Thus, the Greek philosopher Zeno some 2000 years ago found that he could not show that motion is possible; indeed, his famous paradoxes appear to show that motion is impossible. Though nobody believed them to be true, nobody succeeded in showing the fallacy of these paradoxes until the rise of mathematical logic less than a hundred years ago. But meanwhile, the world did not stand still. Indeed, nobody argued that motion should stop because it had not been shown to be logically possible. There is no more reason to abolish the death penalty than there was to abolish motion simply because the death penalty has not been, and perhaps cannot be, shown statistically to be a deterrent over and above other penalties. Indeed, there are two quite satisfactory, if non-statistical, indications of the marginal deterrent effect of the death penalty. .

             Once a punishment exceeds, say, 10 years in prison (net of parole), there may be little additional deterrence in threatening additional years. We know hardly anything about diminishing returns of penalties. It would still seem likely, however, that the threat of life in prison deters more than any other term of imprisonment. The threat of death may deter still more. For it is a mistake to regard the death penalty as though it were of the same kind as other penalties. If it is not, then diminishing returns are unlikely to apply. And death differs significantly, in kind, from any other penalty. Life in prison is still life, however unpleasant. In contrast, the death penalty does not just threaten to make life unpleasant, it threatens to take life altogether. This difference is perceived by those affected. We find that when they have the choice between life in prison and execution, 99% of all prisoners under sentence of death prefer life in prison. By means of appeals, pleas for commutation, indeed by all means at their disposal, they indicate that they prefer life in prison to execution.