Fighting Crime in Lawless Areas: Evidence from Slums in Rio de Janeiro

Abstract

We use Rio de Janeiro’s slum pacification program initiated in 2008 to analyze the effect of policies fighting crime in lawless areas. We correct the bias from the unobserved rise in crime reporting via the use of a proxy variable and bounded variation assumptions. We find that the program reduced the murder and robbery rates, but strongly increased the assault and threat rates. We explain these results by providing evidence that increased enforcement weakened the security service gangs provide on their turf, and may incentivize criminals to switch from serious to less-serious crimes.

Keywords: Crime, Law Enforcement, Drug Gangs, Proxy Variable, Reporting Rate, Brazil

JEL Classification: D74, H76, K42