Two of my brightest colleagues, Pedro Souza and Marcelo Medeiros, have recently published a study on 'The contribution of wage differentials between public and private sector workers to the overall household income inequality in Brazil'.
The paper is published in Portuguese only the title translation is my bad but you may read the abstract:
We estimate the contribution of the wage differential between workers with the same attributes in the public and private sectors to the household per capita income inequality in Brazil. The estimate is based on counterfactual simulations and the contribution to inequality on a factor decomposition of the Gini coefficient. Data comes from the Brazilian National Household Survey PNAD 2009. The differential corresponds approximately to 17% of the wage bill of workers in the public sector, is regressive and highly concentrated. However, because it amounts to a small share of the total income (1%) its contribution to the total inequality is of 3%. The sector composition effects on inequality are times higher than the segmentation (price) effects. These conclusions are robust to changes in the definition of the sectors and to different estimation techniques.
