ICANN/DNSO
DNSO Mailling lists archives

[ga-full]


<<< Chronological Index >>>    <<< Thread Index >>>

Re: [ga] Re: iCANN's protection


On Wed, Apr 18, 2001 at 05:15:27PM +0200, Alexander Svensson wrote:
> Hello Kent!
> 
> From memory, I think ICANN was called "a totally new
> institutional animal" by the Economist, so I agree
> that it is useful to compare it to various existing 
> institutions and to see what we learn from it:

It's good to think "outside the box" on occasion :-)

> Kent Crispin wrote on 17.04.01, 16:46:37:
> > I'm sorry, but I think your analysis of ICANN-as-monopoly is *very* far
> > from describing ICANN's real behavior or its real nature or even it's
> > possible nature.  I think a far better model for ICANN would be a
> > professional society, or a industry standards body.  I'm thinking of
> > organizations like the American Bar Association, or the American Medical
> > Association, or the American Institute of Certified Public Accountants,
> > or the Underwriters Laboratories (Sorry these are all american
> > institutions -- just the ones I'm familiar with).  
> >
> > These kind of
> > organizations have quasi regulatory powers; they are frequently
> > recognized by law; they are monopolies.  However, NONE of them have an
> > "at-large membership" that makes them "accountable" to the public at
> > large.  Instead, they rely on the fact that the entities involved
> > compete against each other, but have a common interest in standards.
> 
> Hmmm. It is a bit difficult to use U.S. examples,

Please forgive my provincialism.  I don't know similar examples 
elsewhere, though I'm sure they must exist.

> because
> frankly I don't know all that much about e.g. the American 
> Institute of Certified Public Accountants (AICPA), but from what
> I gather on the web, Certified Public Accountants enjoy a
> government-protected monopoly of providing certain accounting
> jobs.

Well, they are certified; they get licenses, in the same way that a
doctor or lawyer is certified and gets a license.  While you could call
that kind of certification/licensing  "a government-protected 
monopoly", it's a fairly weak monopoly.  There is a fairly stiff test 
you have to take to be certified; anyone can stand for the test.  I'm 
not at all sure about the processes in detail.  At a very generic 
level, the other examples I mentioned (the american medical 
association, the american bar association) share this characteristic.  
The government really isn't the best agent to define the certification 
requirements.

It is certainly true that organizations like this have a pervasive
effect on the ecomomics of their respective professions.  The american 
medical association has been criticized for artificially restricting 
the supply of doctors by setting the standards for medical schools too 
high.  It is also true that organizations like this can be intertwined 
with the government in interesting ways -- it was news recently when 
the Bush administration let it be known that the American Bar 
Association would no longer be consulted on candidates for federal 
judges.

The basic point is that the government grants to these trade 
organizations certain "powers" that look much like regulation.

A similar model is one provided by standards bodies.  Imagine that
Cisco, Juniper Networks, and several other Internet infrastructure
companies got together and produced a bunch of proprietary standards 
that they cross-licensed to each other.  Anti-trust authorities would 
be very suspicious.  But if Cisco et al participate in the IETF, where 
the standards are open, then it is cool.

> If AICPA decides do drop their professional standards
> yet at the same time doesn't want non-CPAs to do certain jobs,
> those others would seek protection against AICPA overstepping 
> their role from (a) the U.S. legislator, (b) the U.S. courts.

Through anti-competition authorities.

> (I would assume that this also roughly applies to the other 
> professional organizations you mention.)
> 
> Problem regarding ICANN: We don't have a world legislator 
> or a world court,

Fortunately, none is needed.  The rational for government oversight of
trade organizations is concern about competition; trade organizations
are basically controlled through anti-trust authorities.  And ICANN
falls under the anti-trust jurisdiction of every country in the world in
which it has a significant impact on competition.  That is, for all
practical purposes, ICANN is under the thumb of every anti-trust
department in the world.  The idea that there is no oversight over
ICANN, that ICANN is "unaccountable", is a pernicious myth.  The US 
dept of justice, the EC's DG-4, and other anti-trust agencies all 
provide oversight.

> and we probably both agree that it would not 
> be a good idea to leave the decisions on the global Net to a 
> single national legislator/court. Some may argue that the 
> closest we get to a world legislator is an assembly of all
> legislators in the world, but I have seen little support
> for the idea that the GAC becomes a supervisory board...
> So how else do we keep such a public check? E.g. by electing
> representatives to the Board.

Election of representatives to   the board is not necessary for oversight,
and there are tremendous practical difficulties with the "ICANN as
government" model.

Also, in terms of fundamental legal and control relationships, ICANN is
much closer to a trade organization than it is to any kind of
government.  *All* it's contractual committments are with the "trade",
not the public at large.  Moreover, the sheer mechanics of contracts and
negotiation give a great deal of direct and immediate influence to the
contracting parties -- contracts, even contracts with governments, are
never negotiated publically.

More important, though, is that it is clear that the "elected
representative" model simply won't work.  This was demonstrated with
crystal clarity in the at-large elections -- they demonstrated without
question that capture on that scale was possible, and they also
demonstrated without question that the elected board positions became
purely political.  Both of these problems are death for any idea
ofICANN as technical coordinating body.

> I assume the common interest you would refer to when talking 
> about ICANN is a stable Internet. What about the other White 
> Paper principles: competition; private, bottom-up coordination;
> representation? What do we do if they are in conflict, and
> who makes the choice?

The ICANN board.  If they make the wrong decision, DOJ and DG-4 haul 
them into court.

> > The decisions made by these bodies most certainly have an impact on the
> > general public, but the general public doesn't have direct
> > representation.
>
> Because there are other mechanisms in place and people are
> reasonably happy with them. This does not seem to apply here.

No, I would argue that this applies here, as well.  The mechanisms are
certainly there.  And indeed there are unhappy people, but there are
people unhappy with trade associations, as well, and on the whole, in
either case, most people simply don't care.  

Moreover, ICANN is still in its formative stages.  It is causing lots 
of change, and there are fundamental policy choices that do have to be 
made -- this is an inevitable consequence of birth; that is why the USG 
keeps its hand in -- consider the analogy of the *beginning* of the 
Board of Accountancy:  there would have been unavoidable serious policy 
decisions, and they would impact the practices of current accountants.  
There would be lots of controversy, lots of angry people.  But after 
the fundamental policy decisions are made, things change dramatically.  
The Board of Accountancy is utterly boring to most people nowadays.

-- 
Kent Crispin                               "Be good, and you will be
kent@songbird.com                           lonesome." -- Mark Twain
--
This message was passed to you via the ga-full@dnso.org list.
Send mail to majordomo@dnso.org to unsubscribe
("unsubscribe ga-full" in the body of the message).
Archives at http://www.dnso.org/archives.html



<<< Chronological Index >>>    <<< Thread Index >>>