Social distancing and COVID-19 under violence: Evidence from Colombia

Did violence increase social distancing and decrease COVID-19 cases? We investigated the effects of massacres on social distancing and subsequent impacts on COVID-19 cases in Colombia. Using an augmented synthetic control method model, we find that massacres reduced human mobility toward parks by six percentage points compared to unaffected areas. However, we did not find significant changes in workplace mobility. Moreover, alterations in social interactions caused by the violence had minimal effects on the spread of COVID-19. Following the occurrence of the first massacre, there was a decrease in 35 new cases per 100,000 inhabitants in the subsequent four months, with no evidence of changes in COVID-19 deaths. By leveraging an exogenous shock unrelated to the fear of the disease or its previous levels, we demonstrate the effect of social distancing and offer insights into social dynamics and public health.