Crime dynamics in Buenos Aires : evidence from a quasi-experimental design

Abstract

In July 2018, Buenos Aires City Government reorganized police districts of the recently created Buenos Aires City Police, reducing the number of police stations from 54 to 43 and reshaping its borderlines to match them with the boundaries of the comunas and neighborhoods. The new districts, unlike before, do not have in many cases multiple police stations simultaneously patrolling one neighborhood. This paper utilizes the exogenous variability introduced by the reform to study the impact of the police district restructuring and the police station downsizing on crime dynamics in Buenos Aires. Given that the boundaries of comunas and neighborhoods were drawn from administrative and political issues not related to crime, and since the new distribution of police districts happened only in some neighborhoods, this papers uses a difference-in-differences approach to evaluate the potential effect of the intervention on five different types of crime at a neighborhood and street segment level in City of Buenos Aires during the 2017-2019 period. The results show that modifying police station districts and the number of police stations led to a decrease of 16.4% in the number of thefts committed in the treated neighborhoods, together with an increase of car thefts of 11.3% when compared with control neighborhoods. The negative impact of the intervention holds when analyzing results at a street-segment level.

Keywords: police districts, crime, difference-in-differences, quasi-experimental design.