[council] Response to Pelage Audit

["Harold J. Feld" <hfeld@mediaaccess.org>]

I understand from ICANNWatch that Michael Palage conducted an audit of 
the "representativeness" of the various DNSO consticuencies.  I have 
posted the following response, which I share here.

I believe that the basic premise of the Palage audit is flawed.  It 
proceeds from a premise that for a consticuency to have value and be 
"representative," it must *contain* a substantial numerical 
representation and cross section of the effected community, not merely 
have the *potential* for any concerned member to join and participate 
equally.  This fundamentally misconceives the only way in which the 
ICANN consticuency system can function.

While I do not mean to cast aspersions on Michael Palage's efforts, I 
would observe that this report has the effect of maximizing the 
legitimacey of the Registrar consticuency and its most closely aligned 
consticuencies (gTLD and ccTLD registries), while deligitimizing the 
consticuencies with the least confluence of common interests (end-users).

Under the assumptions of the audit, only three consticuencies can be 
truly "representative" because the universe of entities is clearly 
identifiable, sophisticated, small enough to be organized, and with a 
sufficient vested interest in ICANN to ensure maximum participation by 
potential members.  (1) gTLD registries; (2) gTLD Registrars; and (c) 
ccTLD registries.

The rest of the consticuencies have issues of "representativeness" 
because they are open ended places for general alignments of interest. 
How many busniesses are in the business consticuency as compared to the 
total number of businesses in the world?  How about IP organizations or 
practitioners? Or ISPs?  And, of course, the one consticuency that 
ALWAYS gets blasted for this, Non-coms.

So this has the effect and appearance, bluntly, of an attempt by the 
registrars to push end users out.  End users will never be 
"representative" by any of these criteria.

The fallacy of this is immediately apparent, since the only other option 
is to simply eliminate the end user voice completely.  In modern policy 
making, we rely on representation- individuals come together in groups 
and the groups advocate.  In addition, in modern nation states, the 
government theoretically acts "in the public interest."

But ICANN doesn't have a public interest mandate.  It is explicitly a 
"consensus building body."  That means that it derives its decisions 
based on "stakeholder participation."  i.e., if you don't show up, you 
don't have a voice in the process and you can't complain later.

This ignores the problem that, as a practical matter, it is simply flat 
out impossible for 99.99% of the people potentially effected by an ICANN 
decision to meaningfully participate.

One alternative is to use the consticuencies as a proxy for the parties 
who should be there.  By this logic, we look to see if the consticuency 
is representative in the sense that it appears to have the right 
collection of interests rather than numerical participation.  This is 
hardly ideal, but the alternative is no voice for these interests at all.

If these interests are dismissed, however, you will end up with a train 
wreck.  This has already happened with other closed orgs,like the RIRs. 
  Anyone remember when the RIRs decided they were going to outlaw 
multihoming IP addresses?  Well, turned out LOTS of people who weren't 
included in those deliberation cared ALOT because they used this 
technique for a number of things that they thought were important and 
the insulated interests around the RIR table didn't.

Happily for the RIRs, they pulled back pretty damned fast and have stuck 
to their knitting every since.  But ICANN is on its way to making many 
such mistakes.  If it moves without a public interest voice because 
those voices are deemed "not sufficiently representative," ICANN will 
find itself walking smack into a moving train without ever knowing what 
happened.

Harold