The Problem of Hung Juries -- and How to Solve It

This column originally misstated that no other modern country used the unanimous decision rule for criminal jury trials, but has since been corrected at the author's request to note that Canada and parts of Australia use the rule. -- Ed.

Another story of jury "deadlock" has made the papers. The defendant is Osama Awadallah, accused of lying to a grand jury investigating the 9/11 attacks. The jury charged with adjudicating whether Mr. Awadallah in fact lied failed to reach agreement on a verdict -- because, according to press reports, a lone holdout refused to convict. Now the government faces the choice of dropping the case against Awadallah, or using considerable resources on a retrial.

We need to stop giving holdouts the power to prevent convictions. We don't require unanimity for any other important decision in our pluralistic polity, so why this one? No other modern country (save Canada and a few jurisdictions in Australia).

My point isn't that we should make it easier to convict: It's that we should make the process of reaching a decision fairer to all parties concerned and more democratic. Accordingly, I believe that just as we must relax the unanimity decision rule for conviction, we must also render it much easier for juries to acquit.

Currently, both acquittals and convictions must be unanimous -- in federal court and in forty-eight states. That's not fair to the defendant: He fails to get the benefit of a clear outcome that would allow repose, and he faces the risk of retrial, even if eleven jurors thought either that he was innocent, or that the government had failed to prove its case.

I recommend that a supermajority be required to convict, and a mere majority be required to acquit. These reforms would effectively abolish the hung jury.

Beyond Hung Juries: New Jury Decision Rules Would Have Other Benefits Too

Hung juries -- while costly for all involved (prosecutors' offices, judges, defendants, and witnesses) -- are not especially common. We hear about them frequently because they seem to happen most often in high-profile cases, perhaps when their cost to the justice system is highest. But the elimination of hung juries is not the only reason to rethink our jury decision rules.

As noted above, adopting a supermajoritarian rule for conviction would remove an anomaly in our system: It would harmonize our jury decision rule with the decision rules prevalent in political and constitutional life more generally (as well as with the rules prevalent in other common law countries). We don't expect our Supreme Court justices or our appellate judges to agree unanimously to render a decision, so the idea that citizens are expected to come to complete agreement is rather bizarre.

To be sure, there are unique features of the jury that may counsel in favor of departure from the decision rules prevalent in political and constitutional life. But the jury is undoubtedly a political institution through and through -- and it is becoming even more so as we recruit more diverse members of our communities to serve, as we make it more difficult to avoid jury service, and as we make it harder and harder to challenge jurors lawyers may not like.

There is no question that the stakes are high when the criminal jury issues verdicts: a person's liberty is often at stake (if not his life). Yet, the general laws under which law-abiding citizen live undeniably coerce us every day, and many decisions jurors make are as often value choices as they are simple "fact" determinations.

Unanimity Shouldn't Be Equated With Certainty

One might argue that unanimous verdicts are more "legitimate" because they are more certain to be correct. But evidence suggests that such verdicts offer only the appearance, not the reality, of certainty.

First, most empirical evidence concludes that verdicts reached under various decision rules do not vary in any dramatic way: unanimity requirements seem to produce the same number of convictions that supermajority requirements do.

Second, there is empirical evidence that people assigned a unanimous decision rule will often agree to decide by supermajority (albeit unanimously) -- meaning that in such cases, there is no difference between the two rules, except that some jurors agree to pretend falsely that they agree with a verdict from which, in fact, they may strongly dissent.

Finally, there is evidence corroborating what we all know: that many people are willing to falsify their preferences by changing their vote, even without such an agreement -- and the incentives for doing so are especially high when one is a holdout juror. Peer pressure is real -- and not all holdouts are as headstrong as David Lipschultz, the lone juror unwilling to vote for conviction in the case of Osama Awadallah.

For all these reasons, our security in the purported certainty afforded by unanimity should be questioned. In fact, because supermajorities may have fewer incentives to compromise to win the support of holdouts, greater accuracy might be accomplished through supermajority rules. Suppose eleven jurors believe the government has made an airtight case for first-degree murder, but there is a single holdout. A compromise might result in a second-degree murder conviction -- even despite the presence of clear evidence of premeditation. Such a verdict may reflect jury bargaining, but it departs from the evidence, accuracy, and justice for the victim.

To the extent that one likes unanimity not for actual help in getting the "right" answers, but because one thinks it creates good public relations for the jury system, there is no easy rebuttal. It may be true that the public associates unanimity with legitimacy -- but as I have just argued, this is not a well-founded association to make. We don't have such stringent rules in the context of the civil jury or courts martial, yet both those systems are respected and deemed legitimate by the public; and the few states that have experimented with relaxed jury decision rules for the criminal jury -- Oregon and Louisiana -- do not show evidence of especial deficits in legitimacy or the appearance thereof (nor have they prompted popular efforts to return to unanimity).

Finally, unanimity's proponents must address the truth that unanimity is often an illusion. Besides the reality of compromise verdicts, there is also the reality of retrial. Suppose a jury hangs, leading to a retrial. The second jury empanelled to try the defendant then convicts unanimously. Despite this second "unanimous" verdict, the truth is that of a total of twenty-four jurors who heard the evidence, fewer than twenty-four were persuaded to convict. Indeed, it's possible in theory that only thirteen -- a bare majority -- were persuaded, if in the first trial, the holdout favored conviction, not acquittal.

The final tally in such a case, where a trial is followed by a retrial, cannot be considered unanimous. Yet no advocate of unanimity truly feels that the ultimate supermajoritarian result is illegitimate. How, then, can unanimity supporters claim that unanimity is necessary for legitimacy, or the appearance of legitimacy?

Supermajority Rules Need Not Lessen the Quality of Jury Deliberation

Quite apart from the results a decision rule produces in the way of verdicts, one can reasonably be concerned with the quality of deliberation that would take place under a given decision rule. On this score, the unanimity requirement fares somewhat better than a supermajority rule -- but other changes to the system could close the gap.

To begin, it's true that the relevant empirical evidence on this matter, while not univocal, suggests that unanimity (as compared to less exacting decision rules) is more conducive to better (that is, longer and more satisfying) deliberation -- with jurors feeling better about themselves and their fellow citizens. It is certainly possible that deliberation could suffer under a relaxed decision rule -- with a straw poll ending debate, and with jurors in the minority feeling it was a waste of time to voice their views since the requisite majority was already against them.

But there are two possible ways to improve deliberation under a supermajority rule: First, one could simply have courts adopt a policy of refusing to accept any verdict in the first few (say, four) hours of deliberation. This is effectively what the British did when they relaxed their unanimity requirement in 1967. When jurors know they must deliberate for several hours, they won't just take a straw poll and sit around; they will allow all jurors to air their views, and carefully consider them. It would be a callous jury that used its four hours merely to chat.

Second, we can also stimulate deliberation by simultaneously relaxing the acquittal decision rule -- as I advocated above. Under a majority acquittal rule, dissenting jurors will be encouraged to engage in the proceedings because their preferred verdict seems much more attainable; bringing a few of their colleagues around will not only stave off conviction but may actually result in acquittal. That engagement can lead to higher- quality deliberation.

Ultimately, supermajoritarian rules for conviction, combined with majority rules for acquittal, actually render it easier for these minorities ultimately to "win." Far from giving minorities only empty votes, as some critics of supermajority rules charge, the hybrid rule actually empowers minority jurors because it gets them closer to their preferred verdicts.

More, even when superminorities lose, being able to be "on the record" as dissenters would actually give them concrete voices. Under the current system, the members of a headstrong superminority can only hang a jury -- meaning that no real record of their views remains and that their voices get thwarted: the accused often either stands trial again or pleads guilty to a lesser offense.

Finally, the feedback the government could potentially reap from dissenters going "on the record" could lead to better laws and policies. Making political decisions with the input of a pattern of juror dissents would ultimately be empowering to the dissenters -- and would make for more deliberative policies than using public opinion polling and interest-group lobbying alone. We are a culture that prizes dissent: we expect it from our legislators and our courts, and there is no good reason not to reinforce the chance to voice dissent on the record in our criminal juries. At the very least, such on-the-record dissent is better than a false consensus.

In the end, although I believe both the acquittal rule and the conviction rule should be changed, I also deem the unanimity requirement for an acquittal to be a much greater affront to our system of justice. This rule should be changed even if the unanimity rule for conviction is not. We purport to presume the innocence of our defendants; and we often take pride in our justice system because it puts a high burden of proof on the state. But by requiring unanimous agreement to acquit an individual of a crime, we undermine that central commitment of our free society. We must make it easier for jurors to acquit.