Rai Chatterjee

Brazil has been one of the most prominent country case studies in academic research charting the relationship between crime and economic development. The issue of violent crime, in particular, has taken centre stage in the past year, as record high murder rates have coincided with the election of President Jair Bolsonaro. His campaign promised of tough anti-crime policies, which resonated with Brazil's track record of a mano dura-heavy-handed-approach to public security. The city of Rio de Janeiro, specifically, has undertaken a process of militarisation and implementation of police operations in its favelas to combat violence and improve security before the 2014 FIFA World Cup and the 2016 Olympic Games. Using an instrumental variable regression , we control for endogeneity in crime rates across the Integrated Circumscriptions of Public Security (CISP) in the municipality of Rio de Janeiro. This is done by utilising the implementation of the 38 Unidades de Policia Pacificadora (Police Pacification Units, UPPs) to study the effects of crime on nightlight data, a proxy for income at the local level. Our regression results suggest that, when controlling for homicide rates and income levels in neighbouring CISPs in the prior month, UPPs had a statistically significant effect on reducing crime rates. Contrary to the academic literature, however, our primary findings indicate that decreases in crime rates actually corresponded with statistically significant decreases in future nightlights. We explore possible reasonings for this, potential for future studies, and the effects of other types of violent crime, such as rape and theft, robberies, and deaths by police officers on future income levels.