In 2007, the Wall Street Journal published an article claiming that each execution saves more than 70 lives. This example is used to show how easy it is, using simple or advanced econometric techniques, to produce results that do or do not support the deterrence hypothesis. Moreover, we also point to some puzzles which have not been satisfactorily solved so far. We then present a critical survey of the papers published in the last ten years. It is shown how simple changes can produce quite

more »

... rent results using the same data. Finally, we draw some conclusions about the usefulness of statistical arguments in policy debates, but also on the moral questions involved in this particular debate. . 17. This is the coefficient in the long-run equation corresponding to relation (3a"). 18. If we regress the homicide rate in states without executions on the first component, we get a similar picture, with identical values for the R 2 and the t-statistic of the principal component as well as the Durbin-Watson and Q-statistics. To take account of the autocorrelation in the residuals we use for all following models heteroscedasticity and autocorrelation corrected standard errors to calculate the t-statistics.