ICANN constituencies were always a bad idea. It’s time to get rid of them.
Constituencies flout the fine pronouncements by ICANN about transparency and representation, they are highly inefficient, and they serve to ensure that consensus is rarely achieved. The net effect is to concentrate power in the ICANN Board.
Bottom-up decision-making? Not even close. This is what constituencies do:
- Discourage participation — many ICANN constituencies have fees and membership criteria. If you don’t fit the mould, where do you go?
- Flout democratic principles of representation — you know the old slogan: “One constituency, one vote!” Er, right…
- Protect and promote special interests — by definition. That’s what functional constituencies are: special interests.
- Limit participation by new groups. For instance, it’s glaringly obvious that domainers are an important part of the domain industry, and yet they are rarely seen at ICANN. Instead, they are having their own meetings, bypassing ICANN and reducing its appeal.
- Dilute the effect of unanimity — one small undemocratic constituency can cast doubt on consensus achieved by all other parties.
- Promote factionalism — the primary disease of this system. The battle lines are drawn even before an issue is discussed. Fluid coalitions based on a particular issue are difficult because of entrenched longer-term interests. Distrust and dislike are institutionalized.
- Bring decision-making to a near standstill — in a community that values consensus, constituencies are an invitation to be intransigent in hopes of a better deal.
- Prone to capture — when ICANN matters only occasionally (e.g., for the business constituency), a few full-time people can effectively hijack a constituency.
- Go out of date — when was the last time you heard the words “ISP” and “ICANN” in the same sentence? A good argument could be made that domainers are a far more important interest group than ISPs. Why don’t they have a constituency?
The GNSO — the “open to anyone” body of ICANN — is divided into six constituencies (formerly seven). They are based on the power centers at the time of the formation of the GNSO (called the DNSO at that time) in Singapore in February 1999. I was co-chair of that meeting and argued against the formation of constituencies at that time, noting the gridlock of Hong Kong politics, which was (and remains) constituency-based.
In case anyone thinks I’m making an original critique, I’m not. Functional consitituencies (as opposed to regional constituencies, a feature of most democracies) are well-known as a Bad Idea.
Harvey Feldman, career U.S. Foreign Service expert in East Asia and former U.S. Ambassador to Papua-New Guinea, described Hong Kong constituency politics this way:
Only half of the Legislative Council’s members are elected from [regional] constituencies. The other half are chosen by “functional groups” [the equivalent of ICANN's 'functional constituencies'] in a method pioneered by Mussolini back in the 1920s and applied to Hong Kong in its Crown Colony days as an effective means of divide and rule. The functional constituency members assure the SAR government (and China) of their ability to control the Council. [emphasis mine]
Similarly, a recent scholarly book, Functional Constituencies: A Unique Feature of the Hong Kong Legislative Council, has revealing jacket copy:
provides… a full list of functional constituencies and the size and make-up of their electorates including how certain major companies may control large numbers of votes…. Then the impacts of the functional constituency members on social policy-making and on economic policy are examined, and deleterious effects on economic efficiency of its entrenchment of vested interests argued [emphasis mine].
Reasonable people can debate whether this what was intended when we set up this structure within ICANN. I was there, so I can say with some certainty that the less-politically-savvy groups in the audience were attracted to the idea of constituencies because they seemed at the time a fair way to distribute power. Others, less idealistic and naive, probably understood the long-term consequences and were pleased with what they saw.
Many have noted that ICANN’s constituencies give a lot of power to businesses, and exclude civil society interests. Just look at the list:
- Business (= big business)
- Intellectual Property (= big business)
- Internet Service Providers (= business)
- Registries (ICANN-specific business)
- Registrars (ICANN-specific business)
- Non-Commercial (anything else)
The Business and Intellectual Property constituencies serve the same masters, though their inflections are different. The ISP constituency has no more relevance to ICANN policy than, say, DNS providers, but as fossils will, they continue to make an imprint. Registries and registrars have the most excuse for existing, as they are a creation of ICANN itself, and need a forum. Then there is the lone “catch-all” non-commercial constituency, which is known for its collection of kooks, endless and pointless debates about procedure, and the inability to agree on much at all. In the time-honored tradition of dealing with trouble-makers, ICANN has provided this supposed “progressive” constituency with a forum to eviscerate itself, which it effectively has done.
Getting rid of ICANN’s constituencies would break down the special-interest silos that prevent consensus, and would strengthen the voice of the GNSO, which is the only place within ICANN where there is even a promise of participation and representation for all the people ICANN spends so much money doing outreach to. As it is, the ICANN constituencies are anti-democratic, anti-Internet, and anti-quated.
Tags: ICANN, ICANN constituencies, intellectual property constituency, business constituency, ISP constituency, non-commercial constituency, registrar constituency, registry constituency, domainers, functional constituency, Hong Kong constituencies, Harvey Feldman, GNSO, DNSO, ICANN meeting Singapore