Thanks, could you respond to my reply to dlthomas, as well?

That's the conclusion I keep coming to, but I have trouble justifying this to others. It's just such an obvious built-in response that bad people deserve to be unhappy. I guess the inferential difference is too high.

Follow Up: What is your opinion of prisons? How unpleasant should they be?

Is the answer to the second question something like "the unpleasantness with the best [unpleasantness] to [efficacy in discouraging antisocial behavior] ratio, while favoring ratios with low unpleasantness and high discouragement"? (Feel free to tell me that the above sentence in unintelligible).