What To Do About Punitive Damages

Many complain that punitive damages have reached "crisis proportions." Yet the true problem is not the size of such awards, but the method by which we determine them -- a system that confuses the role and procedures of civil and criminal law.

• Exxon paid $125 million in criminal fines for the Valdez oil spill. A civil jury then punished Exxon again, with a $5 billion punitive award -- the largest ever made by any jury or affirmed by an appellate court.

• The drug firm Richardson-Merrell faced over 1,500 personal-injury actions for fraudulent marketing of the hazardous drug MER-29. Juries awarded punitive damages in three cases. The threat of punitive awards nudged the firm to settle most of the rest and helped make the settlements larger than they would otherwise be. In effect, the company was punished repeatedly for the same course of action.

• BMW sold Alabama doctor Ira Gore a new car without telling him it had been damaged in shipping, then refinished. This reduced the car's value by $4,000. A jury gave Gore $4,000 in compensatory damages. Then, to send BMW a "message," it added $4 million in punitive damages. The award was later reduced to "only" $2 million, and finally vacated altogether by the Supreme Court. BMW did not break Alabama law. But if it had, the criminal fine would have been just $2,000.

Imagine that the defendants in these cases had been common criminals -- muggers or drug kingpins. Punishing them twice for the same conduct would be unthinkable. Having each victim bring a separate case -- each demanding punishment for all the defendant's crimes -- would be strictly forbidden. So too would be a punishment of 2,000 or 4,000 times greater than the legal fine for their conduct.

Restitution and compensation are fairly easy to define. But punishment goes hand in hand with stigma or moral blame. Civil, or "tort," law, is not designed to punish people for morally reprehensible conduct. That's why we have criminal law -- which comes with limits to prevent abuse.

To protect the innocent, the criminal law requires high burdens of proof, hard-wires predictable punishments into the law and guarantees against double jeopardy.

Criminal fines paid by guilty defendants go to the state, not to injured parties. Decisions to prosecute are made by public officials who have the discretion to decide what is worth prosecuting and what isn't, and are accountable to the public for those decisions.

Tort law, with its system of punitive damages, has no such safeguards. If we mean to continue with punitive damages, we should have some limits:

• First, only allow punitive awards in cases where the defendant was reckless, or intended to cause harm or break the law. Leave punishment for those actions that truly deserve it.

• Second, make sure each state's punitive damages system has at least the safeguards mandated by the Supreme Court -- clear jury instructions, post-verdict review by the trial court and appellate review.

• Third, make punitive damages look as much like criminal fines as possible. Raise the burdens of proof. Divert damages awards to the state. And avoid punishing people repeatedly for the same conduct.

• Fourth, make defendants pay the true deterrent amount -- the economic benefit they realized by engaging in their reprehensible conduct. If someone saved $10 million by being reckless, his or her total fines should be at least that high. If compensatory damages, criminal penalties and regulatory fines already exceed $10 million, no punitive award is necessary; the wrongdoer has suffered enough already.

We should avoid thinking in aggregates. A fine that's $100 too high and one that's $100 too low don't cancel each other out. The size of any award doesn't tell us the facts of the individual case.

We should also avoid thinking in absolutes. Litigation is neither "bad" nor "good." Limiting punitive awards to an arbitrary number, or to an arbitrary multiple of compensatory damages, just assumes that high levels of punishment are inherently evil.

Instead, we should focus on what constitutes a "correct" result, and work on how to achieve such results in individual cases. That is, after all, what justice is all about.