[discuss] NTIA "Oversight" (was Dear ICANN - Feedback)

Milton L Mueller mueller at syr.edu
Wed Apr 16 22:48:01 UTC 2014


-----Original Message-----
> But I don't agree that oversight "does not need formal enforcement mechanisms".

I don't either, it does. I was shocked by Mike's casual dismissal of the need for external accountability.  This is one reason why I hate the term "oversight" in the context of Multistakeholderism, it seems to imply that some powerless committee reviewing your decisions ex post is enough. We've seen time and again that it is not.

> For example, an oversight mechanism (with a contractual enforcement mechanism) 
> might be charged with assuring that ICANN can make binding rules only when 
> supported by consensus and only on topics related to preserving the stable and 
> secure operation of the net. Such a standard would not enable the "oversight" 
> mechanism to second guess specific decisions or icann operations.

In fact, all accountability really amounts to the power of someone to second-guess or override a decision. Think of an electorate "throwing the bums out" or a higher court overturning a lower court for violating a right, or someone being fired for bad performance. Effective accountability ensures that a decision maker tries very hard to make sure that there will be no need to attempt to second-guess. Obviously, the infinite regress problem that Mike flags is real, but only if your accountability mechanisms are poorly designed.

> It would not rely on embarrassment (e.g., investigative reporting) but 
> something like a global rule of law.

Yes.

> What Brazil could achieve is a clear statement that there is a real global internet 
> polity, that it wants to rely on multi-stakeholder policy development to make 
> binding rules to which even states should defer, and that the institutions 
> charged with that political function will establish rule of law oversight 
> mechanisms that constrain potential abuses of the powers granted by that polity.

The current document we are working with is, unfortunately, pretty far away from such an insightful and visionary approach. Perhaps we can move it closer 




More information about the discuss mailing list