The dependent variable for this study is per capita total expenditures for police in 1971, including expenditures for salaries and operations, as well as for capital. Both linear and nonlinear models test the relationships among the variables. A simultaneous model which recognizes the interdependence of the police expenditure function with the crime and total city revenue functions is tested. Findings show significant predictors of police expenditure to be total city revenues, percent black, population density, size, civil rights mobilization activity, and poverty. The relative size of the black population was a significant predictor of expenditures for police salaries, operations, and capital items, while the level of black political mobilization activity was a signficant predictor of capital expenditures. Evidence suggests that police expenditures are a resource which is mobilized when a minority group appears threatening to the dominant group. This finding underscores the need for an examination of the role of the police in determining the political balance between members of minority and majority groups. If the acquisition of policing resources is in part a response to perceived threat, the use of these resources may affect the civil liberties of minority group members. Four tables, two figures, and a list of 61 references are included.