Diferencial salarial público-privado: Uma análise em painel de dados com a PNAD contínua entre 2016-2019 ^♦

Abstract

This article estimates the wage differential between public and private sector employees from 2016 to 2019 using the Pesquisa Nacional por Amostra de Domicílios Contínua. To evaluate the premium, we apply (1) regressions by least squares and the decomposition by conventional Oaxaca-Blinder and (2) a panel analysis regressed by fixed and random effects and the decomposition by an extension of the Oaxaca-Blinder model to panel models. In the cross-sectional methodology, a difference of 62.3 % favorable to the civil servant was found, with 75 % explained by observable variables. The other 25 % represent effects of unobserved variables, that is, an unexplained differential of approximately 12%. When considering the panel data, our decomposition points out that 73% of this differential is explained by fixed effects, 18% by agent characteristics and only 11% represented by unobservable variables representing a difference of approximately 6%. The results still indicate an increase in the unexplained percentage of the differential over time.

Keywords:
Salary difference; Fixed effects; Random effects; Oaxaca-Blinder