The model shows that police expenditure in neighboring communities, as well as other factors, affects the crime level and thereby the community's police expenditure. The model consists of a general equilibrium of police expenditure and crime which is analyzed for all communities in the metropolitan area. The reaction curves which result reflect the responsiveness of a community to the action of neighboring communities. By using comparative statistics, multipliers are developed which show the total effects of changes in variables. Regression models show that interjurisdictional crime spillover is found for aggregate property crime, particularly for breaking and entering and auto theft. No crime spillover is found for larceny and violent crimes. This study confirms the expected deterrent effect of police on crime which is seldom recorded in other empirical studies. Three tables, extensive formulas, and 18 references are included. (Author summary modified)