic and, if necessary, international law enforcement agencies. A further distinction is that the gravity of the crime depends on the domestic criminal legislation: an offense that would be dealt with through lawyers in a democracy might imply capital punishment in an autocratic country. The series of distinctions could continue.

The third set of factors is the multiplicity of stakeholders (Keohane and Nye, 2000; Herrera, 2002; Giacomello, 2003). There are three main sets of stakeholders on the Internet: national governments (and their agencies), the information and communication technology (ICT) business sector, and non-governmental organizations (NGOs). The NGO set may include a vast (and loose) array of individual users’ groups, civil liberties advocates, and consumers’ organizations, which operate to advance their specific interests. NGOs and ICT firms may or may not be transnational. NGOs are non-state actors that may pursue perfectly legal ends such as protecting human rights, the environment, or privacy. A minority of NGOs, however, namely terrorist groups or organized crime, are involved in assassination, destruction, or illegal trafficking (Josselin and Wallace, 2001). These NGOs use the Internet mostly as a communication medium, exactly as the “legal” NGOs do. At different stages, the diverse, individual actors that constitute the three stakeholders have formed alliances to increase or to resist control on the Internet. In this respect, September 11, 2001, was the watershed. Before that date, especially in the United States, civil liberties NGOs and private industry joined forces to resist the US government’s attempts at increasing control on the Internet. As Chapter 2 shows, that “odd alliance” was successful. After September 2001, however, the US government was apparently able to convince the ICT industry to privilege security, even if at the expense of civil liberties.

So, why do governments want to control the Internet? The more advanced a country is, the more dependent on computer networks (like the Internet) it will be. Governments want to control the Internet to protect their information infrastructure networks that are so crucial to the survival and well-being of their countries. As early as the 1970s, the new information technologies were thought to be likely to increase states’ vulnerability. A report to the Swedish government, the Tengelin Report (Tengelin, 1981), emphasized the main risks of a networked society (including dependence on foreign vendors and the threat of hackers’ raids). Once national governments realized the actual extension of the Internet and the potential reach of individual users, they started to consider increasing their control over the network. After all, controls over inflows and outflows of people, goods, and information have been vital for states to assert their authority, and thus their sovereignty.

These viewpoints require some further specifications. There are two distinctive faces to the problem of “control to protect”: (a) controlling the national information infrastructure (NII), of which the Internet is an integral part, and (b) controlling on-line content. In case (a), the NII should be controlled for protection in the same way as governments control and try to protect their sea-lanes or railways. In case (b) that control may easily turn into abuse, censorship, and violation of privacy. It is not accidental that some less developed countries such as China, which are less dependable on computer networks, tend to focus on controlling content first and securing the NII second. For most countries, (a) and (b) cases are placed on a security continuum that goes from national security to law enforcement and public order. On that continuum, advanced countries (which happen to be, in most cases, democracies) have a propensity to see (a) and (b) within the domain of law enforcement and public order rather than of national security. For these democratic countries, cybercrime, perhaps cyberterrorism, is what they perceive as Internet threats more than hostile governments trying to wage cyberwar on their NII. Law enforcement agencies should primarily deal with those threats. The majority of case studies presented in this volume support these findings.

Protecting the NII is hence one of the principal reasons why national governments want to control the Internet, and it is also an integral part of the national security policy of most countries (national security can be seen as a counterpart of social control, according to McMahon, 2002). But states are now substantially different institutions from what they were when international trade (and the business sector) was considered to be subordinate to national security (Strange, 1988). Today, the mission of national security, particularly in the area of ICT, cannot be fulfilled without close cooperation between the public and the private sector (Bendrath, 2001; Rathmell, 2001; Bruno, 2002; Dunn and Wigert, 2004a). The largest portion of the NII is owned by private enterprises and the necessary software and communication tools are also produced by companies. This growing (and indispensable) presence of the private sector in the most exclusive government function, namely national security policy, represents an epochal change. One may object that the defense industry has always been crucial to national security or that mercenaries, as a form of private enterprise, have been offering their services for centuries.⁹ Most importantly, this new role for ICT firms in the shaping of national security policy implies that the private sector would yield an influence on national governments never before seen in history.

It is common that private interests of firms collide with the requirements of national security. In the current circumstances, however, national governments cannot simply overrule whatever opposition the ICT sector may have with reference to protecting the NII. Ironically, countries that have experience with telecommunications and utilities (traditional government monopolies) should be better off to some extent. If push comes to shove, these governments would have some expertise on how to take most of the NII under their direct jurisdiction and control. For example, small, neutral countries such as Sweden or Switzerland adopted a “total defense” model during the Cold War that required a very close partnership between the public and private sector. In Israel, this integration is even further advanced. Although the United States has always cherished pragmatic, business-like approaches to problem-solving, close cooperation between the public and private sector is not the rule. Some large industries, such as automobiles or steel, have undoubtedly benefited from strong government support, but the American business culture has traditionally been wary of government intervention.

National governments of advanced countries then have a new “partner” in the decision-making process of national security policy. But if the sancta sanctorum of national governments (the national security policy) is necessarily open to non-governmental actors like software and telecom companies, it also becomes more accessible to other non-governmental actors. In the 1990s, Internet users and consumers, independent software developers, and civil liberties advocates created a network of NGOs. These organizations are mostly concerned with content control and free access to the Internet, but with computer networks it is often quite difficult to keep content separated from the infrastructure. As the case of cryptowars¹⁰ presented in Chapter 2 shows, this state of affairs may have considerable implications for the policy-making of national security.

National governments want to control the Internet because they need to protect their NII and they need the active contribution of the private sector. The difficult distinction between content and infrastructure makes it somehow easier for autocracies to control the Internet. Autocracies all over the world have a long-standing praxis of monitoring content on radio and television networks. The literature on how autocratic countries control the Internet no less than other media is now abundant (Saich, 2001; Chase and Mulvenon, 2002; Kalathil and Boas, 2003; Reporters Sans Frontières, 2003; Kurlantzick, 2004).¹¹ Indeed, Kurlantzick (2004: 21) notes how easy “authoritarian regimes have controlled and, in some cases, subverted it.” If, for all autocracies, controlling Internet content is a sort of “second nature” (Gibbs, 2004), China has been (thanks to Western technology) the most efficient controller (Saich, 2001; Chase and Mulvenon, 2002; Kalathil and Boas, 2003; Kurlantzick, 2004).¹²

Autocratic governments claim that, in this era of terrorism, they need to control the Internet (with no distinction between content and infrastructure) to protect their NII and to fight the (greatly exaggerated) “cyberterrorists.” The national security communities have al...

Table of contents

• Cover Page
• Title Page
• Copyright Page
• Dedication
• Illustrations
• Foreword
• Acknowledgments
• Abbreviations
• 1. Introduction
• 2. Sometimes Security Just Does Not Prevail: The Case of the “Cryptowars”
• 3. What Democracies Do?: An Overview
• 4. The United States: The Sole Information Superpower
• 5. Das Netz Über Alles: Germany On-Line
• 6. “Internet Per Tutti!”: Italy’s Elusive Information Society
• 7. Conclusions: Digital Winners, Virtual Losers
• Bibliography