Graduation Year


Document Type




Degree Granting Department


Major Professor

Dwayne Smith


A developing body of research within the fields of criminology and rural sociology has emphasized the importance of considering geographic place in the study of interpersonal violence, and domestic violence in particular. Exploring how place is related to domestic violence lends itself to considerations of geographic variation in socio-structural conditions. A handful of studies since the 1980s have explored structural correlates of intimate abuse largely rooted in one of two theoretical contexts: social disorganization or gender inequality/patriarchy. However, knowledge regarding the relationship between place, social structure, and specific types of violence remains limited. The present study is intended as an examination of the relationship between place, social structure, and intimate homicide. Specifically, this study explores the influence of rurality, social disorganization and gender inequality on male perpetrated-female victim intimate partner homicide (femicide). Analyses are also conducted on non-domestic homicide to serve as a comparison to femicide findings. Several research questions are explored including, (1) does rurality have a significant relationship with femicide rates, (2) does structural gender inequality have a significant relationship with femicide rates, and is this relationship conditioned by rurality, (3) does social disorganization have a significant relationship with femicide rates, and is this relationship conditioned by rurality. All research questions are also explored for non-domestic homicide rates.

The data come from several sources including the 2000 U.S. Census (theoretical indicators and control variables), the U.S. Department of Agriculture's Economic Research Service (measure of rurality), the North Carolina State Center for Health Statistics (measure of homicide), and the North Carolina Coalition Against Domestic Violence (measure of femicide). A unique contribution of this study is the use of non-official data sources for homicide measures which are not bound by the same limitations (e.g., accuracy, voluntary reporting) that limit the commonly utilized UCR and SHR data. Negative binomial regression is used to analyze county-level rates of femicide and non-domestic homicide for the population of North Carolina counties (N=100).The results indicate that (1) place does matter, as illustrated by significantly higher risk of femicide and non-domestic homicide victimization in rural counties compared to non-rural counties; (2) increasing female equality in rural counties may serve as a protective factor against femicide victimization, but this relationship is mediated by social disorganization; and (3) increasing social disorganization in non-rural counties is associated with higher risk of non-domestic homicide.

The present study has several implications for femicide and disaggregated homicide research. First, the findings demonstrate the importance of considering geographic location in modeling structural theoretical indicators and processes. Second, the significance of certain theoretical indicators representing both gender inequality and social disorganization contribute to the development of a matrix of risk that can be used to encourage and/or justify the more arduous task of testing fully specified models of the theories across place. Third, the present study contributes to the literatures extending social disorganization to rural places and to domestic violence, and the role of structural gender inequality in gendered violence. Future research exploring structural explanations for intimate partner homicide are urged to make comparisons across disaggregated homicide types and, most importantly, consider the influence of rurality.