Twelve years ago, a New York jury acquitted Lemrick Nelson of stabbing Yankel Rosenbaum and set him free. This week, following Nelson's stunning confession that he had in fact stabbed Rosenbaum, a federal jury convicted him of violating Rosenbaum's civil rights (though they also seem to have concluded that the stab wounds did not cause Rosenbaum's death). In other words, the first jury blew it.
Five years ago, an Indiana jury convicted Richard Alexander of multiple sexual assaults and sentenced him to 70 years in prison. Three years later, two other men were charged with the same crimes after one was linked to the assaults by DNA evidence and the other confessed. By then, Richard Alexander had served three years in prison for a crime he did not commit. In other words, that jury blew it too.
But the members of those juries will never be punished for their errors. That means they never had the right incentive to get their verdicts right in the first place.
That's not to say the juries were intentionally malfeasant. It's only to say this: Weighing evidence is a difficult job. It requires a lot of attention and a lot of energy. And it would be a good thing if juries performed that job with diligence. The way to make workers diligent, as every manager knows, is to reward them when they succeed and punish them when they fail. It would be easy to apply that principle to juries: When subsequent evidence reveals that jurors got the verdict right, send each of them a big fat check. When subsequent evidence reveals they got it wrong, hit each of them with a big fat fine. And if you worry the associated risk will discourage people from serving on juries, pay them each a big fat fee for serving in the first place.
With that system in place, there would be a lot fewer jurors falling asleep in the courtroom, or ignoring evidence that might take some intellectual effort to decipher, or allowing others to cow them into submission in the jury room. There would still be some manifest unfairness—cases where jurors do the best possible job, still get the verdict wrong, and are punished despite their good efforts. But manifest unfairness is part of any good incentive system. You can spend years learning the restaurant business, carefully line up your investors and decorators and kitchen staff, brilliantly fill a market niche—and still fail because of a stray rat, a random terrorist attack, or a sudden fad for cooking at home. That's unfair in a sense—but we accept such unfairness as part of a system that encourages entrepreneurs to do the best they can with the limited information they've got, and yields, on average, better restaurants than we could get any other way.
Likewise with juries. Of course juries will make mistakes, and of course some of those mistakes will be unavoidable. But if we punish the juries who acquit the Lemrick Nelsons or convict the Richard Alexanders, we'll get better verdicts on average.
Of course, the goal is not to get jurors to convict everyone they think is probably guilty; it's to get them to convict everyone they think has been proved guilty beyond a reasonable doubt. But it's easy to adjust incentives to get whatever results you want. If there's a general sense that juries are too quick to convict, we can either raise the penalty for a false conviction or lower the penalty for a false acquittal. If they're too quick to acquit, do the opposite. In fact, that's one of the hallmarks of a good incentive system—it's easily tweaked when you want it to work a little differently.
One problem with this scheme is that cases like Nelson's and Alexander's, where the truth emerges eventually, may be too few and far between to concentrate a juror's mind. Rewards and punishments are less effective when they're less likely ever to be delivered. One solution might be to hold occasional "mock trials" of defendants who are known in advance (say through confessions or incontrovertible DNA evidence) to be either guilty or innocent. Withhold the key piece of evidence from the jury, let them deliberate, and then ignore them if they get it wrong and reward them if they get it right. Jurors would never know whether they were sitting on a real trial or a mock trial, so there's always the prospect of a big reward and an incentive to be accurate. The mock trial system would be horribly expensive, but it would also have benefits in the form of better verdicts. Is the cost worth the benefit? I haven't a clue, but it's at least worth thinking about.
In the meantime, there's no reason we couldn't use the occasional Lemrick Nelson case to send jurors a message. Every assembly line worker in America, every cab driver, every doctor and lawyer and magazine columnist, reaps financial rewards and punishments that depend on his performance. Only jurors are excepted. You can justify that exception only if you believe that getting court verdicts right is the least important job in America.