Comments on gTLD plan

Old Delusions and New TLDs

Comment Submitted by David R. Johnson and Susan P. Crawford[i]

The draft "Plan For Action Regarding New TLDs" calls for the creation of three new sponsored TLDs next year. The plan also initiates a study of numerous questions regarding the TLDs that were selected in November 2000 -- and an even more elaborate and lengthy process to determine whether ICANN should establish a semantic taxonomy for the domain name system.

You might think this plan is a thoughtful and responsible step forward -- until you consider its implications for ICANN's future role. This plan suggests that ICANN will regulate, in advance and in the abstract, the way the meaning of domain names evolves. It implies that ICANN can decide what an appropriate "community" of domain name registrants is and how that community can best be served. It suggests that, instead of always favoring competition, and imposing rules only when most affected agree they are necessary, ICANN can deny new competitors entry into the marketplace without any showing that such entry will create technological or other concrete harm.

The plan reflects a delusion that is at the root of most of ICANN's problems: the myth that the Board speaks on behalf of a "Global Internet Community" and that it is the job of the Board to determine when and how and by whom various subsidiary (presumably subordinate) "communities" may be served and even what strings of characters will have meaning for them. The plan speaks of "determining by objective criteria whether a proposed TLD sponsor stands as an accountable, legitimate policy entity…" -- as if "legitimacy" could be a matter of "objective criteria" or as if ICANN could plausibly judge some other organization's "accountability." In fact, autonomous and decentralized actions create meanings and communities. These relationships cannot be dictated top-down. Strings of characters may have entirely idiosyncratic associations with the online activities that grow up around them, and communities are created by their members, not by ICANN's blessing. Markets are created by competition and marketing, not by central planning.

The implications of Mr. Lynn's plan for competition and growth in the domain name market are bleak. The plan suggests that sTLD sponsors must be non-profit entities, as though the stability of the DNS might be threatened if a new registry had a profit motive. It cites concerns by ISPs about whether "a large number of popular gTLDs" would create "a large customer-education and support burden," as if the very popularity of new TLDs would not itself demonstrate that the resulting "burdens" were worth the costs. (This same principle would suggest that we defer to used car dealers who seek to prevent manufacturers from introducing new models.) The plan declares the current batch of sTLDs (.aero, .coop, and .museum) a success, despite the fact that their small scale and high verification costs inherently make registration in sTLDs more costly for registrants and less profitable for registries and registrars. It cites concerns by IP holders about cybersquatting, without noting that new TLDs can be (and have been) required to provide remedies for this diminishing problem. In short, the plan appears specifically designed to make sure we only get small TLDs in the foreseeable future. Large "communities" of potential registrants need not apply.

The implications of the plan for ICANN's institutional future are even bleaker. If it is to become the narrow technical coordination body most want, ICANN cannot afford the misadventure of trying to build a global taxonomy. It is one thing to develop consensus-based policies needed to prevent market failures. But it is quite another to shut down an entire marketplace until some consensus has emerged regarding the meaning and proper role of language. It should not take years of study to decide that trying to create a taxonomy in advance, for every script in the world, is a supremely bad idea. How long should it take us to conclude that the Internet will not be well served by calling in the French Academy?

If ICANN is to limit its policy development to matters that closely relate to sound operation of the domain name system, it cannot afford to get into the business of judging which "communities" deserve to be served by the DNS, or whether they are sufficiently "tightly" or "loosely" defined, much less whether the registries that try to serve those communities are adequately "legitimate." Instead, ICANN's Board should propose promptly to eliminate all the overreaching provisions that crept into the existing contracts, due to staff's illegitimate exercise of leverage, and that unnecessarily constrained the existing sTLDs at their births. Communities of registrants of particular TLDs may come together to create some global policies, if allowed to do so by ICANN. But the way to make that happen is not to create a complex licensing scheme requiring ICANN's permission before any such communities can arise in the first place.

If the Department of Commerce meant what it said in connection with the recent renewal of ICANN's MOU ("ICANN should not be 'the government of the Internet'"), DOC should promptly make clear to ICANN that adoption of this plan would constitute a large step in the wrong direction. The Board of Directors should spare DOC the trouble by rejecting the plan in December and, instead, deciding to open new TLDs at a reasonable pace under rules that require only that (1) applicants possess a reasonable set of minimum technical and financial qualifications and (2) agree to future ICANN policies that are supported by documented consensus.

The key issue is how ICANN approaches the creation of new TLDs -- this question is much more important than how many are created this year. We hope that the Board (or DOC, or registries, registrars and registrants who finally realize that the proposed plan will prevent competition that would deliver better products and services) will awaken from the central ICANN delusion.

[i] David R. Johnson is teaching a class at Yale Law School about Internet policy and governance. Susan P. Crawford is a partner with Wilmer, Cutler & Pickering in Washington, D.C. This comment was written in our personal capacities, and not on behalf of any client of Wilmer, Cutler & Pickering.