Thursday, April 23, 2020

Unanimity and hung juries

In the wake of Ramos v. Louisiana, students of American military justice are once again pondering such basic questions as the proper or required size and voting requirements for court-martial panels. A frequently expressed concern in conversations about unanimity is the prospect of hung juries, necessitating retrials that may be costly and a diversion of personnel from regular duties. These conversations would be more productive (and potentially helpful to Congress and the courts) if some empirical data were available. What data would you want to have?

American military justice currently has a unanimity requirement for findings and sentence before a capital sentence may be adjudged in a case that has been referred capital. In what percentage of capitally-referred cases is unanimity not achieved at findings or sentence?

Unanimity is required in federal trials. In what percentage of federal criminal cases do juries hang? Some data can be found here. Between 1980 and 1997 the rate averaged 2.5%. Hang rates vary considerably from jurisdiction to jurisdiction, but they remain low. It would be interesting to know if the hang rate is affected by jury size in those states that use juries of fewer than 12. Intuitively, one would think the larger the jury, the greater the chance of a holdout, and hence, the higher the risk of a hung jury. By this analysis noncapital courts-martial, with 4 or 8 members, would presumably be less likely to hang than a civilian jury of 12. (Capital cases probably have a different dynamic.)

What percentage of federal criminal cases in which there has been a hung jury are retried? The Editor has not yet been able to find data on this question. Of those that are retried, some number are resolved by guilty pleas in accordance with pretrial agreements. Again, no data have been found thus far on this question.

If courts-martial performed like civilian federal juries, how many hung courts would there be over the course of a year? In FY18 1491 special and general courts-martial were tried. (Of these, there were convictions in 1268 cases.) If unanimity were required and the hang rate was 2.5%, that would mean 37.275 cases would potentially have been subject to mistrial and retrial. That, then is the maximum impact of a unanimity requirement for FY18.

Why "potentially"? In fact the number of retrials would almost certainly be less than 37.275. In some cases, convening authorities would have decided a retrial was unnecessary or not worthwhile or unlikely to produce a conviction the second (or third) time around. Of those that were retried, some number would be resolved by guilty plea with or without a pretrial agreement, and of those that were contested, some number would be bench trials, resulting in either case in less expense and diversion of personnel for service as panel members.

What is more, the results of such retrials that did occur out of that population of 37.275 cases would have been more accurate. Consider this abstract of a paper that appeared in the International Review of Law and Economics:
Relaxing the unanimity requirement for verdicts in a given criminal trial leads to fewer hung juries and more verdicts of all four types: correct and wrongful convictions, and correct and wrongful acquittals. A criminal proceeding, however, does not necessarily end when a jury hangs. We demonstrate that if retrials occur until a verdict is reached, a unanimous verdict rule is generally more accurate than a nonunanimous rule with respect to the probabilities of all four types of verdicts. Thus, a tradeoff between hung jury costs and verdict accuracy exists when considering unanimous versus nonunanimous verdict rules.
William S. Neilson & Harold Winter, The Elimination of Hung Juries: Retrials and Nonunanimous Verdicts, 25 Int'l Rev. of L. & Econ. 1 (2005). It would thus seem that retrials produce greater accuracy. If so, that would be a dividend of any shift to unanimity in noncapital courts-martial.

Recognizing that federal criminal trial data may not be a perfect surrogate for what would happen if unanimity were required in noncapital courts-martial, and that the rate of hung juries may be higher in a military setting (of course, it might also be lower), and counterbalancing the benefit of increased accuracy in the handful of cases that would go to verdict on a retrial, how do the benefits and costs of such a change stack up?

Comments invited. (Real names only, please.) 


  1. Gene, great analysis. I would suggest the number would be lower than 37. The 1491 number of courts-martial include guilty plea cases that obviously would not lead to hung juries. I believe at least half the courts-martial are historically guilty pleas, so the number of litigated cases would only be around 750. As a result, it would seem very likely there would be less than 20 hung juries a year. I'm guessing this is in line with the number of cases the military retries after findings or sentence is overturned on appeal. Considering the military did almost 10,000 courts in FY90 and 5000 in FY00, it is hard to imagine even a 100 hung juries a year would really be a logistical challenge. In fact, that would still be less than the 2000 courts the military did in FY15. In the Air Force at least the number of JAGs on active duty is almost exactly the same as the early 90s. In other words, there are plenty of JAGs to do a few more cases and they would benefit from the experience.

  2. An afterthought: should unanimity also be required for noncapital military commission cases? See 10 U.S.C. § 949m.


Comments are subject to moderation and must be submitted under your real name. Anonymous comments will not be posted (even though the form seems to permit them).