This study aims at evaluating the returns to additional years of education for male workers in Brazil. It applies Heckman's procedure to measure the differences between wage returns to schooling in urban and rural areas, based on data from the National Household Sample Survey (PNAD), for the year 2009. A significant methodological contribution to the literature on returns to schooling has been made by incorporating data sample into the estimation. Results show wage differentiation due to age, race and geographic region; as to education, wage returns in urban areas are bigger than in rural areas. Additionally, the use of a complex sample has produced results that are considerably different from those based on simpler sample. Therefore, it is deemed necessary to incorporate those aspects into new researches based on data from PNAD.
Returns to education; qualification; complex sample