Socioeconomic inequalities: an analysis of the determinants of infant mortality rate in Brazilian municipalities

Abstract

This study investigated the main socioeconomic determinants of infant mortality rates (IMR) in Brazilian municipalities for the 1991, 2000 and 2010 census periods. As such, four methodologies were applied. Of which, three econometric methodologies were tested: Linear Data Model in Panel, Poisson model in Panel Data and Quantum Regression model in Panel Data. The fourth methodology tested was the Exploratory Analysis of Spatial Data (AEDE) in order to observe the distribution of the infant mortality rates and its spatial correlation. As for the three econometric results (the linear panel, the Poisson models, and the quantized regression with fixed effects) each identified that socioeconomic variables had an influence on the determination of infant mortality rates, this is relevant for the improvement of the health status of the Brazilian population. Results revealed, however, that income has a more persistent effect on infant deaths and fecundity compared to educational variables in Brazilian municipalities. For the exploratory spatial data analysis (AEDE), the overall index showed that there is evidence of positive spatial autocorrelation, indicating that municipalities with high (low) IMR are surrounded by municipalities that have a high (low) IMR, thus revealing a contagion effect or overflow of IMRs.