Punitives and Policy-think

Put to one side your views about whether the Philip Morris v. Williams majority misrepresents precedent or takes another misguided step into the land of substantive due process.  Instead, I think it is worth asking whether Breyer's distinction between (1) allowing a jury to consider the potential harm to other parties as probative of reprehensibility and (2) allowing a jury to consider actual harm to others directly makes any sense.  Indeed, the very distinction "eludes" the very bright John Paul Stevens.  I think -- from a pure policy perspective -- there is something to recommend Breyer's approach (though my views here are only provisional). 

To be sure, punitive damage awards are meant to punish -- and the distinction Breyer forces onto jury instructions may be more subtle than a jury could realistically follow.  Still, the distinction does not elude me: It is reasonable to believe that the degree of punishment (and magnitude of damages award) that a given action warrants should be tied somehow to how many others the given conduct puts in harm's way; but there remains a difference between awarding a plaintiff punitives for the very fact of a defendant's action harming others and awarding a plaintiff punitives for the reprehensibility of an action that had the potential to cause harm to many others. 

Take Stevens' own example: the murderer who throws a bomb and kills dozens of bystanders -- who we all agree should be punished more severely than the murderer who kills only his intended victim.  Stevens too quickly draws his conclusion from the analogy to the criminal justice system.  Suppose we were considering not punishment by the criminal law but wrongful death civil lawsuits.  There, I don't think the distinction should "elude" Stevens.  Suppose the "intended" victim's family initiates a civil suit in the case where there are dozens of other dead bystanders.  It seems perfectly plausible to imagine that we'd want to limit punitives somehow -- and not enable the first plaintiff to file or the first plainiff to judgment to get the benefit of all of the murder's assets.  Perhaps the best way to accomplish this is never to allow punitives to go directly to the plaintiff and force a state administrator to compensate all victims proportionately.  But assuming a second-best world of punitives paid out to individual plaintiffs, trying to find some instruction to disallow the first plaintiff to collect for all the other harm caused to third parties is reasonable.   This may be why Stevens spends some time in his dissent focusing on the reality that some of the award was going to the State rather than to Williams directly -- to help convince us that Williams isn't getting a huge windfall that is being taken away from other harmed parties, the very fear that animates the majority's newfound test.

Is Breyer's instruction perfect?  Of course not.  It still seems to allow a jury to consider the act of bombing a crowded street to assess reprehensibility.  But it at least minimizes the opportunity for Williams to collect for all dead Oregonian smokers, some of whom may wish to litigate their own claims while there is still money to collect.

One final point:  Breyer's distinction might enable the punitive damage system to avoid some of the effects of luck associated with tortfeasor conduct.  By focusing a jury's inquiry on the potential effects of conduct upon third parties rather than on actual harm caused, the tortfeasor becomes responsible for all its egregiously negligent and reprehensible conduct, not merely the conduct that causes harm.  The gap between conduct and injury is often a matter of luck -- and Breyer's distinction helps juries, in assessing punitives, to focus only on conduct and not whether the conduct itself led to harm.  This could be a good thing, since the tort system all-too-often has no choice but to demand payment from negligent AND unlucky defendants, whose negligence actually led to injury, rather than merely the negligent.  Of course, to some extent, this point undermines my other observation about the desirability of Breyer's distinction because, in practice, it may lead to higher awards than a focus only on actual harm caused.  In any case, both of these thoughts suggests that there is, after all, something to Breyer's distinction.

UPDATE:  Dedalus (in the comments) is right. What I have called "Breyer's distinction" is not really the distinction Stevens is criticizing. Indeed, upon another reading of the opinion, I am no longer convinced that the distinction is attributable to Breyer, for he suggests elsewhere that punitives may be awarded only for potential harm to the plaintiff. That isn't quite what his quote from BMW suggests, but I explicitly wanted to leave precedent out of it. Worse, Breyer does make clear that actual harm seems to be allowed before the jury to assess reprehensibility. So much for that. I guess the distinction eluded me after all!