Effect of police on crime and arrests: Are police deterring or incapacitating criminals?, The

Aim: To estimate the causal effect of police numbers on rates of crime and arrests. Method: The data used in the present study consists of monthly Local Area Command (LAC) level counts of police officers and selected violent and property crimes for the period July 2000 - April 2003. These crimes include: break and enter, theft, motor vehicle theft, robbery and homicide. Using these data I exploit variation in police numbers driven by a massive recruitment campaign in the lead up to the 2003 New South Wales (NSW) State election to estimate the causal effect of police numbers on rates of crime and arrests. Results: I find that a one per cent increase in the size of the police force generates a 0.8 per cent reduction in theft and a 1.1 per cent reduction in motor vehicle theft. This roughly equates to one additional police officer preventing 17 thefts and four motor vehicle thefts each year. My estimates are too imprecise to draw any definitive conclusions with respect to other types of crime. I do not find evidence that an increase in police numbers generates any significant change to the arrest rate. This indicates that police reduce theft and motor vehicle theft through deterrence rather than incapacitation. Conclusion: The implications of the present study are threefold. First, an increase in police numbers generates a substantial reduction in property crime. Second, an increase in police numbers has no significant effect on the arrest rate for property crime. Finally, the cost of an additional police officer is almost definitely offset by the benefit she provides to society in the form of crime reduction.