A short abstract I submitted in a journal

Low tests scores and low graduation rates are examples of educational challenges in Brazil, especially if we refer to secondary students. Public policies have been created to improve these indices and the ``Jovem de Futuro'' program is one example.We analyzed details of this program to understand if and how it works. ``Jovem de Futuro'' has been a program focused on support for public secondary schools. It provides training for principals and teachers to improve school management and funding for schools.

A randomized controlled trial was designed for selection of treated schools, which ensured our ability to perform a causality analysis. Our results indicate that the program improves student performance in language (Portuguese) and mathematics tests. The scores of treated students increased about $0.2\sigma$. Furthermore, we did not find any difference in the performance of students when we grouped them, for example, by sex or race. Our analysis also indicate that students perform better if they are from schools with better previous results. Because repetition and dropout rates are a real problem in schools of secondary level in Brazil, we estimated how effective the program was at changing these rates.

Our results indicate repetition and dropout probabilities fell if students are from schools that received goals to improve these rates. On the other hand, if schools did not receive any goal, dropout and repetition probability from students in treated schools increased. Finally, because of these mix effects - better test performance and worse repetition and dropout rates - we calculated and estimated the impact in the Index of Development of Basic Education (IDEB). This index put together: performance in tests and rates of students improved. Our results show that the impact in this index was lower than in the student tests performance in schools from areas with lower approbation rates.