Re: [wg-c] breaking up (names) is hard to do

[Jonathan Weinberg <weinberg@mail.msen.com>]

At 09:36 AM 8/28/99 -0700, Kent Crispin wrote:
>The distinction between "for-profit registries" and "non-profit
>registries" is not precise enough for our purpose.  You mention that
>CORE could be considered a "for-profit", because it is controlled by
>a group of registrars, most of which (but not all) are for-profit. 
>There is also the distinction between the "registry" and the
>"registry operator" to consider, and the fact that the total
>operation could involve several contracts with for-profit entities. 
>We also have to distinguish the business model of the registry with
>the nature of the entity making the proposal to run the registry. 
>
>For example, IBM could propose to run a registry.  IBM is a
>for-profit entity, but the proposal for the registry could still
>include provisions that cause the registry to be run on a
>controlled-cost basis (non-profit/cost-recovery/fixed-price/etc).
>
>I think many of us tend to use the term "non-profit" registry as a
>shorthand that covers this possibility.  But if you think about it,
>under this broad definition of "non-profit", NSI is currently a
>"non-profit registry" -- it's profits are limited (theoretically)
>through contract to the USG.  [Of course, the issue of the *efficacy*
>of that control is quite debatable.] Perhaps the term "limited-profit
>registry" would be better, but I suspect that would simply confuse
>matters more. 
>
>So, what is meant by the term "for-profit"? To me, and to some
>others, I believe the term "for-profit" refers to a situation where
>the registry has freedom to raise SLD prices as it wishes.  That is,
>we don't mind downard pressure on SLD prices, but not upward
>pressure.  This follows from a belief that the competition in
>registry services should be driving the price of SLD registrations
>towards the cost of maintaining them -- any significant mismatch is
>evidence of monopoly effects in registry services.
>
>> The big problems are twofold:
>> 
>> First, in a steady-state system, in which TLD's are delegated and
>> subject to competitive redelegation, how will we level the playing 
>> field between non-profit and for-profit registries?
>> 
>> Second, during the rollout, how will we guard against the creation
>> of more NSIs (monopolists) if the creation of new TLDs/Registries 
>> stops short of the addition of <largenum>?
>> 
>> The third problem (how to prevent a registry from renewal gouging)
>> is actually a simple one: insert a contractual mandate limiting the
>> increase of registry charges for SLD name renewals, based on some
>> objective metric.
>
>This actually, in my terminology, requires that all registries be 
>"limited-profit".
>
>> And actually enforce the provision, expressly
>> providing that other prospective registry operators and SLD name
>> registrants have standing to challenge any failure to enforce
>> forfeiture of the TLD for renewal gouging. 
>
>"Standing" is not the same as "enforcement".  For enforcement to work
>you actually have to have clauses in the contract that require timely
>escrow of registry data, and the technical ability of ICANN to
>transfer that data to a new registry without undue difficulty.  It 
>also requires that the registry operator have no intellectual 
>property claims that can encumber that transfer.


	Is there the core of a compromise here?  For some of us, part of the
attraction of a mixed system of "for-profit" and "non-profit" registries
has been that the "for-profit" ones might have greater incentive to do
(innovative and socially useful) things that would lead more people to
register SLDs in those TLDs; the incentive lay in the fact that, assuming
prices were held constant, more registrations would mean more money for the
registry.  OTOH, few of us see social value in the registry operator making
more money simply by jacking up prices (as opposed to making the service be
-- or appear -- more valuable and thus getting more customers), and a bunch
of folks have expressed the obvious concern that for-profit registries are
more likely to seek to do that.  As a result, there's been a lot of traffic
on the list regarding contractual conditions that might, among other
things, restrain for-profit registries' ability to raise prices.

	Kent's formulation suggests that -- just maybe -- many of us might agree
on a system in which a registry might or might not be operated on a
strictly cost-recovery basis, but *all* registries would be subject to some
sort of meaningful limitation on their ability to raise prices.
(Obviously, formulating such a limitation isn't straightforward, and some
folks have suggested to the list that it can't be done.  Yet I wonder.)
What do people think?

Jon


Jonathan Weinberg
co-chair, WG-C
weinberg@msen.com