Re: [gnso-vi-feb10] Question to WG on RAA

[Eric Brunner-Williams <ebw@xxxxxxxxxxxxxxxxxxxx>]

Independent of how this topic entered the problem area of implementing
the Board's Paris decision to eventually accept some new gTLD
applications, and subsequently allow the applicants to operate some
new gTLDs, our discussion of Vertical Integration is informed by a
larger, and longer discussion of Competition Policy.

This, the CP discussion, starts with the single-source,
non-competitive contract for no-cost provisioning and publication
services, and changes (SRI epoch), next to a competitive contract,
NetSol I epoch), next to a competitive cost plus contract (NetSol II
epoch), next to a renewable monopoly publication (VGRS I epoch) and a
testbed regime of competitive provisioning (NSI+6 epoch), next to a
plurality of renewable monopoly publications (new gTLD I epoch), next
to a negotiated least largest monopoly divestiture (PIR epoch), and
finally to a second round of renewable monopoly publications (new gTLD
II epoch).

The single greatest event which created registrations outside of the
control of the circa-2000 monopoly publisher (VGRS) was the
divestiture of the .org registry. All other non-VGRS registration
creating events combined match that single divestiture. For a variety
of reasons the .net registry was not divested, and since 2003 the
Board has not re-examined its founding premise, that competition
policy need not consider the structural choice of "competitive
non-monopoly registrars" and "concurrent monopoly registries".

The question before us is bigger than "shall we inform the Board that
they are erratic dunces, and that a change from the generally true 15%
cap on registry ownership in registrars had the wrong polarity and
magnitude values?"

We are not trying to make certain that a 2% market share entity is not
advantaged, or disadvantaged, relative to a 0% market share entity. We
are not trying to make certain that a 0% market share entity is not
advantaged, or disadvantaged, relative to another 0% market share entity.

It was my assumption after the pre-Seoul webinar, and after the Seoul
issues discussion between Richard and Brian, that there are some
issues all of the position advocates share:

 - a desire to grow the market, and not merely the portion held by
VGRS, though that would likely grow as well;
 - a desire to maintain, or increase, the utility of domain names to
the users of domain names.

Perhaps there are more shared values, but these two, relative shared
growth of use (seasonally corrected for economic downturns, natch) and
relative shared growth of value, were sufficient for me to have asked
then, and ask again now, if we can agree on goals and non-goals.

The first goal I have is for policy not to be made erratically, and
without information. Whatever it is we have, it is not greatly
illuminated by pretending that it is aluminum ingots (Alcoa) or fire
alarms (Grinnell) or peripherals (IBM) or parcel delivery (UPS) or
cameras (Kodak) or ... We don't gain a great deal by having some
Chicago School ideologues inform us that ... they are Chicago School
ideologues, shopping the same answer to every question. Whatever it is
we are doing, it is not greatly improved by radical rule changes at
every point of departure.

Not stasis, but if we're going to have revolutionary change and blood
in the streets, I'd like a smarter, and less provincial gang of
neo-Leninists than what has clawed its way to momentary intellectual
dominance in the Anglo-American political economy.

The second goal I have is for policy not to negate policy. We have a
process, started at Paris, and a series of opportunities to chase
after shinny objects to fill our idle time until ... this resembles
waiting for DNS Godot. If there are business models which cannot be
made to work until the rules change, this is not a sufficient motive
to halt the process started at Paris.

Non-goals are things like perfection, such as coming up with a rule
that won't need to be tested for utility and necessity in 10 years.

I want a policy recommendation that goes to the GNSO that is roughly
consistent with expectations prior to the CRAI report. I don't mind if
more than one policy recommendation goes to the GNSO, making other
claims of utility or necessity. I'd like this to happen "sooner"
rather than "later", and for subsequent policy recommendations, even
those as radical as redelegation or registry services vs registrar
services demarkation change or ... to originate from the GNSO to the
Board. Not from out of a hat at the wave of a wand.

Eric