Punitive Awards After BMW, a New Capping System, and the Reported Opinion Bias

Capping punitive damages awards is a centerpiece of the tort reform movement. According to the American Tort Reform Association, as of June 30, 1996, forty-three states allowed punitive damages awards. Of these, twenty-nine states impose no caps on punitive damages and fourteen impose some form of cap. In states that cap punitive awards, the preferred method is to employ a simple multiple of the compensatory award. Eleven states rely on a multiple of the compensatory damages award. The most popular multiple is three times the compensatory award, but this is used by only five states. The capping multiples range from one to five. Two developments highlight the importance of rethinking punitive damages caps. First, the Supreme Court\u27s decision in BMW of North America, Inc. v. Gore shifts the range of argument in the punitive damages area. The Court relied on, inter alia, a breathtaking 500 to 1 punitive to compensatory damages ratio to invalidate an Alabama punitive damages award. Since BMW, several courts have relied on it to reduce punitive damages awards. Thus, the central question is no longer whether there will be review of punitive-compensatory ratios. Rather, the question is whether caps will be systematically defined by legislatures, or applied on an ad hoc basis by courts as a matter of federal constitutional law. Second, systematic knowledge of the patterns of compensatory and punitive damages awards is now available. The available data suggest that businesses, insurance companies, and defense lawyers cannot support the claim that punitive awards are frequent, or that they follow a crazy pattern with little or no relation to compensatory awards. The strongest claim that the data support is that some punitive damages awards have been extreme. The mass of punitive awards satisfy a surprisingly regular pattern. Together, these two developments suggest a new approach to capping punitive awards. The approach is not based on arbitrarily chosen multiples of compensatory damages, the dominant current practice, but on experience. Using publicly available data, caps can be constructed that leave unaffected the mass of punitive awards while reining in, or triggering more active review of, the few extreme awards. We present here the empirical case for a system that caps punitive awards at ten times the compensatory award. The focus here is narrow. This Article concerns one aspect of caps on punitive damages, the relationship between the compensatory and punitive award. This relationship is important because the punishment imposed ought to relate in some measure to the harm caused