4.4.1. Consensus between stakeholder groups

We will therefore now consider the possibility of requiring that consensus be reached between the stakeholder groups on any proposal that has been democratically agreed within and across those groups, thereby essentially giving each stakeholder group the power to veto it. Since multi-stakeholder governance networks are inherently consensual structures anyway, and possess only soft power with which to enforce their decisions, the recognition of a formal requirement of consensus amongst stakeholder groups is also a more natural fit for this mechanism of governance than democratic voting.

This could address the problem of government unilateralism in two ways. First, it would provide an alternative to the use of the effective power of veto that governments already possess in many issue areas. Instead of denying the competance of the governance network to make even non-binding decisions, and overriding its authority through the use of the coercive force of law, it will be possible for governments to formally veto any proposal the network makes without thereby undermining its authority as a forum for ongoing collaborative policy development.

Secondly, and alternatively, even if governments do continue to deny the independent authority of the governance network to develop soft law and purport to relegate it to an advisory role (which seems more likely, in the IGF’s case),[1] the formal requirement that consensus be reached between all stakeholder groups would institutionalise a power of veto for the other stakeholders, that would go some way towards equalising their power with that of governments. The analogy of the “KILL SYSTEM” command of the ITS operating system, referred to at Section is an apt one, in that giving all equal power to undermine the governance process stimulates the development of norms to regulate the use of that power. Consociationalism

Consociationalism is the theory of a form of organisation designed to institutionalise the reservation of power to distinct stakeholder groups within a consensual decision-making forum. First and most famously described by Arend Lijphart,[2] it describes an ideal form of what may more broadly be called consensus democracy, which includes various other forms of democratic governance characterised by the sharing of power between stakeholder groups at the executive level.[3]

In its ideal type as identified by Lijphart, a consociation exhibits four characteristics:

As the above characteristics may suggest, consociation has been most commonly studied as a form of government, in which the various stakeholder groups of a pluralistic community are divided along territorial, ethnic, linguistic, religious or cultural lines. It is designed to reduce conflict by preserving the independence of groups entitled to be represented in government but which may not be willing to cede control to a majoritarian democracy.

Consociationalism can also be applied to other fora of governance, including multi-stakeholder governance networks.[5] However, some adaptation of Lijphart’s ideal type would be required for it to provide an appropriate model of a consociational network for the Internet governance regime. In particular, since the relevant stakeholders would be governments, the private sector and civil society, it would be difficult to provide for proportional representation within the grand coalition, because the size of each stakeholder group’s membership is not easily commensurable.[6]

There is however an alternative method of ensuring proportional representation, and thus also fairly institutionalising the power of mutual veto, without restricting the composition of the grand coalition. This is for an executive council of the governance network to be formed, to which each stakeholder group would elect an equal number of representatives. This executive council would be required to ratify all decisions of the grand coalition by consensus. If consensus could not be achieved, the objecting stakeholder group or groups would effectively thus exercise their power of veto.

Although this could be seen as elitist for removing direct power from the grand coalition to a meritocracy, the advantage of doing so is that it would enable the grand coalition to deliberate more freely and less strategically, knowing that ultimate political power lay at a higher level,[7] whilst at that higher level, formal decisions could be made more quickly.[8] By analogy, the grand coalition would become the public sphere of the Habermasian model, and the executive council its parliament.

Another possible objection to the above model of a consociational Internet governance network is that it is not deliberatively democratic, at any of three levels: within the grand coalition where decisions are made, within the executive council where they are ratified, or in the process by which the executive council is elected.

This objection is however quite easily answered, as there is no reason why Lijphart’s ideal type could not be adapted to incorporate deliberative democratic principles in the process by which decisions are made by the grand coalition. It has even been said that deliberative democratic principles are already “embedded in the very structure of consociational democracies,” since the process of forging consensus between stakeholder groups is naturally much closer to that of democratic deliberation than the process of majoritarian rule in a representative democratic parliament.[9]

The consensus of the executive council on whether to ratify a proposal could also be facilitated by democratic deliberation, though it is less important that it should be, since the executive councillors would already have participated in the process of deliberation within the grand coalition, and the role of the executive council is a more limited one simply designed to institutionalise the stakeholder groups’ right of mutual veto. As for the process by which the councillors are elected, it would be cumbersome to attempt to do so by any other method than voting, however it would be possible for a multi-stakeholder nominating committee to deliberate upon a shortlist of candidates to be presented to the grand coalition for election.

This model of a consociational governance network has much in common with the variant of consociationalism found in the European Union. The European Parliament, containing multiple directly-elected MEPs for each country in rough proportion to their size, is the equivalent of the grand coalition. The Council of the European Union, containing only one minister from each country, equates to the executive council. To pass an EU law using the co-decision procedure, the Council is required to reach agreement on the proposed text with the Parliament, much like the process of ratification proposed here. Depending on the issue area in question, it is required to do so either by unanimous consensus, or by the “rough consensus” of qualified majority voting.[10]

Thus the option of consociation for a multi-stakeholder governance network, although relatively novel, is certainly not untested; the ILO with its sharing of power between governments, employers and workers provides another example,[11] and yet another is found in the Forest Stewardship Council (FSC); a non-governmental forum responsible for the development of standards for sustainable forestry.[12] It also accords with the conclusion of the discussion of hierarchical organisational forms at Section 4.2.6, that a meritocracy established by a democratic or consensual process could be the most effective and legitimate structure for the organisation of a governance network, if only an appropriate such process by which for the meritocracy to be established and supervised could be found.

Such a process having been found in deliberative democracy, the resulting meritocratic/democratic/consensual hybrid—the consociational governance network—may be the fulfilment of that earlier prediction. Better yet, it also offers a solution to the problem of government unilateralism that deliberative democracy alone cannot provide.



See Section 6.2.


Lijphart, Arend, Consociational Democracy (1969)


Lijphart, Arend, Patterns of Democracy: Government Forms and Performance in Thirty-Six Countries (1999)


Lijphart, Arend, Democracy in Plural Societies: A Comparative Exploration (1977)


Skelcher, Chris, Jurisdictional Integrity, Polycentrism, and the Design of Democratic Governance (2005)


See Section


Dryzek, John S, Deliberative Democracy in Divided Societies: Alternatives to Agonism and Analgesia (2005)


See Section 4.2.2.


Steiner, Jürg, Bächtiger, André, & Spörndli, Markus, The Consociational Theory and Deliberative Politics. A Conceptual Framework for a Cross-National Analysis (2002), 77


See Section


See Section


Although the FSC (see http://www.fsc.org/) does not conform to Lijphart’s ideal type of consociationalism, it follows similar principles that reinforce the equality of all stakeholder groups and the mutuality of their endeavour (see generally Lipschutz, Ronnie D & Fogel, Cathleen, “Regulation for the Rest of Us?" Global Civil Society and the Privatization of Transnational Regulation (2002), 136 and Schmidt, Eleonore, The Forest Stewardship Council: Using the Market to Promote Responsible Forestry (1998) ).

Its stakeholders are categorised into three chambers according to their predominant interest: Social, Environmental, or Economic. Each chamber is then further sub-divided into North and South for members from the developed and developing worlds respectively. These meet together once every three years as the General Assembly, where any motions for new or amended statutes, by-laws, principles or policies of the FSC are debated and voted upon.

Votes are weighted to give each chamber an equal third of the vote regardless of the number of members it has, and similarly within each chamber’s voting bloc, half is allocated to each of North and South. In order to pass, a motion must achieve consensus, which “is defined as the absence of sustained opposition but does not require unanimity”: FSC, By-laws (2005), Article 15.

The Board of Directors of the FSC contains three members from each chamber, including at least one from each sub-chamber, who are elected by the General Assembly by postal ballot for three year terms. These are chosen from a slate of candidates put forward by a nominating committee headed and appointed by the Chair of the Board, which is in turn made up of one member from each chamber, including at least one each from the North and the South.