Declines in Police Reporting of Domestic Violence during Shelter-in-Place: Evidence from American Cities

We measure the effect of shelter-in-place (SIP) orders on the police reporting rate of domestic violence during the COVID-19 outbreak. We find police reporting of domestic violence decreases by an average of 9.3 percent with the imposition of SIP orders. We identify these effects by exploiting the unique variation in the timing of SIP orders across 19 cities in the United States. These results are present when testing different treatments including school closures and voluntary household social distancing. The established literature, focusing on household responses, has found an increase in emergency calls for service when cities implement shelter-in-place orders. Our focus on police response demonstrates the divergence of the effects of these SIP orders between household and agency behaviors. SIP changes the environment for both households and police, albeit in different ways. Our findings suggest police reporting of domestic violence decreases under this new environment. We note implications for city officials who must consider how such orders uniquely impact both households and police responsiveness.