Deterrent Effect of the Death Penalty: An Extended Time Series Analysis

Abstract

Prompted by a number of important recent Supreme Court decisions, the last few years have witnessed a renewed research interest in the deterrent effect of the death penalty for murder. There has been an attempt by sociologists and economists, to overcome the shortcomings of the classic investigations in this area [1–6]. Recent studies, however, are not without some important limitations [7–10]. The possible implications of one of these limitations–the level of aggregation used in these studies–is discussed in this paper, and an alternative type of analysis is presented in examining the deterrence hypothesis of an inverse relationship between states' levels of execution and homicide rates. Contrary to the hypothesis, a series of multiple regression analyses, utilizing different models of the execution rate-homicide rate relationship fails to provide general support for the deterrence argument. Analysis shows a few select sociodemographic factors (income, education, unemployment, urban population, race) to be better predictors of states' homicide rates than use of the death penalty. Some policy implications of these findings are discussed, as well as limitations of the analysis.