This Sunday, the New York Times ran a standard "plaintiffs' lawyer feeds to the Times a cherry-picked set of documents and opening statement alleging a heinous conspiracy that we repeat credulously with a one-line disclaimer saying the defendant denies it" story, this one about Lilly and weight-gain from the anti-schizophrenia drug Zyprexa. Because the Times only gave one side of the story, we can't evaluate it; the evidence adduced seemed much more to me like 20/20 hindsight rather than a smoking gun. But we noticed a striking procedural detail that the Times elided; so did the two defense attorneys behind the new Drug and Device Law Blog. To wit, Lilly claims that the documents were released in violation of a protective order. It's far from clear that this is necessarily so, as James Beck & Mark Herrmann acknowledge, but they note:
Doesn't this happen in virtually all of the mass torts? Courts enter protective orders; defendants produce documents subject to the protective orders; documents are leaked to the press; night follows day.
What surprises us is how infrequently courts seem to care about this. Although documents are often leaked in violation of mass tort protective orders, courts rarely conduct full-scale investigations to find and punish the person who violated the court's order. Why not?
We'd like to see an academic (who might have the time to pursue this) research the mass tort cases (since about 1985 or so, when modern mass torts sprung into being) to determine the percentage of cases in which protective orders have been violated and the number of times courts have tracked down and punished the violator. If, as we suspect, protective orders are routinely ignored, but violators are rarely punished, that fact should influence how courts conduct future cases.
Let me further note that if there is an academic out there interested in studying this empirical question, they should contact me at the AEI Liability Project; I'd be interested in hearing their proposal and possibly funding and/or co-writing a paper.