Ensaios sobre pobreza no Brasil

Abstract

The overall objective of this thesis is to analyze poverty in Brazil from an integrated approach, combining the construction of a multidimensional poverty index with the investigation of individual, household, and contextual determinants of poverty, from both a monetary and multidimensional perspective. The study consisted of three essays. The first aimed to construct a multidimensional poverty index for the years 2019 and 2023, based on the methodology of Alkire and Foster (2011), and to analyze the situation of deprivations using subgroup decompositions and a temporal analysis based on Shapley decompositions grounded in Roche's proposal (2013). The results indicate a decrease in multidimensional poverty in the country during the analyzed period and highlight the unequal distribution of deprivations faced by the poor in different regions and population subgroups. The second essay aimed to investigate the determinants at individual and contextual levels of monetary and multidimensional poverty in Brazil, based on the application of the multilevel logistic regression method. Data aggregated by geographic strata were used to analyze contextual determinants, allowing the identification of the extent to which the risk of poverty is associated with differences between these locations. The results highlight the importance of contextual factors in determining poverty; furthermore, it was observed that the effects are differentiated for rural and urban areas. The objective of the third essay was to investigate the individual and contextual factors specifically associated with rural poverty, measured using the extreme poverty indicator, emphasizing the role of nonagricultural occupation in poverty reduction. This study also applied a multilevel logistic regression model in conjunction with data from geographic strata. The results indicate that in structurally poorer strata, non-agricultural work plays an important role in mitigating poverty.