Comments on sTLD Proposal (non-HTML version, if better)

[Karl Manheim <karl.manheim@xxxxxxx>]

Comment on Criteria to Be Used in the Selection of New Sponsored TLDs
Submitted by Karl Manheim, 3/25/2003

At the invitation of the Board, I respectfully submit the following
comments on the proposed selection process for a limited number of
sTLDs.

My comments fall into the following 5 categories:

1. The number of new gTLDs is artificially low
2. The decision to limit new gTLDs to sponsored/restricted groups
suppresses competition in the name space
3. The criteria for selection are unnecessarily cumbersome and costly to
meet; as a result only well-financed western-based submissions will
qualify
4. The proposal to outsource the evaluation process, so as to avoid
interest group politics, is a recognition that the “beauty contest”
model for allocating scarce gTLDs is inherently flawed;
5. The suggestion that this selection process should be made “routine,”
and provide the template for future expansion, indicates that ICANN has
moved beyond “proof of concept” to a permanent and total exclusion of
new unsponsored/unrestricted gTLDs.  This “static root” policy locks in
incumbents in their quasi-monopoly positions and locks out everyone
else, to the special detriment of users outside the United States and
Western Europe.

I appreciate that the first 2 points are technically outside the request
for comments; I will therefore keep my remarks very brief.  These and my
other points are also addressed elsewhere.  Please see gTLD-auctions.net
for further resources and analysis of the proposed gTLD allocation
method and alternative models.

1. The number of new gTLDs is artificially low.

There continues to be a concern among some parties that root expansion
is a perilous process that needs to proceed slowly and with painstaking
evaluation.  Others, including the engineering community, feel that the
pace of expansion could pick up substantially without jeopardizing the
root or DNS.  I see no evidence that the initial expansion in Nov. 2000,
or the regular addition of ccTLDs before and since, has had any
detrimental effect on technical operations.  Rather, the snail’s pace
(now approximately 1 per year) of new gTLDs is unnecessary and serves
merely to inhibit the growth of the Internet.  As the Final NTEPPTF
Report stated:  “A complete evaluation of the new gTLDs is a formidable
undertaking that could stretch out indefinitely and could be
extraordinarily expensive.”  New TLD Evaluation Process Planning Task
Force, July, 2002.  That seems to be the limbo we are in now.  I believe
it is time to move ahead with a reasonable number of gTLD adds.

2. The decision to limit new gTLDs to sponsored/restricted groups
suppresses competition in the name space.

Sponsored gTLDs are, by design, used only by small sets of registrants
“from a well defined and limited community.”  As a result, persons,
firms and interests that do not belong to a cohesive non-profit
organization are shut out from any expansion of the name space.  This
seems to follow the Business Constituency Position Paper on new gTLDs:

“The BC sees no value in new unsponsored/unrestricted names and would
need to be convinced otherwise by the results of the evaluation process.
The BC’s current position is that all new names should be
sponsored/restricted within the ICANN categorization.”  A Differentiated
Expansion of the Names Space, December 2002

Surely, this can’t be what ICANN agreed to in its MoU with the
Department of Commerce for the “development of robust competition in the
management of Internet names … in a manner that will permit market
mechanisms to support competition and consumer choice.”  Rather, the
freezing of the existing top level domain space, except for limited
purpose non-profit communities, will stifle competition and keep
consumer prices high.

3. The criteria for selection are unnecessarily cumbersome and costly to
meet; as a result only well-financed western-based submissions will
qualify.

 The proposal suggests 6 broad criteria, with a total of 17
sub-criteria, for selection of new sponsored gTLDs.  These range in
value from 4 to 30 points each.  Some of these are beyond dispute, such
as “ensur[ing] stable registry operation.”  But others purport to
evaluate, hence intervene in, the internal affairs of applicant
organizations (e.g., “Appropriateness of the sponsor;” “Responsiveness
to community”).  These are not matters for an organization charged with
management of Internet names and numbers.  Rather, they are policing
functions, undertaken by ICANN solely because it so greatly restricts
entry into the root.

 Yet, the larger objection to the selection criteria is that they
require huge resources merely to qualify, and then additional resources
to compete in the ensuing “beauty contest.”  Perhaps well-heeled
organizations such as museums can survive the gantlet, but the Getty,
the Armand Hammer, the Guggenheim, etc. hardly need special treatment by
ICANN.  The proposal acknowledges that an applicant’s “economic
viability” is an important criterion.  And the need to serve “broad
global communities” means that only organizations with large presences
in the developed world will qualify.

4. The proposal to outsource the evaluation process, so as to avoid
interest group politics, constitutes recognition that the “beauty
contest” model for allocating scarce gTLDs is inherently flawed.

 The last expansion round was hardly a model of objective process.  Nor
is it realistic to expect that from a constituency-dominated
organization.  ICANN is comprised of interest groups, as it should be.
Each group, also as it should, represents its own interests.  But there
is inherent tension between insiders and outsiders, especially when the
former are expected to judge the latter.  Accordingly, the Proposal
correctly suggests that evaluation be accomplished by “external
evaluation teams … not involved in ICANN activities and … not subject to
ICANN political nuances and pressures.”

 This statement speaks for itself.  It reveals an inherent flaw in the
use of subjective evaluation criteria.  The “beauty contest” method,
employed so disastrously at the FCC for 75 years, despite its mandate,
similar to ICANN’s, to promote the public interest, is simply the wrong
way to go.  Perhaps it is unavoidable in this case, especially where the
name space is kept artificially restricted, but it ought not to become a
model for future expansion.  That leads to my final point.

5. The suggestion that this selection process should be made “routine,”
and provide the template for future expansion, indicates that ICANN has
moved beyond “proof of concept” to a permanent and total exclusion of
new unsponsored/unrestricted gTLDs.  This “static root” policy locks in
incumbents in their quasi-monopoly positions and locks out everyone
else, to the special detriment of users outside the United States and
Western Europe.

 There is a lot at stake in the dispute over gTLD expansion.  Incumbent
registries have a quasi-monopoly position which allows them to extract
monopoly rents from registrants.  Even though they may be paying too
much for registration services, incumbent registrants also have
quasi-monopoly positions in their unique second-level domain names.
Addition of new gTLDs jeopardizes both monopoly positions.  Exclusion of
competitors, especially those without a voice in the bottom-up consensus
driven ICANN process, is the best way to preserve the advantages of
incumbency.

 Such anti-competitive behavior might be tolerable for a short while, as
the dominant participants test out “proof-of-concept” measures.  But,
the Proposal presages a permanent regime of exclusion.  This prevents
the root from being put to its highest and best use; it locks out
innovation and new entrants, especially those from Less Developed
Countries; and it serves to artificially inflate the price of
registration services.

 The Board does not need to be reminded that ICANN is a California
not-for-profit corporation.  As such, it has a fiduciary duty, not to
its members, but to the broader community.  See Cal. Corp. Code § 5510.
A regime to permanently exclude new unrestricted gTLDs is inimical to
the public interest, and should be swiftly rejected.


Respectfully submitted, Mar. 25, 2003

--
Karl Manheim
Loyola Law School
919 S. Albany St.
Los Angeles, CA  90015
Tel:   213-736-1106
Fax:   240-414-7747
Web:   http://faculty.lls.edu/manheim

The contents of this email and attachments may be confidential,
copyrighted and/or
protected by the Electronic Communications Privacy Act, 18 U.S.C. §
2510, et seq.