What’s a “Jury,” Anyway?
It’s Election Day! So it’s time to distract myself from impending national doom by writing about the constitutional law of what a jury is.
Justice Gorsuch, you see, has argued in a dissent from a denial of certiorari that the Constitution requires twelve-person juries. This is a fascinating little issue that I’ve actually thought about before: I think it’s one where the different modalities of constitutional argument yield wildly different results. (Hat tip to Evan Burnick for bringing the matter to my attention!)
We can start with history, which is what Gorsuch is using and which, in this case, actually backs him up! To the best of my knowledge, juries of fewer than 12 people were unknown to the Anglo-American common law tradition as it stood in 1791. So too were juries that could reach verdicts without a unanimous vote. I think the historical argument that, when they put the word “jury” into the Constitution in a few places, this is how they understood it is quite strong.
If we just look at the text and not the history, though… well it’s pretty hard not to notice that it just doesn’t define what a jury is. All the Sixth Amendment gives us by way of modifiers is “impartial;” the Seventh Amendment and Article III don’t even give us that. (Has anyone ever tried to make the quite obvious, albeit really dumb, textual argument that jury impartiality is not required in civil cases??) You can easily imagine that the text might specify that there must be twelve jurors, or that the verdict must be unanimous; it does neither. Now the originalist will respond that they didn’t need to specify because that went without saying, and fair enough. But that’s a clash of modalities. I think it’s totally fair for the textualist to say that these contextual glosses are not actually part of what the word “jury” means.
As I said in response to Evan on Twitter, I’ve always been inclined to think that the structural perspective supports the requirement of a unanimous verdict in criminal cases, but does not necessarily constrain the size of the jury. The unanimity rule, from this perspective, is related to the requirement of proof beyond a reasonable doubt. If there’s a dissenting juror, doesn’t that sound like reasonable doubt?
One thing I like about this structural perspective is that it easily allows us to distinguish between criminal and civil juries. Proof beyond a reasonable doubt applies only in the former: if the unanimity rule is derived from that requirement, then logically it too should only apply to the former. Indeed I think the structural perspective would say that, as proof in civil cases may be by a mere preponderance of the evidence, so too can the verdict be by a mere preponderance of the jurors, i.e. a simple majority rule. (The technical point that a tie is meant to go to the defendant would be quite a bit more meaningful with numbers of jurors than with quanta of evidence!) Neither the textual nor the historical mode, I think, is capable of making this distinction: if a “jury” simply meant twelve and unanimous at the Founding, then it means that for the Seventh Amendment as well as the Sixth.
Now, technically you can say that one dissenting view on the question of whether there is reasonable doubt does not imply that there is, in fact, reasonable doubt. (This is, I guess, a doctrinal perspective? Heh.) But maybe we should think of the reasonable doubt requirement not so much as a technical doctrinal matter but as an expression of part of the constitutional ethos: to wit, that we are to be exceedingly careful with the awesome power of criminal punishment. “It is better to let ten guilty men go free than to imprison one innocent man,” that sort of thing.
I’ve heard it said that, in the old days, another way that “reasonable doubt” was expressed in jury instructions was the phrase “to a moral certainty.” This is entirely my own thinking, but I’ve always liked that. What it connotes to me is the idea that we should act such that, if we were later informed that we had made the factually incorrect choice, we should still say that we did what was right at the time. Reasonable doubt expresses this perfectly: we cannot really avoid convicting people as to whose guilt there is some nonzero doubt, not if we wish to have a society meaningfully governed by law. So when we convict someone who turns out to be innocent, but whose guilt was not “reasonably” in doubt at their trial, we have acted only as we should; mistakes of that form are unavoidable. That is not true if the doubts were reasonable — and in fact I think we can give content to what makes a doubt “reasonable” through this very notion.
The unanimity requirement is another way of implementing the same requirement of moral certainty, at an institutional level. We want to go as far as possible in the direction of avoiding wrongful convictions while respecting the necessity of basically having a criminal law, and a nice clean way to do that is to say that we shall only inflict punishment in cases where no juror doubts that punishment is needed. It is again not a matter of tight analytic logic, but of doing as much as one can to avoid mistakes.
But none of this has anything in particular to say about the size of the jury. Any particular size requirement must needs be arbitrary: the cleanness of the unanimity rule cannot be found here. We could always say “well but perhaps we should ask one more person, just to be sure,” with no end in sight, and therefore we cannot be generally obliged to say that in the first place. There would, I suppose, come a point at which a jury was so small that the unanimity rule did not feel meaningful — and now we have lapsed into the register of prudence. What this point is, exactly, is hard to say. But I see no particular reason why the requirement would be specifically twelve.
Of course doctrine is not necessarily averse to arbitrary standards, when any standard would be arbitrary. And even if we do not think the history is binding, we could view it as providing a convenient anchor. In other words, if there must be some lower bound on the size of the jury, so as to prevent the absurdity of a one-person “jury,” then why not choose the number with a historical resonance?
Conversely a prudentialist might worry that requiring a full twelve jurors would undermine the right to a jury trial itself, by making such trials that much more expensive and thereby encouraging the state to avoid them via plea bargaining.
Returning to the question of unanimity, I think that a doctrinalist would likely favor that rule, if only because any lesser standard — or at least any standard intermediate between unanimity and simple majority — would have to be chosen somewhat at random.
As for prudence, well, on this as on so many issues I think it can point in any number of directions. The argument I made above about doing as much as possible to avoid convicting the innocent has a bit of a prudential flavor to it. Conversely Akhil Amar has argued that the pluralism of contemporary society makes it unreasonable to expect unanimity from our juries, and in fact that to make that requirement vaguely workable we have had to actively sculpt our juries in a manner most inconsistent with their original conception as a broad cross-section of the community.
(Certain political economists have argued that the unanimity rule is actually bad for defendants. This I suppose qualifies as a prudential argument, though on the merits it is a very very bad one: the model used to make that claim bears almost no resemblance to real-world criminal juries.)
On both questions, then, we get a split verdict — as it were — among the modalities. But the case for requiring unanimity seems to me generally stronger than the case for requiring twelve jurors. Unanimity is favored by history, doctrine, structure, and ethos, with prudence characteristically split. Conversely the twelve-person requirement is really favored only by history and doctrine — and the latter not entirely independent of the former — while finding little or no support in text or prudence or structure or ethos. If I were Supreme Chancellor I would probably say that criminal juries must be unanimous but that there is no particular size requirement (other than the prudential rule against the literal reductio ad absurdum). But I could not fault anyone for saying either that both are required or that neither is.
