A spatial analysis of the effects of gender, social structure and culture on lethal violence: Examining regional differences using an integrated model of homicide and suicide.

摘要

This study uses spatial autocorrelation techniques to examine regional patterns in male and female offender rates in the United States from the perspective of the integrated model of lethal violence. Historically, researchers working in the integrated model paradigm have paid particular attention to regional and cross-national variations in lethal violence in an effort to explain why suicide rates tend to be highest in Northern latitudes while homicide rates are concentrated in the South. Additionally, the integrated model approach to homicide and suicide provides a distinct method of thinking about the relationship of lethal violence to culture and social structure. To help measure culture, counties within United States will be grouped into twelve regions based on earlier work done by Zelinsky (1951), Gastil (1975) and Whitt et al. (1995). Within this theoretical approach, social structural variables or the forces of production such as income, Medicaid, unemployment and inequality will be used to determine the amount of lethal violence within each region. Cultural variables or the forces of direction such as age, gender, divorce and church membership will be used to establish the direction (either homicide or suicide) of the lethal violence contained within each region. In this study, spatial autocorrelation procedures are used to gain a better understanding of the spatial nature of these two forms of violence. Results indicate that patterns of lethal violence continue mirror earlier findings of Northern and Southern boundaries. Additionally however, new regionalizations begin to emerge between the Southern and Western regions within the United States. Finally, regarding the gender of the offenders, spatial diagnostics detect a diffusion process, or lag model, for male perpetrated homicide and suicide. This means that the male lethal violence in one county spills over and influences the male lethal violence in nearby locations. Contrastingly however, spatial dependency diagnostics suggest that the female lethal violence analyses are misspecified, or an error model of estimation is required. This means that the incident of female lethal violence is best explained by independent variables beyond those presented in this study.