The dynamics of neighborhood property crime rates

We investigate theoretically and empirically the inter-temporal dynamics of neighborhood property crime, a prime contributor to the quality of life and the vitality of markets in urban areas. We develop a microeconomic model of year-to-year changes in crime rates that incorporates endogenous relationships between the recruitment of criminals and deterrent effects spawned by responses of neighborhood residents and/or police. We operationalize the model using annual panel data for census tracts in Cleveland, Detroit, and Seattle, and use dynamic panel econometric procedures to estimate parameters. Although the details vary across cities, all estimated models demonstrate a time path that converges to a stable state within 10 years and often sooner, ceteris paribus, regardless of the size of disequilibrating shock. The Cleveland and Detroit models provide evidence of nonlinear, endogenous deterrence response effects. The Detroit model provides evidence of an endogenous net fear response producing threshold instability for large increases in neighborhood property crime rates