The recent exchanges about the deliberate-infliction-of-pain punishments remind me that I've long wanted to write a few thoughts about the death penalty — and especially about why even people who are generally conservativish-libertarianish like me may want to oppose it. I should note at the outset that many people have thought much more than I have about the subject, so I'm not sure how much I can add. Nonetheless, at least some of these points aren't, I think, disucssed as often as they should be.

First, let me mention again that I do support the death penalty, because I think it's the just punishment for sufficiently heinous murders. I'm not sure how much of a deterrent effect it has, and I don't think the financial arguments alone (even if the cost of the death penalty appeals and delays is reduced) suffice to justify it. So the ultimate and in my view adequate justification for it, I think, is retribution.

Nonetheless, I think there are some very important arguments against it. The risk of error, and the irreversibility of error, are two obvious ones. In practice, I think errors in death cases are on the aggregate more likely to be corrected than in life imprisonment cases, because death cases draw much more attention from lawyers, judges, and others. But the availability of new technologies (such as DNA evidence) might change the balance here; likewise, greater acceptance of the death penalty might change the balance, too. In any event, I will set this point aside just because it's so obvious. Let me mention instead three arguments that are less commonly heard, but that conservatives and libertarians should take seriously:

1. The Utility of the Death Penalty as a Means of Silencing Dissenters: The death penalty is an especially powerful tool for repressive governments, because it can let them easily — and with seeming legitimacy — dispose of dissenters. This is so even if the death penalty is limited to murder; they can trump up a charge of murder, and quickly put the dissident to death. Had he been allowed to live, he might have eventually been freed when the government changed, by revolution or just by softening; he might also have become a focal point for public agitation, both in the country and outside it. (That happens to martyrs, too, but under many plausible conditions a live imprisoned dissident attracts more attention than an executed one.)

Now this is likely to matter only to governments in the middle. We don't worry much about decent governments doing this; and the really heinous ones will kill whomever they want no matter what the legal rules surrounding the death penalty have been. Moreover, even if the death penalty is illegal, and an oppressive government cares about legality, it can always change the law. And some oppressive governments may find themselves politically constrained not to use the death penalty for dissidents even if the penalty is actively used for murderers (consider the Soviet Union in the 1960s and later).

Still, one can certainly imagine many governments (1) that are oppressive but not completely unresponsive to popular opinion, within the country or outside it (2) that find it too politically costly to reverse a firm, well-entrenched, and broadly agreed-on no-death-penalty rule, and (3) that would be willing to trump up capital charges to dispose of dissidents even if they aren't willing to just kill the dissidents extrajudicially. When those governments are in power, and they may one day be in power even here, a firm traditional rule that the most they can do is lock up convicted criminals may provide a check on them.

2. The Utility of an Anti-Death-Penalty Rule in Free Countries: Moreover, even if we don't worry too much about possibly oppressive future regimes here — or if we have confidence (whether or not misplaced) that the error rate for the death penalty in our country won't be too great — we may want to discourage the death penalty in other countries, such as emerging but fragile democracies or mildly oppressive autocracies. Even if we trust our system, we may not trust theirs.

Naturally, we could try to persuade them that the general rule should be "No death penalty unless your legal system is really good; yours isn't as good as ours, so you shouldn't have the death penalty, even though we do." We might even believe that this argument is both morally and factually accurate. But it may well be quite unpersuasive nonetheless. The most effective way to deter the death penalty in such countries — a death penalty that, as I argue above, could be a tool for political oppression as well as posing a risk of normal error — might be to have a flat "No death penalty, either in our country or in yours" rule.

Those who are hostile to molding U.S. law to European norms (and who don't think that there's much of a risk of internally oppressive government in the U.S.) might look at it this way: Don't think of abolition of the death penalty as surrendering to European views. Rather, think of it as a tool for us to help protect Europeans the next time some European countries turn internally oppressive.

3. A Precedent for Limiting Government Power: Recall that in the modern state the government has very broad constitutional authority. It can take our property, either because of our crimes or for a variety of other reasons. It can lock us up. It can put is jail for saying certain things (relatively few things in the U.S., but more in other democratic and otherwise liberal countries). It can broadly interfere with our professions and businesses. Maybe it shouldn't be able to, but in fact it can.

The governments or free countries often refrain from exercising these powers, and are often legally constrained in exercising them. But they're always available, either within the existing law or with a few changes to the existing law. People may grow up hearing that they have an inalienable right to liberty, but when they see the world as it is, they recognize that their own countries — whose legal systems and institutions they are generally taught to respect or even love — restrict liberty in all sorts of different ways.

Clear, simple, and consistently adhered-to rules that limit the government's power can help fight this sense that the government is all-powerful. We can't have a clear rule that the government can never lock people up, or can never take their property, or can never restrict their speech, or even can never kill people. But as to killing, we can at least have something close — at least during peace-time, government officials may not deliberately kill people except in actual self-defense (or defense of others) against imminent crime. Both the exceptions (war and self-defense) are intuitively and historically understood. Both involve the most urgent of necessity. The rule isn't perfectly simple (few rules for behavior can be); but it's probably as simple as possible, and if it's broadly accepted and internalized it reminds both the governors and the governed that there are strong limits to legitimate government power.

4. Radical Distrust of Government: Finally, for really hard-core libertarians (and I'm not one), the "risk of error" argument isn't simply "all legal systems have risks of error" or even "we have particular problems in our legal system that magnify the risk of error." Rather, hard-core libertarians believe that government is naturally extremely prone to error, both moral and factual. They believe (and this is an oversimplification, but I think not a gross one) that government posts tend to attract not very good people; that they tend to make good people worse, by a combination of bad incentives and the corrupting effect of power; and that the institutions tend to limit good officials' power to do good, and magnify bad officials' power to do bad.

If this is so, then one might oppose the government-imposed death penalty because government is simply inherently corrupt, and not to be trusted to do anything but the absolute bare minimum to prevent murderers from roaming the streets. (The true anarchists wouldn't even have the government do that, but I'm sticking with hard-core libertarians here.) In fact, one might wonder why many hard-core libertarians do support the death penalty, as in my limited personal experience they tend to do. My guess is that many libertarians believe there's a basic human right to retribution for crimes done to yourself or your close relatives, and that if the government is to take it away from us, it has an obligation to provide us with some government-conducted substitute.

These, it seems to me, are powerful arguments that conservatives and libertarians need to seriously consider. (I stress again that they are only a subset of the arguments in the debate, chosen by me precisely because in my likely idiosyncratic view they're especially likely to be appealing to conservatives and libertarians.)

In the face of these arguments, it seems to me, the strongest reason to support the death penalty is a belief that it is morally wrong to allow some people to live — the view that every day that (say) a mass murderer can live and enjoy life is a continuing wrong to his victims and to the victims' loved ones. That is my strongly held view, and I suspect that it is the reason why most supporters of the death penalty ultimately support it. Yet I must recognize that there are substantial potential practical and moral costs, including costs that especially resonate with me because of my conservative and libertarian views, to accepting this moral imperative.