Abstract: Some critics oppose ICANN’s authorization of closed gTLDs, claiming they could be anticompetitive because closed domain registry operators would have monopoly control over generic TLDs of potentially broad appeal. But this misstates competition on the Internet. But closed gTLDs would provide the most innovative alternatives — and strongest competition — not only to incumbent TLDs like .com, but also today’s most popular domain names.
Today’s market leaders won’t be beat by simply copying them, no matter how much money is spent on ads. New entrants must offer consumers something new and different. Closing the TLD may sound nefarious, but it gives the registry operator the incentive to invest not only in marketing the TLD, but also in innovative new business models that may change the paradigm of how TLDs function. The operator of .hotels would no more “monopolize” the hotel booking market than the owner of hotels.com does today. But it could turn the domain name system into a more useful and accessible form of navigation, while offering new features like added security or thematic consistency across the TLD.
The future of the domain space will inevitably be messy and unpredictable in the best sense. But it is precisely that messiness — that unpredictability, that constant shifting of basic paradigms — that will most benefit consumers. Forcing new gTLDs to replicate the paradigm of .com will not. There may end up being legitimate concerns about particular registries’ abuse of market power, but such concerns should be handled by those best positioned to evaluate them: national competition authorities. There is no justification for ICANN to evaluate TLD applications on the basis of speculative competition concerns, as a practical matter conferring on ICANN the ability to enforce a per se rule against closed gTLDs; competition authorities are far more capable of making proper determinations ex post. ICANN should be a coordinator of the domain name space, not the global regulator of the Internet.
To download this paper in full, go to: