We were discussing how governments could take over ICANN. This was in response to a new meme gaining popularity within the Beltway: the idea that we need to retain the Commerce Department’s leash on ICANN (the JPA) because if we don’t, other governments (autocratic Russians, Chinese commie hordes, turbaned Islamo-fascists or languid, dirigiste Europeans) will somehow “interfere” with DNS.
It should be obvious that the future of the JPA has virtually nothing to do with ICANN’s subordination to other governments. It's about ICANN's subordination to the US government. Yes, the relationship between ICANN and the world’s national governments is important — as CDT and others have suggested, we do need to come to grips with whether we are going to “transition” ICANN to non- governmental governance. But to use this as an argument for retaining the JPA is truly one of the silliest, most off-target and logically-inverted distortions of the dialogue around this topic I have ever encountered.
In order to understand how bizarre this argument is we must first note that the JPA is not in any way connected to the IANA contract, which is the real source of ICANN’s authority over the DNS and the real unilateral whip the U.S. holds over ICANN. Whatever happens to the JPA, the US still retains a critical form of oversight authority over ICANN, it just doesn't get to specify particular policies and practices.
In an reflection of the lack of thinking that has gone into this meme, the people advancing it have not specified how this government takeover is supposed to happen, much less how retaining the JPA would prevent it. As I started to argue yesterday, it is hard to conceive of a single plausible scenario for governments to exert more power over ICANN that isn't either a) already happening, and being aided and abetted by the USG; b) more likely to happen the longer the US stays in control or c) require the US government’s consent.
To prove this point, here are some plausible scenarios and mechanisms through which national governments could put pressure on ICANN. By examining these, we can better appreciate just how far off base the CDT and others are.
Scenario 1: The Enemy Within (GAC)
The most important vehicle for governments to exert influence over ICANN is already alive and kicking within ICANN. It is called the Governmental Advisory Committee (GAC). The GAC is a bizarre entity: it is a mini-United Nations in which all of the world’s national governments and intergovernmental organizations are eligible for membership – but it is nominally constructed as an “Advisory Committee” within a private corporation. As an intergovernmental body GAC has the full potential to reproduce the alliances, coalitions and politics of the UN. But unlike a “real” international organization, it is not based on a formal treaty, and its rules and powers were never ratified by any democratically elected legislature. Thus, as a vehicle for arbitrary governmental interference in DNS, GAC is far more dangerous than the UN. Ever since 2002, it has been practically mandatory for ICANN to follow GAC’s “policy advice;” if the Board does not it must jump through some procedural hoops and negotiate with GAC. No other policy making organ within ICANN has that power. Contrary to naïve views about US government’s status as a bulwark against governments, it is a demonstrable fact that the US government has been among the most active and aggressive in insisting that the “private” ICANN take into account and follow the advice of governments via GAC. It was “policy advice” from the GAC that gave us the rule that all top-level domains must be subject to censorship to ensure their conformity with standards of “morality and public order.” The US government was completely in accord with that policy, and may even have helped to draft it. It was “policy advice” from the GAC, driven by the U.S., that effectively overturned a supermajority vote from ICANN’s GNSO that would have introduced some privacy protections into Whois. It is not hard to project a future scenario in which GAC’s power is increased: GAC can start to insist on voting board members instead of mere liaisons, for example. And it is not at all implausible to envision a GAC in which governments hostile to the US begin to actively organize to use its powers to interfere with the ICANN Board.
Simple question for CDT and others supporting the JPA: If you’re concerned about governmental interference in DNS, why aren’t you talking about the GAC?
Scenario two: A binding international convention
Another way governments could exert more control over ICANN would be by formally entering into negotiations about “globally applicable public policy principles” for the critical internet resources administered by ICANN. They could attempt to collectively agree on a set of rules, laws or principles to which ICANN would be subordinate by virtue of its need to comply with recognized international law. But this scenario does not seem so threatening, for a variety of reasons. First, any such treaty would be the exact opposite of arbitrary “governmental interference in DNS.” It would be an attempt by governments to formally and deliberately clarify the rules and principles ICANN has to follow, in a transparent way that commanded worldwide assent. In contrast to GAC “policy advice,” such a treaty or convention would have to be ratified by national legislatures, affording the public significant opportunities to object to any rules that would enable repressive or arbitrary interference. And since it would be a multilateral effort, any government that did not agree with the resultant rules could simply refuse to go along with it. If the US government and the broader US polity did not like the result, it could simply refuse to ratify the convention and since ICANN is located here it would not be bound. Any Internet governance convention that did not include the US would likely be a non-starter to begin with. But the JPA with ICANN affords us no leverage over this situation. If a large number of the world’s governments wanted to start such negotiations they could do so (or not) regardless.
Simple question for CDT and others supporting the JPA: if you’re concerned about arbitrary governmental interference with DNS, why aren’t you calling on the US government to support international negotiations with other government to obtain their support for a binding international convention that limits state interference and ratifies ICANN’s status as an independent entity?
Scenario three: Secession and fragmentation
Renegade or authoritarian governments could attempt to influence or destroy ICANN by seceding from the ICANN root and setting up a competing one. Of course, the tremendous value of universal connectivity imposes huge penalties on such a strategy. But if anyone really wanted to try to do this, they could do so with or without the JPA. Indeed, the overt US dominance symbolized by the continued presence of the JPA can only increase the likelihood of such a scenario. As a matter of fact, China has already done something close to this. It has not fragmented the root but it has pre-empted any attempt to establish a Chinese-character .gongsi by setting up a national alternate root in which those TLDs are occupied. Russia could easily do, or try to do the same. Whether or not ICANN is subjected to Commerce Department nitpicking over its policy objectives and checklists has nothing to do with this scenario.
When it comes to the geopolitics, the JPA is really a sideshow. On its face, ending the JPA eliminates one part of ICANN’s subordination to the only government that has any real authority over it, the USA. So in that respect ending the JPA has to be considered a step away from “governmental interference in DNS.” The real issue is not whether the JPA gives us leverage over bad behavior by other governments. The issue is whether it gives us any real leverage over bad behavior by ICANN.