Retribution

Mark Kleiman has a good post on retribution. It changed the mind of a least one blogger, Eugene Volokh. I'd like to commend Eugene on admitting he was wrong on this issue.

Eugene was arguing for retribution and Mark argued against it. Now that is not entirely true since Mark did make some good points on why things considered retribution by some should be allowed.

On reflection, I'm not sure "retribution" is really the most helpful term to use, because it conflates three related but distinct ideas.

One is that the suffering of the person punished is a good in itself, because he deserves to be punished. That's a view with which I disagree, though I'm a little surprised that none of the bloggers denouncing it as "barbaric" or "medieval" has bothered to notice that it was the view, for example, of Kant, the beacon of the Enlightenment. (Kant goes so far as to say that the criminal is "owed" his punishment, and that withholding it is an injury to the criminal.) Let's call that view "just deserts."

A second idea that goes by the name of "retribution" is that the punishment of the offender is in the interest of the victim, the victim's family and friends, and those who share characteristics with the victim. If you regard the satisfaction of someone's preferences as being in that person's interest, then this version of the retributive idea -- call it "vindication," as opposed to "just deserts" -- is obviously true. Most, though not all, victims and their intimates want the perpetrator punished, and punished severely. How much weight the satisfaction of that preference deserves in policy-making is a different question, but that vindictive punishment benefits the victim in this limited sense it would be hard to deny.

But vindication also serves the victim's interest in a more material sense: being the victim of a crime is a blow to self-esteem and a source of the loss of social standing, and the punishment of the perpetrator helps reverse those losses, with more severe punishment being more effective in that regard.

It's not hard to see why that should be true. As a matter of sociological fact it is safer to aggress against lower-status people, and riskier to aggress against higher-status people; that's part of what status means. Donald Black has documented what seems to be a universal rule that the severity of punishment tends to rise with the status of the victim. So victimizing someone signals to that person, and to others, that the victim is someone who can safely be picked on.

If the victim, or his kin, or the state, inflicts damage on the perpetrator, that demonstrates that the victim was not in fact someone who could be picked on, and the more severe punishment the more effective the demonstration. That tends to make the victim feel better about himself, and to increase the esteem in which the victim is held by his community. (The fact-value distinction, however valid philosophically, has little weight psychologically: that the perpetrator isn't punished, or isn't punished much, suggests that he didn't deserve much punishment.) Severe punishment might also make that victim, and similar victims, less attractive targets for future victimization. That's not the same as general deterrence, because it's specific to the victim or victim class, though obviously the phenomena are related.

A third meaning of "retribution" is that punishment serves an expressive purpose as well as more immediately practical ones. Punishment announces the judgment of the community about the wrongfulness of the act, and the extent of that wrongfulness.

I think these are important distinctions that are often lost on people who are anti-retribution.