Moral Certainty and the Constitution

Criminal defendants are innocent until proven guilty beyond a reasonable doubt. That's the standard formulation, the modern one at least. The idea is pretty simple: criminal convictions carry severe penalties, usually imprisonment, these days for lengthy terms, and so if there's doubt as to whether or not the defendant is guilty, we shouldn't impose those penalties on them. Except that there's always some doubt about everything, so we limit it to reasonable doubt. Makes sense. Another version of the same idea, which I believe may have been used in criminal cases in the old days, is the idea of proof "to a moral certainty." Moral certainty is an old idea that seems to have been inspired as a defense against, like, philosophical skepticism. The idea is, well, okay, yes, we never really know anything absolutely for certain, but we can know things to a degree of certainty that's enough to act on. Moral certainty is basically, I would think, that degree of certainty upon which it is morally acceptable/defensible/correct to act. And of course, this standard will be more or less strict depending on the action being contemplated. Sometimes we might think it correct to act on little more than just a hunch or a suspicion, because it's not that bad if we're wrong, or because it could be very bad if we don't act and should have. In, say, the criminal law context, however, we really don't want to convict someone who doesn't deserve it, for the reasons given above, and so it takes a lot of proof to prove guilt to a moral certainty. It's all one standard, but the work it does varies by context.

I've just been reading McCleskey v. Kemp for my Capital Punishment class. That's the case where the Supreme Court rejected a challenge to the Georgia death penalty as racially discriminatory and hence a violation of the Equal Protection Clause despite an incredibly sophisticated statistical study demonstrating that death-eligible killers whose victims were white got the death penalty a lot more often than those whose victims were black. The Court was basically applying standard (twisted) Equal Protection doctrine, which holds (under the Washington v. Davis standard) that a challenge under the Equal Protection Clause bears the burden of demonstrating intentional discrimination on the basis of race or some other improper factor. Merely showing that a policy has a "disparate impact," i.e. that it's worse for black people/women/etc. than for whites/men/etc., is not enough, although a sufficiently stark disparity can be circumstantial evidence of discriminatory intent. So in McCleskey, the Court says that this study shows only the risk that race may infect capital sentencing decisions, not that any given sentence was the product of intentional racial discrimination. After all, the authors of the study conceded that they cannot prove "to a moral certainty" that race influenced any particular case.

And when I read that phrase, I just had to think, like, what does moral certainty require in this case? I mean, I actually think that the idea of moral certainty ought to doom the death penalty altogether,* but let's set that aside. The Court was asked in this case to set aside McCleskey's death sentence as having been improperly influenced by race. If McCleskey is wrong and his sentence was totally race-free, but the Court found for him anyway, well, horror of horrors, he would spend the rest of his life in jail. That's, I mean, if we consider that a by-hypothesis fair jury thought he should hang, then I guess that's not a great result, but certainly, Warren McCleskey's being imprisoned rather than executed isn't in any way repugnant to our fundamental laws or principles of justice. If, on the other hand, he was right, and his sentence was discriminatory, but the Court didn't act, well then, a black man would be executed for the specific crime of killing a white person. And that is a constitutional abomination. And so it seems to me that this is a case where the "moral certainty" standard is very low. In fact I think the "moral certainty" standard should basically be on the other side, that unless we could say to a moral certainty that race was not the deciding factor in McCleskey's case we should grant him his reprieve.

And this, I think, should hold true across constitutional law generally. People often say that our Constitution, with its multiple veto points between the bicameralism-and-presentment legislative system, the President's pardon and prosecutorial discretion powers, judicial review, etc., is designed to minimize the amount of legislating that goes on, that it is inherently biased in favor of government inaction. As Akhil Amar I think rightly notes, however, the system is not really designed to prevent any laws from passing but to prevent unconstitutional laws from passing: everyone gets a chance to disapprove a law if they think it's unconstitutional. And if we buy that understanding of the Constitution's overarching priorities, then shouldn't we have a general rule that we won't accept the risk of a constitutional violation? We already do this in places: the Miranda rule, for instance, is a clear case where the constitutional violation is not interrogating a suspect without having first informed them of their rights, it's tricking someone into confessing who doesn't know that they have a right not to, and we can't know that this isn't happening unless we provide a warning. There are other examples, but there are also lots of places where we take the opposite approach, where the Court in essence seems to view the wrongful exercise of judicial review as this terrible tragedy that we must be so very careful to avoid. But that just seems backwards. The tragedy is letting constitutional violations happen right under our noses, when we know that they might be happening, just because we can't be sure that they are. We should err, in other words, on the side of not violating the Constitution, and judges are supposed to play their part in making sure that we do.^

This is, I think, one of the strongest forms of my problem with modern equal protection doctrine and the Washington v. Davis standard: discrimination on the basis of race is, like, the greatest constitutional evil there is. It's America's original sin, we literally fought a war that culminated in a series of amendments designed to eradicate it, and the only thing that lets us even pretend to be a decent country is that beginning in 1954 we actually started trying to live up to those amendments. So why the bloody hell is the burden on the challenger to prove that there was intentional racial discrimination? Maybe some of the time when policies have a racially disparate impact it will be completely innocent, but like, so what? If we assume ad arguendo that there's this vast swath of racially disparate things governments do where we just can't tell whether or not they have discriminatory purpose, shouldn't we just not do any of those things? Like, even when we're wrong, I'm not too cut up about having governments no longer do things that disparately hurt black people. If that is the cost of my false positives, and the cost of false negatives is allowing deliberate racial discrimination to proceed, like, even under the Washington v. Davis theory that the Constitution only forbids the latter, shouldn't we still be willing to swallow an awful lot of false positives to avoid false negatives? Now, I actually think we can be somewhat more, ahem, discriminating between different disparate-impact policies, and sort out which ones are okay and which ones aren't (basically by putting stronger and stronger burdens on the government to justify its policy choices as the disparities grow wider), but the prime directive of our equal protection doctrine has just got to be making sure that, as far as is possible, we not let organs of the American government do racially discriminatory things.

And in general that we not allow constitutional violations to happen. We, the American people acting through our federal and state governments, don't have to do things. We could not do things. And so, I think, we shouldn't require moral certainty, in any very stringent sense, before courts will stop us from doing things in the name of the Constitution; rather, we should demand moral certainty that our actions are constitutional before we do them. And courts should be similarly hesitant to allow any government action, particularly action that deprives any person of life, liberty, or property, to proceed if they cannot satisfy themselves that it is constitutional. Perhaps there should be some prudential exceptions to that rule, where the practical consequences of inaction in a field of constitutional uncertainty will be truly dire, or--if this is a thing--where the potential violations feel somewhat de minimis. And the rule isn't that all legislation must prove its constitutionality beyond a reasonable doubt, as we mean that in the criminal context. But the judicial attitude should be one of vigilance, of caution not toward wielding the awesome power of constitutional review but toward failing to wield that power when the Constitution's defense required it.




*Once we admit that there's any possibility of wrongful convictions, which at this point is on par with admitting that the earth is round, then, well, we just don't have to execute people; even in cases with the most overwhelming proof of guilt of the most reprehensible crimes, we can just lock people up, presumptively forever, but then retain the ability to at least partially fix our mistakes if it turns out we made one.
^Of course, judges aren't the only ones with this responsibility; in at least some senses, legislatures have primary responsibilities for not passing unconstitutional laws in the first place. This doesn't mean, however, that legislatures should decline to pass any law that they worry the courts will invalidate; doing so would result in a gap between those laws that the courts actually do think are unconstitutional and those that the legislature can actually pass, which would be a needless limitation on the people's ability to self-govern. Rather it means that a legislature should obviously not pass any law which it itself believes to be unconstitutional, and (zooming in one level) legislators oughtn't vote for laws they think are unconstitutional. That includes laws that the courts would predictably uphold!