Net law expert slams domain dispute process

The process for deciding ownership of Internet domains is flawed, biased and in drastic need of reform, an expert in Internet and e-commerce law has concluded in a study released today.

Professor Michael Geist of the University of Ottawa has extended a previous study last year into how ownership of domain names is decided and concluded that, if anything, the situation has got worse.

The study is particularly timely given that Internet overseeing organisation ICANN is hosting an international meeting in Ghana next week in which the process (called the Uniform Dispute Resolution Process or UDRP) is to be reviewed.

Last August, Professor Geist accused two of the four domain arbitrators - the National Arbitration Forum (NAF) and the World Intellectual Property Organisation (WIPO) - of actively choosing judges who favoured complainants (trademark holders). This ensured them more business as the complainant in domain disputes is allowed to decide where the case is heard.

This accusation was rebuked by arbitration judge Scott Donahey, the arbitrators themselves and ICANN, who all claimed Geist's conclusions were flawed as he had not included undefended (default) cases when comparing NAF and WIPO to the other main domain arbitrator, eResolution.

In the new study, all default cases have been included as well as the most recent cases up to 18 February 2002, and Geist's conclusions are even more striking. With undefended cases included, eResolution chooses in favour of a complainant in 64 per cent of cases when one judge is presiding (63 per cent last time).

Since the arbitrator itself decides who will decide who sits on a one-panel case, the significantly higher percentages for NAF and WIPO (86 and 83 per cent respectively) raise questions over just how uniform the dispute resolution process really is.

It is no coincidence, Geist argues, that NAF and WIPO have the lion's share of the market - 34.5 and 59.2 per cent respectively. A system which rewards those who chose in favour of one party in an arbitration dispute is unlikely to be perceived of as fair. And, as if to confirm the damaging effects of this policy, eResolution announced its bankruptcy at the end of 2001.

Adding further fuel to the fire is the fact that the "random" selection process for judges has seen just six panellists at NAF decide 56.4 per cent of all one-judge cases. Of these cases, 95.1 per cent were in the complainant's favour.

WIPO is better in this respect in that 121 judges have now decided at least five cases apiece, although there remains a mathematical discrepancy between a random selection of judges and those actually chosen to preside over single-panel cases.

The single-judge bias is further demonstrated, Geist argues, by the fall in decisions made in the complainant's favour when a three-judge panel is used. In this situation, the complainant chooses one judge, the respondent another and a third is agreed to between the two parties. When a three-judge panel is used, decisions in favour of the complainant fall 37 per cent in the case of NAF and 21 per cent for WIPO (a 25 per cent drop in total).

However Geist's study has not met with universal agreement. A spokeswoman for ICANN told us that "other well-respected scholars don't agree with Geist's analysis" and points to a review of UDRP by the Max Planck Institute.

This 78-page review is also cited by Geist's critic Scott Donahey, who claims the study is "nothing more than a recasting of data". Both people point to the Planck Institute's first conclusion that: "As a matter of principle, the UDRP is functioning satisfactorily. No major flaws have been identified in the course of the evaluation."

In reality, however, the Planck review actually supports Geist's observation that there are significant differences between the domain arbitrators, although it stops short of providing an explanation for these differences, claiming it is beyond the scope of its review.

The Planck second conclusion states: "It was however confirmed that considerable differences exist with respect to the outcome of decisions handed down by eResolution Panels on one hand and NAF and WIPO Panels on the other."

Within the review itself, the Planck Institute points out first that the judges themselves are independent and unswayed by the arbitrator itself but then continues: "There remains however a certain risk that panellists [judges], by handing down decisions which are markedly in favour of right holders might seek to increase their popularity in the trademark community at large, and to profit from this effect for their regular work, in particular as a lawyer or trademark attorney."

This situation, it says, "remains remarkable or even troubling, and it calls for further consideration".

However, NAF has provided a stronger case as to why Geist may be mistaken in his views, telling us that by focusing purely on the outcomes of cases he is looking at the wrong end of the process.

The assumption that NAF and WIPO are knowingly deciding in favour of complainants since eResolution makes far fewer decisions in their favour is flawed, NAF Assistant Director Timothy S. Cole told us.

Since eResolution asked for all submissions to be delivered online, additional material (case exhibits, trademark registrations, case exhibits) - which may often tilt the balance in a case - are not delivered by paper-loving lawyers, Cole suggests.

The steep fall in complainant-winning cases when a three-judge panel is used, he argues, is simply due to the added effort and supposed legitimacy of a domain by a respondent that is willing to pay for the cost of the extra judges.

WIPO failed to reply to our requests for comment.

Despite claim and counter-claim, however, one thing that every independent observer can agree upon is that UDRP urgently needs to be reviewed in light of the last three years.

ICANN assures us that "Professor Geist's views are welcome and will be considered in the [upcoming] evaluation". Others have also submitted their reviews of UDRP. We can only hope that a fairer and more equitable system will result at the end of it.

For more information on the study and an extensive breakdown of arbitrators' and judges' decisions, visit Professor Geist's Web site at UDRPinfo.com.

Related Stories

Professor tears ICANN domain dispute policy to pieces
What the hell is UDRP?