Mortalidade violenta no Brasil: abordagens espaço-temporais

Abstract

Brazil has experienced persistently high rates of violent mortality, a phenomenon that, while present globally, reveals unique characteristics in the country. Despite significant social and economic progress in recent decades, homicide, traffic accidents, and suicide remain at alarming levels. This dissertation aims to contribute to the understanding of these three major causes of violent death by analyzing their spatial and temporal dynamics, identifying patterns, trends, and possible future directions should current behaviors persist. The first article focuses on the evolution of homicide rates across Brazilian microregions and presents projections up to 2022 based on a spatial hierarchy (States > Major Regions > Brazil). The results point to an almost generalized upward trend, particularly in the Northeast, and reveal the presence of spatial clusters. These findings suggest that geographic space plays a significant role in explaining the distribution of lethal violence, emphasizing the need for immediate action. While the growth trend is confirmed, the spatial concentration of homicides raises additional questions about the underlying territorial dynamics. The second article addresses traffic accidents, a theme less frequently explored in national literature. The study identifies the socioeconomic determinants of mortality rates from transport-related accidents at the microregional level, incorporating the spatial structure of the data. The results suggest that different regions are situated at varying stages of the inverted U-shaped relationship between income and traffic mortality. Less developed areas still face rising mortality rates, while more economically and structurally advanced regions appear to have surpassed the peak and show declining trends. Spatial analysis also reveals that microregions with similar patterns tend to be geographically close, indicating specific local dynamics that should be considered in public policy. The third article investigates the determinants of suicide rates per 100,000 inhabitants in Brazilian microregions for the years 1991, 2000, and 2010. The central hypothesis evaluates whether local spatial influence—through imitation or contagion effects—plays a role in explaining suicide rates. Exploratory spatial analysis and spatial regression models confirm the importance of geographic proximity, even after controlling for socioeconomic variables, supporting the literature on neighborhood effects in suicide incidence. However, the significance of these socioeconomic factors varies across time periods, suggesting that the drivers of suicide are not static and evolve according to spatial and temporal contexts.