General Counsel's Second Analysis
of VeriSign Global Registry Services' Request for Amendment to Registry
Agreement
To the Board:
Background
This is a follow-up to my
17 April 2002 report on the request of VeriSign Global Registry Services
(VGRS) to amend the registry agreements for the .com and .net top-level
domains to accommodate its desire to provide, during a one-year trial
period, a Wait Listing Service (WLS). At its
meeting on 22 April 2002, the Board adopted
the following resolutions:
Resolved [02.53] that the Board requests the Names Council to coordinate
within the DNSO a comprehensive review of issues concerning the deletion
of domain names and possible solutions for those issues and to submit
to the Board, no later than 10 June 2002, a status report on that review,
with the status report to include any recommendations (with supporting
materials) concerning VeriSign's request to modify the .com and Net
agreements to allow it to provide a wait-listing service, for a fee,
as part of its operation of the .com and Net registries;
Resolved [02.54] that the Secretary is directed to advise the Address
and Protocol Councils of VeriSign's request and the Board's intent to
consider that request at its Bucharest meeting on 28 June 2002, with
an invitation to those councils to submit comments (if any) on the request
before that time;
Resolved [02.55] that the Board invites public comments on VeriSign's
request and directs that a suitable mechanism be established for allowing
comments to be submitted over the Internet to ICANN for at least a thirty-day
period; and
Resolved [02.56] that the Board invites comments on VeriSign's request
at the Public Comment Forum to be held on 27 June 2002 in Bucharest,
Romania.
The proposal was called to the attention of all three Supporting Organizations;
web-based comments were invited;
and the topic was discussed
at the 27 June Public Forum in Bucharest. At that Public Forum, the
Names Council gave a report on a consultative process it was conducting,
but its work was not yet completed. Presentations were given by the Names
Council task force and VGRS that tended to narrow, but not eliminate,
areas of controversy regarding the proposed WLS. Accordingly, the Board
adopted
the following resolutions regarding WLS in Bucharest:
Resolved [02.84] that the Names Council is requested to provide, no
later than 26 July 2002, its final recommendations, with its supporting
rationale and any separate positions of DNSO constituencies, on the
VeriSign WLS request (including the modifications made on 27 June 2002),
so that the Board may act shortly thereafter.
The Names Council (and its task force) continued its work, and on 14
July 2002 the task force issued the Final
Report of the Transfer Task Force on the WLS Proposal. After a public
comment period, the Names Council voted
15-3-1 (one member not voting) to adopt the task
force's final report, including separate statements of the gTLD registry
and intellectual property constituencies, as well as an extensive set
of links to statements from other groups and persons. The task force report
calls for the WLS proposal to be rejected, but also recommends various
conditions to be implemented if it is approved.
Analysis
As noted in the Bucharest presentations, introduction of the WLS would
require various amendments to the .com and .net registry agreements because
(a) it changes the functional specification under which those two
registries are to be operated (it involves a change in the name-allocation
algorithm) and (b) it involves VGRS offering a new registry (sole-source)
service for a fee. As has become apparent in the extensive community discussions,
a broad spectrum of opinion exists in the community regarding whether
these amendments should be made, though as noted above the scope of controversy
has been significantly narrowed in the consultative process.
In analyzing VGRS's request, it is useful to separate it into two issues
(these questions have been prevalent in the community discussion and are
reflected in the Names Council's work):
-
Should the VGRS be permitted to offer some form of WLS?
The principal objection that has been raised to implementation of
the WLS in any form has been that it is a sole-source service
that would tend to displace services offered at the registrar level.
The "displacement" of registrar-level services is due to
the fact that the registry-level service would provide consumers with
wait-list reservations that would be guaranteed effective (in the
event that the existing registration is deleted) because they would
be fulfilled before a deleted name is released to the pool of available
names that registrar-level services can access. (In this connection,
it should be noted that WLS reservations would not be offered by VGRS
directly to consumers, but through registrars on an equivalent-access
basis.)
To some consumers, the resulting registry-level service would likely
be a more attractive and convenient option than registrar-level services
even at a higher price. The "displacement" of registrar-level
services is due to this increased attractiveness and convenience.
In the past, when confronted with a situation where a registry operator
wishes to offer a service, on a sole-source basis, to customers that
might displace competitively offered services due to the sole-source
service's enhanced features, ICANN has permitted the sole-source service
while taking steps to ensure there are safeguards in place so that
the registry operator does not abuse its sole source position. In
particular, in the case of .name, in July 2001 the Board considered
whether the registry operator should
be permitted to offer a second-level e-mail service, which for
technical reasons could be offered only at the registry level. Because
the service simply offered consumers an additional (and potentially
more attractive) option, the Board decided
the service should be permitted, subject to various protective
conditions (equal registrar access, performance requirements, and
a price cap).
Although the specifics of the WLS differ from the .name second-level
e-mail service, many of the same considerations apply:
- In both cases, a registry operator wishes to offer an optional
service to consumers that some consumers may find attractive.
- In both cases, some consumers are likely to prefer the proposed
service. (It should be noted, however, that in both cases other
consumers may not find the proposed service more attractive than
alternatives. In the case of the .name second-level e-mail service,
consumers had many other alternatives for e-mail services under
other names. In the case of the WLS, some consumers may prefer certain
existing services that charge only when the name is deleted and
then successfully registered, in contrast to the WLS which requires
a payment whether or not the existing registration is deleted allowing
the WLS reservation to be successfully fulfilled.)
- In both cases, only the registry operator is situated (for technical
reasons) to offer the service.
- In both cases, consumers may be diverted from purchasing other
services offered competitively.
The approach followed in the case of .name suggests that the WLS should
be allowed, thereby giving consumers an additional option, provided
there are appropriate safeguards in place to prevent the registry operator
from abusing its sole-source position to the detriment of consumers.
In other words, the focus should be on promoting consumer choice, and
registry services that offer consumers additional choices should ordinarily
be permitted, subject to any conditions needed to prevent abuse of the
sole-source position.
-
If so, what conditions, if any, should be required?
Assuming VGRS is permitted to offer some form of WLS, the question
arises what conditions (if any) should be required. In its report, the
DNSO task force suggested several, and some others have been agreed
by VGRS:
1. The WLS should only be introduced after the Redemption Grace Period
is in place. The task force's report recommends that the introduction
of WLS should be "dependent on the implementation and proven (for
not less than six months) practice envisaged in the proposed Redemption
Grace Period for Deleted Names policy . . . ."
The Redemption Grace Period was authorized
by the Board in Bucharest and is currently in the implementation
process. The reason stated for this condition is so that existing registrants
are afforded an appropriate mechanism to mitigate the effects of inadvertently
deleted names, and are not required to purchase, as a matter of course,
WLS reservations on their own registered names to avoid deletions that
are caused by errors in the process that are currently occurring. (Registrants
would be able to purchase WLS subscriptions on their own names, but
once the Redemption Grace Period is in place this should not be necessary.)
All constituencies support implementation of the Redemption Grace Period,
but the gTLD registry constituency believes
that its implementation should be unrelated to WLS. Particularly because
the implementation of the Redemption Grace Period will require cooperation
of the registry operators, however, it seems appropriate to condition
the WLS on prior implementation of the Redemption Grace Period. Otherwise,
registry operators would have the incentive to implement WLS before
their implementation of the Redemption Grace Period, resulting in existing
registrants being induced to purchase WLS reservations as "insurance"
against the too-frequent inadvertent deletions that are occurring. It
also seems appropriate to promote consumer awareness by requiring a
period of months between the introduction of the Redemption Grace Period
and the WLS; this might be three months (as originally proposed by the
task force) or six months (as finally proposed by the task force).
The DNSO also recommended that the WLS not be introduced before "establishment
of a standard deletion practice." The rationale for this proposed
condition is not well explained in the task-force report. Moreover,
because the establishment of a standard deletion practice will likely
be an extended process (because it may require cooperation from every
registrar), this condition may have the effect of delaying WLS for an
extended period.
2. No preferential treatment should be given to existing SnapNames "SnapBack"
reservations. VGRS originally proposed to preference its technology
partner, SnapNames, by a special provision under which presently existing
SnapNames' registrar-level "SnapBack" reservations would not
be permitted to be reserved within the WLS. This special provision had
the effect of preferentially exempting SnapNames subscribers (but not
subscribers of others) from having their reservations affected by the
introduction of the WLS. This special preference prompted extensive
community criticism. VGRS has responded to the criticism by revising
its WLS proposal so that there will not be any preference or
exclusion given because of SnapBack or any other existing registrar-level
reservation service.
3. Avoidance of registrar preferences through advance knowledge of deletion.
Some of the discussion regarding WLS involves the concern that the WLS
might have the effect of benefiting registrars that are currently sponsoring
desirable names. Thomas Roessler has pointed
out that the registrar sponsoring the registration of an existing,
highly desirable name might have an advantage in marketing WLS reservations
for that name because it is the first to learn of the upcoming deletion
of the name. To address this concern, it has been proposed that registrars
not be permitted to place (for customers) WLS reservations on names
under registration through their sponsorship at any time after a date
sixty days before the date the name is deleted. This should minimize
the possibility of some registrars having an unfair advantage by virtue
of their knowledge of a likely deletion. VGRS has agreed to this condition.
4. Transparency of reservations. The DNSO task force voted 6-1-1 in
favor of requiring full transparency as to who (registrar and registrant)
has placed a WLS reservation on a particular name. The intellectual
property constituency is opposed and the gTLD registry constituency
abstained.
As reflected in the task-force vote, there are opposing arguments regarding
whether the WLS should provide this transparency. On the one hand, intellectual
property owners are concerned that disclosing their WLS reservations
would disclose their interest in registering names if they are deleted,
and would therefore cause present registrants (including perhaps cybersquatters)
not to allow the registrations to expire, perhaps leading to extortive
activity. Others argue that transparency allows consumers to have maximum
information, which allows for appropriate and well-informed conduct,
and that there are other measures available to deal with cybersquatting.
Furthermore, the existence of a WLS reservation can in any event be
detected by attempting to reserve the same name.
The Board should consider whether a transparency requirement should
be put in place; VGRS is indifferent to this requirement.
5 . Pricing. Much community discussion revolved around the price VGRS
proposed to charge registrars for placing WLS reservations. The task
force voted (5 in favor, 0 opposed, 3 abstentions) for the proposition
that "WLS should be cost based, consistent with previous considerations
for approval of Registry services by the ICANN Board." The intellectual
property, gTLD registry, and noncommercial constituencies abstained
from this task force vote.
The registry agreements provide for price caps for domain-name registrations
and other registry services because the sole-source basis on which those
services necessarily must be provided creates the potential for abusive
charges. Where a registry operator is placed in a position of market
power (particularly customer lock-in) by virtue of its appointment by
ICANN, it has been viewed to be appropriate to guard against abuses
of this market power. This can mean that cost-plus-reasonable profit
price caps are appropriate in some cases. Where market mechanisms are
effective to restrain prices, on the other hand, it should not be necessary
to establish price caps. Whether or not VGRS would have market power
with respect to the WLS at this point in time has been debated, and
there are cogent arguments on both sides of this question.
The treatment of economic/competition issues (such as the establishment
of price caps) is a difficult role for ICANN, because of limitations
both on ICANN's appropriate role and on its economic expertise. Some
have expressed concern that the role of establishing price caps stretches
beyond ICANN's appropriate mission, while others have noted that registry
operators are placed by ICANN in a position of economic power, sometimes
with little or no market-based restraints, that requires protective
conditions including limitations on prices charged for sole-source services.
This dilemma the potential need for price restraints in
some situations coupled with the general unsuitability of ICANN to supply
those restraints has led ICANN in the past to seek to rely
on competitive mechanisms to establish price caps where possible (such
as by allowing prospective operators to provide competitive bids of
prices and then requiring them to adhere to their proposals). Although
economic/competition issues are necessarily implicated by ICANN's activities
from time to time, it has been remarked in the ongoing reform process
that ICANN should establish means to obtain
appropriate expert advice from better-situated organizations to
assist in addressing these issues.
In this case, it may be appropriate to avoid this difficult question
because the WLS is currently being proposed for only a one-year trial
and because it does not present the prospect of lock-in of existing
registrants. The price VGRS now proposes to charge for WLS reservations
is US$24 per year (this is reduced from the US$40 it originally
proposed), and in view of the limited time of the trial, the uncertain
volume of reservations that will be sold, and need to amortize start-up
costs over one year, the proposed price appears to be plausibly cost-based
(accounting for the risk of limited consumer interest). Accordingly,
it may be appropriate to permit the WLS trial to proceed at the US$24
price, while initiating a parallel policy-development process to determine
how, and with what expert assistance, ICANN will make decisions that
rest in significant part on economic/competition issues. This could
include evaluation of the ways in which outside expert bodies (such
as OECD, or national competition authorities) and/or panels (of individual
experts) can be used to provide expert analysis on particular issues.
Furthermore, if the price is too high then WLS will not be accepted
in the marketplace.
6 . Trial period. As VGRS has proposed, the WLS should be implemented
on a one-year trial basis, so its effects on registrants and consumers
can be better evaluated. The WLS should include a sound plan for data
gathering and analysis.
Respectfully submitted,
Louis Touton
General Counsel
Comments concerning
the layout, construction and functionality of this site
should be sent to webmaster@icann.org.
Page Updated
25-Apr-2003
©2002 The Internet Corporation for Assigned
Names and Numbers. All rights reserved.
|