Studies & Opinions

Splitting the Root Unlikely

Alex Tajirian

November 15, 2005

Discussions about the possibility of splitting the Domain Name System (DNS) root have lately gained momentum, as the date for the Tunis summit - 16 to 18 November – is fast approaching. I use some insight from economic theory on platform competition to shed some light on the issue.

The DNS is a platform that facilitates the demand coordination between Internet users and ISPs who host the websites. Thus, the issue of splitting the DNS is tantamount to examining platform competition between the existing DNS and a potential competitor.

The current DNS platform has been correctly characterized as a monopoly. Although the supply side has been restricted to the Internet Corporation For Assigned Names and Numbers (ICANN), forces on the demand side can lead to a monopoly. Thus, an interesting question is whether a free market for DNS, given current technology, would support multiple competing platforms?

For competing platforms to exist, the platform has to regenerate economic profits to its managers. In the case at hand, one might argue that there are strong political interests that might out weigh rational economic considerations. Yet, if splitting the root results in major economic losses to the new entrants - due to the interruption of Internet commerce – the split is unlikely. Nevertheless, an entrant can generate revenue from launching new top-level-domains (TLDs) or devising a mechanism to extract fees from either or both sides of platform users.

Recent research on increasing returns – the phenomenon is also referred to as increasing returns and network externalities – suggests that the demand for product variety is a factor that may tip a market into a single platform. However, for TLDs there is no evidence that such a factor is strong or a new platform owner might be able to extract enormous economic profits.

First, the demand for the recently launched TLDs has fallen below expected projections and the number of registrations constitutes a small fraction of the dot-com. (It should be noted that the dot-info has shown greater interest outside the U.S., but is still within the current DNS platform.)

Second, there has been no lack of entrepreneurs creating competing platforms. The most successful, albeit with limited success, is New.net. However, none has gained enough momentum to make it a viable competitor. Thus, product differentiation between DNS competing platforms is not strong, while differentiation within the DNS exists and can increase with the introduction of .eu early next year.

Thus, from the demand side, it is unlikely that the DNS will split any time soon.