Impact of social distancing due to the COVID-19 pandemic on property crimes in São Paulo: a Bayesian spatiotemporal modelling case study

• Ano de defesa: 2021
• Autor(a) principal: Fukushima, Ian Shinji Ferreira
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Biblioteca Digitais de Teses e Dissertações da USP
• Palavras-chave em Português: Análise bayesiana; Aproximações de laplace aninhadas e integradas; Bayesian analysis; Causal impact; COVID-19; Crime; Impacto causal; Integrated ested laplace approximations
• Link de acesso: https://www.teses.usp.br/teses/disponiveis/96/96131/tde-24092021-152217/

With the goal of stopping COVID-19 infections, several countries adopted social distancing regulations, including closure of non essential establishments and mobility restrictions. This paper estimates the impact of these regulations on property crime in the state of São Paulo. The impact is estimated using a Bayesian spatiotemporal model, and is disaggregated by microregions. The space variability of the impact is used to infer which observed variables characterize places with higher probability of impact. We found that most microregions experienced a decrease in property crime, and that these places are characterized by higher isolation indices and less receipt of emergency cash transfers.