Using a Terrorist Attack to Estimate the Effect of Police on Crime

Two of the main challenges in the literature on crime are to isolate significant causal effects of police on the amount of crime, and to distinguish between deterrence and incapacitation. Following a terrorist attack on the main Jewish center in July 1994, all Jewish and Muslim institutions in the city of Buenos Aires (including schools, synagogues, mosques and clubs) were given 24-hour police surveillance. Thus, in these areas the distribution of police presence after the terrorist attack can be presumed exogenous in a crime-regression. Furthermore, crime reductions can only reflect deterrence effects. We collected data on the location of all car thefts for three neighborhoods of the city of Buenos Aires before and after the terrorist attack to study the deterrent effect of police on crime. Our estimates suggest that there is a large, negative effect of visible police presence on car-theft. We also find evidence of displacement: there is a positive and significant effect of visible police presence on the number of car thefts in the immediate surrounding area. The effects approximately cancel out, so there is no overall effect of police on crime.