Professor Michael Geist from the University of Ottawa has added to the cacophony of criticism against ICANN over its domain dispute resolution policies (UDRP). He has published a paper on the subject entitled Fair.com? An examination of systemic unfairness in the ICANN UDRP.

The professor studied over 3,000 cases to see if the widespread criticism of the policies was justified, where they may have gone wrong and what could be done.

In the end, he turned in as damning a report as any other investigation into domain resolution. It is clear that the accusations levelled at the dispute resolution bodies are backed up by the facts.

The two largest resolution bodies - WIPO and NAF - are hugely more likely to rule in favour of a trademark holder. He accuses many panellists of "clearly misinterpreting" UDRP and clearly demonstrates that the system is bent toward trademark holders. He also criticises the ability of complainants to select in which resolution body the case is heard - giving companies an incentive to rule in favour of the trademark holder.

One of the most intriguing aspects of the report is its breakdown of cases into one and three-member panels. In one-panel discussions, 83 per cent of cases are in the complainant favour; however this falls dramatically to 60 per cent when using a three-person panel.

Through a time-consuming study of which panellist acted in which cases, he was also able to show beyond doubt what many (including ourselves) have accused arbitrators of doing - namely, picking trademark-friendly panellists on purpose on single-panel cases.

Although the roster of judges is diverse, some panellists have been picked an extraordinary number of times. The six busiest at WIPO have accounted for 17.2 per cent of cases at WIPO; of 53 per cent at NAF.

Top of the list comes James Carmody - who has sat on an incredible 140 single-panel cases with a remarkable 95.7 per cent in favour of the complainant. James Carmody is the man responsible for ruling in favour of AOL in the Aimster.com case. He decided that "aim" stood for "America Online Instant Messenger" and could not be misinterpreted as standing for anything else (like the verb or noun).

Equally, those that have often voted against trademark holders are very rarely, if ever, selected to stand as a sole panellist.

With a three-person panel, the likelihood of bias is greatly reduced due to the fact that the complainant gets to chose one member and the respondent another. This simple balancing sees the number of cases won by complainants fall by a quarter.

However, many complainants do make a point of asking for a three-person panel since, although is yields fewer results and costs more, it is seen to be fairer and less erratic than the one-person panel.

Professor Geist concludes, among other things, that mandatory three-person panels are essential if fairness is to be retained. He argues there should be minimum and maximum caseloads for panellists and that greater transparency should exist in the decisions made by arbitration providers.

The whole report is available in a pdf here. ®

Related Links

PDF of the report
Professor Geist's homepage

Related Stories

Why ICANN's domain dispute rules are flawed: Part I
Why ICANN's domain dispute rules are flawed: Part II
WIPO rewrites UDRP rules yet again