Doctoral Dissertations

Date of Award


Degree Type


Degree Name

Doctor of Philosophy


Political Science

Major Professor

John M. Scheb

Committee Members

William Lyons, David Folz, Anthony Nownes, Bruce A. Ralston


Between the late 1980s and 2001, the Knoxville Community Development Corporation and the Knoxville Police Department implemented policies intended to reduce crime in Knoxville’s public housing. Beginning with pilot programs in two housing sites in 1989, a cooperative relationship between KCDC and KPD emerged that evolved into the KCDC Security Patrol in 1990. By using Arcview software, I was able to separate the yearly violent crime incidents for 1992-2001 that occurred in two separate geographical regions: 1) the area represented by the KCDC housing sites where the policies were implemented, and 2) the remainder of the City of Knoxville. The descriptive statistics indicate that none of the six policies evaluated in this study have an effect upon total crime and an effect upon the occurrence level of aggravated assaults in the KCDC sites. However, the 1999 team-based approach to policing and the new applicant screening process of 2000 seem to somewhat reduce crimes of opportunity and/or property crimes (i.e., burglary, larceny, auto theft, and robbery). A 1996 One-Strike eviction policy has the same effect, but separating it from confounding events is not possible. The demolition of College Homes in 1998 has a marginal effect upon aggravated assaults. Because I could not conduct a survey of Knoxville’s public housing residents, regression analyses in a cross-sectional design are used to gain further insight into other possible variables effecting Knoxville’s crime rate. An individual census tract is the unit of analysis. Socio-economic and demographic indicators of crime are used, as well as a policy dummy variable and a variable representing the degree of public housing within each tract. A methodological problem during data collection within KPD prevents the use of data prior to 1996. The regressions reveal that crime in Knoxville (1996-2001) is consistently associated with the proportion of males, and the proportion of those who are over the age of 25 who do not possess a high school education.

Files over 3MB may be slow to open. For best results, right-click and select "save as..."