Urban Crime and Spatial Proximity to Liquor: Evidence from a Quasi-Experiment in Seattle

Abstract

There is a well-established correlation between retail liquor outlets and crime, but few studies identify causal effects. I exploit a unique source of identifying variation to establish causality: a 2012 privatization of liquor retailing in Washington State that rapidly expanded liquor availability into preexisting grocery and drug store chains. Based on 166,000 police reports from Seattle and a fixed-effects panel model, I find a significant positive effect of liquor availability on neighborhood crime both in OLS and IV estimates. Reducing the distance to the nearest liquor retailer by one mile leads to an average treatment effect of roughly 6 to 8 percent higher monthly crime rates. Violent crime and drug crimes are persistently affected, with more transitory effects on shoplifting and other non-violent crimes. Using an event study framework I investigate whether the results are due to new crime or spatial redistribution of existing crime, finding evidence of both effects. Overall, expanded liquor retailing appears to have had a significant causal effect on crime.