Paul Walker nicely summarizes the David Friedman argument on the law and economics of counting the criminal's utility.
I agree with his analysis entirely when I'm wearing my economist's hat. When doing economics, as economists, we have to set the moral questions to one side. Wertfreiheit.
Now, if the conclusion of a serious study found that murderers' enjoyment of crime were greater than the cost imposed on victims, in other words that allowing murder is Kaldor-Hicks efficient, I'd then take off my economist's hat and put on my amateur moral philosopher's hat and say that murder should nevertheless be illegal because it infringes the victim's rights and because, as amateur moral philosopher, I really don't mind discounting the utility of rights-violators all the way down to zero. But I would be taking off my economist's hat when doing so. If economics gives us the efficiency-based case against murder, as I rather expect it would, so much the better. But if it doesn't, it's far better to present the economics straight up, and then present the value judgments separately, than to pervert the economic analysis by doing things like, oh, declaring at the outset that the murderer gets no utility and that a total discounting of the murderer's utility is consistent with good economic practice.
Aside: wertfreiheit is one of my favorite economics words. Methodenstreit is another. I may currently be engaging in methodenstreit with BERL over wertfreiheit. I like that. Is it strange that I always hear the Sepultura lyrics as "War of Methodology" rather than "War for Territory"?