In which I'm a better libertarian than Jane Galt

Toward the end of a piece in which she professes doubts about the guilt of Kenneth Lay and Jeffery Skilling, Jane Galt has this to say about using prison for criminal punishment:

As for number three, I agree with Mr Delong that deterrence and retribution are legitimate questions of justice--but I also think that jail is lousy, immoral, and highly inefficient way to achieve them.

Lousy because jail makes the criminals cost us money. Yes, courts cost money . . . but what costs money is the troublesome process of sorting the innocent from the guilty. We're spending money on the blameless, not the perpetrators. Once they're convicted, we know (as well as frail humans can) that they're guilty. Why should we spend money to punish them, when they could be making money, or hey, just entertainment, for the society they've wronged? Fastow's skills may not be much, but stick an ankle bracelet on him and set him to painting overpasses or something.

Lousy, I'll grant. Immoral, possibly. And those are both problems with using prison.

And she's right, it's also inefficient. But, as software marketers say, that's not a bug, it's a feature!

Indeed, given that 22% of state and federal prisoners in the year 2000 were doing time for drug crimes, which shouldn't be crimes at all, the problem with imprisonment would seem to be that it doesn't cost the government nearly enough.

The lower the price, the more of the good is purchased; this applies to criminal punishment just as it does to ice cream. So if the price of criminal punishment (to the government, that is) is lowered -- by more cost-effective prisons, or by switching to public whipping, or by selling the organs of executed criminals -- more criminal punishment will be purchased.

The marginal prisoner is a lot more like Tommy Chong than Charles Manson: do we really want the price of punishment to drop?

(Via the mysterious knzn, who has become my favorite econ blogger, even though he's on the wrong side of this issue.)