[council] ICANN institutional structure (long)
Jonathan Cohen wrote:
> If there is such a 'plot'' noone has told me about it .But ,I grant you, its clever! (grin)
I do not believe it is a plot. I believe it is an unintended
consequence of the Board's actions. Economists refer to this as the
theory of rational expectations. When behavoir yields predicatble
results, rational people will shape their behavior accordingly.
A natural consequence of the Board's determination to exercise ultimate
policy authority, and its history of the last two years of exercising
that authority regularly rather than in exceptional circumstance, is
that those who desire to affect policy will focus their efforts on
persuading the Board. Persuading the DNSO is, perhpas, useful, but not
sufficient. Nor is it even necessary, since the Board may over-ride the
determination of the DNSO based on nothing more than the judgment of the
majority of directors.
Furthermore, it effects the flow of discussion below. Since the Board
is the ultimate arbiter, it does not benefit us in the DNSO or the
community at large to fix upon a single recommendation: particularly
when a matter is complex and subject to multiple p.o.v. This is why I
have supported the approach taken by the WLS task force. It provides
the Board, the ultimate arbiters, with a range of options that best
reflect the general feeling of the community on a complex issue. While
it is by no means certain how the DNSO will proceed on this, my own
recommendation is that we discuss the task force report and then send it
to the board with any additions we have.
There are advantages and disadvantages in this model, not least of which
is that it is a fundamentally different model from that proposed by the
White Paper and endorsed in ICANN's bylaws. But it is, IMO, the only
rational response to the last four years of history.
All of us on the DNSO want our work to be useful. We do not want to
waste our time any more than the Board does. My organization is
spending a significant percentage of our telephone budget for me to
participate in these calls. Our colleagues in Australia and Asia are
staying up to God-awful time of the morning to participate. We are all
trying to make it work and not waste our time on things that are
As a result, we need to accomodate ourselves to the fact that, as Vint
Cerf observed in Accra, DNSO recommendations are "advice" and that
"advice is just advice."
> I believe that every Board member wants there to be broad based discussion of all significant
>issues.I know that consensus is what should be the basis for all significant policy decisions
>and the Board wishes that to be the case.Nothing in the new ''world'' changes that.
I do not doubt your sincerity. I think, in no small part because you
are "up there" and we are "down here," that you are not entirely aware
of the consequences of your actions.
Let me take an example. My non-profit and a bunch of others are on a
list that includes some grant officers. One grant officer had the habit
of asking very philosophical and far reaching questions on the list.
Orgs would expend considerable resources trying to answer them and
re-writing their grant proposals to reflect the perceived philosophical
The grant officer was totally unaware of this effect. To the grant
officer, it was just a mailing list and a place to ask interesting
questions. The grant officer was totally oblivious to the fact that
because organizations would live or die by the funding this grant
officer provided, orgs would warp tyheir behaviuor all out of proportion.
Meanwhile, the grant officer's reputation in the community suffered.
People regarded the grant officer as inconsistent and irrational and
straining their organizations to the limit for trivialities. The fact
that the grant officer had never _asked_ for any of this did not change
For good, or ill, the Board has established itself as the center of
decision-making at ICANN. This action has consequences, whatever
individual Board members may wish. Just the Prime Minister of Canada
can't open his mouth without having a dozen tea-leaf readers trying to
figure out the implication for this or that policy, you sit in the same
place. You can chose to ignore this effect as unwarranted and
unjustified, but this will not make it go away.
I have seen the self-censoring that takes place among those who wish to
be effective. This self-censoring has two layers: (a) on the issue in
question, those seeking to have real impact will not propose things that
they perceive (rightly or wrongly) as outside the realm of actions the
Board will consider. And I have seen people not want to "piss off"
Board members or be perceived as geenrally too radical.
And I have seen a number of people, particularly in my consticuency,
completely lose heart and abandon the process all together.
>Hopefully the new format for Support organizations will encourage even broader community
>involvement and the opportunity to channel opinions and discussion into a consensus
>,Or,a clear lack thereof.
I am not saying otherwise. What I am saying is that the locus of
decisionmaking is not at the SO level, with modest oversight from the
Board, but at the Board level. The DNSO's roll is not to identify
consensus and determine name policy. The DNSO roll has become one of
convening the conversation, synthasizing the main threads, and then
reporting up to the Board for ultimate decision.
DNSO recommendations still may carry weight. The conversation is still
usefull. Nor am I suggesting that the Board has a predetermined result
in mind and we are simply going through the motions. To the contrary.
I am trying to realisticly define the roll that we actually play in the
An understanding of how the process actually works is crucial to
meaningful participation at all levels.
To move back to WLS. If the DNSO is actually the arbiter of policy,
with the Board only making an up/down decision and only vetoing in
extrordinary circumstances, then it would be our roll in the process is
to find THE consensus and settle on a fixed course of action. If our
roll is to develop the consensus and ADVISE the Board, then it is our
roll to chronicle the broad range of feeling in the community and
present the Board with a set of recommendations for the Board's
consideration, with the recognition that the Board may reach an entirely
independent conclusion based on subsequent inputs or its own
understanding of the community consensus.
In neither model is the DNSO useless or input fruitless, but it is
DIFFERENT. For reference, I am participating even as we speak in the
U.S. Federal Communication Commission's Spectrum Task Force on spectrum
policy. This task force has no authority to make a decision. It is
simply charged with investigating and making recommendations.
Ultimately, it will issue a report to the Chairman, who may or may not
order the relevant bureaus to prepare an item for the Commission to take
official action. The full Commission will then vote on whether to
approve the item, which will go out for further public comment, analysis
by the relevant bureau, and a final Commission vote on a final Order.
So did I waste my time working on my comments to the spectrum task
force? No. But I do not over-estimate their effectiveness and I
calculate my strategic behavior accordingly, figuring out how much
"ammunition" to spend at this level and abosrbing information from the
other players in the hopes of reaching accomodation, identifying allies,
and countering rival arguments. If I though the Spectrum Task Force had
the authority to actuall decide something, I would act differently,
because I would be trying to influence decisionmakers rather than simply
inject ideas into the debate and sound out positions and arguments.
>If clear con
> sensus comes to the Board it is unlikely to decide against it unless there is a very cogent
>reason for so doing. This same right exists under the current By-Laws.
The manner in which the Board has exercised its authority under the
bylaws is what has created the expectation that the Board is the
ultimate decisionmaker, and will exercise that power in all
circumstances, rather than simply extrordinary ones. All institutions
begin as words on a page, but the manner in which these words play out
is what creates the character of the institution.
Again, I want to stress this isn't a matter of personalities or
conspiracies. It is a matter of institutional evolutionary process. It
happens at my local synogogue, when we decide whether something is
decided by the Board or the religious committee or the rabbi. Practice
and custom create expectations about who will do what, even if the
language of the bylaws support multiple interpretations. It happens in
the dialog between the Executive, legislative and judicial branches.
To take a real world example, the U.S. Constitution says that the
President shall make treaties "with the advice and consent of the
Senate." In the first administration of George Washington (first
President under the Constitution), Wahsington invovled the Senate early
in the negotiation process toreceive their advice. This proved
fractious and difficult. After trying it once, Washington decided that
the executive would handle the entire negotiation and would simply
submit a finished treaty to the Senate for ratification. This is the
model that still exists today. The Senate does not get to re-write a
treaty the President has signed. It only gets to vote yes or no.
Note that the langauge of the Constitution supported _either_ model, or
something else entirely. But custom and practice have now established a
clear course of action from these ambiguous words, and we shape our
At every Board meeting, the Board has applied a de novo standard of
review, carefully probing to determine if this decision is the "best"
decision and if it reflects the community consensus. It is arguable
that this is the role of the Board. It is equally arguable, however,
that the role of the Board was merely to look at the procedure and
supporting evidence and then give an up/down yes/no.
Put another way, was the Board intended as a decisionmaking agency or a
reviewing court? The Board has made it clear that it has adopted the
former, rather than the later, model. In 1998, this was up for grabs.
In 2002, it isn't. This is just a fact of life, and consequences flow
from this reality just as they would have if the "reviewing court" model
>The Board was not and could not reasonably be absolutely bound by the consensus rule,
>or the Board itself would be a 'lame duck'.
I could (and would) argue that this was, in fact, the intent. To
minimize the power of the Board and its roll in decision-making. I
stress again this is an institutional, not a personal issue. The
question in 1998 (and as it has continued to unfold) is whether the
Board are decisionmakers or merely serving a reviewing function.
What is interesting is to see how this plays out as ICANN matures. For
all that four years have passed, ICANN is still a young institution (in
no small part because it keeps altering its institutional structure, so
that decisionmaking processes are not coming to maturity). The .org
divestiture was, I believe, the first real full use of the ICANN
processes as set forth in the bylaws. Our handling of subsequent issues
obvious builds on the understanding of how the process works based on
this experience. Again, let me stress that this is not about RESULT.
It is not about what view "won." It is about institutional processes.
The Board chose to review the DNSO report de novo, and felt free to
modify it to produce a result it perceived as better. The Board could
have used a different standard of review and could have acted
differently when it saw a problem with the DNSO recommendation.
By contrast with the .org divestiture, the WLS is hideously complicated.
As an initial matter, it is unclear what standard should be applied to
the decisionmaking. Should VRSN be required to show that WLS will serve
the public interest? Should opponents have to demonstrate that it will
have a significant anti-competitive effect? Should a negotiation take
place as to how to mitigate harm or benefit the public?
What is clear is that it will be the Board, rather than the DNSO, that
is the ultimate arbiter. This is particularly true here, where
consensus will be elusive. (Although given that ICANN must prove
"consensus" when it makes decisions adverse to VRSN, as set forth in the
1999 MoU between VRSN, DoC, and ICANN, the question of standard of
review and supporting evidence become far more significant than in other
circumstances where no right of review of ICANN Board decisions exists.)
Accordingly, we must shape our roll in light of the fact that it will be
the Board, not the DNSO, that makes the ultimate policy decision.
Furthermore, as I have stated, the plan set forth in the blueprint
further re-enforces that it is the Board, not the GNSO, that is the
ultimate arbiter of consensus, must review all decisions de novo, and
has a responsibility to enunciate its perception of the best decision in
every case. Essentially, the blueprint codifies the model as it has
evolved, removing the ambiguity in the existing bylaws formulated when
the role of the Board and the DNSO in formulating policy and
decisionmaking were unclear.
Thus, a Board liason will, by the pure fact of its existence, influence
the dialog and shape it. The whole _point_of such a liason is to keep
the Board informed and to provide feedback to the consticuencies.
Because of the decisionmaking structure of ICANN, it is _inevitable_
that the Board Liason feedback will channel discussion along particular
lines, simply because the bulk of participants will not want to waste
their time on issues the Board will not consider.
>However, not one Board member has ever expressed the view to me that this is some kind of
>''power and authority'' struggle.
I do not accuse anyone of ill-motives, although I do think many of you
are insufficiently sensitive to the institutional repercussions of your
decisions. (And, like everyone else in the universe, constrained by your
own perceptions and philosophies). The actions of the Board and the
current blueprint make it clear that ultimate decisionmaking authority
rests with the Board. This has long-term institutional consequences.
Ignoring those consequences does not make them go away. Furthermore,
because the Board members are themselves participants in the process, it
is difficult for the Board memebrs to look beyond the current issues and
personalities to the broader institutional implications (I'm not saying
you don't try, but it is difficult. Building institutions is a
specialty in its own right).
>The whole idea of ICANN is input from all sectors of the community
>including all the geographic regions that reaches some consensus
>balance or compromise and then for the Board to generally base its
>policy decisions on such consensus. The Board would not be creating
>So,s with new mandates to help develop policy and new SO,s such as the
>ccTld or the At-Large if it didnt want serious input and a real and
>better effort at finding community consensus ,which has been an
elusive >target at best!
This is not inconsistent with what I am saying. I do not claim, as some
do, that the "fix" is invairably in. To the contrary, I hear rumors of
so many "fixes" being "in" that I have confidence that most decisions
remain open to discussion.
This is purely an institutional issue. Where do decisions get made?
Answer, with the Board. It did not need to be that way. It is not that
way now on decisions made by the RIRs. And, I would argue, it was not
conceived that way in 1998. But it has become that way. The Board is
the ultimate decision-maker. Consequences flow from that institutional
Thus, the changes proposed are designed to help the Board make better
decisions, and that will enjoy greater legitimacy within the world at
large. They are not designed to limit the power of the Board and shift
decisionmaking to different levels of the organization.
Let me stress again, I am not saying that this necessarily produces a
"bad" Board or a Board that is captured by a particular special
interest, or that has pre-determined policy. But the institutional
consequences of moving all final decisionmaking to the Board have
consequences and we should work with them, rather than ignore them. At
the very least, the benefits and costs of this model (rarther than a
de-centralized model) should be as clearly understood as possible.
>Harold you have come up with excellent and positive suggestions time
>and again.You are an eloquent champion for those whose primary concern
>is that ICANN be truly democratic and be responsive to the values and
>opinions of the non-commercial users of the internet.If you continue
>the process of identifying and organizing that community and
>encouraging increased participation internationally...it will be heard
>AND listened to(as it always hasbeen).It will be balanced fairly as
are >the concerns and issues of all constituents..
Thank you for the kind words. As I hope I have made clear, I was not
suggesting (here at least :-)) that Board does not listen. I am simply
recognizing the reality that the decision will be made by the Board
after it listens, rather than at the DNSO level, where I and my fellow
consticuency representatives would make the decision with the other
>.BUT not every argument is successful,as you know.
>We must all realize that and being wary if you must,not necessarily
>assume a systemic bias is or will be at work.
I have not, but there are institutional biases that flow from the Board
being the decisionmaker. The First notable one is that we will have
limited impact due to our inability to get to most face-to-face meetings.
>The Directors are 'people' and most with no direct commercial interest
>in the internet (in the ICANN context)..we worry about privacy and
>individual users etc because thats who we are!
I have never said otherwise.