Estimation of Opportunity Inequality in Brazil using Nonparametric Local Logistic Regression

Abstract

This article measured opportunity inequality in Brazil by combining a series of theoretical and empirical tools. The database was built using a two-sample instrumental variable (TSIV), developed by Angrist and Krueger. After that, the axiomatic approach put forward by O'Neill et al. was used, in which the estimation of children's income distribution function is conditional on their fathers' wages. The inference process was based on nonparametric local logistic regression. The results indicate that Brazil has a high level of opportunity inequality. In other words, in the context of intergenerational mobility, those whose fathers belong to lower income strata have to expend greater effort in order to attain a certain income level.