COMMENTS ON THE PROPOSED REVISION OF THE AGREEMENT WITH VERISIGNThe proposed revision
to the NSI / ICANN agreement will, according to the VeriSign and the ICANN Management,
"dramatically restructure the relationship between ICANN and VeriSign in a number
of positive ways."
That it will dramatically restructure the relationship
is not challenged, that the outcome will be positive for the internet community is
challenged most strenuously.
At first glance it seems strange that so fundamental
a change is being contemplated so soon after the original agreement was signed.
It is particularly surprising since the main objective of the original agreement
was to embark on a process that would both bring competition to the Registry/Registrar
market and ensure the structural change that would sustain it.
VeriSign
and the ICANN Management argue that already the market has changed so dramatically
that it is possible to contemplate the drastic changes that they propose.
They allege that "the introduction of competition in the Registrar business has been
much more successful and more rapidly successful than anyone anticipated" and that
"VeriSign's once-dominant market position has been severely eroded". This proposition,
crucial as it is to the rational for the proposed change, is unsupported by any substantive
economic or legal
analysis.
VeriSign and the ICANN management have merely
pointed to the existence of 180 competitor Registrars, 90 of whom are apparently
already operating under the co-operative shared Registry SRS system and the
fall in the price of a one-year domain name registration by "competing" Registrars
- interestingly NSI Registrar appears to be excluded. The fact that many of the VeriSign
names were the early domain names to be registered means that Verisign has a significant
ongoing competitive advantage. It benefits from the inertia that many registrants
will have to moving Registries and from the tortuous process imposed to move away
from the NSI Registrar. Having attempted this process once, a company is unlikely
to want to repeat the headache to achieve a saving of less than a hundred dollars
per year per domain in relation to other domain names.
Six days after the proposal
for the variation was posted, a further note was posted by the ICANN management which
set out statistics showing NSI/Verisign's declining share of registrations. In that
note the issue is presented to the internet community in this way: "the attractiveness
of the present proposal to the internet community therefore depends in part on whether
the increased opportunities for Registry-level competition outweigh any benefits
of continuing the Registrar-sale condition in promoting Registrar-level competition".
ICANN are regularly accused of posting unrealistically short timescales
in their consultation process and Melbourne was no exception. Most of the internet
community were taken aback by the radical nature of the proposed changes. Certainly
none of them had time to consider these issues properly before the ICANN meeting
in Melbourne. The result was that the only people who could speak confidently about
the issue were those that were involved in the negotiations. The rest, particularly
those for whom English was not a primary language, were left scrambling to catch
up. Not surprisingly, many constituencies felt that the whole proposal was
being pushed through with indecent haste. For those like the Registrar constituency,
supposedly the main beneficiary of the proposal, the haste seemed very indecent indeed.
Whilst
one expected that VeriSign had its own motives for pushing the new agreement through,
members of the internet community in Melbourne were surprised at the aggressive manner
in which ICANN General Counsel Louis Touton tried to push constituencies to a hasty
conclusion. At the Names Council, Louis demanded to know whether the Constituency
wanted to have its opinion taken into account.
On the Monday, the ICANN board had
its public meeting and a number of disgruntled members of the internet community
gave vent to their exasperation. For the management of ICANN due process was
clearly not a very cherished ideal. As board member Mr. Mueller-Maguhn said
"I thought it was normal for the Board to make the decision, then the lawyer would
work out how to implement the decision. Here it seems like the reverse."
Looking
back on the issues raised in Melbourne, and with the benefit of some reflection,
one cannot help but think that this whole initiative was contrived by VeriSign to
hoodwink ICANN and the Internet community. At first blush, this proposal might appear
to have some merits from an anti-trust point of view. But any serious analysis
would confirm that VeriSign remained very dominant and that the current structure
of the market leaves little hope for the emergence of a deep-rooted competitive environment.
The statistics presented by VeriSign and the ICANN management are very superficial.
In the case of the evidence of domain name registrations they were apparently heavily
influenced by the number of names that were deleted. We know that there are
large numbers of names remaining on the NSI Registry that have either expired or
have never been paid for. We also know that deletions are made to suit the
exigencies of VeriSign's commercial policy. In these circumstances, very little
weight can be given to the figures provided by VeriSign and ICANN Management in their
note of 6 March. Nor should too much attention be paid to the figures for VeriSign's
share of total registration. Whilst the trend is undoubtedly encouraging, the
danger is that these statistics may be a superficial indicator that bears little
relationship to the underlying structural competitive advantage held by VeriSign.
Even as a superficial indicator VeriSign, as a Registrar, retains more than 52%
of all registration in the .com, .net, .org Registries, with the next largest competing
Registrar having less than 12%. Indeed, although the ICANN Management / VeriSign
note refers to a 180 Registrars, 90 of whom are already operating under the SRS,
there are in reality only 3 that have obtained more than 5% market share. The
largest of these has less than 12%. The state of competition therefore at the Registrar-level
remains very frail indeed.
However there are two overriding factors
that play a significant role in tightening VeriSign's grip on the Registrar's market.
In the first place it controls over 3 million names on its Registry that were registered
in earlier years, sometimes in the period of its monopoly.
These names have a
curious status in that many of them were applied for by individuals but not paid
for and, in some cases, paid for but not renewed. No ordinary Registrar could
have afforded to keep these names on the Registry because they would have to pay
the annual fee. However, VeriSign / NSI Registrar can do this because whether
they pay the Registry or not makes little difference to their parent shareholders
because it is a neutral transaction in the parent pocket. In saying this, I
am disregarding the marginal cost of making the necessary Registry entry which by
all accounts is trivial. The result is that VeriSign appears to be doing what many
would describe as warehousing on a huge scale. Under RFC 1591, Registries are
required to process all domain Registrars' requests in a non-discriminatory fashion,
and without bias to any particular customer or network provider. Where they
fail to do so, the early drafters of these Internet rules considered it appropriate
that the Registry's continued appointment could be challenged. It would seem, therefore,
that VeriSign is susceptible to a challenge under RFC 1591.
The fact that the
Registry charges are immaterial to VeriSign as a parent of both Registry and Registrar
puts other Registrars at a significant disadvantage in that whereas they would delete
their unpaid domain names and the names would then go back into circulation to be
bid for on a first come first served basis, in the case of VeriSign the effect of
selling the domain name means that they will almost certainly be retained to the
account of the NSI Registrar. Significantly following the row that accompanied the
suggestion that VeriSign would auction its unpaid names, VeriSign purchased GreatDomains
and put a first tranche of what is believed to be a quarter of a million names on
the GreatDomains site for sale.
The lack of separation between the NSI Registry
and the NSI Registrar gives rise to at least one further significant competitive
advantage. VeriSign, through its NSI Registrar, is always at an advantage when taking
over additional names. This is because it will not have to pay the transfer
fee. A recent illustration of this would be the purchase by VeriSign of GreatDomains.
There are other examples where VeriSign has approached domain name holders with an
offer to undercut their existing Registrar fees if the names are transferred in bulk
to VeriSign/NSI Registrar. VeriSign is able to do this without much concern
over the transfer costs. Those fees are merely paid from one pocket to another (the
actual costs for transfer are again relatively insignificant).
In these circumstances,
VeriSign can sensibly adopt a policy that maintains its price premium, and enables
Registrars to get into the market giving some semblance of change in the competitive
environment. Provided none of these new potential competitors gets too big,
there is no serious risk to VeriSign's ascendancy in the market, particularly since
as VeriSign had predicted, consolidation is around the corner. At this point,
VeriSign has an unassailable
built-in structural advantage over all other Registrars.
It also owns the Registry that holds 80% of all domain names. VeriSign can
determine where it strikes and which companies it acquires and in doing so, it has
the significant advantage that it need have regard only to the actual costs involved
in performing a transfer at the Registry and does not need to concern itself overmuch
with the much larger standard fees that are charged by the Registry.
One
of the answers, I assume of last resort, that was being peddled around by VeriSign
/ ICANN Management last week was that even if VeriSign were forced to give up the
.com Registrar, it could always buy it back. I think not. So flagrant
disregard for the spirit of the 1999 Agreement, should by any normal rule of nature
involve a strong response from the Department of Commerce. Its failure to respond
would send shockwaves through the internet community, particularly the international
members. Even if the spirit of the agreement should turn out not to be of any
concern to a company that claimed that it wanted to earn the trust of the internet
community, the prospect that the owner of the largest Registry should purchase the
dominant Registrar would inevitably result in absolute prohibitions from a number
of anti-trust authorities around the world, including the EU - even in the unlikely
event that the Department of Commerce were to consider that there were some reasons
for not intervening.
Only last week the European Parliament in a near unanimous
motion, or one could say in ICANN-speak with an overwhelming consensus, called for
an EU representative to be charged with negotiating matters with the ICANN.
The Parliamentary concern was to find a way to guarantee the independence of the
ICANN from the US.
So in conclusion, the emergence of the 7 new gTLDs provides
in theory more Registry-level competition, however, for the moment this remains in
the realms of conjecture or hypothesis. Interestingly, VeriSign has found its
way into at least one of the consortiums behind these gTLDs. The most impressive
as a potential competitor to .com is that of the Afilias consortium with its .info
domain name. It will not surprise anyone to know that VeriSign is firmly embedded
in that consortium.
There is no anti-trust authority that would be prepared
to sign off on the justification offered for this new proposal. If no anti-trust
authority would be willing to do this, then the internet community should not do
so either.
Clive Stanbrook Q.C. & Nick Lockett
Stanbrook Technology Law
Group
Stanbrook & Hooper, Brussels
techteam@stanbrook.com