Collateral estoppel, or issue preclusion, is the principle that a party that has lost an issue at trial doesn't get to relitigate the same issue at a future trial. Offensive collateral estoppel permits a plaintiff to use the doctrine against a defendant. Once upon a time, the doctrine only applied if it was possible for both parties to use collateral estoppel against one another from a previous trial, but a generation ago courts began to acknowledge the possibility of "nonmutual" collateral estoppel: if Y wins issue A against Z, X can seek to preclude Z from contesting issue A in a future trial.
In the wake of the Cona/McDarby verdicts, we see a number of plaintiffs' attorneys (faithfully reported by the press (also here)) and plaintiff-attorney bloggers suggesting that nonmutual offensive collateral estoppel will now apply to the next 4000 New Jersey Vioxx trials (not to mention the $12-billion consumer-fraud class action), with only the issue of causation and damages to be litigated.
Except that Merck can defend against any such motion for collateral estoppel by noting that there are conflicting jury verdicts out there, because Merck won every single issue (including a consumer fraud claim) in the Humeston New Jersey case.
In my mind, it's common sense that if there are inconsistent verdicts, nonmutual collateral estoppel cannot possibly apply. Any other result would mean that a defendant is essentially playing Russian roulette with every trial, because it could win twenty trials in a row, and then be precluded from defending itself if it lost the twenty-first. And courts agree: they don't grant collateral estoppel in such circumstances. As the Supreme Court noted in the landmark case of Parklane Hosiery, preclusion "may also be unfair" if the judgment relied upon "is itself inconsistent with one or more previous judgments." See especially Setter v. A.H. Robins Co., 748 F.2d 1328, 1330-1331 (8th Cir. 1984) and Hardy v. Johns-Manville Sales Corp., 681 F.2d 334, 345-346 (5th Cir. 1982), products-liability cases with just this fact-pattern, which denied collateral estoppel because the defendant had won some as well as lost some. So, honest question: has there ever been a case (much less a New Jersey case) where a court has imposed a Russian-roulette nonmutual-offensive-collateral-estoppel standard, and imposed collateral estoppel on a defendant who has had some success? And if not, why are attorneys treating it like this is even an outside possibility in New Jersey? And why can't reporters double-check with a single expert who can debunk these patently false legal arguments?