Document Type



Master of Arts


Criminology and Criminal Justice

Date of Defense


Graduate Advisor

Finn Esbensen


Elaine Doherty


Matt Vogel

Janet Lauritsen

Adam Boessen


The purpose of this study is to better understand predictors of neighborhood crime in metropolitan areas through a routine activity theory (RAT) lens. This paper examines whether neighborhood-level aspects of RAT have varying explanatory power for different crimes. This is accomplished by drawing on a sample of census tracts from the National Neighborhood Crime Study (NNCS). Using these data, multiple independent variables are regressed on four different types of crime using an ordinary least squares (OLS) regression. The results do not suggest that there is a strong degree of variation in the ability of specific measures to predict specific forms of offending. Insofar as these measures truly capture elements of RAT, the findings indicate that RAT is efficient in explaining crime rates in general, and does not do an especially good job in differentiating crime acts in disaggregated scales. However, of the different measures, vacant housing is an especially robust predictor of each type of crime, suggesting that features of the physical environment of a neighborhood contribute to criminal behavior above and beyond shared characteristics of neighborhood residents. This suggests that situational crime prevention can be an effective crime control method by increasing security of an area to dissuade potential offenders and using target hardening.

Included in

Criminology Commons