In support of bottom-up determination of second level allocation policies

[Vittorio Bertola <vb@xxxxxxxxxx>]

Hello,

I would like to comment on the proposed framework for the allocation of
single-letter second level domain names in gTLDs where they are actually
reserved. I have to disclaim that I am currently a member of dotMobi's 
Policy Advisory Board, and I have been involved with ICANN in several 
voluntary positions for the last eight years.


I think that the current proposal suffers first of all from a conceptual
error. To which extent can ICANN dictate to the first-level gTLD
registries how they have to allocate second level names? It is correct
and actually advisable for ICANN to put constraints on which policies
can be adopted and on how these policies are to be defined, so that
allocation methods are fair to registrars and registrants, and offer
even opportunities for access to domain names. However, I think that
mandating from the top specific allocation methods, uniform for all
gTLDs, goes well beyond ICANN's coordination role.

ICANN should feel free to forbid allocation methods that do not seem
fair or do not guarantee even opportunities for access, or to require
that policies are adopted through appropriate public consultation and
representation processes, but should not mandate specific allocation
policies; it should in any case allow registries to make their own
proposals and submit them to ICANN for the assessment about their
fairness (and there already are policy processes to that effect), so
that policies can be tailored to the specific gTLD and that diversity in
the gTLD space can be preserved and maximized.


I would then part the issue into two separate ones:


1) What is the allocation method that puts the reserved names to best use?

I think that it should be the method that best guarantees that all names
are developed, host useful services (as opposed to being used to host
pay-per-click advertising pages and nothing more) and become broadly
used by final users. To this purpose, auctions seem to me an
inappropriate method: they maximize the amount of money that can be
squeezed out of the market, but they do not offer any guarantee that the
planned use of the domain name is sound or, indeed, that the domain name
will ever be used. This is even more true in gTLDs that are
community-based, where making money might not be the primary purpose of
the gTLD itself and of its second-level registrations, and the richest
registrants might not be the ones who could actually develop services
that are useful to the target community.

The specific method of allocation should thus vary from gTLD to gTLD;
ICANN could ask each registry to come up with a proposal, then validate
it against a set of predefined criteria to evaluate its fairness and its
ability to put the names to good use, before allowing the registry to
proceed. As I said, there are processes that were established exactly to 
that purpose.


2) Who gets to keep the money?

I understand that wherever there is money to be raised, lots of people
will start to push to get a share of it for their preferred use: the
current proposal really looks like "we're going to give a bit of money
to everyone so that everyone stays happy". However, there is a basic
issue first of all: is this ICANN's money, or is this the registry's /
registrar's / registrant's money?

Until now, the assumption was that as long as a registry pays the
ICANN fee on the registration, all the rest of the price would be the
registry's and the registrar's, while the difference between the
registration price and the actual market value returning from whatever
service is built on the domain name is the registrant's.

Breaking this assumption could have unintended and dangerous long-term
consequences, and is unadvisable unless these consequences have been
carefully considered in the overall. For example, it might be possible
that if ICANN feels entitled to force registries to pay to ICANN higher
fees for more valuable names, then registries would feel entitled to
exact from registrars higher fees for primary sales of more valuable
names, and registrars from registrants, and the entire primary
registration market as we know it today might actually disappear and be
replaced by a secondary registration market, under the assumption that
each level of the DNS hierarchy and value chain gets to make as much
money as it likes before passing the name to the following level.

Moreover, the assumption that all unregistered names are the registry's
(as assumed when some registries started to use wildcards) was strongly
rejected by the community: the sentiment was that the registry is just a
steward of the gTLD, but not its owner, and so it should not be able to
do anything it wants with second level domain names. So, why should
now ICANN allow this kind of behaviour to itself?

All in all, registries should pay ICANN for the services it provides,
and it is hard to understand how a single letter registration could have
a higher service cost than any other one. It is hard to motivate
the view that in this specific case the money should go to ICANN, 
whatever purposes for this money it might come up with. If ICANN needs
more money, then it should prove the need through the established
funding processes.

Incidentally, the very market value that these domains will have is
connected to the investments made by the registry to deploy and develop
the gTLD: in other words, ICANN would be exploiting the registry's
investments to make money for itself and for the purposes it likes.


Finally, I note that the auction model is typical of a business culture
and does not really take into account global diversity, nor the fact
that auctions would basically exclude anyone but major companies from a
few developed countries from having access to these names.


In the end, for all the reasons above, I find the current proposal
excessively top-down and I think that ICANN should accept the proposals
by several registries to allow different allocation methods.

Regards,
--
vb.                   Vittorio Bertola - vb [a] bertola.eu   <--------
-------->  finally with a new website at http://bertola.eu/  <--------