Violence in rural, suburban, and urban areas: Trends, demographics, and measurement

The goal of this dissertation is to improve our understanding of violence in rural communities by addressing substantive and methodological gaps in the literature. I do this by comparing rural, suburban, and urban areas on three components of violence: trends, demographic explanations for trends, and measurement. Specifically, I addressed the following questions: (1) How do Uniform Crime Reporting Program (UCR) and National Crime Victimization Survey (NCVS) aggravated assault trends compare by rural, suburban, and urban location? (2) Do demographic changes in the population account for changes in aggravated assault trends in rural, suburban, and urban communities? (3) Are the same factors associated with police notification and emergency room treatment among victims of serious assaults, and are these associations moderated by rural, suburban, and urban location?;The findings suggest the UCR and NCS/NCVS consistently show substantial rural and non-rural differences in violent crime trend patterns, demographic explanations for trends vary by rural and non-rural area, and there are few rural and non-rural differences in victim and incident characteristics associated with violent crime data sources. First, the UCR and NCS/NCVS exhibit similar aggravated assault rate patterns: non-rural rates decreased from the mid-1990s to the mid-2000s while rural rates increased slightly (UCR) or decreased more slowly (NCS/NCVS). Second, despite rural areas experiencing much smaller rate decreases from 1993 to 2005 than non-rural areas, demographic changes contributed much more to rate changes in rural compared to non-rural areas. Third, analyses reveal few victim and incident characteristics associated with victim emergency room treatment; many associated with police notification; and few associations moderated by rural, suburban, and urban location. In general, incident factors are associated with police notification more often than victim emergency room treatment. The methods used to address these questions and the results of the analyses provide important implications for future research. Particular attention should focus on the utilization of both the UCR and NCVS; accounting for demographic composition in violence studies within and between rural, suburban, and urban areas; and the role of sample selection bias in criminological research that utilizes police and/or emergency room data.