Re: [wg-c-1] WORK: Question #1 New GTLDs

[Milton Mueller <mueller@syr.edu>]

There are a lot of analogies to radio frequency allocation and licensing here that
need to be considered, especially with respect to the property status of gTLD
delegations.

At the dawn of broadcasting, most governments reacted in horror to the notion that
bands of radio frequencies could be "owned" by private, for-profit businesses. A
large part of this rested on notions of scarcity. There were a very limited number of
channels available, and such scarcity gave those who controlled the bands a powerful
influence over communications. This led to a "public trustee" model of license
delegations. The licensee didn't own the channel, it used it subject to stringent
public obligiations and regulations.

As the available spectrum has expanded, the rationale for the public trustee concept
has progressively eroded. If there are 100 channels on cable, 50 on broadcast radio,
and thousands on the Internet, it makes little sense to treat a broadcast license as
a kind of quasi-governmental privilege. The public obligations are being relaxed,
spectrum is being auctioned, and licensees given far more flexibility to exploit the
economic potential of the channel. In some cases they are being given a real property
right.

We seem to be in danger of forgetting that in the DNS name space, there is virtually
no technical scarcity constraint. We can easily add thousands of TLDs to the root,
more than there is feasible market demand for. We can accommodate multiple business
models, and virtually any idea about how to do it that we want.

Kent believes that the real constraint is the policy disputes that might arise over
TLDs. This assumes an environment of artificial scarcity. The mindset is derived from
the NSI legacy model, and the CORE alternative. In both models, the value of
administering a TLD is magnified by restrictions on the number of competitors. Hence,
any grant of a TLD administration right raises political disputes over access to
monopoly profits. Most cybersquatting is also predicated on the fact that a dot com
address has value because there is such a restricted name space. The trademark owners
also fear new TLDs primarily because they think of the current environment of
artificial scarcity.

Eliminate the artificial scarcity, and most of the disputes go away as well--except
in the legacy gTLDs com net and org, which have an "installed base" advantage.

Taking the radio analogy in a separate direction, there is now growing interest in
"unlicensed" bands of spectrum which anyone can share and use at their own risk. It
is a kind of Ethernet of the airwaves, congestion and performance depend on who else
is using the band at the same time. No license needed. There are many applications
that might flourish in such an environment.

Here there is a clear analogy to shared-registration TLDs.

We can accommodate the proprietary model. We can accommodate the shared model. We
could easily assign a TLD delegation to every business and organization that has
shown any reasonable interest and capability in administering one, and have thousands
to spare. The slightly harder task is to set up a procedure that allows enterprises
in developing countries, which are not yet ready to participate fully, the latitude
to stake claims to TLDs (or to create shared ones) on an ongoing basis in the future.

And no, we do not want a committee of the "haves" to decide whether "the Internet
community" "needs" new TLDs. Such public interest determinations have been proven
time and again, in a variety of contexts, to be nothing more than the members of a
cartel deciding whether they want to permit new competition. The number of TLDs
should be determined exclusively by the willingness of businesses or organizations to
administer them, and by the willingness of users to patronize them.

Why look for and create resource incompatibilities when there really aren't any?

What I cannot and will not support is any policy that deliberately attempts to
maintain scarcity so that regulatory leverage can be exerted over registries and
registrars.

--MM