Donald Boudreaux, chairman of George Mason University's economics department, is an excellent economist. I'm a devoted reader of his blogs, Cafe Hayek and Market Correction. Boudreaux also writes a twice-monthly column for PittsburghLive.com. His most recent column reminds me of this old joke:
An engineer, a priest and an economist are trapped together at the bottom of a well. The engineer looks around for a while and thinks he has a solution.
"Given the materials we have amongst us, I should be able to build a contraption to get us out of here," says the engineer.
The priest replies, "No. That will never work. We must pray, and put our faith in God. He will show us a way out of here."
To that, the economist says, "You're both wrong. Assume a ladder."
Boudreaux, who has both a J.D. and a Ph.D. in economics, fancies himself an expert in the "game" of sentencing. In the course of his recent column he asks why we don't "punish rape even more severely -- say, by executing convicted rapists?" His answer
is "if rapists were punished as severely as murderers, the number of murders would rise."
Put yourself in the place of a man who is a threat to rape women. If you learn that rapists will no longer merely be locked in prison for years but, instead, executed, you're a bit less likely than before to rape. That's good. But suppose that this higher "marginal" cost of committing a rape isn't sufficient to prevent you from raping a woman. So you rape a woman. Once you commit the rape, you are subject to being executed if you're caught and convicted.
What will you now lose by becoming also a murderer? Nothing. In fact, you have everything to gain by killing your rape victim. If you let her live, you run a real risk of being identified, captured and convicted -- and then executed. But if you murder the woman after you rape her, you reduce your chances of being caught and convicted. (The chief eyewitness to your heinous crime, after all, will be in her grave.) So with nothing to lose and much to gain by killing your rape victim, you're more likely to kill her than you would be if the penalty for rape were lower than is the penalty for murder.
Punishing rape less severely than murder ensures that rapists still have something more to lose if they kill their victims.
Notice the critical assumption that Boudreaux makes so glibly: "If you learn that rapists will no longer merely be locked in prison for years but, instead, executed, you're a bit less likely than before to rape." (My emphasis in bold italics.) How does Boudreaux know that the threat of capital punishment would make rape just a bit less likely? And just how much is a bit? In short, Boudreaux has arrived at his preferred answer -- don't execute rapists -- by making a critical and unsubstantiated assumption, namely, that the net effect of imposing capital punishment for rape would be more murder.
It is also possible, of course, that the effect of imposing capital punishment for rape would be a lot less rape and, therefore, a lot less murder (to borrow from Boudreaux's arsenal of imprecise terminology). That is, the threat of capital punishment for rape would, at the margin (as Boudreaux is wont to say), deter rape and therefore avert many instances in which rapists would otherwise kill their rape victims because they have "nothing to lose" by doing so.
The correctness of Boudreaux's assertion is an empirical question, not one to be decided by assumption. If, on balance, the threat of capital punishment deters murder, it ought to deter rape. There is empirical evidence that capital punishment does deter murder, just as crime generally declines as the certainty of punishment rises. Score one for me.
Boudreaux isn't finished, however. He goes on to offer us his theory of finely calibrated punishments:
Of course, the same logic applies also to other crimes. We don't execute armed robbers not because we don't want to further reduce the incidence of armed robbery; it's because we don't want to strip armed robbers of incentives to let their victims live.
And likewise for the entire range of criminal sanctions. For all of its imperfections, our current criminal law generally -- and sensibly -- punishes crimes of lesser significance less severely than it punishes crimes of greater significance. Pickpockets impose real costs on society, but (because pickpockets are both unarmed and don't invade the privacy of people's homes) these costs aren't as high as those costs imposed by robbers and burglars. So the law recognizes that it would be a fool's gambit to attack pickpocketing by increasing the severity of its punishment so much that pickpockets shift into robbery and burglary.
I find it hard to believe that Boudreaux believes this. If he does, it's because he's been locked in the ivory tower for too long. For one thing, not all criminals are capable of or interested in committing all crimes; a pickpocket, for example, isn't necessarily a repressed murderer. For another thing, criminal law, which varies from State to State, is roughly calibrated "to let the punishment fit the crime." The kind of precision posited by Boudreaux simply doesn't exist.
The loose and variable hierarchy of punishments imposed by various States, courts, and judges owes more to Biblical precedent ("an eye for an eye") than to Boudreaux's game-theoretic version of sentencing guidelines. Some States impose capital punishment but others do not, for example. If the effects of punishment on crime were as well understood and agreed as Boudreaux makes them out to be, every State would impose capital punishment. And every State would impose it -- with fierce certainty -- in cases of rape as well as in cases of murder.
Does Capital Punishment Deter Homicide?
Libertarian Twaddle about the Death Penalty
Crime and Punishment
Abortion and Crime
Saving the Innocent?
Saving the Innocent?: Part II
More on Abortion and Crime
More Punishment Means Less Crime
More About Crime and Punishment
More Punishment Means Less Crime: A Footnote
Clear Thinking about the Death Penalty
Another Argument for the Death Penalty
Less Punishment Means More Crime