The marginal deterrence theory suggests to me that all killers are ruthless, and cannot be rehabilitated.
It only requires that the criminal is rational. We know that is not always the case. We also know that psychology often makes the assumption of homo economicus suspect (e.g. the distinction between rational behaviour and rationalising behaviour). However, we don't need a binary: everyone is rational or irrational. We just need for rationality to be a significant feature. Self-preservation effectively ensures that is the case.
But there's also a decrease on the one end(deterrence), and as you say a increase on the other. Could be it evens out.
I couldn't agree for the two reasons I've given. First, the creation of any perverse incentives will ensure the punishment regime is inappropriate. Second, there is likely to be a distinction between the behavioural characteristics of the single and the multiple murder. Given the assorted explanations for murder (many of which are not quite consistent with self-interest), deterrence is likely to be more suspect for the single murder. Take an extreme such as infanticide. Would the death penalty be able to control such behaviour? I very much doubt it.
The multiple murder, in comparison, is much safer ground for homo economicus. Self-preservation is classic self-interest in play and therefore consistent with the marginal benefit and marginal cost comparison.
I've read that some criminals actually prefer jail because they are able to be more comfortable than in their real lives on the outside.
The perception of canny comfort in the prison system is largely a myth encouraged by folk that want to avoid the nasty reality of institutionalisation. I think there are indeed other issues that need to be examined: for example, cross-country comparison into the size of prison populations suggests that either countries have more general social problems (encouraged, for example, by income inequality levels) or have too much reliance on the prison system.
But it would have to work all the time by your estimation, and it doesn't. You won't factor in all the possibilities, and that's why this theory isn't totally right.
No theory is 'totally right'. I'm just not prepared to give up the notion that a significant percentage of criminals are rational agents.
Not all murderers get the DP. Most don't from what I've read. Race, sex economics those are all factors in who gets the DP.
There certainly are other variables at play. A rational criminal would of course include those variables in his expected benefits and costs. We could argue that multiple murders would be less likely, other things remaining equal, amongst socio-economic groups with a lower risk of receiving the death penalty. I'm not aware of any analysis into these specific issues though.
That's where I have an issue. If it can't be dished out fairly, then it's not a good practice.
Inequity is always a worry. However, the death penalty fiend can remark that those problems can, over time, be eliminated. The inefficiency, however, cannot. Any punishment that creates perverse incentives should be avoided like the plague.
