Although in most states and in the federal system, the law’s answer to the title question is “yes,” Youngjae Lee’s answer—with a qualification it will take the rest of this jot to explain—is “no.” To be more precise, his answer, surprisingly, is that it depends on the issue that is liable to disagreement. Making certain assumptions, Lee argues that unanimity is the best rule to adopt for juries reaching decisions about empirical facts in criminal cases. In these circumstances, requiring unanimity among jurors is both most faithful to the beyond-the-reasonable-doubt requirement for conviction and most faithful to the justification of this requirement. But juries must make decisions on all of the elements of crimes (and sometimes on affirmative defenses, I might add); to do this, often juries must make decisions on issues that are at least partly evaluative. (Lee calls them “moral issues.”) Some of his examples come from the core of criminal law: rape (reasonable belief in consent or a reasonable expectation that defendant recognize lack of consent) or homicide (depraved-heart murder, reckless homicide, self-defense). For these decisions, Lee argues, unanimity is not the rule to adopt.
He arrives at these conclusions by assuming a principle of rationality that has lately attracted attention from epistemologists: the “equal weight view.” That view says that if there is disagreement among persons with equal cognitive capabilities and equal access to information (“epistemic peers”), each belief is equally reasonable, and so, has equal weight. Each person should adjust his belief in the direction of those with whom he or she disagrees. In a simple case of 11-1 disagreement where eleven have the highest confidence about the defendant’s guilt, the equal weight view requires that they lower their confidence. Under some circumstances, lowering by the eleven results in an insufficient average level of confidence among all the jurors—insufficient to satisfy the requirement of being beyond a reasonable doubt—so a unanimous verdict of not guilty should be reached. If the sole dissenter is not very confident in his opinion for acquittal, the average belief in the probability of guilt may remain high enough to satisfy the standard of beyond a reasonable doubt and so, a unanimous verdict of guilty should be reached. But not if the level of confidence satisfying the beyond-a-reasonable-doubt standard is very stringent. Then any amount of dissent regarding conviction leads on the equal weight view to acquittal.
Lee then adds other assumptions. One is that jurors are likely to fail to apply the equal weight view consistently—i.e., they do not always adjust their confidence levels in the face of disagreement with those they recognize as epistemic peers. When this happens, he shows, assuming the equal weight view is correct, a supermajority voting standard will sometimes result in a false conviction. A unanimity rule would lead to either an acquittal or a mistrial, due to a hung jury. Something similar happens under Lee’s next assumption: that it is likely that jurors who are very confident of a defendant’s guilt and applying the equal weight rule will not recognize dissenters as epistemic peers. In both cases, given the undesirability of convicting the factually innocent, the unanimity rule leads to better results when jurors disagree. It generates decisions that approximate ones that jurors would reach if they were more rational, Lee claims. Plus, it is a way of enforcing the beyond-a-reasonable-doubt requirement.
But only for the finding of factual matters. On most of Lee’s earlier assumptions, in an 11-1 split on moral issues, the equal weight thesis would require acquittal. But moral disagreement is common. Lee thinks many splits among jurors on moral issues, with various numbers of dissenters, would, on the equal weight view, have to end in acquittals.
The mechanism that generates this outcome, however, seems wrong. It is inappropriate for disagreeing jurors to alter their opinions on moral issues in accordance with the equal weight view. Lee contends that doing so conflicts with the justification for the criminal jury: the jury reflects the community morality and is the community “conscience.” Lee takes the latter word seriously and tries to explain why respecting a juror’s conscience conflicts with instructing the juror to revise a moral judgment in the face of controversy. Simply put, in moral disagreement, it is not rational to treat another’s conscience and one’s own as equally reasonable.
I don’t think he convincingly pinpoints why, for reasons too lengthy to explain here. However, the case for the inappropriateness of an alternation-and-unanimity requirement for moral decisions can be strengthened. If there is a moral truth to which the community is committed, and if exposing that commitment requires advanced moral skills, then the alteration requirement is inappropriate; for rarely will there be twelve jurors with equal moral abilities. It is unlikely that disagreeing jurors are epistemic peers, contrary to one of Lee’s assumptions. (Lee has misgivings about this not-epistemic-peers response.) If, on the other hand, the question of community morality is about the application of a social norm, it is likely the jurors are epistemic peers. However, social norms are indeterminate at points. (Lee remarks that the evaluative terms in question are “vague.”) If the disputed issue falls into this region, a decision must be made, a precisifying. One can argue for the appropriateness of a majority, or a supermajority, on democratic grounds, perhaps; however, given that there are always deviants from social norms, there is no reason to require unanimity.
I said that Lee answers “no” to the title question, with a qualification. Lee ends his article by suggesting that if beyond a reasonable doubt requires the equal weight view (recall that he has made assumptions that are merely plausible), then it may turn out that jury decisions on moral issues should not be required to be beyond a reasonable doubt, after all.