Convergência espacial de homicídios no estado de São Paulo no período de 1980 a 2020

Abstract

The growing concern about crime in Brazilian society has made it a central topic in recent political and economic discussions. This is because it affects the well-being of society and hinders economic development, resulting in numerous social costs, such as financial losses due to increased public and private expenditures to combat it. Thus, this study aims to analyze the spatial patterns of crime in São Paulo and, concurrently, test the hypothesis of convergence in homicide rates for the state's microregions. To achieve this, tools like Exploratory Spatial Data Analysis (ESDA) and the Absolute β Convergence model are employed to examine the distribution of crime in microregions of São Paulo during the period from 1980 to 2020. The results of exploratory spatial data analysis revealed positive autocorrelation of criminality in microregions of São Paulo throughout all analyzed periods. The LISA indicators showed significant transformations in criminal clusters, with a shift of criminal activity from the capital region to the interior of the state. Additionally, a notable reduction in Low-Low clusters was observed, along with the emergence of new High-High and High-Low clusters in the interior of São Paulo in the recent period, signaling a process of crime internalization. The Absolute β Convergence model pointed to the occurrence of convergence in homicide rates in São Paulo, indicating a homogenization of homicide rates among microregions over the years. The convergence analysis also revealed an acceleration in the convergence process in the most recent period, potentially leading to a violence equilibrium for the state in the near future.