Addendum

Addendum: Options for At-Large Governance

As the NAIS team moved towards its final report, it produced a diverse set of options for consideration. Our final recommendations are explained in the body of the report above, but consideration of other ideas for representation of the public voice in ICANN continues to be valuable. These options for the ICANN's At-Large Directors and Membership, and public participation in ICANN more generally, range from relatively minor process points to wholesale revision of some of the organization's basic operating principles.

We present below a template of the larger set of options developed for consideration in the preparation of this report. These options were set forth in the Interim Report published by the NAIS group prior to the ICANN meeting in Stockholm in June, 2001.

Discussion of options is organized into two basic categories:

1. Options Based on the At-Large Directors Model-Many options retaining At-Large Director seats distinct from those of the Supporting Organizations. Within this strategy, there are multiple dimensions of questions for resolution, including:

Number of At-Large Directors

Selection Mechanism

Membership Criteria

Membership Role

Regional v. Global Structure

2. Options External to At-Large Board Directors - In addition or in conjunction with At-Large Directors, issues of public participation might be addressable through other mechanisms:

Limiting the ICANN Mission/Slate of Activities

Reforming the Supporting Organizations

Establishing new Bodies as a Check on Board Authority

Within each of these strategies, there are persistent questions about implementation and process; as a result, a number of viable options for public participation have emerged, ranging from peripheral changes to the 2000 election process to wholesale rethinking of the organization and its mission. And while a number of such options are discussed below, no single option alone will act as a panacea for the problems in ICANN. The best solutions for an effective public voice will probably come through a considered recombination of the elements listed here.

I. Selecting At-Large Directors to the Board. Discussion in the ICANN community has frequently emphasized the importance of At-Large Directors as a means of public participation in Board activities. Many of those interviewed believed that an efficient, reliable, fair process for selecting At-Large Directors would provide the most likely path to successful reform of ICANN.

Within the approach of selecting Directors, there are a number of critical issues that need resolution. The 2000 election, in particular, displayed evidence of many problems that must be solved if we are to build a lasting, effective system for public participation. With that in mind, we offer here a list of options areas for consideration.

Implementation Questions

Number of At-Large Directors

Continue status quo (No. At-Large Directors = No. of S.O. Directors)

Reduce number of At-Large Directors

Increase number of At-Large Directors

Selection Mechanism

Direct election

Indirect election

Hybrid election

Delegate to existing membership organizations

Membership Criteria

Open membership

Nominal membership fee

Digital certificates

"Webs of trust"

Knowledge/experience-based criteria

Membership Role

Electorate

Policy-making

Oversight/review

Advisory

Regional Structure

Five-region model

Global model

Expanded regional model

Hybrid model

Number of Directors. In the current model of the ICANN Board, At-Large Directors constitute a portion of the Board equal to the total number of Supporting Organization Directors: Nine At-Large Directors = Three ASO Directors + Three DNSO Directors + Three PSO Directors-plus a nineteenth seat reserved to the President/CEO. The status quo is seen as providing the Board with a certain balance. Other options expressed frequently in our consultations included:

Reducing the At-Large Directors, for example to five (tracking to the five-region model of geographic representation), or to three (matching the representation of a single Supporting Organization). Implementing either of these could include commensurate reductions in the number of Directors selected by the Supporting Organizations.

Increasing the number of At-Large Directors to ten (two from each of the five geographic regions), or more. And though it might be more accurately considered an external option, it would be possible to have the entire Board selected by some public membership-such a change would obviously entail the elimination of Board-level representation for all three Supporting Organizations.

Selection Mechanism. In 2000, as is discussed above, five At-Large Directors were selected by direct election in each of five geographic regions. The strengths and problems of that election have already been discussed, and continuation of the 2000 model remains a viable option for future public participation. Other possibilities for a selection mechanism include:

Indirect election. While indirect elections can raise questions of both accountability and transparency, they may provide a means not only to select qualified Directors, but also to facilitate a persistent role for the At-Large Membership itself.

Hybrid election. If some kind of "council" that acts as intermediary between the user community and the Board proves desirable, it might be possible to establish such a council without sacrificing the legitimacy that comes with direct election of Board members. Some kind of combination election could be held, either with separate elections for Directors and council seats, or a proportional voting system could be used.

Existing membership organization(s). Some of the difficulties that were encountered in voter verification, fraud protection, and outreach might be avoided by capitalizing on the membership structures of existing organizations, mapping their memberships onto ICANN's At-Large Membership. Clearly, fair identification of such organizations could be difficult.

Membership Criteria. Again, as discussed above, the 2000 election used extremely open criteria for membership, requiring only that voters 1.) be age sixteen or over, 2.) have a verifiable postal address, 3.) have an e-mail address. No fee was required. However, problems with the postal return system and other difficulties in voter verification may have complicated certain aspects of the election. Revision of the membership criteria might help.

Nominal membership fee. This could not only help the At-Large membership become self-supporting, but could discourage frivolous registration and/or certain types of voting fraud. However, membership fees run the risk of placing membership out of reach for would-be members in developing nations.

Digital certificates. These could assist with certain aspects of election administration, but in the absence of a robust, worldwide public-key infrastructure, they seem unlikely to assist with voter verification in its most intransigent forms.

"Webs of trust." Webs of trust have been used effectively to authenticate limited groups of users for purposes of public key exchange and in other instances of identity verification, but they are basically untested. Questions of scalability and robustness remain. Also, webs of trust may trend towards basically closed memberships.

Knowledge/experience-based criteria. In the interest of promoting an educated (and therefore presumably responsible) electorate, some have proposed that potential members be required to demonstrate their knowledge of/interest in issues of addressing and naming. One particular proposal in this category would be to restrict At-Large membership to those users that own domain names. Again, this clearly tends towards a closed membership.

Membership Role. The role of those Internet users who registered as At-Large Members in 2000 has been a matter of significant controversy since the election's conclusion. While some have claimed that the, the election being over, those users no longer have a specialized role in ICANN, others believe that the At-Large Membership is a lasting community that should have a persistent role in ICANN. Clearly, a user-oriented membership could easily play several of the roles listed below.

Electorate. The membership would play an obvious role in selecting Directors to the Board. This is the role that the 2000 At-Large Membership most obviously played.

Policy-making. As the S.O.'s serve to propose and review ICANN policies, so could a public membership. Some have raised concerns that a policy-making or policy-review role for the user community would be needlessly duplicative of the S.O. process, while others maintain that the S.O.'s do not adequately include a distinct user perspective.

Oversight/review. The membership could play some role in validating Board decisions prior to their implementation. It could also/alternatively serve as an independent reconsideration authority, to resolve Board actions contested by parties in the ICANN process.

Advisory. Finally, the membership could play a purely advisory role not only to the At-Large Directors but to the Board as a whole, or to the Supporting Organizations.

Regional Structure. The 2000 election elected five directors, one from each of five geographic regions. As is discussed above, the model probably enhanced the election's overall efficiency and resistance to capture, but may not have provided certain Internet communities with ideal representation.

Global At-Large Directors. Future Directors could be chosen without reference to any kind of regional model. This model provides simplicity and some baseline fairness, but might result only in capture of Directors by populous or well-organized nations.

Expanded regional structure. ICANN has no especial commitment to the five-region model. More regions, or a sub-division of regions, could be established, and Directors adjusted accordingly.

Hybrid structure. Questions of geographic representation are not either-or models. It would certainly be possible to select some Directors on a regional level, and others globally.

II. Other Strategies

While the selection of At-Large Directors remains an important and compelling strategy for promoting the public interest, it is not the only one. Other approaches to bringing ICANN's activities closer to public interest ideals could defuse certain problems in the selection of At-Large Directors, or even reshape ICANN in ways that would permit us to thoroughly rethink our approach to DNS administration. Below, we attempt to list a broad range of possible answers to the questions before ICANN.

Limiting the ICANN mission/slate of activities. If the public's interest in ICANN's activities stems from the policy implications of ICANN's decisions, then one solution might be to redefine ICANN's mission so that it is more closely confined to pure technical management.

Changes to the Articles of Incorporation. While the ICANN Articles of Incorporation charge ICANN with administrative responsibility in four specific areas, they do not establish meaningful limits on ICANN's activity within those areas. Many of those contacted for this report were concerned that ICANN may be entering areas of policy-making beyond what it was designed for. Changes to the Articles that limit ICANN to technical coordination might stave off the broad effects that inappropriate policy-making could have.

Disband ICANN and establish a new organization. If ICANN in its current form is so resistant to reform that meaningful public participation can't be implemented without compromising ICANN's basic mission to efficiently administer the systems in its charge, then it might be advisable to scrap ICANN and establish a new starting point. With the flexibility that comes from a fresh slate, it might be possible to build an entirely new organization, learning from the successes and mistakes of the ICANN experience.

Reforming the Supporting Organizations. Some of those interviewed felt that changes to the Supporting Organization substructure might address the need for public representation. However, the Supporting Organizations are sometimes thought of as providing representation to those affected directly by ICANN policy-a group that includes some, but not all users, since many users are affected mostly by secondary effects.

Changes to S.O. constituency structures. Any of the Supporting organizations, but most notably the DNSO, might benefit from review of their internal organization. The DNSO's lack of an Individual Domain Name Owners' constituency, in particular, has been the subject of criticism from many sources, but addition of an IDNO alone is unlikely to provide a long-term solution. A new constituency structure that more accurately groups like interests together could greatly increase the efficiency and effectiveness of the S.O. process.

Addition of New Supporting Organizations/Radical S.O. Reform. New supporting organizations-such as a ccTLD or User S.O.-might increase participation in the Supporting Organizations and provide a new policy role for the user community. Some have expressed concern that the S.O. structure as it currently exists fails to make allowances for important divergences of opinion on key ICANN issues, or that it underrepresents certain groups. By adding new S.O.'s or recalibrating the existing ones, it might be possible to bring the ICANN process more in line with the public interest. Clearly, this option would include a significant review of the Board's own internal structure.

Divest ICANN of the Addressing and Protocol Supporting Organizations. Presently, controversy about public participation has been heaviest in and around the Domain Name Supporting Organization, since the area of domain name policy includes issues of obvious public concern. Addressing and protocol issues, on the other hand, have to date been less controversial and in any case seem to demand a specialized approach in fostering the public interest perspective. Addressing, protocol, and domain name policy work all demand different expertise, working styles, and priorities. In light of that fact, one may question the wisdom of placing all three under the authority of a single ICANN.

Establishing new bodies to counter Board authority. The current ICANN model has sometimes been treated as a top-down one, with significant authority centralized in the nineteen-member Board. One possible reform would be to temper that authority either by dividing it with another, as-yet-envisioned ICANN body (likely including a strong public interest perspective) or by establishing a meaningful oversight body capable of reviewing Board decisions and, in special circumstances, reversing or altering them.

Separation of Authority. The authority of the ICANN Board could be at least partially decentralized, and a new deliberative body established to share that authority through a series of checks and balances. That body might be selected by a radically different method from the Board itself.

Oversight Body. Some kind of ICANN "judiciary" could be established, capable of reviewing decisions of the Board and comparing them both to the corporation's appropriate mission and to demonstrated consensus in the community. However, this option, like tie one before it, would constitute a major change in ICANN's operating procedure, and could even run the risk of conferring inappropriate, pseudo-governmental legitimacy

table border="0" cellpadding="2" cellspacing="2" width="80%" align="center">4.2 Timeline of At-Large Activities TABLE OF CONTENTS