Modelagem estatística da impunidade: métodos inferenciais e preditivos para dados de crimes no estado de São Paulo

Abstract

High levels of violence in Brazilian society have been interpreted by political scientists as indicative of a legitimacy crisis - namely, a widespread lack of public trust in the State. This context is exacerbated by impunity, a critical social issue that undermines the principle of equality before the law. Due to the inherent challenges in quantifying impunity and in order to investigate the efficiency of the judicial system in the country, the analysis focuses on the state of São Paulo and proposes a novel indicator: the time elapsed between the filing of a police report and the initiation of a formal criminal investigation. The analysis focuses on intentional homicide cases in the state of São Paulo from 2017 to 2022. A survival model with Intrinsic Conditional Autoregressive (ICAR) spatial frailty is employed, incorporating municipal-level socioeconomic variables such as the Human Development Index (HDI) for education, per capita income, and the Gini index. Also, a simulation study was conducted and concluded that the model is capable of accessing the true parameters appropriately. The study is thus relevant as it contributes to the literature on the subject by proposing a novel approach to the analysis of impunity in Brazil.