Our discussion last week on capital punishment generated a lot of comments that are worth discussing in more detail. Since capital punishment is so controversial, we decided to continue the same subject this week. First, let me correct a misunderstanding in some of the comments. I never claimed the evidence is anywhere near conclusive that capital punishment has an important deterrent effect. I stated that the evidence from quantitative studies is decidedly mixed, yet I concluded that "the preponderance of evidence does indicate that capital punishment deters". Although the weight of the positive evidence should not be overstated, the frequently stated claim that these studies prove that capital punishment does not deter is clearly false. My belief in its deterrent effect is partly based on these limited quantitative studies, but also because I believe that most people have a powerful fear of death. David Hume said in discussing suicide that "no man ever threw away life, while it was worth living. For such is our natural horror of death‚Ä¶". Schopenhauer added also in discussing suicide "‚Ä¶as soon as the terrors of life reach a point at which they outweigh the terrors of death, a man will put a an end to his life. But the terrors of death offer considerable resistance‚Ä¶". Nevertheless, the main point of my comment last week was not to try to prove that capital punishment deters murders, but rather to argue against the view that it is "immoral" for the State to take lives through capital punishment even if we assume that the deterrent effect on murders is sizeable. Indeed, I believe that deterrence can be the only reasonable basis for capital punishment. Revenge, retribution, and other arguments sometimes made to justify capital punishment are too subject to government abuse, and have been abused. Some readers interpreted my views as implying that a major goal of government policy should in general be to save lives. That is not my belief. I am against governments interfering, for example, with the rights of people to overeat even when that causes obesity, disease, and possibly early death because overeaters are primarily "harming" themselves. In my view, people should have the right to do that. Murder, on the other hand, involves taking the lives of others, and any reasonable discussion has to distinguish such behavior from individuals taking actions that affect only their own lives. In economists' language, murder involves the most severe negative externalities. If we assume for the sake of this discussion that there are two fewer murders for each murderer executed, the State would reduce two of these severe externalities for each murderer that it executes. This issue of the effect of capital punishment on innocent victims has to be confronted by even those most opposed to its use. And I frankly do not see how any reasonable and relevant philosophy could oppose the use of capital punishment under the assumptions of this example. Admittedly, the argument gets less clear-cut as the number of lives saved per execution falls from two to lower values, say, for example, to one life saved per execution. In this case, I compared the qualities of the life saved and the life taken, to the dismay of some readers. In particular, I wrote that "wouldn‚Äôt the trade-off still be desirable if the life saved is much better than the life taken, which would usually be the case?" I do not see how to avoid making such a comparison. Consider a person with a long criminal record who holds up and kills a victim who led a decent life and left several children and a spouse behind. Suppose it would be possible to save the life of an innocent victim by executing such a criminal. To me it is obvious that saving the lives of such a victim has to count for more than taking the life of such a criminal. To be sure, not all cases are so clear-cut, but I am just trying to establish the principle that a comparison of the qualities of individual lives has to be part of any reasonable social policy. This argument helps explain why capital punishment should only be used for some murders, and not for theft, robbery, and other lesser crimes. For then the trade off is between taking lives and reducing property theft, and the case in favor of milder punishments is strong. However, severe assaults, including some gruesome rapes, may approach in severity some murders, and might conceivably at times call for capital punishment, although I do not support its use in these cases. A powerful argument for reserving capital punishment for murders is related to what is called marginal deterrence in the crime and punishment literature. If say perpetrators of assaults were punished with execution, an assaulter would have an incentive to kill the victims in order to reduce the likelihood that he would be discovered. That is a major reason more generally why the severity of punishments should be matched to the severity of crimes. One complication is that capital punishment may make a murderer fight harder to avoid being captured, which could lead to more deaths. That argument has to be weighed in judging the case for capital punishment. While marginal deterrence is important, I believe the resistance of murderers to being captured, possibly at the expense of their own lives, is really indirect evidence that criminals do fear capital punishment. Some readers asked whether I also favor public executions of convicted murderers, mangling of their bodies, and other methods used in some countries still, and in most countries in the past? I do not because they seem unnecessarily abusive of convicted murderers without any compensating gains. However, I admit I would reconsider this position if it were demonstrated that such added punishments have a large effect in reducing the number of murders. For those who find such a position "barbaric", I would ask how many innocent victims are they willing to tolerate before they might take a more positive position on these additional punishments? Of course I am worried about the risk of executing innocent persons for murders committed by others. In any policy toward crime, including capital punishment, one has to compare errors of wrongful conviction with errors of failing to convict guilty persons. My support for capital punishment would weaken greatly if the rate of killing innocent persons were as large as that claimed by many. However, I believe along with Posner that the appeal process offers enormous protection not against wrongful conviction but against wrongful execution. And this process has been strengthened enormously with the development of DNA identification. However, lengthy appeals delay the execution of guilty murderers, and that can only lower the deterrent effect of capital punishment. So to summarize once again my position on this controversial question, I favor capital punishment because and only because I believe it has "sizeable" deterrent effects. I would join the anti-capital punishment side if this view turns out to be wrong, if it were proven that many innocent persons are wrongly executed, or if it is administered in such a racially biased manner as to wrongly convict many black persons, and to be little used against white murderers. But I do not believe that the available evidence strongly supports any of these arguments against the use of capital punishment.