Working Paper on the ICANN Structure and the Nominating Committee Concept
Posted: 9 May 2002

Committee on ICANN Evolution and Reform
Working Paper on the ICANN Structure and the Nominating Committee Concept

This is one of a series of papers in which we offer thoughts for further discussion. These are not conclusions, so these thoughts may or may not actually form the basis of specific recommendations to the Board or the community as a whole. We hope they will provide a focus for more detailed discussion by the community in the next few weeks. The most useful comments will be those received by 19 May 2002.

Click here to read comments on this working paper.


The discussion that follows rests on the notion that the evolution and reform discussion involves several distinct, but interrelated issues: what ICANN's work should be (mission statement); how ICANN's work should be done (policy-development process); how that work should be paid for (funding); how to organize ICANN's internal structures (structure); and how to populate ICANN's Board (the NomCom process). All are interrelated, but each deserves separate consideration from the other. In this paper we deal with only the latter two issues – how to organize ICANN's internal structures and how to populate ICANN's Board (and Steering Committees).

I. Structure Issues

As earlier noted, we have chosen to take as the starting point for analysis the Stuart Lynn proposal. Nevertheless, we recognize, as Dr. Lynn himself has made clear, that his proposal was intended as a starting point, not a finished product. For convenience, we use in this document the nomenclature of the Lynn proposal, but we do not mean to signify that we have adopted those particular words or titles.

There appears to be widespread agreement that the problems identified in the Lynn proposal are accurately described, and we accept that description as a base for examining proposed changes. In general, the Lynn proposal has drawn the most criticism on two fronts: government involvement in Board selection and funding, and the perceived closed nature of the NomCom. The other recommendations included in the proposal have not received nearly as much attention, or for that matter generated as much criticism. Hopefully, this document will encourage more focus on some of these other issues.

The rationale for the emphasis on governments in the Lynn proposal was in large part the need to have them more involved, as a way to reflect the public interest in ICANN's activities. In this working paper, we assume that in the relevant timeframe there will be neither government funding nor direct government participation in filling specified seats on the Board, other than an ex officio role for the GAC chair. This should not be taken as a conclusion that either or both of these ideas for increasing explicit government involvement in ICANN are undesirable; the rationale for each was laid out in the Lynn proposal, and has significant merit. In particular, greater government involvement is one way to fill the vital need to reflect the public interest on ICANN's Board through mechanisms that are practical, valid, affordable, and not readily subject to capture. In this working paper, however, we explore alternatives to direct government involvement in Board selection because of the practical difficulties that have been expressed in implementing that idea in the near future.

Further consideration of the Lynn proposal and the various comments received suggests that adjustments to the Lynn proposal similar to the following may broaden the base of support for those proposals. We invite community comment on this perception, and on the following concepts:

A. Organizational Structure

The Lynn proposal suggested a structure parallel to what exists today, with intermediate policy-development bodies (the forums and the Policy Councils) providing recommendations to the Board for action. With respect to structure, the Lynn proposal essentially collapsed the ASO and PSO into a single body, created a new Generic TLD Names Policy Council (GNPC) to replace the DNSO, and suggested a new Geographic (perhaps better called Country, to avoid acronymic confusion) TLD Names Policy Council (CNPC). It also created a Technical Advisory Committee to provide technical advice.

The major structural innovations in the Lynn proposal are (1) the elimination of the fixed number of constituencies in the DNSO, and (2) elimination of the direct election of both the Steering Committees (today's Councils) and the Board of Trustees. Those are replaced within the proposed Policy Councils by (1) self-organizing forums that are solely devices for policy discussion and advocacy to the Policy Council Steering Committees and the Board, and (2) a totally separate process, part ex officio and part NomCom, for selecting both members of the Board and the persons who serve on the Steering Committees (the equivalent of today's Names Council, for example). The Lynn proposal treats the various substantive policy areas relatively uniformly, by establishing for each a Steering Committee that is about half ex officio representatives of the major groups in each area and the other half generated by the NomCom. The forums were assumed to be cross-Council, participating to the extent relevant in each of the three Policy Councils.

Address-Allocation Issues. The nearly uniform reaction to the notion of combining address-allocation and protocol numbering policy in one body has been to the effect that these do not necessarily mix well, and that the rationale for the suggested combination is weak. Bottom-up policy development seems to be working effectively within the ASO framework, and to the satisfaction of the affected community. For these reasons, the existing ASO structure could be retained essentially as is, and remain responsible for formulating address-allocation policy in the same way as does the ASO today. Following the Lynn proposal, this would mean that the Chair (or designee) of the ASO would have an ex officio seat on the Board. This working hypothesis seems consistent with the principles just published by the RIRs, but we welcome their further comments, as well as those of the rest of the community.

Protocol Issues. If the ASO and PSO are not combined into a single body, and since we agree that electing Board members should no longer be a role of these policy advisory bodies, the PSO could be reconsidered.

The Lynn proposal recommended the creation of a Technical Advisory Committee; the PSO could, as a practical matter, be transformed into that entity, but as a technical advisory body, not a policy body, and with a membership chosen on the basis of relevant technical competence and experience. As pointed out in the Working Paper on ICANN Mission and Core Values, the scope of ICANN's "policy" responsibilities in the protocol area is quite constrained, because ICANN follows the long-established policy of simply administering the registries of parameter and protocol values according to the guidance developed by the "creator" of the particular registry involved, in nearly all cases the Internet Engineering Steering Group (IESG) and the Internet Architecture Board (IAB). Comments are invited on the appropriateness of this allocation of responsibility.

Presumably the TAC would become the principal vehicle for technical advice to the Board; the Lynn proposal suggested that its Chair (or designee) would have an ex officio seat on the ICANN Board. As outlined in the Lynn proposal, this would provide a minimum of two obvious sources of technical competence through ex officio seats on the Board – the TAC Chair or designee (voting) and the IAB Chair or other designee (non-voting). The IAB designee would serve in effect as a technical liaison to the IAB and the IETF.

Names Policy Issues. If there are to be serious efforts at consensus policy development in the names area below the Board level, there should be some vehicle through which that policy-development effort can be undertaken (essentially replacing or through evolution of the DNSO). In the Lynn proposal, this was called the Generic Names Policy Council (GNPC), but the name is obviously not critical.

There seems no doubt that names-policy issues will continue to require considerable attention within ICANN. Some have argued that those issues should simply be decided by the Board, and advocated eliminating the subordinate policy-development bodies altogether. We are skeptical that this approach is desirable; it is not clear why it would expedite the process to have the entire debate at the Board level.

If there are to be intermediate names policy-development structures in ICANN, and if the ccTLD community is to be part of ICANN, the establishment of a policy entity associated with geographic or country names (the CNPC – see above) should be carefully considered. There are many complex issues related to this general area that require considerable discussion, and thus we will defer a full discussion of this particular subject to a later, comprehensive analysis. If the CNPC is to be part of a new ICANN structure, it will be necessary to clarify exactly what its policy role will be, who will be eligible for membership, and how it will be managed. For this working paper, we simply assume its existence.

Thus, assuming the basic overall policy-development structure of the Lynn proposal would hold, the intermediate policy-development structure outlined above would have three principal subordinate entities below the Board: for addresses, non-geographic names, and geographic names.

These policy-development bodies would be joined in the Lynn proposal by four other advisory bodies: the Technical Advisory Committee (TAC); the Government Advisory Committee (GAC); the Security Committee; and the Root Server System Advisory Committee (RSSAC). In the Lynn proposal, the TAC would have an ex officio voting seat on the Board, along with the policy-development bodies, and each of the other advisory committees would have non-voting ex officio seats on the Board. In both cases, representatives would receive Board communications and be entitled to attend and participate in all Board meetings and teleconferences.

In sum, this intermediate structure would produce seven ex officio Board seats – four voting (APC, GNPC, CNPC and TAC) and three non-voting (GAC, Security Committee and RSAC), all filled without participation of the NomCom. If we assume a Board with fifteen voting members – which seems desirable, since that is close to the upper edge of effective size – and the CEO retains a voting ex officio seat, that would leave ten Board seats to be selected by some other means. The means of selecting these additional Board seats are discussed in part II of this working paper.

B. Steering Committees

The Lynn proposal suggested that each Policy Council have a Steering Committee partially ex officio and partially produced by a Nominating Committee process. The rationale was to make the Steering Committees more broadly representative of the entire community, and not just of those active in ICANN. We accept this structure as a working hypothesis for the GNPC and CNPC, but we recognize that this must be subject to getting general agreement on NomCom composition and process. This topic is addressed in detail in part II of this working paper.

It should be noted that the Lynn proposal's more fluid forum concept, where forums could be self-organized even around specific issues and range from permanent to very temporary, is designed to make it easier for all interested persons and groups to participate in ICANN. This concept has been put forward (but not adopted) in previous discussions about ICANN, but the evolution of ICANN warrants its reexamination.

Under the forum concept, it is not feasible either to have elections for NomCom positions or for every forum to have a seat on any Steering Committee or to determine how any specific seat is assigned, since that would create perverse incentives for forum creation and potentially unworkably large numbers of such bodies. One way to address this concern would be to have both permanent and temporary forums, with the permanent ones being those that meet certain criteria for continuous involvement in the process (and include all or most of the significant stakeholder interests today) and having automatic representation on the NomCom, and the temporary ones being all other forums, which can to have a role in Steering Committee selection after they have participated meaningfully for some period and/or meet any additional criteria deemed appropriate. Further, a way must be established for the eventual creation of new permanent forums and for the elimination of initial ones that outlive their usefulness.

We invite community comment on all these questions, or any other workable – with an emphasis on workable – options.

C. Independent Review Panel

Comments to date have been mixed on the Lynn proposal's suggestion that some alternative means be established to meet the goals of an independent review panel. Some have noted that the notion of independent review carries certain dangers. There is certainly a risk of significant disenfranchisement if some less representative entity than the ICANN Board has the ability to override Board decisions on the merits. On the other hand, some have stated that the mere existence of such an entity would provide a comforting (to some at least) disincentive to Board decisions outside ICANN's bylaws – if, and this is a very large if, an independent review panel was populated with persons who had both clearly delineated and limited authority and were widely viewed as credible persons throughout the entire ICANN community. For some commentators, the need for some type of review process is fundamental and the establishment of an Ombudsman can fill the need. The creation of an independent review panel would also deal directly with a potential legal issue, since the existing registry and registrar agreements provide that the contracting party may avoid compliance with certain newly developed ICANN policies until an independent review panel is in place. There have been various proposals on this subject, but very little detailed discussion or analysis of the implications, ramifications, and pros and cons of various ideas.

Another issue that has been raised by some is the scope of review by an independent review panel. The current Independent Review Policy calls for the scope of review to be limited strictly to "compar[ing] contested actions of the ICANN Board to the Bylaws and Articles of Incorporation and to declare whether the ICANN Board has acted consistent with the provisions of those documents. The IRP is also responsible for reviewing determinations regarding the presence of a consensus on which an adopted policy is based, to the extent provided in agreements between ICANN and third parties."

We believe that any review by an independent review panel should be subject to very careful constraints. To broaden independent review, for example, to involve a review of the merits of a particular decision seems fraught with potential problems. It does not seem appropriate to provide some non-representative body with the ability to override decisions by the Board (assuming a Board selection process that produces a Board which is representative of the community as a whole) simply because an independent review panel disagrees with the Board on the merits. Indeed, the inclusion in the current policy of a role in evaluating consensus would have to be conformed to the decisional standards ultimately adopted by ICANN (for more detail, see the Working Paper on the Policy-Development Process). In addition, we note our view that any independent review panel must have a clear mandate to dispose summarily of frivolous issues so as not to allow the mere existence of the process to be misused to impede the exercise of the community will.

II. The Nominating Committee Concept

The Lynn proposal suggested a fifteen-person Board, with five ex officio and ten At Large members – with five of the At Large members to be selected, somehow, by governments. The rationale was two-fold: (1) a less directly representational Board would likely be a better Board in the long run, and (2) the government-selected At Large trustees would provide a voice for the general public interest, and thus respond to the accountability/legitimacy issues that generated much of the motivation for At Large online voting. The Lynn proposal further suggested that the five At Large trustees not selected by governments (and about half of each Policy Council Steering Committee) be selected by a Nominating Committee (NomCom) made up of three Board members and four other persons selected by the Board.

As indicated earlier in this paper, for purposes of this discussion we assume that governments will select no Board members in the relevant time frame. Thus, the NomCom on this assumption would be responsible for filling vacancies in the ten At Large Board seats, not just five as suggested in the Lynn proposal.

A. The Rationale for a Nominating Committee

The major criticism of the NomCom idea to date has been that, since it was made up of Board members and Board appointees, it was too controlled by the Board, and thus formed a "closed circle" that was too self-reproducing, preventing needed renewal, corrections, and adjustments as the Internet evolves. This criticism appears to the Committee to have some merit, and thus we solicit comment on a modification designed to respond to that criticism while retaining the perceived benefits of a NomCom process.

We would like to solicit comments on the following concept: a NomCom consisting of delegates from identified constituencies (possible examples would be such groups as registries, registrars, RIRs, large business users, consumer organizations, academic or other non-commercial organizations, etc.) Such a NomCom might have a non-voting chair appointed by the Board (possibly a Board member, possibly not), and it might also include one or more liaisons (IAB, GAC, etc.) that could either have voting rights or not. These could serve to provide insights about particular needs on either the Board or a Policy Council Steering Committee, and about the strengths or weaknesses of particular people being considered based on prior knowledge or experience. Another way to accomplish this goal is to have some ex officio members (voting or non-voting) on the NomCom from the Board.

We use the word "delegate" intentionally. For ICANN to be the most effective organization it can be, its Board should be as neutral and objective, and as capable, as is possible under the circumstances. We do not believe that a Board composed primarily of representatives of specific interested groups is the best way to reach this goal. By contrast, we believe that a Board composed of persons with diverse skills and perspectives and the right personalities, that can act collegially and effectively to make decisions that are grounded in the best interests of the overall communities affected would serve both the public interest and the long-term interest of all direct participants in ICANN. This view suggests that on the NomCom, the general public interest could be represented by those delegates selected by consumer organizations and user groups, while the providers would also supply delegates to the NomCom. We do believe it is both appropriate and useful to have those constituencies that are directly and consistently involved in ICANN to have an important role in the selection of Board members, but we respectfully suggest that it is more consistent with the goal – a consistently high-quality, effective Board – that all constituency representatives on the NomCom (whether from providers, consumer groups, or users) think of themselves as "delegates," not as "representatives" – in other words, as bringing a particular perspective and knowledge to the joint process of selecting good ICANN Board members, rather than "representing" the narrow interests of the particular constituency that brought them to the NomCom.

This is not, in our view, metaphysical semantics. The only way that an ICANN-type entity can be effective is if it advances the common good. Now, this is of course a composite of the legitimate interests of all participants (as well as non-participants), but it may in any particular case not directly reflect any one of those interests. What is needed in ICANN, not only at the Board level but also in the policy-development bodies, is less political log-rolling and more genuine efforts to find consensus solutions. This effort will be advanced if we establish a culture and tradition of looking for the common good of the community, even if this occurs through the perspective of a particular interested party or group.

B. Specific Issues on Which Comments are Invited

The comment period has ended. Comments regarding these issues may be read here.


This working paper presents various ideas, based on Dr. Lynn's proposal and the comments the Committee has received, on ICANN's overall structure and the Nominating Committee concept. The ideas presented have varying levels of appeal to the members of the Committee and ICANN's President, but they are clearly not conclusions or recommendations. They are offered to the community as the basis for further discussion, in an effort to focus the community debate on more detailed and particular issues given the extreme time pressure that we are all working under.

We encourage prompt and detailed comments on any or all of the topics discussed in this working paper.

ICANN Board Evolution and Reform Committee
9 May 2002

Click here to read comments on this working paper.

Questions concerning the layout, construction and functionality of this site
should be sent to webmaster@icann.org.

Page Updated 20-Aug-2003
©2002 The Internet Corporation for Assigned Names and Numbers. All rights reserved.