This study analyses the impact of a new admission policy on patterns of school segregation in Rio de Janeiro. Termed the “Lottery System”, the new legislation randomly allocates children to schools, based on a list compiled by parents with their preferences (of up to five schools). Segregation in terms of two different indicators of potentially disadvantaged students were calculated using the Dissimilarity Index and the Segregation Index, for parents' education and student ethnic background. Rio de Janeiro city has one of the largest public school networks in Latin America. There are around 1,400 schools providing Preschool, Primary and Secondary Education. This study focuses on Primary Education (1st to 5th grade; children age 6 to 11), analysing the total population – around 250,000 pupils in a total of 700 schools Previous studies in Rio de Janeiro public schools have described patterns of school segregation for public and private schools and tested the plausibility of the school composition effect theory. School admissions policies could be seen as an important factor to remove barriers to disadvantaged pupils getting into desirable schools. There is a wide debate across different countries about the potential impact of lottery schemes to reduce pupils’ selection. Here we briefly summarize the international evidence related to lotteries schemes and present new evidence for the city of Rio de Janeiro, using secondary data for all public schools from 2006 to 2016. The initial hypothesis states that the implementation of lottery schemes in oversubscribed schools will reduce the capacity of school bureaucracy to select pupils and, therefore, reduce the overall levels of schools segregation in the public network. In order to analyse the potential impact of the new admission policy, the study presents a quasi-experimental design and Interrupted Time Series. Our analyses compare the school segregation patterns before and after the introduction of this system in the initial years of primary school. Complementary data using regression analysis are presented to enhance understanding about the potential impact of the new policy. Results refute the initial hypothesis. The evidence suggest a rupture in the segregation patterns when comparing the students enrolled in the 1st grade with those enrolled in other grades of Primary Education. School segregation becomes greater considering only the 1st grade pupils with the other pairs in Primary Education as of 2011. The results raise the issue of how similar education policies, implemented in different contexts, could have opposite effects.

None