Civil Asset Forfeiture, Crime, and Police Incentives: Evidence from the Comprehensive Crime Control Act” (with Shawn Kantor and Carl Kitchens), Economic Inquiry. 2020. [Link]

The 1984 federal Comprehensive Crime Control Act (CCCA) included a provision that permitted local law enforcement agencies to acquire up to 80% of the proceeds derived from civil asset forfeitures obtained in joint operations with federal authorities. We investigate how this rule governing forfeited assets influenced crime and police incentives by taking advantage of pre‐existing differences in state‐level civil asset forfeiture law and the timing of the CCCA. We find that after the CCCA was enacted crime fell about 17% in places where the federal law allowed police to retain more of their seized assets than state law previously allowed. (JEL K42, K15, H76)

Working Papers:

“Streetlights and Crime: Evidence from Chicago” (Job Market Paper) [Link]

Can repairs to basic public infrastructure lead to declines in criminal activity? In this paper I utilize streetlight outages as a natural experiment to estimate the relationship between streetlight outages and crime. Using hourly, geocoded crime and streetlight level data from the city of Chicago, I find that streetlight outages lead to a 16 percent increase in non-violent crime at night. There is no evidence that the increase is driven by reallocation across space; rather, it seems to be a result of increased opportunity. Within the non-violent categories, the largest impacts are found in the subcategories of motor vehicle theft, criminal damage to property, and burglary which increase by 13,8, and 18, percent respectively. No statistically significant impact is found on violent crime.

"Early WARNing: The Impact of Unexpected Mass Layoff Events on Subsequent Crime"

I use administrative data on mass layoffs and plant closings to examine the effect of negative employment and news shocks on crime outcomes. Exposure to stressful events may cause the marginal citizen to commit crimes they otherwise would abstain from. I find no evidence that an unexpected mass layoff events cause increases in violent or property crime in aggregate. To ensure true effects are not being masked by aggregation, I stratify the data by city size and focus specifically on domestic violence. In rural subsamples, I find there is no change in property or violent crime; however, unemployment shocks are associated with a positive and statistically significant increase in reports of domestic assaults, corresponding to an additional 2.7 reports per 1,000 population, representing a 10% relative increase in the days following a mass layoff announcement.