Social ecological theories of crime have recently been extended to explain spatial variation in police behavior. Although these theories successfully identify community characteristics that affect local policing, they fail to acknowledge the class-based origins of formal social control and the relative autonomy of the police. For this study, cross- sectional data were obtained from 25 police agencies' vice divisions and their corresponding jurisdictions to test the integrated hypothesis. Four social ecological variables and a fiscal measure of relative autonomy were examined as police- behavior predictors. The fiscal autonomy variable explained a statistically significant proportion of the variance in proactive policing. This suggests that poorly subsidized agencies rely on inhouse income generation facilitated by proactive policing tactics such as asset seizure. Additionally, two social ecological variables were statistically significant predictors of proactive policing: the proportion of families below the poverty line and the proportion of nonwhite residents. Fiscal autonomy and the proportion of nonwhite residents were also statistically significant predictors of police behavior. Agencies financed by government funding are able to rely on citizen-initiated actions; whereas, agencies not subsidized must rely on officer-initiated detection of vice crime. Overall, the resultant integrated model provided support for relative autonomy as an indicator of proactive and reactive approaches to policing. 2 tables, 5 notes, and 82 references