6.3.2. Structural reform

Common to the reforms proposed above to the plenary body, Secretariat, Advisory Group, open consultation meetings and dynamic coalitions is the need for a new, democratically or consensually appointed multi-stakeholder body to exercise the following substantive functions:

Such a body would essentially constitute the “effective and cost-efficient bureau” referred to in the Tunis Agenda and in early contributions from Forum hawks that preceded the formation of the Advisory Group.[2] Once it had become apparent, following the expiration of its initial mandate from the Secretary-General, that the Advisory Group possessed neither the capacity nor the legitimacy to fulfil the roles that the Tunis Agenda and the hawks demanded, some of the hawks began to renew their calls for a multi-stakeholder bureau,[3] and these are calls that would also be addressed by such a body as outlined above. Consociationalism and the IGF

The establishment of a new multi-stakeholder bureau for the IGF will however have to be acceptable not only to the Forum hawks, but to the doves also, lest they withdraw from the governance network altogether.[4] This will require some thought to be given to their motivations for opposing an IGF with the structural capacity to fulfil its mandate to develop public policy recommendations, and to how those objections can be countered without compromising the democratic principle or the equality of all stakeholders within the network.

There is very little difficulty in explaining the doves’ opposition in terms of the realist school of international relations theory.[5] From that perspective, it is a truism that the first priority of actors in the international system is to preserve their own political and economic power.[6] The status quo in Internet governance favours the forum doves, in that the United States holds authority over the global DNS root (with varying degrees of support from its allies), and has strongly supported the “private, bottom–up coordination”[7] of the Internet technical community, in which the private sector has heavily invested. Ipso facto, on the realist view, there is no reason for them to concede any ground to the Forum hawks who seek to disturb that status quo, particularly since the hawks—developing country governments and civil society—wield comparatively little political and economic power with which to do so.

Whilst the realist school has been found overall to offer a simplistic account of the status and motivations of international actors (particularly in understating the significance of non-state actors), it does plausibly explain the doves’ opposition to the development of the IGF’s capacity to make soft law through a consensus process, since such a process magnifies the power of minorities, as WSIS demonstrated.[8] Although allowing “rough consensus” addresses this to some extent, governments in particular are likely to be no more comfortable with rough consensus than they are with full consensus, since it may require them to accept politically unpopular concessions if they are left in a small minority; quite a far-fetched expectation if the US government’s position on oversight of the global DNS root is taken as an example.

In designing a multi-stakeholder bureau for the IGF it will therefore be necessary to balance the need to ensure that the bureau is capable of effectively performing the roles outlined above, with the risk of overtly challenging the existing political and economic power of the Forum doves and thereby inducing them to leave or to undermine the IGF. A liberal institutionalist gloss upon the realist scholar’s position would allow that even the Forum doves will have reason to support an empowered IGF if it will contribute towards a more sustainable international Internet governance regime for the long term, provided also that the soft power it exercises offers no significant threat to their own.

Such support from the doves would not be forthcoming for a multi-stakeholder bureau that acted by voting or rough consensus, because such a bureau might reach decisions that would challenge the authority of particular stakeholders or stakeholder groups; most obviously the sovereignty of governments. To continue the previous example, consider the (admittedly far-fetched) prospect of the bureau recommending the use of an alternate DNS root, if the only dissent to this recommendation came from the United States government and ICANN.

There can be nothing wrong with the IGF in plenary session reaching such a position through a process of democratic deliberation, even if only rough rather than full consensus on it is achieved in the end. Neither would (or could) the effect of the plenary’s rough consensus be prevented from carrying its own normative resonance, which might independently influence the actions and shape the expectations of participants in the Internet governance regime. However, there are good reasons why the bureau ought not also in such a case elevate the plenary’s rough consensus to the level of a formal recommendation of the IGF:

Kenneth Cukier, moderator of two sessions at the Athens meeting, had also earlier written of the need for new multi-stakeholder structures not to reiterate old power struggles over Internet governance:

The battle over the institutional design becomes a proxy for a much narrower interest one wants. In 1996 with the IAHC, this was for new TLDs by Internet entrepreneurs; in 1998 with ICANN it was for privately operated TLDs by NSI (now VeriSign); in 2005 with WSIS it is for more power by governments. As in previous cases, any arrangement that leaves other parties unsatisfied is bound not to endure long. Every party employs the term “multi-stakeholder” to mean that they will enjoy predominant power but leave a few, merely symbolic crumbs for others.[10]

The alternative is a structure which institutionalises the distinct but complementary authority of each of the stakeholder groups. Such a structure has already been described as a consociation.[11]Apart from procedural benefits to be discussed later,[12] the principal benefit of a consociational bureau which is segmented into stakeholder groups, over a bureau in which the stakeholders are required to make decisions only as a uniform entity, is that it formally and publicly institutionalises the equality of the stakeholder groups, thereby acting as a balance to the political and economic inequality of the stakeholder groups outside the IGF. This in turn fosters the value of deliberation between equals, which is one of the key requisites of deliberative democracy, and one of the most difficult to realise.[13]

To put this in context, whilst the power of mutual veto would allow governments to formally block a proposal supported by the other stakeholder groups (thus addressing their inability to abide the diminishment of their existing political and economic power), this only reflects their effective power to do so through domestic or intergovernmental rules in any case. However what is unique and valuable about the consociational structure is that this same power of veto is placed in the hands of the civil society and private sector members of the bureau, who could just as easily formally block a proposal supported by governments. This is a power that those groups lack within the present Advisory Group, since they sit in that group in their personal capacities and any balancing of stakeholder interests within it takes place informally and behind closed doors. Nominating committee

Having concluded that a multi-stakeholder bureau is required, the next question that arises is how its members should be appointed. There are three basic alternatives. The first is for members to be appointed by the UN Secretary-General. This is the manner in which the Advisory Group is currently appointed, and is also the method of appointment of GAID’s 60 Strategy Council, and before that the 55 members and Panel of Advisors of UNICTTF.[14]

It might be thought that this alternative was strongly ruled out at Section, and indeed it was, except to the extent that it would be possible (and politic) during the term of the IGF’s initial five-year mandate, for the bureau to continue to be officially appointed by the Secretary-General, but acting on the recommendations of the stakeholders. This however begs the question of how the stakeholders should make those recommendations.

In GAID’s case, nominations for the private sector representatives were coordinated by the ICC, and those for the civil society representatives by CONGO through a volunteer nominating committee.[15] But to rely on external organisations to provide nominees simply shifts responsibility to them to ensure that the nominations are made in a democratically legitimate manner. Therefore the other two methods for the multi-stakeholder bureau’s appointment (whether by the Secretary-General directly or by some peak body on behalf of an entire stakeholder group) must still be examined.

The second of them is voting. This is the method upon which APNIC, auDA and the GKP rely to elect their executive committees. However all of these organisations comprise a fixed body of members, whereas as Nitin Desai succinctly put it at the February 2007 consultation meeting, in the IGF’s case “there’s nobody who can elect a bureau, because there’s no membership.” A closer match for the IGF in this respect is ICANN, which of course attempted to hold elections for the At Large positions on its Board of Directors in 2000, but wrote the experiment off as a costly failure.[16]

The third possible method for the appointment of the IGF’s multi-stakeholder bureau is to attempt to reach consensus on the appointments through democratic deliberation. After all, this is the process that was found to allow the IGF to fulfil its policy-setting functions despite its plenary body lacking a defined membership. However, it is not without reason that none of the other exemplar organisations utilise this time-consuming method of appointment, since it presents the risk that consensus might not be able to be reached.

Since none of these three basic alternative methods of appointment is appropriate, the answer lies in a hybrid approach, which includes elements of the hierarchical, democratic voting and consensual methods. A model for such a hybrid is found in the IETF’s nominating committee (the Nomcom), which makes appointments to the IAB and to the IETF’s IESG.[17]

The IETF Nomcom is a committee of at least fifteen, comprising a Chair, ten volunteers, three liaisons representing related organisations and an advisor.[18] The ten volunteers are chosen by random selection from an open pool of nominees, whose only qualification for membership is that they have attended three of the past five IETF meetings. The Chair is appointed by ISOC following a process of consultation with members of the IETF community and the board of ISOC.

The Nomcom’s deliberations are to be “based on its understanding of the IETF community’s consensus of the qualifications required,” as well as being guided by criteria provided by the bodies to which candidates are to be appointed, though the final decision to put a candidate forward is put to a vote of the ten volunteers.

The Nomcom’s recommendations are not final. Through a process described as “advice and consent,” its recommendations must be ratified by ISOC’s Board of Trustees in the case of appointments to the IAB, and by the IAB in the case of appointments to the IESG. When the confirming body withholds consent to a nomination, the confirming body and the Nomcom discuss the matter, and the Nomcom makes a new recommendation for any position left empty.

A similar process could be employed by the IGF for the appointment of its multi-stakeholder bureau. The main modification required to the structure of the nominating committee is that there should be an equal number of seats for each stakeholder group, and that whilst the committee would deliberate as a whole, only members from a given stakeholder group would select that group’s nominees, in order to preserve its autonomy within the bureau. This equates to the system of proportional representation within the grand coalition that is one of the hallmarks of consociational ordering.

A second but related related reform to the IETF model is that each stakeholder group within the nominating committee should be empowered to employ consensual procedures for the nomination of its bureau members, rather than being limited to voting. In particular, it is typical that seats on the executive bodies of intergovernmental organisations (for example the ISO Council and Presidency of the EU) will be held by member governments in rotation, typically with one third (initially selected by lot) stepping down every year. There is no reason in the short term why governments ought not to be able to agree that this also ought to prevail within the multi-stakeholder bureau of the IGF.

Another consideration is that of diversity: both of the nominating committee itself, and of its nominees for the bureau. In this regard, there are a number of alternative models to inform the IGF’s approach. The IETF goes no further than prohibiting more than two members with the same primary affiliation (for example, the same employer) from acting on the Nomcom, and imposes no requirements on the diversity of the nominees.

Other organisations utilising nominating committees have different procedures; ICANN for example is not concerned with the NomCom’s own diversity, but requires its slate of nominees to exhibit diversity in geography, culture, skills, experience, and perspective.[19] The FSC upholds diversity at each level; requiring both its multi-stakeholder nominating committee and the candidates it nominates to include members from the North and South constituencies, and for those candidates also to reflect regional and gender balance.[20]

For the IGF, it is also appropriate that diversity be upheld at both levels, as a measure of redressing the impediments of disadvantaged groups in participating in the IGF’s democratic processes on an equal footing, which is a substantive democratic value.[21] However as to exactly what measures of diversity should be institutionalised in the composition of the nominating committee and its slate of nominees, there is room for disagreement.

Whilst it is clear enough that the digital divide and its underlying economic divide impede universal participation in Internet governance, do issues such as gender and disability present similar obstacles that must be overcome in order to achieve substantive democratic equality within the IGF? As there is no a priori answer to this question, the criteria to be satisfied by candidates for the nominations committee and the bureau, and the appropriate size for those bodies to make them sufficiently diverse, should be determined in the first instance through a process of democratic deliberation in which all those potentially affected can discuss the issue in the light of relevant background material.

Since this presents a chicken-and-egg scenario (as it would normally be the bureau that would take a decision on such issues, drawing upon the deliberations at open consultation meetings), herein lies the place for an hierarchical hand in performing the following “bootstrapping” functions (which are analogous to, but narrower than, those performed by ISOC and the IAB for the IETF’s nominating committee):

It is suitable that these tasks fall to the UN Secretary-General, after having solicited recommendations from the IGF in plenary session at an annual or open consultation meeting. In future, they would fall to the bureau.

Thus in summary, the first significant structural reform that is required for the IGF is the establishment of a nominating committee comprising equal numbers from each of the stakeholder groups, who would periodically deliberate upon the appropriate composition of a multi-stakeholder bureau for the IGF, and then separately nominate candidates from their own stakeholder groups either by voting or through consensual means. This nominating committee would be chosen by random selection from an open pool of volunteers, subject only to the fulfilment of criteria designed to ensure that the process is also substantively democratic. The establishment of these criteria, along with similar criteria for the bureau, and the appointment of the nominating committee’s non-voting chair, liaisons and advisors, would be performed in the first instance by the Secretary-General, and thereafter by the bureau itself, in both cases acting upon any consensus that may emerge during deliberation by the IGF’s plenary body. Multi-stakeholder bureau

The second significant reform is the establishment of the multi-stakeholder bureau itself. The broad parameters for the structure of this body have already been established above, when it was determined that it should be a hybrid between a standard deliberative democratic organ and a consociation:

This immediately raises the question, how are those stakeholder groups to reach a coordinated view on when to exercise their right of veto? Whilst a more detailed answer will be given when considering the bureau’s processes,[24] either consensual or democratic means may be used, and they may be tailored to the stakeholder group in question. This is another of the main benefits of the consociational form, which allows for the governmental group, for example, to require full rather than rough consensus to be reached on a recommendation if its veto is to be withheld. In fact, “there is no reason why a UN multi-stakeholder process should exclude an inner intergovernmental process, between governments, using existing UN rules, within [scil with] a defined membership.”[25]

It follows that unlike in the present Advisory Group (but in common with all eight exemplar organisations, and almost all other UN bodies), members of the bureau will be appointed in a representative rather than a personal capacity, at least to the extent that the decisions they make will be for a particular stakeholder group. The importance of this is that stakeholders are appointed to the bureau not simply because of their personal merit, but because their participation provides a balanced base of legitimacy for the bureau that is drawn from the values of each stakeholder group.[26]

Two justifications have been given for the appointment of Advisory Group members in their personal rather than representative capacities. The first, that this allows for “more full and active participation of all members” (as claimed by ISOC at the February 2007 consultations), will arise for consideration in the following section on the IGF’s processes. However there is a second, structural, justification which is that appointing members in representative capacities could risk “enshrining a fixed system of representation.”[27]

This risk is minimised by the fact that each new bureau will be appointed by a new, randomly-appointed nominating committee. It is further reduced by appointing bureau members as representatives of stakeholder groups rather than of particular stakeholders, so that their mobility between employers, governments or organisational affiliations will not affect their seats. Appointing members to represent stakeholder groups also allows unaffiliated individuals to sit on the bureau within the stakeholder group that they are best qualified to represent (most likely civil society).

The next question of structure that has yet to be addressed relates to how the chair of the bureau should be appointed. Rather than being a policy-setting role, the appropriate role of the chair is simply to preside over meetings of the bureau, to coordinate its work programme, and to act as a point of contact for the IGF’s Secretariat and other bodies with which the IGF is required to coordinate. Even so, it is inevitable that the substantive values of its incumbent will influence the manner in which the role is performed, as was found above to have been the case for the Secretariat and Advisory Group.

For this reason, a number of the earliest submissions on the appropriate structure and composition of a bureau for the IGF, even from Forum doves, stressed the importance of ensuring the diversity and accountability of the chair. ISOC, for example, suggested that rather than a single chair drawn from a given stakeholder group, there should be a panel of rotating co-chairs.[28] There is no example of this from amongst the exemplar organisations, but the appointment of co-chairs from different stakeholder groups has long been a practice of consultation panels in the Australian telecommunications regime.[29] Another instructive example is that of the CS-IGC, which although not a multi-stakeholder group, has two Coordinators who are elected for alternating two-year terms. Whilst, since August 2007, the Advisory Group has had two co-chairs, neither comes from outside the governmental stakeholder groups.

Three best practices can be drawn out from these suggestions and examples:

In like manner, the bureau should be entitled to appoint its own advisors and liaisons, who would provide information to guide the bureau in its deliberations and act as a conduit for communication between the IGF and other participants in the Internet governance regime. Amongst the exemplar organisations, similar arrangements exist in ICANN (whose Advisory Committees each appoint a liaison to its board) and CGI.br (which includes a non-voting Internet expert), as well as the IETF’s Nomcom (which contains three liaisons and an advisor) and even the IGF’s Advisory Group (to which the Chairs have appointed their own Special Advisors).

Despite the fact that these positions would not participate in decision-making, due to their privileged position of influence upon the policy-setting process, the principle of subsidiarity indicates that ultimate responsibility for their establishment and appointment should rest in the bureau at large, though there is no reason why it could not delegate those tasks to the chair and even institutionalise that arrangement in standing rules that future bureaus could follow.

The same applies to the case of standing or ad hoc subcommittees that the bureau may wish to form, such as drafting committees (like those into which WGIG was divided), an appeals committee (such as that of the CS-IGC), and liaisons for appointment to other organisations. It is unnecessary for present purposes to be prescriptive about these internal structures, provided that the means by which they are established are themselves the product of multi-stakeholder democratic deliberation.

So in summary the multi-stakeholder bureau of the IGF is to be a balanced group of individuals appointed as representatives of their stakeholder groups, who are to deliberate on its operational programme together, but to exercise a power of veto of its formal recommendations within the stakeholder groups. Its chair, subcommittees, advisors and liaisons are to be appointed by the bureau itself. In the case of the chair, who exercises a special coordinating role, this should be done by voting, subject to the position’s rotation through all of the stakeholder groups and the appointment of two co-chairs to provide continual stakeholder balance.



Though while the IGF remains a UN-affiliated body, this will be on the nomination of the UN Secretary-General.


Such as those of the IGP and MMWG referred to above: IGP, Building an Internet Governance Forum (2006) and MMWG, Internet Governance Forum Input Statement (2006) .


See Section


A prospect made most explicit by Chris Disspain of auDA in May 2007: see Section


This may be compared with the stated reasons for their opposition at Section


See Section 3.1.2.


NTIA, Management of Internet Names and Addresses (1998)


See Section


Examples of this are given by the gTLD-MoU, over which the NTIA rode roughshod, and that of its successor the IFWP, which similarly at the hands of IANA and NSI “was ultimately bypassed and superseded by more powerful forces impatient with the transaction costs of an open, democratic process:” Mueller, Milton, Ruling the Root: Internet Governance and the Taming of Cyberspace (2002), 5.


Cukier, Kenneth N, Slouching Towards Geneva: Ten Unappreciated Axioms of Internet Governance (2005), 4


See Section

The proposal of ENSTA and EUROLINC put forward in May 2007 for the formation of a “four components bureau” for the IGF is close to the consociational ideal: see EUROLINC, Thoughts for Rio: a Bureau for the IGF (2007) (and compare the proposal for three WSIS-style bureaus made by the G77 and China in February 2006: G77, G77 & China Paper on the Proposed Internet Governance Forum (2006) ). Although the proposed four components bureau would only have a procedural mandate, it would be necessary for consensus to be reached not only within, but also between its four components, with much the same effect as granting them a power of mutual veto: Muguet, Francis, A Legal Analysis of the Internet Governance Forum Process (2007), 20. The main distinction between these proposals and a consociational bureau is that the division of power between the stakeholder groups proposed here would not create separate bureaus, which defeats the essential purpose of the bureau as a forum for multi-stakeholder deliberation.


See Section 6.4.1.


See Section


ECOSOC, Information and Communications Technology (ICT) Task Force (2002), 6


Martens, Jens, Multistakeholder Partnerships: Future Models of Multilateralism? (2007), 29 and see at 39 some other methods used by other UN-affiliated multi-stakeholder networks to nominate civil society and private sector representatives.


See Section 5.4.3.


See Section


IETF, IAB and IESG Selection, Confirmation, and Recall Process: Operation of the Nominating and Recall Committees (2004)


ICANN, Bylaws (2006), article VI, section 5.


FSC, By-laws (2005), Articles 50–53.


See Section


The exact number should be determined in open consultation, balancing need for diversity against the fact that deliberation is easier in smaller groups: see Section


Including preparing briefing documents, meeting agenda and synthesis papers and reports, and reducing the consensus of the plenary body on substantive matters into draft recommendations, and on procedural matters into proposals for reform of the IGF’s structures and processes.


See Section 6.4.1.


Muguet, Francis, A Legal Analysis of the Internet Governance Forum Process (2007), 19


See Section


Government of Canada, Canadian Comment on First IGF (2007), 2


ISOC, ISOC Statement, Internet Governance Forum Meeting 17 February 2006 (2006), 1. See also Government of Australia, Programme Committee—Internet Governance Forum (2006) , 2.


Such as Telstra’s Consumer Consultative Council (TCCC) and the former Consumer Advisory Council (CAC) of what is now Communications Alliance.


Rotation between geographical regions, although not required as a matter of principle, could also be justified on substantive democratic grounds if it were to meet with the consensus of an open consultation meeting.


Accountability will be discussed further at Section