Defense Date


Document Type


Degree Name

Doctor of Philosophy


Public Policy & Administration

First Advisor

Dr. Laura J. Moriarty


This study investigates both space and time aspects of neighborhood crime distributions using social disorganization as a theoretical framework in the City of Richmond, VA. Neighborhood crime, in this study, might be considered as any type of index crime aggregated to neighborhood level. For the purpose of the present study, however, neighborhood crime only includes "homicide" categorized as an index crime in the Uniform Crime Report (UCR). Homicides in neighborhoods have been realized as rare events, and have become problematic to establish robust statistical models in the literature. With the focus of neighborhood homicide, this study questions the consistency of Social Disorganization Theory (SDT) by the longitudinal research setting. It, therefore, constructs and verifies seven hypotheses (residential mobility, race/ethnic heterogeneity, family disruption, socio-economic status, population density, youth, and vacancy) to test SDT, while it establishes and further confirms its main hypothesis "Neighborhood homicide increase is likely to be associated by the increase in neighborhood social disorganization over time."This study constructs a longitudinal research design with 10 years, uses Census 1990, Census 2000 and homicide data (From the City of Richmond Police Department) as secondary data. Nonetheless, this study uses only two main census decennial years to calculate the other years' structural covariates by the linear interpolation technique such that this study is able to include additional years to construct the essential difference models. Population includes all neighborhoods in the City of Richmond such that this study works with entire population, but no sampling procedure. As an analytical strategy, this study constructs eleven different binomial logistic regressions, whereas it constructs multinomial logistic regressions as difference models to verify the main hypothesis for neighborhood homicide. Once this study realizes clustered neighborhoods with respect to experiencing homicide hotspot(s), it constructs a stepwise multiple regressions model to explore the most important social disorganization variables for the most problematic neighborhoods.In terms of findings, the most important social disorganization variables attributed to homicide distribution in the City of Richmond are: The low SES (Socioeconomic Status), residential mobility, vacancy, population density (across only the concentrated neighborhoods), and family disruption.Accordingly, this study has successfully contributed to the literature around SDT, social crime prevention, and spatially integrated crime policy analysis.


© The Author

Is Part Of

VCU University Archives

Is Part Of

VCU Theses and Dissertations

Date of Submission

June 2008