Bringing Predictability to the Chaos of Punitive Damages

Punitive damages remain unique in the American legal system. Awarded in the civil context with none of the protections offered in criminal law, courts levy punitive damages to punish and deter. The Supreme Court of the United States has clearly stated that courts may only seek to achieve these two goals when imposing punitive damages. A closer reading of the Court’s punitive damages jurisprudence, however, reveals another goal that has largely been ignored: predictability. Unlike punishment and deterrence, predictability is not a purpose for which to award punitive damages. Instead, the Court requires that, when awarded, punitive damages must be predictable. Failure to provide fair notice of the penalty for which a defendant may be liable amounts to a violation of the Fourteenth Amendment’s Due Process Clause. Though the Supreme Court has articulated that predictability is on par with punishment and deterrence in its punitive damages jurisprudence, the question remains as to whether this goal has been achieved. This Article answers that question with an empirical analysis of 167 blockbuster punitive damages awards—punitive damages awards over $100 million. These extreme awards offer a unique and important opportunity to investigate whether the Court’s current approach to predictability has worked. Unfortunately, it has not. Our analysis reveals that blockbuster awards remain unpredictable. In general, the amount of blockbuster punitive damages awarded is roughly as predictable as deaths caused by catastrophic floods, and few would argue that devastating natural disasters are readily predictable. Given the continued unpredictability of punitive damages, as evidenced by the random occurrence of blockbuster awards, this Article takes the next logical step of articulating a new constitutional framework. We argue that the Supreme Court should abandon its complicated system based on three “guideposts,” which eschews bright-line rules. In its place, we offer a clear limit on punitive damages. We propose that punitive damages may not exceed the compensatory damages awarded against the same defendant by more than a factor of three. This 3-to-1 ratio limit stems from the same type of empirical evidence that the Court has historically embraced in its punitive damages jurisprudence. The lone exception to this limit applies when a person is killed or injured. In that case, the value of statistical life serves as the lodestar for determining the total damages payment, thus guiding courts toward predictable punitive damages awards. If the Court takes predictability as seriously as it has stated, it can adopt our approach to take meaningful steps toward this important goal.