The paper examines the determinants of the wage differential between workers in the formal and informal sectors of the Brazilian labor market. An econometric method of repeated cross-sections (pseudo-panel) is used, that allows the researcher to control for observed and unobserved characteristics of the individuals in the sample. We present strong evidence of self-selection bias, meaning that the wage differentials are the consequence of better individual non-observable characteristics of the workers in the formal sector and not to the characteristics of this sector.
