DP:  Of course I completely agree:  the limitation regarding consensus policies has done absolutely nothing to curb Board power.  But that's not because it "means absolutely nothing" - I have a pretty good idea what it means, I think you have a pretty good idea what it means, and I think most people in the community have a pretty good idea what it means.

MM: :) Sorry, can't agree - after 18 years in the ICANN I have less and less an idea what "consensus" means. The term has been abused so frequently in this context I wince every time I hear it. I know what it used to mean - no objection from any engaged party (Quaker consensus). But neither ICANN nor IETF nor anyone except perhaps Quaker meetings actually operate that way. So then we're down to "rough consensus" or what I prefer to call "declared consensus" which means in essence that some Working Group chair gets to decide which people/opinions he or she is going to ignore, with no precise specifications of numbers or criteria. Which makes for a very weak basis for legalistic appeals before an IRP, imho.

On the GNSO Council they have avoided this problem by defining various supermajority levels as "consensus" but then we're just voting really. So do you mean there are precise vote levels that have to be obtained by a policy, and those can be used to appeal if they are ignored? Or do you mean reliance on "declared consensus?"

DP:  But surely the primary vehicle for "overturning GAC advice" should be the IRP - that is, if the Board adopts a policy on the basis of GAC advice, or GAC + ALAC advice, why wouldn't the (newly constituted) IRP - if it's really doing its job - say:  "you can't do that, because you're only empowered to implement consensus policy, and you don't have a consensus here?"  You seem to have given up on that, and I don't understand why.

MM: I would think that a much stronger basis for challenging a board/GAC decision would be that it is outside the boundaries drawn by the mission statement and core values. One could claim "by making this decision ICANN is using its authority over DNS to regulate content." And the IRP panel could decide whether that was a true claim or not. I suppose that an IRP panel could also determine whether there was consensus or not at the time a policy was passed, but again I find any definition of consensus that does not rely on precise quantities to be too easily gamed.

DP:  I get it that there are lots of governments out there that want to dictate policy here, and that we're trying to make sure that doesn't happen.  I think an ENFORCEABLE requirement that the Board can only implement consensus policies relating to the security/stability of the DNS is going to be the only real protection against it happening, regardless of the voting allocation formulas and/or the provisions regarding GAC advice.

MM: What if there _is_ a rough consensus among the active constellation of actors in the ICANN regime that it should do something that is outside its mandate? There is, e.g., often consensus and/or a democratic majority in the U.S. that an unpopular speaker's freedom of expression rights should be suppressed.