4.2.2. Oligarchy

The principal advantage of oligarchical rule, in which power is restricted to a small defined group, is that by reducing the number of participants, decisions can be made more quickly and with less contention amongst the decision-makers than in a system, such as democracy, with a broader power base. These benefits of oligarchy accrue at the cost of it being less representative and accountable to the governed than democracy, and therefore offering them less protection against tyranny.

However, the force of this objection is lessened by two observations. First, there is no transnational demos to which the interests of all three stakeholder groups can be traced. Therefore, at least until such a polity is brought into being, democratic representation is not an option, leaving oligarchy as perhaps the only practical alternative.[1]

Second, there is empirical evidence that oligarchy may be the natural state of any organisation, no matter how it is initially structured; whether as a bureaucracy, democracy or anarchy. This is Michels’ “iron law of oligarchy”: that in any small group there is a tendency for power to be concentrated in the hands of an elite who have both the will and the means to organise others.[2] ICANN as an oligarchy

This latter point is evident in the context of Internet governance. Because ICANN for example, whilst claiming to operate upon consensual principles, has in practice been driven largely by its professional staff and select corporate and government stakeholders, it has been described as an “authoritarian–pluralist” institution.[3] Former ICANN board member Karl Auerbach explains, “ICANN is an oligarchy. ICANN claims it’s a private organization yet it claims immunity from things like antitrust because it derives its powers via contracts with the government.”[4] Weinberg agrees that “ICANN demonstrates Aristotle’s fear of the degeneration of aristocracy into oligarchy.”[5]

One of the factors to which the ICANN board’s unilateralism in policy formation is attributed is that there has been widespread disengagement from and apparent apathy towards many of the issues in question by the Internet community. However this may be explained in part by the fact that the constituencies for community participation in ICANN’s policy processes were established in a top–down fashion,[6] that the avenues for the receipt of input have been unduly limited,[7] and indeed that the community’s input, when provided, has been largely disregarded.[8]

In any case, if Michels’ Law holds, then attempting to reform ICANN to redress its perceived lack of public accountability is a misguided endeavour. An alternative, putting idealism aside, is to recognise ICANN for the oligarchy that it perhaps inevitably is, and allow it to make policy according to its own best judgment, without being distracted by the need for token efforts at public engagement. After all, nobody (not even ICANN) ever claimed that the organisation was a democracy. To the extent that its board exercises a representative function at all, it is more in the nature of guardianship than agency. Franda writes:

For business people generally, the idea of a board member’s representation is not the public representational function of someone duly authorized by an election or other legitimizing process to speak for a large constituency. Rather, it is the idea that someone will know and understand a specific business interest and be able to speak for that interest in forums where such interests are being challenged.[9]

Thus the EFF, for example, was often described as a representative body long before it actually had an open membership, by dint of its demonstrated ability to advocate for the rights and interests of online communities.[10]

On this view, an “autocratic approach might well be most efficient way to structure ICANN in order to carry out its mission.”[11] The IGF as an oligarchy

The same argument applies in principle to the IGF. If power within a governance network such as the IGF would inevitably evolve (or devolve) into an oligarchical form anyway, concentrated in the hands of those who have the initiative to put the most into the process, then perhaps it is quixotic to insist that the IGF should ever pretend to anything other than an oligarchical structure. If those whose power was most critical to the IGF’s establishment were those who would lead it, the IGF would perhaps be an oligarchy of the United States and European Union, the ICC, ISOC and the ITU for each of the respective stakeholder groups. No doubt they would do at least as good a job of policy development between themselves as ICANN has in its domain.[12]

But there are two problems, one procedural and the other normative. The procedural problem is that we still have no answer as to how the governance network is to operate, save that its composition is to be limited to a small number of powerful members. How are those participants to resolve disagreements between themselves? The specification of the oligarchical form alone provides no answer. Perhaps the power of certain large stakeholders will exceed that of others. For example the ICC had much less to do with the final form of the Tunis Agenda than did the other three stakeholder representatives nominated in the previous paragraph. What would prevent those other representatives from vetoing the power of the ICC within the IGF oligarchy, and thus effectively excluding the private sector as a stakeholder altogether?

The second and more fundamental problem is that there are no normative criteria implicit in the oligarchical form by which to determine who is to be privileged with the status of membership. Indeed, this question is entirely exogenous to the oligarchical structure, which simply entrenches existing power relations. Whilst it may be true that organisations will tend towards oligarchy in any case, it is fatalistic simply to allow that “those who will rule, will rule” without questioning how it came about that they should do so. Would those who might form an oligarchical governance network be the most public spirited amongst the stakeholder representatives, or simply the most politically and economically powerful?

Even the bureaucratic organisation, whilst already rejected as unsuitable for a governance network, provides normative guidelines for the progression of suitable candidates up the hierarchy of authority, by requiring that promotion should be on the grounds of merit, rather than on other grounds such as nepotism or political influence. But can the concept of merit be applied to a governance network, and if so, how would it fall to be assessed? These are questions central to another form of hierarchical governance: meritocracy.



Though this question will be revisited at Section 4.3.


Michels, Robert, Political Parties: A Sociological Study of the Oligarchical Tendencies of Modern Democracy (1962)


Palfrey Jr, John G, The End of the Experiment: How ICANN’s Foray into Global Internet Democracy Failed (2004), 40, 45


Koman, Richard, Karl Auerbach: ICANN “Out of Control" (2002)


Weinberg, Jonathan, Geeks and Greeks (2001), 327


See Section


For example, a 2007 request for public comments on ICANN’s performance was held open for less than one month: see http://www.icann.org/announcements/announcement-08may07.htm.


Palfrey Jr, John G, The End of the Experiment: How ICANN’s Foray into Global Internet Democracy Failed (2004), 34, 50–52


Franda, Marcus F, Governing the Internet: The Emergence of an International Regime (2001), 67. To similar effect is the proposition discussed within the IGF’s Advisory Group that its “members should be chosen on the basis of how large and diverse a community they connect to (which is different than ‘represent’)”: IGF Secretariat, Advisory Group Discussion 6 December 2007 to 15 January 2008 (2008) , 3.


Godwin, M, The Electronic Frontier Foundation and Virtual Communities (1991)


Palfrey Jr, John G, The End of the Experiment: How ICANN’s Foray into Global Internet Democracy Failed (2004), 57


The IGF’s stakeholders have indeed divided into a small number of factions, but not quite along the lines suggested here: see Section