New generic Top-Level Domains (gTLDs) appear to be headed for introduction next year, finally. That’s a good thing for many ICANN constituents who have been waiting for them to become available. Important questions persist about how new gTLDs will affect ICANN and its constituents, however, despite a lot of effort to resolve concerns. Pressing those questions should not be taken as criticism of the basic wisdom of making new gTLDs available to many constituents under many circumstances. But too much is at stake not to get it right.

Questions center on ICANN’s budgeting and protections for entities holding or claiming rights to particular names. In this post I’ll discuss the budget issues.

ICANN stands to make a lot of money from approving new gTLDs. Estimates range around $92 million (500 applicants at $185,000 each). By itself, a new source of revenue for ICANN isn’t necessarily controversial. But fair questions may be asked of how ICANN is raising that money and how it will be spent.

ICANN claims that the $185,000 application fee covers its projected costs and no more. Yet $60,000—about a third of that fee—is earmarked for a contingency reserve, to cover risks like litigation and administrative costs for greater-than-expected numbers of applications. The Joint Supporting Organizations/Advisory Councils Working Group on New gTLD Applicant Support rightly “questions if ICANN really expects a total of US$30,000,000 (US$60k x 500 applications) in unknown or variable costs to surface.” Even if this figure were a reasonable estimate of unknown or variable costs, no one has adequately explained why insurance can’t cover it. Insurance is how other organizations customarily plan against risk. Why not ICANN?

Digging even deeper, ICANN’s estimate of $92.5 million in projected revenues may be artificially low. Contests over strings like .bank probably will be resolved through auction, and it would not be commercially unreasonable for the final price of such valuable online property to exceed the application fee by orders of magnitude. It’s unclear why the estimated number of first-round applications has stayed at 500 when they will include both commercial and noncommercial applicants like .berlin. Total revenues from the first round of applications will vary widely, depending on how many applications are actually submitted and how much individual applicants end up paying for the rights to especially valuable TLDs. That variability matters because $92 million is a substantial sum by itself—ICANN’s total revenues in 2005 amounted to less than $18 million—and because even that figure may be considerably lower than ICANN actually recovers. With insure.com fetching $16 million last year, it’s anyone’s guess what global dominance of .news will get on the auction block. If revenues can be expected to exceed current estimates, a reduction in application fees ought to be explored.

That brings me to the last point I want to make about the gTLD budget. As a nonprofit corporation, ICANN has always claimed to charge fees on a cost basis. But there is no relation between the value of .news or other unusually valuable TLDs and the cost of administering the application process for that name, much less the cost of facilitating the operation of a single new gTLD. When cost and value become disconnected there is a windfall, and the question is how a multimillion dollar windfall will affect ICANN as an institution. As things now stand, its board of directors has not decided how to handle unused portions of the contingency fund or surplus revenues created from auctions over highly valuable names. The gTLD budget process was designed to stand apart from ICANN’s ordinary operating budget, but that separateness leads one to ask what ICANN will do if the first round of new gTLD applications produces a profit, much less a windfall. An agreement to consult the community if the gTLD application process yields profits is not a commitment to use those profits in a particular way. Nothing prevents ICANN from ultimately pouring its gTLD profits into any project it chooses. Perhaps including in the application a written commitment by ICANN’s board of directors not to use such profits from the gTLD process for ordinary operational expenses would be a helpful assurance to the many constituents that plan to apply for a gTLD during this exciting but uncertain first round.

Tomorrow I discuss protections for entities holding or claiming rights to particular names in the gTLD application process (Click to Read Part 2).