4.2.5. Anarchistic–hierarchical Internet governance

What has been described is essentially a hybrid of anarchistic and hierarchical governance; but the distribution of power in this hybrid is much more even than in the case of co-regulation described previously. It is closer to the kind of voluntary association that anarchist Colin Ward describes in stating

that there are at least two kinds of organisation. There is the kind which is forced on you, the kind which is run from above, and there is the kind which is run from below, which can’t force you to do anything, and which you are free to join or free to leave alone. We could say that the anarchists are people who want to transform all kinds of human organisation into the kind of purely voluntary association where people can pull out and start one of their own if they don’t like it.[1]

How, if at all, is this insight applicable to Internet governance?

To answer this will require the implications of the “open source-style” hybrid structure to be isolated; specifically, the assumptions that must be satisfied to ensure that the stakeholders in a governance network that is structured in hierarchical form are not at risk of being oppressed.[2] These assumptions can be reduced to three:

If these assumptions sound familiar, it is because the first two are amongst those that underlie the ideal of the perfect free market (along with additional assumptions not needed here, as we do not require that the governance network be effective, only that it not be oppressive),[4] and the third is one of those underlying deliberative democracy, to be discussed in the next section.[5] With these three criteria satisfied,

[i]t does not matter whether online discussion groups or even entire networks of such groups are internally autocratic, since individuals can always choose “their own more congenial online homes.” Cyberanarchists, then, see cyberspace as a market of alternative rule regimes. It is the ease of exit and the abundance of alternatives—in essence consumer choice in conditions approaching perfect competition—that bring to fruition the liberal ideals of liberty and consent.[6]

The next question then is the extent to which these assumptions can be fulfilled in the case of a governance network as they are in the case of open source software development. Examining each of them in turn:

An example will put these observations in more concrete terms. Let us assume that the IGF has an hierarchical leadership, which drafts a code governing the issue of Internet interconnection costs. This code is entirely satisfactory to all of the other stakeholders, except for the private sector who claim that interconnection costs should continue to be left to the free market (an alternative mechanism of governance, which it is costless for them to substitute for that of the IGF).

Are the three criteria satisfied? Yes, there is a perfect (or at least a costless) substitute for the code of the IGF. Yes, there is freedom of exit from the IGF so that even if its hierarchical leadership required all IGF members to subscribe to the code, the private sector would be at liberty simply to withdraw from the network. And as to whether the private sector could be coerced to accept the code, regardless of its departure from the IGF—well, this depends on whether the authority behind the IGF is governmental or not. If it is, then it can ignore the private sector’s concerns and pass the code into international or domestic law regardless, which defeats the very purpose of developing it through a governance network.

Let us change the scenario a little. In this case, the hierarchical leadership of the IGF has managed to address the concerns of the private sector in a new code on interconnection costs that is now acceptable to all. A few private sector stakeholders however decide to opt out of the IGF regime and revert to reliance on the market to set interconnection prices. They immediately find that the success of the IGF’s code has permanently lowered market prices for interconnection, and that the costs of differentiating their service so that it can be sold at higher prices are insurmountable. The hierarchical leadership of the IGF (if not composed of governments) not only cannot coerce these private sector stakeholders into accepting the code, but it does not need to. The IGF’s very success has made it self-governing.

Granted, less extreme examples could be given in which the applicable transaction costs would vary markedly from these. But the lesson from open source software remains that if conditions are right, the question of how to impose hierarchical ordering on a governance network (for example by attempting to select a meritocracy through objective or consensual means), becomes redundant. If the hierarchical leadership, however selected, does not act in the best interests of the IGF, then its output will be ignored (the more so, the greater the segment whose interests are disregarded) and it will become powerless. If it does act in the best interests of the IGF, then its power will grow.

It can best do this by engaging all stakeholders in the process of public policy development using a participatory process, much as open source developers collaboratively work on open source software projects. In doing this, the stakeholders will become empowered, the social capital of the governance network will increase, and its effectiveness will grow.



Ward, Colin, Anarchism as a Theory of Organization (1965), 171


That is, being subjected to hierarchical power that has not passed the normative tests of either objective merit or consent.


Or in anarchistic terms, other than by the “natural and legitimate influence” of other stakeholders: Bakunin, Mikhail A, God and the State (1970), 35.


See Section 1.4.3.


See Section


Netanel, Neil W, Cyberspace Self-Governance: A Skeptical View from Liberal Democratic Theory (2000), 404–405