SEVERAL PREVIOUS STUDIES THAT HAVE DEALT WITH POLICE EFFECTS ON CRIME HAVE BEEN CRITICIZED FOR VARIOUS REASONS INCLUDING THE MEASURES USED AND THE ESTIMATION PROCEDURES EMPLOYED. DATA PRESENTED IN THIS STUDY ARE CONSISTENT WITH THE VIEW THAT POLICE PATROL STRATEGIES HAVE AN EFFECT ON THE RATE OF ROBBERIES COMMITTED, THAT THIS EFFECT IS A CAUSAL ONE AND NOT THE RESULT OF A SPURIOUS STATISTICAL CORRELATION, AND THAT THE EXISTENCE OF A GIVEN PATROL STRATEGY IS AFFECTED BY BUREAUCRATIC DECISIONS AS WELL AS BY LEVELS OF RESOURCES. THE MEASURES OF POLICE RESOURCES (PATROL UNITS ON THE STREET) AND POLICE ACTIVITY ON THE STREET (MOVING CITATIONS ISSUED) ARE MORE PRECISE THAN ANYTHING PRESENTLY AVAILABLE IN SIMILAR STUDIES AND PERMIT THE USE OF IDENTIFICATION RESTRICTIONS FOR STRONGER INFERENCES ABOUT THE CAUSAL EFFECT OF ARRESTS ON CRIME RATES THAN HAS BEEN POSSIBLE. THE RESULTS OF THE SIMULTANEOUS EQUATIONS ARE CONSISTENT WITH THE FOLLOWING THEORY: CITIES THAT EXPERIENCE HIGH RATES OF PERSONAL CRIME HAVE HIGHER THAN AVERAGE TAX BASES AND WILL HIRE MORE POLICE EMPLOYEES IN PROPORTION TO THEIR POPULATIONS. IN THOSE CITIES THAT MANAGE TO PRODUCE HIGHER ARREST RATIOS BY EMPLOYING THE AGGRESSIVE PATROL STRATEGY OR INCREASING THE NUMBER OF PATROL UNITS, THE ROBBERY RATES ARE LOWER THAN ONE WOULD PREDICT USING JUST THE SOCIOECONOMIC COMPOSITION OF THE CITY AND THE DENSITY OF CRIMINAL OPPORTUNITIES. FURTHERMORE, A CITY WITH A 'REFORMED' OR PROFESSIONALIZED MUNICIPAL MANAGEMENT SYSTEM IS MORE LIKELY TO HAVE A POLICE DEPARTMENT WITH AN AGGRESSIVE PATROL STRATEGY, THOUGH NOT NECESSARILY ONE WITH MANY PATROL UNITS ON THE STREET. THE IMPLICATIONS FOR AND LIMITATIONS UPON POLICY ARE DISCUSSED. REFERENCES AND SAMPLE DATA ARE APPENDED. (AUTHOR ABSTRACT MODIFIED)