Decomposição da desigualdade de gênero no Brasil: uma análise a partir da função de influência recentrada

Abstract

The objective of this dissertation is to investigate gender wage discrimination in the industrial and service sectors during the period 2014-2017. The model proposed by Firpo, Fortin and Lemieux (2018) was used, which combines the refocused influence function (RIF) with a reweighting strategy to break down differentials throughout the wage distribution. Thus, the model was used to decompose the income differential between genders according to observable characteristics (composition effect) and the return of these observable characteristics (structural effect). In general terms, gender discrimination is asymmetric throughout the distribution, being more intense at the upper end of the distribution. Sectoral results have shown that gender inequality comes from different natures. For the industrial sector, gender inequality is higher in the lower tail of the distribution (10% lower wages), potentially related to secular issues, while in the services sector the gap was higher in the upper tail of the distribution (90% percentile), indicating that inequality is accentuated in positions of greater human capital in the sector.