Some Further Evidence on Homicide and a Regional Culture of Violence

Abstract

Gastil's theory and investigation of homicide and the Southern tradition of violence are briefly examined [1]. Although his attempt to construct a quantitative index of Southerness and relate it to states' homicide rates provides a decided improvement over the usual practice of simply comparing “Southern” and “Northern” states, his analysis suffers from the use of homicide figures issued by the U.S. Public Health Service as an index of his dependent variable, criminal homicide. Inspection of these figures shows them to encompass “causes of death” ranging from premeditated murder to legal executions. To avoid this difficulty and provide a more refined analysis, alternative data for first and second degree murder and murder and nonnegligent manslaughter are examined. These data and Gastil's Southerness index along with eight socioeconomic and demographic variables are fit into a correlation analysis. Results of this analysis show a very substantial positive association between Southerness and rate for all three offenses. While on the surface these findings appear consistent with Gastil's hypothesis, when a number of socioeconomic and demographic factors are introduced into the analysis as control variables, Southerness proves to only have a slight independent effect on rate. That is, the substantial correlation between Southerness and offense rates is primarily a result of the association between Southerness and factors like educational attainment, income, percentage nonwhite population, population density, etc., all of which have been previously linked to levels of homicide. Overall, little support is found for Gastil's and other's notion of a Southern Culture of violence.