Key points are not available for this paper at this time.
Abstract This article analyzes daily homicide data to address whether the occurrence of executions exerts a short-term deterrent effect. The data analyzed are drawn from computerized death certificates from California during the early 1960s, the last period during which frequent executions were carried out there. The analysis of data from one legal jurisdiction helps overcome criticisms of earlier studies based on data aggregated over executing and nonexecuting jurisdictions. Furthermore, the detailed information in the death certificates allows one to analyze the murders of victims of different race and sex separately, which in principle should allow one to detect possible race- or gender-specific effects too small to be discerned in aggregate data. Poisson and compound-Poisson regression models are estimated, to control for seasonal fluctuations and to measure the effect of executions on daily homicide counts. Two-week and four-week periods surrounding the dates on which executions took place are examined for possible deterrent effects, and a large-sample variant of Scheffe's procedure is used to construct the appropriate critical regions for the resulting multiple comparisons. The data are seen to provide no support for the notion that executions deter homicides in the short term.
Jeffrey Grogger (Fri,) studied this question.