Ítalo Oikawa

This thesis on economics from external causes in Brazil, addresses the economic impacts and factors related to mortality and hospitalizations caused by specific injuries. Three essays were presented to address these issues. In the first essay, the aim was to measure the economic loss (consumption, production and added value) and to assess the magnitude and extent of intersectoral linkages (cascade effects), due to deaths from traffic-accidents, suicide and homicide in Brazil, in 2017. Therefore, the Inoperability Product-Input Model (MII) was applied. The descriptive analysis results showed regional disparities in number of deaths and mortality rates due to these causes in Brazil, with higher occurrences with men, residents of the northeast region and at an economically active age. MII estimates, indicated that these three external causes resulted in significant global economic losses in terms of production loss (R$ 4,475.24 million), consumption (R$ 2,330.99 million ) and added value (R$ 2,673.92 million), and the three most affected sectors in terms of loss of production were “the automobile, truck and bus manufacturing sectors, except parts” sector (R$ 860.43 million), “manufacturing of other transportation equipment, except motor vehicles” sector (R$ 161.604 million) and the “manufacturing of parts and accessories for motor vehicles” sector (R$ 230.135 millions). Analysis of the value added components pointed out that the Brazilian economy lost R$ 1,186.19 million (44.4% of VA) in wages, R$ 1,088.30 million (40.7% of VA) in gross operating surplus and mixed corporate income. In addition, the governments (federal, state and municipal) lost R$ 399.42 million (14.9% of VA) in taxes net of subsidies on production and imports. The second essay of this thesis analyzed the spatial factors (economic, demographic and health management) associated with the mortality rate of motorcyclists among the microregions of Brazil in 2016. Spatial econometric techniques (Exploratory Spatial Data Analysis - ESDA, estimation of spatial econometric models and Geographically Weighted Regression - GWR) were used. Based on Grossman's health production function, in which the product corresponded to the mortality rate, it was found that the regional disparities in this rate were explained by socioeconomic factors (or inputs) - income and employment (negative association), demographic - demographic density (global and local negative association) and motorization rate (global and local positive association), and health management - number of specialist doctors (global negative association), as well as the spillovers effects ( spatial lags in mortality and motorization rates) between the country's microregions. In order to analyze, in the period from 2008 to 2018, the spatial factors (economic, demographic and health) of the hospitalization rates (smoothed) of SUS due to land transport accidents between the municipalities of Paraná, Grossman's theoretical model was tested, and the Spatial Empirical Bayes smoothing technique and spatial econometrics (Exploratory Spatial Data Analysis - ESDA and the Spatial Panel Data model) were applied. Regional disparities were observed in terms of the hospitalization rate in Paraná. This rate (in time and space) was positively influenced by the real GDP per capita and motorization rate of vehicles, and negatively by the demographic density. In addition, spillover effects of the demographic variables related to the neighborhood were detected on the hospitalization rate of the municipality under analysis. The health factor (prevalence of transport accidents) was not statistically significant. Therefore, external causes of mortality and hospitalizations are a serious public health problem for Brazil. In addition to worsening health status, those variables generated great economic losses (production, consumption and added value) and human capital in the short and long term. Moreover, economic, demographic and health management factors affected the geographic and temporal behavior of these variables, with relevant local effects.