con256135
Continuum: Journal of Media & Cultural Studies Vol. 9, No. 3, September 2004, pp. 403±410
Working with the Resources at Hand: Constraints on Internet Institutional Design Hans Klein
Legitimacy has emerged as a major issue in debates over global governance institutions. The growth of border-crossing systems in business, ®nance, trade, communications and other areas has led to the creation of supra-national governance institutions to de®ne common rules and to perform system coordination. While most decisions made by these governance institutions are administrative in nature, they often have a public policy component as well, affecting the well-being of people, the environment, and private ®rms. As awareness has grown of these institutions' role in policy making, questions have arisen about who are the policy makers. Who is making global public policy? From whence do they derive their authority? Do global governance institutions possess legitimacy commensurate with their decision-making powers? In what follows I examine this question of legitimacy as it plays out in processes of institutional design. To the extent that a legitimacy de®cit is recognized and attempts are made to correct it, it becomes important to understand factors that in¯uence institutional design for legitimacy. My thesis is that designers of institutions do not operate in a vacuum but must construct institutions from the resources available to them. When designing for legitimacy, they often must choose from the existing set of legitimacy-giving bodies. With the set of building blocks thus given, design choices are `chunky', allowing for fewer bases of legitimacy to be realized in practice than might be possible in theory. As a result, the outcomes of institutional design may be different from might be predicted. By identifying such constraints on institutional design, students of global governance can better understand outcomes of design processes, while policy makers can better formulate strategies for realizing their goals. Hans Klein is Associate Professor of Public Policy at the Georgia Institute of Technology. Klein holds a BSE in computer science from Princeton University and an MS and PhD in political science from the Massachusetts Institute of Technology. Correspondence to: Hans Klein, Assistant Professor, School of Public Policy, MC:0345, Georgia Institute of Technology, Atlanta, GA 30332-0345, USA. Tel.: +1 404 894-2258; Fax: +1 404 894-0535; Email:
[email protected] ISSN 1030±4312(print)/ISSN 1469±3666 (online) ã 2004 Taylor & Francis Ltd DOI: 10.1080/1030431042000256135
404
H. Klein
I examine these issues in the context of the global governance of the Internet. In 1998 the US government transferred signi®cant responsibility for Internet governance to a newly created private corporation, the Internet Corporation for Assigned Names and Numbers (ICANN). ICANN was to coordinate technical resources, making global policies for the allocation of unique domain names and numeric addresses and for the operation of Internet addressing services. As has occurred with other global governance institutions, the policy dimension of ICANN's administrative activities quickly became apparent. ICANN de®ned and enforced new forms of intellectual property in Internet domain names, it regulated market entry and set prices in relevant industries, and it possessed (but did not immediately exercise) the awesome power to disconnect entire country domains from the Internet. With widespread recognition of its policy role came growing calls for a restructuring of ICANN to endow it with a degree of legitimacy appropriate to its powers. In the ®ve years following its creation in 1998, ICANN underwent two episodes of institutional design for legitimacy, and 2003 saw the initiation of yet another round of design. In each episode the constitution of its high-level authority structure was subject to review and in some cases it was explicitly modi®ed. I primarily examine the third round of design, which began at the UN World Summit on the Information Society (WSIS), and which is not yet complete at the time of this writing. This most recent round clearly illustrates the historically conditioned choices facing designers as they seek to correct the legitimacy de®cit of a global governance institution. Legitimacy Without claiming to offer a philosophically or legally rigorous de®nition of legitimacy, I offer here some summary concepts about this important characteristic of global governance institutions. Legitimacy can be understood in relation to two additional concepts: power and authority. Power is the ability to realize one's intentions, translating an envisioned future state of the world into an actual state of the world. Relative to public policy, it usually refers to the exercise of power over societal entities, be they people, organizations, or nation-states. Authority is power endowed with legitimacy (Friedman, 1990). With the addition of legitimacy, power ceases to be the brute exercise of will and becomes the rightful exercise of will. In an earlier work (Klein, 2002), I analysed ICANN's powers of regulatory promulgation and enforcement. The discussion here addresses the means by which that power is rendered legitimate, thereby creating authority. Broadly speaking, legitimacy can be de®ned in philosophical and psychological terms. Philosophically, legitimacy refers to the grounding of power upon some foundational principle. Such a de®nition is less helpful than it might seem, since it simply shifts the question about legitimacy to a question of foundational principle, but the set of established principles is limited and therefore sets some limits to the discussion. Two types of principles particularly relevant to global governance are democratic and technical. The foundational principle of democratic legitimacy is that power is
Continuum: Journal of Media & Cultural Studies
405
exercised only with the consent of the governed. Governance rests on a social contract, in which citizens affected by a governing entity must consent to being governed by that entity. This principle is most frequently implemented through voting and elections, whereby citizens either decide directly on public policies or select representatives who will make those policy decisions. In political-administrative processes of public policy this is often conceived of in terms of interest representation: groups with an interest in a policy ought to be represented in the policy process. The technical foundation for legitimacy derives from the idea of expertise. Experts with relevant training and experience in specialized domains appropriately make certain classes of policies. In its early years, ICANN often sought to establish its legitimacy on the basis of its expertise, claiming that the technical expertise of its staff and board of directors legitimated its exercise of power. As that claim was challenged, calls increased to ground ICANN's legitimacy in democratic principles. A related source of principled legitimacy is the setting of limits on the exercise of power. Not only must power be grounded in principle, it must be constrained in its reach. Majority rule may devolve into a tyranny of the majority, and to protect against that absolute limits can be set on what powers can be exercised. In the case of democratically based legitimacy, the rights of individuals may trump the will of the majority. In the case of expertise-based legitimacy, the power of experts is appropriately limited to domains where expertise is required. Psychologically, legitimacy can be grounded in recognition. If the groups affected by governance institution's policies recognize its right to the exercise of power, then the institution is considered by them to be legitimate. Such `recognized right' (Krasner, 1999) manifests itself in willing acceptance of policies. Recognition may be the result of grounding power in principle, in the case of democratic or expertisebased legitimacy, but recognition may also derive from other sources, such as history or personal charisma. If policy `has always been made this way'Ði.e. the governance institution has an established history of exercising powerÐthen affected parties may accept the continued exercise of such power. Likewise, if policy is made by a respected and trusted individual, then affected parties may again recognize a right to govern. For example, in the pre-ICANN era much Internet governance was performed by Dr Jon Postel, the leading administrator of the domain name system, whose long history of policy making and solid reputation for trustworthiness legitimated his policy decisions. Designers of institutions who seek to endow a governance institution with legitimacy may draw upon all these different kinds of legitimacy. Democratic legitimacy can be realized by locating decision-making power in representative bodies and by de®ning limits on the scope of policy. Expert-based legitimacy can be realized by giving power for technical decisions to organizational units populated by quali®ed experts. Historical legitimacy can be realized by grafting policy-making powers onto an existing governance institution (assuming such an institution exists). Personal legitimacy may be realized by involving a recognized community leader to make the decisions affecting that community.
406
H. Klein
Early History of ICANN The history of global Internet coordination manifests an evolution in the legitimacy of governance. Before the creation of ICANN in 1998, coordination was performed by a small group of technical experts joined together in the Internet Society (ISOC) and headed by Dr Jon Postel. They de®ned new domain names, delegated authority to administrators to perform Internet addressing operations, and oversaw the operation of root addressing servers. Both Postel's personal reputation and the technical expertise of the community in which he worked legitimated his decisions. Furthermore, over time his role became endowed with historical legitimacy. Postel had performed the function since being a computer science graduate student in the 1970s, and in the course of two decades the Internet user community came to recognize Postel's right to make such decisions. Finally, for much of his tenure his exercise of power was limited by the narrow diffusion of the Internet. The number of users affected by his decisions was relatively small, since the Internet did not experience explosive growth until the 1990s, while the ®nancial effects of his decisions were also limited, since commercial use of the Internet was allowed only after 1994. The result of all this was that Postel's exercise of governance functions was widely accepted by affected parties around the world. With the creation of ICANN in 1998 and the near-simultaneous death of Postel, Internet governance was put on a new basis of legitimacy. In this ®rst round of institutional design, the authority previously vested in an individual came to be vested in an institution. Policy-making power was located in ICANN, speci®cally in its 19member board of directors. This legitimacy-giving body combined both democratic and expertise-based legitimacy. In ICANN's original design, nine directors were to be selected by technical communities with expertise in Internet identi®ers. Nine more directors were to be selected by Internet users through global elections. (The remaining position was reserved for ICANN's top administrator.) In terms of expertise, this arrangement combined expertise in technology and expertise in user issues. In terms of interest representation, this arrangement balanced the interests of users and those of the Internet supply industry. As designed, the board of directors would have ®nal authority. Unlike other global governance institutions, ICANN was not to derive its legitimacy from national governments. Indeed, governments were barred from having representatives on the ICANN board and were instead relegated to an advisory body, the Governmental Advisory Committee (GAC). In what was presented as a temporary arrangement, the US government would exercise top authority over the Internet's core technical resources, but only until ICANN's design was fully implemented. However, ICANN's board was never fully implemented. The full complement of user directors was never seated, and four years later ICANN's bylaws were modi®ed to eliminate user representation altogether. The ensuing outcry gave evidence of ICANN's loss of legitimacy. Representatives of the user community, industry, and government questioned ICANN's right to make policy, denying ICANN the
Continuum: Journal of Media & Cultural Studies
407
recognized right to govern. ICANN's policy making was decried as the exercise of power rather than as the expression of consensus in the Internet community. ICANN's revised institutional design of 2002 brought to the fore the role of the US government. Despite its original commitment to cede its authority to ICANN, shortly after ICANN's founding the US Department of Commerce published a contradictory notice that it had `no plans to transfer to any entity its policy authority to direct the authoritative root server' (NTIA, 1999). Decisions by ICANN would have to be validated by the United States before being implemented; ultimate power lay with the United States and would remain there. As the perceived legitimacy of ICANN's board eroded, the US government was increasingly seen as legitimating ICANN's decisions. This constituted an important shift. Whereas ICANN's original design had embodied a novel form of legitimacy, based in large part on the direct election of user representatives from around the world, the emergence of the United States as the supreme authority endowed ICANN with a more traditional source of legitimacy: the nation-state. The ICANN experiment in private, non-governmental authority was stillborn, and traditional political authority reasserted itself. For the purposes of this analysis, I consider the emergence of the United States as the supreme authority as a second episode of institutional design. The original ICANN design gave way to one in which a national government legitimated the exercise of power. US authority suffered from its own weaknesses, however. ICANN's powers were global in scope, but US political institutions were those of a single nation-state, representing the will of a subset of the world's people. ICANN lacked adequate democratic legitimacy. US control of the Internet extended that country's policymaking powers well beyond its borders, arguably constituting an illegitimate exercise of national power at the global level. Internet governance remained without a solid basis in legitimacy. The 2003 WSIS marked the beginning of a third round of institutional design for legitimacy. Concerned about US control of a global system, national governments from around the world endorsed a plan in which the United Nations would make recommendations about Internet governance. Although couched in diplomatic language, the intent was clear: the powers of ICANN would be reviewed with an idea towards rendering them more legitimate. Other national governments wanted a new institutional design that would give them representation in the decision-making process. It is this third round of institutional design that is of particular interest here. It illustrates the historically constrained nature of institutional design for legitimacy. Historical Constraints on Design: the Choice Set In the months immediately after the WSIS, policy makers from around the world gathered in conferences and workshops in Geneva and New York to consider a variety of governance issues. The process of institutional design had not been formalized, but it would be centred on a UN committee appointed by Secretary General Ko® Annan
408
H. Klein
and would draw on input from a myriad of interested parties. While formal discussions addressed the detailed issues like unsolicited e-mail, electronic signatures, and network security, informal discussions at these early meetings focused on questions of legitimacy. The overriding but often unspoken question was how to render ICANN legitimate. The original ICANN was unusual in that it had been designed on a `blank sheet' and did not build on existing governance institutions (Mueller, 1999). By 2004, however, history had caught up with the policy process, and ICANN faced integration into the global system of political authority. ICANN existed and would almost certainly not be completely abandoned, but its crisis would likely be addressed by subsuming it under an existing legitimacy-endowed body. As such, much discussion focused on the available set of existing legitimacy-giving bodies. Three main choices existed: the US government, ICANN's Governmental Advisory Committee (GAC), and the UN International Telecommunications Union (ITU). All of these were alike in that they derived their legitimacy from national political institutions. ICANN's original design, in which legitimacy was derived independently, was dead. Yet each of the three varied in how national governments, industry, and civil society were represented. As such, each one offered a different foundation of legitimacy for Internet governance. The ®rst option was the status quo: the United States would continue to exercise supreme authority. For other governments, this option remained problematic, since it would extend one country's authority over an infrastructure critical to each of their needs. Furthermore, as manifested by US policy in the Middle East at that time, US policy had grown decidedly more unilateralist in this period, and this served to heighten concerns about sole US authority over Internet governance. For industrial interests, many of them centred on the most developed countries of North America, Europe, and Asia, continued US authority offered the not unwelcome prospect of business-friendly oversight. However, that could be compromised by US national security concerns, which had in the past justi®ed policies contrary to free trade (e.g. technology export controls). For civil society groups, many of them based in the United States, continued US control would locate governance in a familiar framework of national law, but, with their emphasis on principle, these groups would be unlikely to support a design that concentrated global authority in one national government. Overall, this option seemed unlikely to attract broad support in a UN-based design process. The second option was to locate authority in ICANN's Governmental Advisory Committee (GAC), changing the GAC from an advisory committee to an authoritative body. The GAC membership was open to all national governments as well as multinational governmental organizations (notably the European Union) and treaty organizations (e.g. the ITU). Locating authority in the GAC would ground Internet governance in the world's national governments, providing a broader foundation for global policy. Despite its inclusive membership rules, some governments exercised a greater role in the GAC than others. The most industrialized countries held the greatest in¯uence.
Continuum: Journal of Media & Cultural Studies
409
The position of the GAC secretariat gave some indication of which members were most in¯uential: initially provided by Australia, by 2004 the secretariat function was hosted by the European Commission in Brussels. Developing countries were represented in the GAC, but their limited resources meant that ICANN participation was expensive in terms of ®nancial costs and in terms of their limited stores of expertise. Were the GAC to become the top authority for Internet governance, in¯uence would expand from the United States to other developed countries, but it would not likely be shared equally by all countries. Public comments by European of®cials suggested that this was an attractive option for them (McCarthy, 2004). The third option was the ITU. The characteristics of this UN agency seemed most favourable to developing countries. As a global coordinator for telecommunicationsÐ and one with a formal mission to support developmentÐthe ITU's representative structure gave comparatively greater voice to developing countries. Their governments had a long-established presence in the ITU and were able to participate in its deliberations. Given their limited resources, developing countries could economize by adding new responsibilities to a body in which they were already effectively represented rather than by trying to participate in another body such as the GAC. For industrial interests, the ITU was associated with the `old' world of telecommunications, based on regulated national monopolies rather than the more recent regime of competitive service providers. It was perceived as being hospitable to national telecommunications monopolies, particularly those in developing countries. The interests of large multinational ®rms would likely be better served by an authority like the GAC, because it could better represent the interests of the most developed countries. For civil society groups, the ITU represented an ambiguous option. On the one hand, it seemed to allow for a more equitable distribution of authority among governments from around the world. On the other hand, formal participation by nongovernmental organizations in the ITU required the payment of fees amounting to thousands of dollars, which could effectively exclude many civil society organizations. In summary, institutional design for legitimacy consisted largely of a choice among existing institutions. Each had its own system of representation, and each distributed in¯uence differently among national governments, ®rms, and civil society organizations. The ICANN model of 2003 concentrated power in one government, the GAC concentrated it in a few, and the ITU dispersed it among many. References Friedman, R. B. (1990) `On the concept of authority in political philosophy', in Authority, ed. J. Raz, NYU Press, New York, pp. 56±91. Klein, H. (2002) `ICANN and Internet governance: leveraging technical coordination to make global public policy', The Information Society, vol. 18, no. 3, pp. 193±207. Krasner, S. (1999) Sovereignty: Organized Hypocrisy, Princeton University Press, Princeton, NJ. McCarthy, K. (2004) `EC tells Europe and ICANN to make peace', The Register, 28 Apr., Available at: http://www.theregister.co.uk/2004/04/28/ec_icann_warning_shot/ (accessed 5 May 2004).
410
H. Klein
Mueller, M. (1999) `ICANN and Internet governance: sorting through the debris of ``selfregulation''', Info, vol. 1, no. 6, pp. 497±520. NTIA (National Telecommunications and Information Administration) (1999) Domain Name Agreements between the US Department of Commerce, Network Solutions, Inc. and the Internet Corporation for Assigned Names and Numbers (ICANN), Fact Sheet', 28 Sep., Available at: http://www.ntia.doc.gov/ntiahome/domainname/agreements/summary-factsheet.htm (accessed 5 May 2004).
Author Query Sheet QUERY NO.
QUERY DETAILS
NO QUERIES
