August 31, 2001
As the Internet becomes increasingly important to people around the globe, a critical question has been raised: How can the voice of a growing community of stakeholders be appropriately included in new mechanisms for the coordination or "governance" of key Internet technical functions? This question is posed today in one such coordination body, the Internet Corporation for Assigned Names and Numbers (ICANN), which is tasked with managing aspects of the Internet's naming system and other critical technical functions.
In 2000, ICANN completed an unprecedented global election for five "At-Large" Directors of its governing Board. ICANN is now in the midst of a major debate over the future of broad participation in its activities. That debate will be closely watched by policy makers around the world and could have profound implications for both the legitimacy of ICANN and the future of public participation in a new breed of non-governmental, global coordination bodies.
This report seeks to address key questions: Why is there a need for a public voice within ICANN? What role does and should ICANN's At-Large Directors and Membership play in providing that voice? What lessons can be learned from the At-Large election with regard to the At-Large Membership and the structure of the ICANN Board? And what practical steps can ICANN take to establish an appropriate and effective public voice within its structure, and hence improve its legitimacy?
This report presents the findings of the NGO and Academic ICANN Study (NAIS) , a collaboration of ten research teams from around the world. The findings reported here began with an elaborate series of studies conducted within each of ICANN's five designated geographical regions. The methodology of these regional assessments varied based on local conditions, but included personal interviews with experts and leaders, intensive review of supporting documents and media coverage, surveys of key stakeholders, and the analysis of experienced researchers. Cross-regional comparisons were made, coupled with a conceptual investigation of ICANN's governance structure. Our final analysis and recommendations are the product of intense consultation, deliberation and collaboration since the formation of NAIS in early 2001.
The NAIS Interim Report (June 2001) presented initial observations and explored the need for a public voice in ICANN's activities. This Report refines those findings, and proposes a structure for ICANN that we believe will provide a longer-term answer to the question of ICANN's legitimacy. Our approach is rooted in the creation of an inclusive Membership, in order to facilitate informed participation in ICANN's activities, and with elections to allow for representation of broader interests on ICANN's Board. In addition, we present suggestions for structural reform, constitutional constraints, and increased accountability.
This fundamental finding is based on the following key observations:
Efforts have been made to confine the scope of ICANN's activity to technical coordination of certain naming and numbering and protocol functions. But many of ICANN's "technical" decisions - such as the creation of new top-level domains - are inextricably linked to "policy" decisions - such as intellectual property protection, competition policy, or the allocation of Internet resources. ICANN faces great pressure to address these policy issues and, by necessity, to make decisions on grounds other than technical merit.
Much here involves the future of ICANN. Its potential authority is largely unbounded. Policy authority over the root server system and other central Internet functions make it possible for ICANN to exert much broader control over Internet activity. To date, many of ICANN Directors and staff have expressed little interest in doing so and have publicly disavowed such a broad scope of power. But ICANN is likely to face increasing pressure from those seeking ways to control a range of behavior online. The broader this pressure, the more sweeping the potential impact, and the greater the need for public participation and representation.
Based on this conceptual model, ensuring inclusive public participation and substantial Board representation should be overriding priorities for ICANN. The At-Large Membership (ALM) remains the great - and as yet unrealized - opportunity for such needed public involvement. Broadly conceived, the At-Large membership could serve a diverse set of goals including‚
Developing an ALM is a necessary precondition to successful At-Large Director selection. Institutionalizing the other roles of the ALM in providing participation and accountability will have additional important benefits for establishing the global public legitimacy of ICANN.
From its inception, ICANN recognized that some form of representation for broad public interests was needed. The creation of nine At-Large Directors - nearly half the 19-member Board - to be selected "At-Large" was a placeholder for that broader representation. But it was only a placeholder, as the exact method for selecting directors was not established at ICANN's founding and has been a subject of great debate.
The October 2000 ICANN election for five At-Large Directors, one from each region, was unprecedented. Potential voters in the election broadly included anyone in the world with an e-mail address and postal address who completed the web registration. Over 153,000 people worldwide registered, and over 30,000 ultimately voted. Up to seven candidates were on the ballot in each region, selected by nomination or a petition-like member nomination process. Voting occurred online, through preferential balloting, and five new directors were elected to the Board.
The election has become a point of contention among those who question the need for public participation in ICANN at all and those for whom the At-Large participation was a basic premise of ICANN's existence. In many ways the 2000 election was perceived to be successful in selecting directors by a broad membership. However, the election also presents challenges for the future.
The 2000 election, though experienced regionally, was developed and implemented globally. All regional elections had core issues in common:
Region-by-region analyses are provided in the NAIS Final Report and summarized separately. They provide a rich picture of both the benefits and challenges‚particular to each region‚of a new global election for ICANN. Major cross-regional observations of note include:
ICANN serves a vital global public trust. It must therefore achieve a level of legitimacy appropriate to that trust, by structuring itself to take account of the public's interest in its activities. As described above, its existing structures have not done so to date.
Our observations and analysis of the 2000 elections indicate that the At-Large Directors and the At-Large Membership (ALM), mandated at the founding of ICANN, have so far been inadequately defined and institutionalized in ICANN's framework of operations. The 2000 elections highlight many of the challenges with global membership and elections. We believe that they also indicate possibilities for meaningful involvement and representation.
We believe that incorporation of the public interest into ICANN can be achieved through the creation of an inclusive membership, resting on two overarching principles: The public membership should be given structure, and the public membership should be given representation.
We also put forward additional proposals for structural reforms and accountability mechanisms that should be enacted in parallel with ICANN's membership structures.
ICANN will gain requisite legitimacy from the creation of a membership that allows for substantial participation and representation by interested members of the public. Since ICANN's activities have the potential to impact all Internet users and in fact the public in general, we favor an inclusive approach to membership. We believe that membership should be open to all who express interest by completing a relatively simple registration process online combined with postal return confirmation.
Online registration and postal return appears to strike a reasonable balance between practicality, security appropriate for a narrow membership organization, and inclusiveness on a global basis. Postal return, while imperfect, provides a measure of authentication and security. While the costs of postal return are non-trivial, these are likely one-time costs for many members. Many of the problems encountered in the postal return in the 2000 election could also be avoided with greater time for registration and better planning.
Problems with alternatives: NAIS considered many other criteria for membership. We found major alternatives far less attractive in achieving inclusiveness and participation in a practical fashion.
For example, limiting membership to holders of domain names was one alternative considered but which ultimately fails basic tests of fairness and practicality. Interested ICANN stakeholders include many Internet users who are not necessarily domain name holders, and holders are a highly imperfect match for those directly interested in ICANN. Domain name holders as a group are dominated by parties already well-represented in the SO structure. VeriSign has estimated that over 80% of current gTLD registrants are commercially-oriented organizations - creating a membership pool that under-includes individuals and is heavily skewed towards commercial groups and organizations. Substantial practical questions remain in determining whether those who own more than one domain name get more than one vote. And capture is still possible through the registration of many names.
Similarly, NAIS rejected the use of a fee for membership. Imposing a fee raises serious equity issues, even with substantial sliding scales for poorer regions. While likely to weed out less committed members, a fee still allows the possibility for over-representation of wealthy interests or countries. NAIS is cognizant of the need to pay for membership activities, but we believe that such costs can be borne by ICANN for at least the next several years, and question the extent to which an equitable fee would raise sufficient funds without jeopardizing inclusiveness and legitimacy at this time.
Legitimacy for ICANN depends on both the opportunity for and reality of meaningful participation. An At-Large Membership that lacks enabling structures is likely to lie dormant, gradually losing its coherence‚described by many as a major disappointment following the 2000 elections. We therefore propose a four-part At-Large Membership (ALM) structure to facilitate participation and representation of a productive and informed membership.
This structure is designed with the goal of limiting, to the extent possible, the negative effect that excessive intermediary structures can have on Members' ability to directly interact with ICANN's policy development structures. For that reason, the NAIS team does not propose any kind of "Policy Council" (like the Names Council for the DNSO) purporting to represent all Member interests. While there may be a need for such a structure in the future, it should be implemented only after consideration by the Membership.
In order to provide the public voice with meaningful representation, At-Large Directors should balance the policy authority of the Directors chosen by Supporting Organizations. Therefore, the number of At-Large Directors should equal the number of SO Directors, even if the size of the Board should change from its current nineteen members. This balance has also been part of the Board structure since ICANN's formation and many participants have cited it as part of ICANN's basic structural balance.
Because changes to the by-laws require a two-thirds majority, this balance on the board would permit only those changes that command at least some support from both the At Large and SO Directors. In our view, this would provide an appropriate level of "checks and balances" on the board, and would provide the At-Large with enough authority to block changes to the bylaws that might negatively impact the public voice. (While we recognize that Directors to date have seldom acted in monolithic voting blocks, we anticipate that future boards may find At-Large Directors sharing certain common perspectives inherent to their positions.) We believe that reducing the At-Large Directors to less than one-third of the Board would seriously undermine the legitimacy of ICANN.
Properly managed, direct elections can provide high levels of representation, diversity, and accountability. Because they eliminate an intermediate layer between members and the Board, they encourage a higher level of participation than other selection mechanisms. They also help to lessen the risk of capture when compared to indirect elections, since the electorate with real decision-making power is significantly larger in size.
The creation of an At-Large Membership structure and the direct election of At-Large Directors is a model with both benefits and problems. As was discussed in the context of the 2000 election, direct elections present challenges in voter registration and education, funding, security, and capture. These shortcomings, however, are outweighed by the system's benefits - and by the problems with alternative models - especially when based on a conception of ICANN that recognizes the organization's unique mission and character.
ICANN's legitimacy is of a fundamentally different character, and should be measured with different standards, than that of traditional governments. ICANN makes important policy decisions, but it is not a government. It does not raise armies or jail people, though it does have the potential to make far-reaching policies about the Internet. The point of elections for ICANN directors is not to confer legitimacy on representatives who will then exercise the full weight of governmental power. Rather, the point of ICANN elections is to choose directors who are representative of the public's perspective to act within the important, but presumably narrow, confines of ICANN's mandate.
The danger of capture-‚through overrepresentation of particular interests or domination by nations-‚is real but somewhat overstated. The 2000 elections provided little evidence of corporate domination of the elections, and improved elections rules and monitoring could mitigate this fear. Nationalism remains a concern as well, but we note that "capture" may arise as little more than a vigorously sponsored effort at voter registration and participation. The response, then, to fear of capture is not less elections, but more democracy and clearer rules. Alternatives to elections appear at least as susceptible to domination by particular interest groups or nations.
Alternative models of selection. A number of alternatives to the direct election model have been proposed by members of the ICANN community. These proposals seek to address some of the shortcomings of the direct election model, but all sacrifice critical elements of legitimacy, accountability, or security.
3.5a ICANN should adopt inclusive authentication mechanisms based on postal returns, while also pursuing alternative models. The At-Large Election's long-term viability will depend on the system's security, accuracy, and resistance to fraud. These questions all lead eventually back to the problem of authenticating users in an election with such a broad electorate. The task before ICANN is to provide reasonable protection that each person voting is a real person, and that no person can vote more than once.
Postal return system. Used in 2000, the postal return system was successful in some ways but problematic in others. Despite its problems, we believe electronic registration with postal return strikes the best balance among security, inclusiveness, and cost. With more lead-time and greater experience, we expect lower failure rates and greater opportunities for errors to be remedied. Reforms such as longer registration periods, better addressing of PIN letters, and opportunities to re-send lost or misplaced PINs should be adopted. A long-term membership means that many members will use the system only initially. Clearer election rules, combined with improved auditing (perhaps on a statistical basis periodically), would provide greater levels of security. Publication of membership rolls‚an option to be pursued by the ALM‚could further reduce fraud and boost confidence. Finally, serious member education is needed to diminish confusion.
Authentication technology is advancing rapidly, though we believe at this time none of the major alternatives provide an acceptable tradeoff in deployment or accessibility. ICANN should experiment with these systems in the next few At-Large elections, perhaps offering them as an alternative to the postal return system.
3.5b At-Large Directors should be elected both geographically (by region) and on a global basis. Measures should be taken to reduce the risk of national dominance. The election of directors on a regional basis, from among voters within the region, was cited as a critical benefit of the At-Large system that guarantees a baseline of regional balance. We recommend ICANN continue to elect one Director from each region (recognizing that the number of regions may change over time.)
Many with whom we spoke also felt strongly that geography should not be the only basis for representation. The election of some directors on a global basis could have the very desirable effect of promoting organization, debate, and communities of interest across regional lines. ICANN should therefore elect the remaining At-Large Directors on a "global" basis, through a separate ballot cast at the same time for regional and global elections. In order to diminish risks of regional domination, we also recommend that no region be allowed to secure more than 25% (one under the current board makeup) of these "global" seats.
3.5c ICANN should refine its election policies, creating more independent Election and Nominations committees and improving educational efforts surrounding preferential voting. As it did in 2000, we recommend that ICANN continue to delegate responsibility for the development of election rules and the identification of some candidates for election to Election and Nominating Committees.
3.5d ICANN should commit to funding the Membership from its operating budget for at least the next several years, while pursuing alternative methods. To ensure that the Membership's development continues unimpaired, we recommend that ICANN commit itself to funding the membership's full expenses out of the ICANN operating budget for a period of at least the next several years. While we recognize that this constitutes a serious budgetary obligation for ICANN, it is also an integral component of ICANN's legitimacy.
A pillar of ICANN's funding structure has been that those groups deriving financial or other benefits from ICANN's operation should shoulder the responsibility for supporting ICANN's activities. These groups indisputably benefit from a legitimate and stable ICANN. The At-Large Membership is an essential component of that legitimacy, and should be supported. We recognize that money paid by the registries and registrars to support these expenses would ultimately be passed on to the customers of those entities - namely, the community of users who purchase and hold domain names.
At-Large Membership and elections would not be cheap, but we believe that the costs of Membership and elections should be manageable. The 2000 election cost on the order of US$350,000 and the largest variable costs-from the postal return system-should decrease sharply as Membership stabilizes if ICANN can institute a reasonable system for "one-time" authentication of members. Even a Membership costing US$1 million (US$500,000 in fixed costs, and US$500,000 in variable costs - enough to register more than three times the 2000 membership), would cost less than US$.04 per domain name registered in ".com" alone.
Self-sustainability is a laudable long-term goal for the Membership, and should also be explored in the future. Possible options include:
The ALM and At-Large Directors are only one method - though an essential one - for enhancing ICANN's legitimacy and stability. Other reforms should be pursued in parallel to the changes outlined for the ALM.
3.6a ICANN should develop structural constraints on Board authority. ICANN should directly confront the reasonable fear that it will venture into an even broader policy agenda than it now does. To do so, ICANN should find a way to constrain its jurisdiction in a binding fashion. The limitations currently in ICANN's bylaws do not effectively serve this purpose, both because ICANN has shown a distressing tendency to amend its bylaws casually, and because there is little public confidence in the restraints that already exist.
ICANN should therefore take steps to limit the Board's mission and powers to provide the organization with additional legitimacy:
3.6b ICANN should create additional accountability mechanisms. Accountability promotes responsiveness by enabling stakeholders to hold the organization responsible for its decisions. To renew accountability we suggest:
3.6c ICANN should pursue Supporting Organization reform. While not discussed in this report in detail, the ICANN community has recently focused significant attention on the proper role of the Supporting Organizations. We note above many concerns about the effectiveness and openness of the SOs, especially the DNSO. Public participation and representation through the ALM does not end the need for sound bottom-up policy development processes in the functional areas represented by the ASO, PSO, and DNSO. We therefore support the notion of reevaluating the SO structure, processes, and board representation.
Having established a theoretical framework for ICANN's legitimacy and proposed a set of structures to make that legitimacy real, we believe ICANN still needs to institutionalize the At-Large Membership, establish its relationship with theICANN community, and set in motion an election. Reform must happen quickly; the terms of office for the five elected At-Large Directors on the Board will expire in November 2002, while four other At-Large seats are filled by Initial Board Members appointed in 1999. Their replacements should be chosen and ready to serve by November 2002 at the latest. Considering the complexity and length of the election implementation, the Board will need to take decisive action at its Annual Meeting in November 2001.
This necessitates major bylaws changes, as well as a commitment by ICANN to adhere to a rigorous schedule of At-Large Membership-related activities over the next year. The full NAIS report proposes a detailed plan of bylaws changes, as well as a timeline for community action. For timely implementation, we feel it is essential for the Board to make substantial progress in this direction by November.
ICANN is an experimental, quasi-public global organization with important responsibilities for a critical worldwide infrastructure that must be managed as a public trust to serve the public interest. As such, its real legitimacy will come only when ICANN establishes robust and effective structures for a public voice to participate and be represented within the institutions of ICANN's internal governance.
Thus it is essential for ICANN to establish an inclusive, open At-Large Membership, with a clear means to participate in the decision-making process and substantial direct representation on the board. Such steps acknowledge the broad potential impact of ICANN's activities, and would significantly legitimize ICANN's operation. They would go a long way towards the basic tenet of good governance that those affected by policies should have a role in deciding those policies. We understand that the difficulties in establishing a functional At-Large Membership are not trivial. However, we also believe these difficulties are outweighed by the benefits to ICANN of such essential legitimacy and stability.
As long as ICANN fails to strongly establish the public voice in its activities, it will increasingly risk that the Internet community will reject the legitimacy of its decisions. The terms of office for the five current At-Large Directors will expire in November 2002; ICANN should make it a high priority to develop an implement a system for their replacement by then.
We note that while SO reform is important it need not and should not delay progress towards establishing an At-Large Membership and At-Large elections next year. To that end, we have presented recommendations and systems that can be set in motion immediately. These approaches are structured in terms that are flexible in the face of changes that seem possible, whether sooner or later, in ICANN's SOs, regions, or board structure.
The Internet is about empowerment - about giving information, and giving voice, to individuals around the globe. What more appropriate opportunity than for ICANN itself to use the power of the Internet to harness the public voice needed to make its own work legitimate. By solving the riddle of its internal governmnent, ICANN will take a great step toward a more stable grounding to ensure the success of its own, larger experiment.
We welcome comment and feedback on this document. For more information and a copy of the full NAIS report, please visit our web site at http://www.naisproject.org/. Comments or questions can be addressed to comments@naisproject.org.
The NGO and Academic ICANN Study (NAIS) is a collaboration of experts from around the world, formed to explore public participation in ICANN and the selection of At-Large Directors on ICANN's governing board. NAIS mirrors ICANN's own study effort, and was created to provide an independent examination, global in scope and grounded in a belief in the importance of public representation. NAIS team members include:
This study was conducted through a generous grant from the Markle Foundation (New York, USA), and through an additional travel grant by DENIC (Frankfurt, Germany).
© 2001 NAISProject.org
Privacy Policy
webmaster@naisproject.org