Guilt, Innocence, and the Death Penalty

Gin and Tacos has more details on the Troy Davis case that point to the perils of judging too hastily. My conclusion earlier in the post on "closure" and Ed's conclusion are basically the same, although we arrive there from different angles. Here's his take:

...Troy Davis isn't the best cause celebre. His legal team chose to risk everything on actual innocence and they appear to have lost. More evidence tends to suggest his guilt than his innocence. But that is the problem with capital punishment, the idea that Good Enough is good enough. Well, we convinced a jury of his guilt, so now let's apply a punishment that can't be reversed if we later realize that a mistake was made.

This case is not the best example of the kind of Innocent Man on Death Row scenario that capital punishment opponents like to publicize. It is a great example, however, of how the process of determining guilt simply does not allow us to be as certain as we would need to be to apply an irreversible sentence. The limited ballistic evidence is disputed by opposing experts. Some witnesses have recanted. Other witnesses failed to definitively identify Davis as the shooter with certainty. I don't expect that the legal system will release people back into society at the slightest doubt of their guilt. But with the death penalty, the slightest doubt should be enough. Irrespective of the choice of legal tactics on the part of his defense, the simple fact remains: If there is any doubt, you can't pull that switch.

Precisely. There are basically two general moral frameworks for arguing in support of the death penalty: a deterrence theory approach and a retributive justice approach.

The former, deterrence, rests on the premise that the existence of the death penalty deters would-be murderers from committing murder because it is possible they could be caught later, convicted, and eventually put to death. It's notoriously difficult to say with any certainty that the existence or non-existence of a given policy has prevented acts that have not occurred given that any basis for proof comes in the form of those acts not occurring. I suppose someone could say in all honesty, "I was going to kill the dude, but then I thought that, hey, I'm in Texas, I could be put to death for killing this dude. Better not do it." But we can say that pretty much anything at all prevented something that didn't happen from happening: e.g. "It's a good thing I had milk in my coffee this morning or the Russians would have invaded Kazakhstan."

Empirical studies on the question of deterrence seem to be a wash, with a new study coming on the scene every few years purporting to demonstrate one or the other, that capital punishment deters crime or that it does not.

Regardless, two points are especially troublesome. First, if the death penalty were found conclusively to deter capital crimes, then why not chop off hands for stealing to prevent that crime too, or guillotine people who don't pick up their dogs' crap from the sidewalk?

Second, if executing someone every so often did indeed deter violent crime among the public, and the moral justification and practical objective of capital punishment were deterrence, then the people executed from time to time would not need be actually guilty of any crime at all. The goal and justification here would be a kind of aggregate social utility - maintaining a society relatively free of violent crime for the rest of us. Just pluck someone randomly off the street (best if he looks like a "bad guy"), conjure up a supposedly airtight case that he's a horrible murderer, and put him to death in an execution covered minute-to-minute by the national media. Bonus social effects might be included by selecting the person to be executed on the basis of characteristics exemplifying various social fears and prejudices and/or political goals of the state.

A deterrence argument is a very shaky argument.

Justification of the death penalty on the grounds of retribution is sometimes referred to as a matter of "restoring moral balance" in society. Frankly, I'm not sure how to parse that claim philosophically. But it seems basically to mean that a society that wishes to uphold basic principles of justice must live according to those principles and moral-legal rules. Those who violate them must be separated from society as an act of restoring society to that state of upholding its principles of justice and, as a consequentialist add-on, securing the safety of the society. Punishment as a means of acting on and upholding a society's conception of justice and moral values implies that the those who live their lives justly (in whatever sense that's meant by a given society) deserve to be members of that society and those who violate the values no longer deserve by their very actions to live in the society. Indeed, the criminal deserves his punishment where that punishment, whatever form it takes, serves to withdraw him from society.

Nothing here says necessarily that the death penalty should be the form of punishment for a retributionist. A criminal can be taken out of society through imprisonment, exile, or launching them into space, whatever.

The philosopher Jeffrey Reiman once made a sharp little Kantian argument for the death penalty (though his own position is anti-death penalty, primarily for practical reasons). Without going into a tedious recounting of the categorical imperative and such, let's say roughly (and borrowing only partly from Reiman, and not ascribing these views/interpretations to him) that the idea on this account is that if we are acting morally, then we are acting as we would presumably expect everyone else to act. Each action a person makes suggests that the act is morally justified for that agent if the rule or "maxim" it implies can be applied universally. If I steal a pack of gum, that act in itself suggests the maxim that stealing is morally permissible (for everyone).

Kant thought that "reason dictates morality," in that hypothetically applying such maxims universally would show us their logic and thus morality or fundamental contradictions and thus immorality (or incoherence, really). One of Kant's own examples was the act of not repaying a loan. The maxim implicit in that act is something like, "it is not necessary to repay our loans." The maxim is incoherent, contradictory, because if we imagine it to apply universally, it says for everyone that it's not necessary to repay loans. But if this were universally the case, there would be no loans in the first place because no one would loan anything to anyone and no one would be borrowing anything, which entails repaying what's borrowed. Such contradictions show us which maxims can be taken to be moral laws and which are not. In this example, it is therefore not moral to not repay our loans.

If someone commits murder, the act itself suggests that the murderer subscribes to the maxim that killing another person is morally permissible. Continuing the Kantian line of thinking, because the would-be maxim applies universally, the murderer morally justifies his own death. Putting the murderer to death is thereby permissible for a society that presumes equal treatment for all its members such that people get what they deserve based on their actions. Society, in "restoring the moral balance," would be justified in executing the murderer.

Now, there are all sorts of problems with this line of argument in favor of capital punishment (e.g., why not rape the rapist, etc?), even beyond my simple portrayal here. But, although I oppose the death penalty, I think this is strongest kind of moral argument that can be made in favor the death penalty.

Assuming for argument's sake that this retributive approach justifies the death penalty, the point I want to return to that started this wordy post is that the guilt of the person to be executed must be absolutely certain for the retributive justification for the death penalty to stand morally at all. Again, the argument is made by combining two basic points:

1) The very basic ethical principle of desert - that we (should) get what we deserve based on what we actually do.

2) The rough Kantian argument that we ought to act such that we could imagine the "maxim" behind the act to apply to everyone. Thus, universalizing the murderer's act of killing someone else justifies someone or society killing him.

The approach concludes that the murderer thereby gets what he deserves with his execution.

What if it is not certain that the person to be executed is truly the murderer? In the retributive argument above, there can be no uncertainties about guilt since it's only the action itself that justifies the punishment. If the person didn't commit the act, there's no moral justification for the punishment. Uncertainty about guilt or innocence runs a morally dangerous game. This seems like an obvious point, but let's play it out one step further.

If an innocent person is put to death, no matter whether legal procedures found him guilty (mistakenly), that act by society/state in turn suggests the universal maxim that the state takes the killing of innocent people to be morally permissible. If so, then, in the Kantian hypothetical, the executed person is guilty of nothing since the state/society establishes with the very act of the execution the moral permission of killing innocent people. We're led to a contradiction that can be phrased in several ways, but I think it's probably clear.

Let's back out of this scenario, back to the world as it actually is, where it's pretty well-established that we believe murder - i.e. killing innocent people - is morally evil. If the strongest moral case for the death penalty is something like the retributive argument above, then only absolutely certain guilt justifies the execution of a person by state/society. Without that certainty, the death penalty is not morally justifiable (at least if one accepts the retributivist argument). If the person is truly innocent of the crime, then the state, by its very act, has not simply made a tragic mistake but has itself committed murder.