External authors:

• Pedro Rodriguez-Martinez ( George Washington University )

Abstract:

We examine how criminal activity responds to a large environmental shock in the urban space. We focus on blackouts, which can abruptly change the provision of light and potentially the distribution of criminal opportunities. Using high-frequency administrative data on more than 374,000 power outage incidents reported in Chile during 2014-15, we analyze how they affect crime along several dimensions. We find no significant effect on the aggregate crime rate, but we also find two offsetting reactions driving this result: a positive increase in burglary and a negative effect on robbery. By exploiting unique features of the data, we analyze several dimensions of treatment effect heterogeneity regarding the magnitude, size, and duration of the blackout. In addition, we find that crime responses differ by municipality socioeconomic status. We validate the lack of ambient light as the main mechanism by conducting a set of tests using daytime power outages as a placebo. Our results suggest that criminals react to changes in incentives by carrying out crimes that yield a higher expected return.