Re: [wg-c] registry contracts

[Milton Mueller <mueller@syr.edu>]

Jonathan Weinberg wrote:

> So here's a possible, adapted, plan for the testbed (drawing on the above, plus the
> suggestions of several other folks):  Each would-be registry submits a proposal to
> ICANN describing its technical and financial qualifications, its proposed TLD string
> and charter, and whether it will operate on a for-profit or a cost-recovery basis.

So far so good.

> If it proposes to operate on a for-profit basis, it must submit its "ceiling price"
> -- the price for registry services that it will not exceed during the delegation
> period.

Two observations.

1. If you believe that "switching costs" are such a serious problem, the problem
applies to non-profits as much as to for-profits. I work at a non-profit university. I
directly observe calculations regarding budget, tuition and other revenue sources.
What constrains the tuition (prices) at a non-profit university (registry)? The
university *definitely* has an incentive to increase the size of its endowment and its
budget. Nearly all organizations do. Students (registrants) have serious switching
costs, but they *can* switch. (Note, a Syracuse U education is not a perfect
substitute for a Temple U or a Michigan education, just as a dot com name is not a
perfect substitute for a .foo name). If we raise our tuition (prices) in order to
exploit captured students (registrants), incoming students (registrations) will
decline, and if the value we offer is not commensurate with the increased price, then
our reputation will suffer. In short, there are severe penalties to opportunistic
behavior.

Now, suppose you don't believe my last statement. Fine. You have just accepted the
fact that non-profits are just as likely--and as able--to exploit consumers as
for-profits. Ergo, they should not be treated any differently. And if you do find the
constraints on opportunistic behavior believable, then they are just as applicable to
profitmaking entities.

In short, I am attacking the relevance of the whole distinction.
It is based on a naive confusion between the legal category of "non-profit" and the
notion of "working selflessly for the public interest." There is no reliable
correlation between these two categories. Indeed, the belief that you solve the
registry problem by licensing only non-profits is just a way of verbally defining the
problem out of existence. Look at the US Olympic Committee, the United Way, the
Educational Testing Service, the World Bank, and other fabulously wealthy non-profits.

Either we regulate prices or we don't. There is no basis for distinguishing between
profits and non-profits in making that decision.

2. "Ceiling price" for what? Services and functions can be bundled and unbundled in
various ways.

Who needs ceiling prices? Here's a salient point: NSI's registry prices are regulated
for the forseeable future. Country codes are available to users. No one will be forced
to register their name in a new TLD. Given the fact that new registries must compete
long-term with a registry that
a) controls 75% of the market
b) commands practically 100% of the mindshare in north america
c) has rates set at a very low $6/year, and
d) is open to any and every accredited registrar in the world,

Why should ICANN care what prices a new registry charges? If its prices and service
are not competitive with NSI and 240 country codes, it won't succeed.