CCWG-Accountability 2nd Draft Proposal

[John C Klensin <klensin+icann@xxxxxxx>]

Hi.

I've started several times to try to make detailed comments on
this proposal, but have concluded that they obscure what, from a
broader perspective, is the most important issue.   While some
of those more detailed observations may be deduced from my
comments on the CWG proposal, most notably the concern that this
proposal will narrow the range of communities and stakeholders
who can actually influence or control ICANN, I am going to focus
this note on that one critical issue.

It seems to me that there is a meta-question underlying the NTIA
request for a proposal for a transition in US Government
responsibilities and authority over IANA.  That question is
whether the basic design and organizational structure of ICANN
is adequate to represent and balance the needs and perspectives
of the global multistakeholder community (or, insofar as they
are different, the multistakeholder community concerned with the
Internet) and to be adequately accountable to those communities.
If the answer to those questions are "yes", then, while some
fine-tuning of ICANN structures may be needed, going ahead with
a transition makes sense.   That is especially reasonable if the
needed tuning is directly and obviously linked to the reduced US
Government role.

However, if the answer is actually "no", then an old engineering
metaphor, one that used to be common in the IETF, becomes
relevant: changing the engines, especially all of the engines,
of an aircraft that is already in flight is not a good idea,
especially if that is avoidable.  Indeed, given the risks, the
only possible justification for such a change is that there is
no alternative, e.g., if the engines are changed, there is risk
of a catastrophic crash but, if they are not, such a crash is
certain.

A review of the CCWG proposal's elements, notably those that
change the fundamental accountability structures within ICANN,
the scope of Board authority, ways that the Board can be
removed, and ways that decisions can be controlled or
overridden, suggests strongly that CCWG has concluded that the
present structure of ICANN not adequate to support a transition
_and_ that it cannot be trusted to develop and adopt reasonable
reforms without US Government pressure and supervision.

If that is really CCWG's conclusion, the broader community
agrees, and we still consider stability and predictability of
those Internet operational and administrative elements that
depend on ICANN and/or IANA to be primary goals, then it seems
to me that the only appropriate option is to suspend IANA
transition discussions, refine and implement the CCWG
recommendations, examine how those changes work in practice and
adjust as needed, and to do so with protections afforded by US
Government oversight.    It would be appropriate to return to
the question of transitioning the US Government role out of IANA
(and/or other elements of the system) only after that process
completes and the changes are judged stable, reliable, and
trustworthy by the multistakeholder community.

Alternately, we have an ICANN system today that is reasonably
well understood.  Even its critics (both generally and of
specific clusters of decisions) believe that it mostly works to
an adequate approximation and can agree that nothing that has
occurred in the last decade and a half has caused the Internet
to melt down (whether due to ICANN decisions or not).  If that
system can be trusted sufficiently to believe that later reforms
are likely to be possible and implemented, then the CCWG plan
should be set aside or at least reduced to those elements that
are clearly and obviously critically necessary to an IANA
transition.  We should get through the transition, let that
stabilize and make whatever corrections are demonstrated by
experience to be needed and then organize a second-generation
"evolution and reform" process to consider whatever other
changes are needed (presumably considering the present CCWG
draft as important input).

Doing both the IANA transition and a major ICANN organizational
and accountability reform at the same time strikes me as
fundamentally unwise and a risk to Internet administrative
stability, akin to that change of all of the engines of an
aircraft in flight.    Unless CCWG is able to make a plausible
claim of omniscience and perfect foresight, no combination of
stress testing mechanisms are going to be an adequate substitute
for either "IANA transition first, evaluation and corrections,
then major structural and accountability reforms" or "structural
and accountability reforms first, evaluation and corrections,
then IANA transition" for the same reasons that laboratory tests
are never a completely adequate substitute for deployment and
evaluation of a system under field conditions and at scale.   In
that regard, even if we believe the ST-WG mechanisms are
completely adequate for the contingencies they have identified,
the contingencies they have not been able to identify remain a
major concern... and no completeness proof has been offered or
is likely to be feasible.

regards,
  John C KLENSIN