Friday, December 30, 2005

Death penalty study methodology

Donohue, of YLS, has a new paper out on death penalty studies. He reviews various studies, testing them for robustness of their conclusions when you change the assumptions slightly, and finds that most are very fragile. He concludes that there's tons of uncertainty, which is certainly true--although he avoids making what, from my reading, seems the plausible additional step. Namely, considering the deterrence effect of the DP, while the effect's uncertainty is quite wide, the actual mean seems to be solidly on the side of positive deterrence. I'm not sure what significance you can draw from that, but there it is.

Not totally unrelatedly, the crazy libertarian David Friedman has a post about arguments for and against the DP--especially that efficient punishments are a mixed blessing.


Post a Comment

<< Home