Abstract
This text attempts to estimate the impacts of the Programa Bolsa Família on grade repetition, using matched data from the Single Registry (Cadastro Único, or CadÚnico), the Attendance Project and annual School Census. The authors use three approaches: i) comparing the results for poor children in the CadÚnico with and without the benefit; ii) estimating the treatment effect; and iii) estimating the impact of the percentage of beneficiary children in a given school on the average repetition level within the school. The results are far from impressive but do show a drop in the probability of grade repetition.