The takedown notice study

A University of Washington study of how one can get invalid DMCA takedown notices has attracted quite a bit of attention. Ed Felten, for example, has written approvingly of it. But after looking at the main page, the FAQ, and the technical report (PDF), I find that the authors didn't do a very good job of it, at least on the reporting end.

The FAQ makes a bold claim: "While others have suggested that the results might not be conclusive, we are the first to provide scientific evidence that people could be receiving DMCA notices today for allegedly illegally sharing content when in fact they were not." Others have previously provided documented evidence that they got invalid takedown notices; one good example comes from the Berkman Center. A single documented example constitutes scientific evidence, but the authors implicitly reject (or are unaware of) all of them.

The paper is vague on how monitoring agencies (misleadingly called "enforcement agencies," though the groups in question are private parties that can't enforce laws) generated their complaints. None of the agencies that issued complaints are identified beyond a general description as "both individual companies focused on monitoring P2P networks and larger industry associations." Perhaps there were policy reasons for withholding names, but the issue should at least have been mentioned.

DMCA complaints are supposed to name infringing content, yet the methods used by the experimenters apparently made no reference to particular copyrighted works. One sample takedown notice is provided, removing all identification of the complaining party, identifying Iron Man as the work being infringed on. How did the creator of the complaint decide that was the work that had been pirated?

Saying that a printer was issued a takedown notice is a nice attention-grabbing claim, but more questions need to be answered before the study can be taken seriously.