RE: [ga] Meaning of "Consensus"

["Gomes, Chuck" <cgomes@verisign.com>]

In my personal opinion, Danny's points are right on target.  It is critical
to get beyond the point where we think consensus can be determined by a vote
of some subgroup of the overall affected population.

Chuck Gomes

-----Original Message-----
From: Patrick Corliss [mailto:patrick@quad.net.au]
Sent: Monday, June 25, 2001 1:20 AM
To: Danny Younger
Cc: [ga]
Subject: [ga] Meaning of "Consensus"


On Sat, 31 Mar 2001 15:46:45 -0500, Danny Younger wrote:
To: <ga@dnso.org>
Subject: [ga] Suggestions


> My preliminary suggestions for DNSO improvement have been posted to the
> Public Forum.  The URL:
> http://forum.icann.org/cgi-bin/rpgmessage.cgi?dnsoreview1;3AC63E2A00000030
>
> I look forward to seeing your comments posted as well.

This describes Danny's views on "consensus".  It is very worth reading.

I quite agree with William X. Walsh that conensus requires more that a
simple majority.  I do not agree that a 2/3 rds majority satisfies
consensus.  What is needed is substantial agreement in the sense that those
who disagree accept the proposal of offer as the best solution they can
reasonably expect or they don't disagree strongly enough to have a fight
over the issue.

Let's say you have a country with two racial groups.  One group has twice as
many members as the other giving them a two-thirds majority.  The majority
decides to overwhelm the minority by taking their land, their personal
possessions and even their lives.   Clearly there is no consensus.

On the other hand, we have a vote about mailing lists.  :Let's say that
everybody except one lone voice agrees that mailing lists are sort of a good
idea.  Some people think there should be more of less of them but, in the
end, everybody agrees on five.   You could say there was general agreement
even if everybody would have preferred a different proposal.  That's more
like consensus.

But please read Danny's comments following:

On Sat, March 31, 2001 at 8:29 PM GMT, Danny Younger wrote:
http://forum.icann.org/cgi-bin/rpgmessage.cgi?dnsoreview1;3AC63E2A00000030

The DNSO recommends domain name policies to the ICANN Board; these policies
purportedly are based on what has been termed a bottoms-up consensus-based
decision-making process.  At issue is whether these recommendations truly
stem from "consensus".   It is my observation that what is commonly held to
be "consensus" by the vast majority of participants in the General Assembly
and by members of the Names Council of the DNSO is not that which ICANN
itself defines as "consensus":

A "Consensus Policy" is one adopted by ICANN as follows:
1. "Consensus Policies" are those adopted based on a consensus among
Internet stakeholders represented in the ICANN process, as demonstrated by
(1) the adoption of the policy by the ICANN Board of Directors, (2) a
recommendation that the policy should be adopted, by at least a two-thirds
vote of the council of the ICANN Supporting Organization to which the matter
is delegated, and (3) a written report and supporting materials (which must
include all substantive submissions to the Supporting Organization relating
to the proposal) that (i) documents the extent of agreement and disagreement
among impacted groups, (ii) documents the outreach process used to seek to
achieve adequate representation of the views of groups that are likely to be
impacted, and (iii) documents the nature and intensity of reasoned support
and opposition to the proposed policy.

To better illustrate this above point, let us consider a recent effort to
arrive at consensus, an effort that began in response to the following Board
resolution:

   Whereas, a proposal has been presented to the Board for various
   revisions in the agreements among ICANN, Network Solutions, Inc.,
   and the United States Department of Commerce that were approved on
   4 November 1999 in resolutions 99.132 and 99.133 and were signed
   on 10 November 1999;

   Whereas, the Board intends to consider what action, if any, to take
   on this proposal in its meeting to be scheduled for 2 April 2001
   at a time to be confirmed;

   It is therefore

   RESOLVED [01.22] that the Board requests all members of the Internet
community, including the Names Council and any of the constituencies and
other participants in the Domain Name Supporting Organization, to provide
comments on the substantive merits of the proposal no later than 31 March
2001;


The General Assembly of the DNSO initiated the consensus process with debate
on this topic and came to the following conclusions (as reported to the
Names Council):

After thorough discussion, the GA has shown rough consensus in favour to
option A, i.e. to keep the current contract.  A straw poll conducted between
the 15 and 20 March has given the following
results:
- 24 in favour of the current contract (option A)
- 2 in favour of the new contract (option B)
- 1 neither of the above

Subsequent to this conclusion being posted, some members of the General
Assembly expressed concern that the issue of the .org charter had not been
fully addressed in the GA statement, to which the Chair replied:

I have spent some time in going again through the 587 messages from
2001-03-08 to 2001-03-25, when I have sent the report.  I might have
overlooked something for the second time, but my findings are the following:

- In favour of the change of the charter of .org: 0
- Against the change of the charter of .org: 7
- Against the removal of names: 5
- No opinion on the subject of the .org charter change: 34

If I misunderstood somebody, can the misunderstood party point out to me the
message that I have overlooked, or misinterpreted?

Of particular note is the Chair's observation on the overall process:

Nowhere I said that there was consensus on changing the charter.  I only
stated that, should the charter be changed, there's consensus that the
current bona-fide registrations be kept.  In all fairness, this is what I
understood.  We have enough problems with the people that estimate that 24
to 2 is not consensus enough, you can imagine the reactions to a consensus
call that was more dubious.

An analysis of the extent of participation in the above-cited debate is
critical to our understanding of how members of ICANN view the concept of
"consensus" or "rough consensus".  Comments on the .org charter topic came
from members of the GA voting registry, and from nine participants not on
that roster.  There are currently 291 registered GA voters, and with these
additional nine participants the total contributory GA pool therefore
consists of a sum total of 300 voices.

Seven of these participants (02.33%) argued against the change of the .org
charter; another five such members (01.66%) argued against removal of names.
Two hundred eighty eight members (96%) chose not to participate in
discussions on this topic.

Similarly, in the context of the GA position on the proposed Verisign
agreement it can be noted that:

Twenty-four participants (12.5%) were in favor of the current contract
arrangements;
Two of the participants (00.66%) were against, and
Two hundred seventy-four (91.33%) chose not to participate in the poll.

It is my view that the above numbers are more reflective of the phenomenon
known as "last man standing" than as any true measure of consensus.

Based on the numbers alone, a casual observer might conceivably conclude
that over 90% of the General Assembly decided that these topics were not in
fact policy issues; that as such no comments were warranted, but this too is
not an accurate assessment.

What is clearly true is that flawed procedures within the General Assembly
have created a situation wherein the expectations of the Board consistently
fail to be met.  Consider the comments of Director Auerbach:

*      As one who does have the job of examining and passing on the output
of the DNSO - I expect well formed policy decisions, including analysis of
the competing views, and backed by procedures that give me confidence that
all parties have had the opportunity to fair participate.

*      I support the determination of group opinion by the use of solid
procedures that include the placement of clearly articulated issues before a
clearly formed electorate who make clear votes that are counted. As it
stands, as a member of the ICANN Board of Directors, I am very unlikely to
give credence to any matter that comes out of the DNSO unless I see
objective data indicating that the DNSO has reached its conclusion by
something better than the hand waving that has to date been called
"consensus".

Having only 8.6% of the participants in the GA determine that which has been
termed "consensus" (no matter what the decision) is comparable to having a
vote of the Board of Directors in which only two members (10.56%) show up to
cast a vote.   One would seriously doubt that anyone would consider such a
vote to be representative of "consensus".

What is missing procedurally is an obligation to establish a necessary
quorum to legitimize the decision-making process in the General Assembly.
Yet even with a quorum, unless decisions are accompanied by a substantive
"analysis" of the kind to which Director Auerbach has referred (and which
are stipulated within the context of a uniform policy declaration), the
pronouncement of "consensus" will never be considered sufficiently
justifiable, nor will it withstand judicial scrutiny.

An agreement on the requirements of "consensus" is clearly needed, as it is
most obvious that many participants in the DNSO confuse simple "majority"
with "consensus".

Even members of the Names Council obfuscate the meaning of "consensus" in
their resolutions. by way of example, note that the "Names Council
resolution on the proposed revision to the ICANN/Verisign (NSI) agreement"
states:

Parts B and C passing by majority and A and D by a 2/3 majority thus
representing a consensus policy within the definitions of the NC by-laws and
referenced in the 1999 ICANN NSI agreement.

"Consensus Policies are those adopted based on a consensus among Internet
stakeholders represented in the ICANN process, as demonstrated by (1) the
adoption of the policy by the ICANN Board of Directors, (2) a recommendation
that the policy should be adopted by at least a two-thirds vote of the
council of the ICANN Supporting Organization to which the matter is
delegated.."

Significant by its absence is the remainder of this quote:

"and (3) a written report and supporting materials (which must include all
substantive submissions to the Supporting Organization relating to the
proposal) that (i) documents the extent of agreement and disagreement among
impacted groups, (ii) documents the outreach process used to seek to achieve
adequate representation of the views of groups that are likely to be
impacted, and (iii) documents the nature and intensity of reasoned support
and opposition to the proposed policy."

Also significant is the absence of the required "analysis" in an NC
presentation that was not much more than a mere compilation of the views of
the constituencies.   Most obviously, not "all" substantive submissions were
included in the supporting materials, there was no written "report", only
the declaration of a resolution with accompanying position papers; the
outreach process was not documented, and an assessment of the relative
weight or fervor of opinions was never prepared.

There are no shortcuts to getting the job done right.  If we intend to
contribute within the ICANN process, we have an implied responsibility under
the Bylaws to guard against the possibility of judicial challenge that may
contest "the presence of such a consensus".

It is our obligation therefore to provide for the requisite "document trail"
to substantiate any claims of "consensus".  This has not been done.  Only
one constituency actually reported on the degree of its outreach and
response to such outreach, the ccTLDs.  Out of 245 current ccTLDs, it was
reported that:

1. Four (4) members are in favour of "A"  (01.63%)
2. Five (5) members are in favour of "B"  (02.04%)
A result based on replies by less than 4% of the constituency members can
hardly be called consensus.

Whereas the remaining constituencies (Business, Non-Commercial, ISP, IP, and
Registrars) didn't even report on their outreach process, it is impossible
to properly assess their conclusions.
There are 180 registrars, but how many actually participated in this
process?  There are 66 voting members in the business constituency.  Where
are the records to indicate that any of them were even contacted?

I, personally in Melbourne, re-submitted my application to join the Business
Constituency (after having faxed it in three weeks earlier). I can attest to
the fact that I was not contacted by the Business Constituency, nor by their
NC members, nor notified in any manner regarding a need to comment on the
Verisign agreement.

I am not, however, disputing that outreach occurred; at issue is the failure
to document outreach and response.  What is missing is a set of publicly
accessible constituency mailing list archives that may be used to quantify
results.  The ICANN Bylaws stipulate that "The Corporation and its
subordinate entities shall operate to the maximum extent feasible in an open
and transparent manner and consistent with procedures designed to ensure
fairness."  Without such archives we do not have the transparency and the
openness that is stipulated.

Solutions to our problems:

1. EDUCATION -- As the concept of "consensus" appears to be thoroughly alien
to most ICANN participants, the ICANN staff should present a workshop on
this topic on every occasion that ICANN convenes in session.  Guest
speakers, such as David Johnson and Susan Crawford, should be invited as
expert commentators.     Documents on the topic of the consensus process
should be provided as a link on the main DNSO website.

2. PROCESS - Those on the voting roster of the General Assembly should be
required to vote; failure to vote will place voting privileges at risk.  A
quorum should be established.

3. OUTREACH - As decision-making will be based on both outreach and
quantifiable response to such outreach, each constituency must have its own
publicly archived mailing list.

4. WORK-PRODUCT - That which will ultimately take the form of a uniform
consensus policy must be rigorously documented.
<end Danny Younger>


I agree generally with these views.

Regards
Patrick Corliss



--
This message was passed to you via the ga@dnso.org list.
Send mail to majordomo@dnso.org to unsubscribe
("unsubscribe ga" in the body of the message).
Archives at http://www.dnso.org/archives.html
--
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