IPR Research Notes

How Accurate Are Jury Verdicts?

Winter 2008 , Volume 30, Number 1

Juries across the country make decisions every day on the guilt or innocence of defendants, ideally convicting those demonstrably guilty and acquitting the innocent.  Wrong verdicts do occur, but how can one measure the errors when the truth is unknown?

According to IPR Faculty Fellow Bruce Spencer‘s research, one can use statistical methods to estimate incorrect verdicts on average. In a study of 271 cases from four state courts, he found juries might have delivered wrong verdicts in at least 1 in 9 and as many as 1 in 6 of these particular cases.

“This small study is not representative of a larger set of cases,” cautioned Spencer, a professor of statistics. “But what it does demonstrate is that larger, carefully designed statistical studies would have much to tell us about the accuracy of jury verdicts.”

Spencer relied on data collected by the National Center for State Courts, which obtained verdicts from the judge as well as the jury on the same case. In other words, as a jury was deliberating about a particular verdict, the judge on the case filled out a questionnaire to report the verdict he or she would have issued had it been a nonjury trial.

Spencer found that juries and judges disagreed on the verdict 23 percent of the time for these cases. “This means that we know at least one of them must be wrong at least 11.5 percent of the time for these cases, and there is evidence suggesting that juries are less accurate,” Spencer said.   A more complex statistical analysis suggests that the actual error rate in these cases is closer to 17 percent (or 1 in 6) for juries and about 12 percent for judges.

“Some of the errors are incorrect acquittals, where the guilty defendant goes free, and some are incorrect convictions,” Spencer continued. For the cases under study, Spencer estimated a ratio of 1.3 wrongful acquittals to wrongful convictions by the jury, compared with a ratio of only 0.1 for judges.

Spencer is planning a new, larger study on the comparative accuracy of verdicts in different sets of cases.

“Estimating the Accuracy of Jury Verdicts” by Bruce Spencer was published in the Journal of Empirical Legal Studies 4(2): 305-29 in July.