School shutdowns were adopted worldwide to contain the transmission of the Covid-19. To benchmark the impacts of the current outbreak on learning, this paper explores the 2009 H1N1 pandemic in the state of São Paulo, Brazil, where all the schools managed by the state government had their winter break extended. Thirteen most hit municipalities adopted the same measure in their local schools, while the other municipal authorities maintained the school calendar as usual. We take advantage of the variation between municipalities to employ a difference-in-differences approach on the set of locally-managed schools. We also explore the variation within municipalities to employ a triple difference-in-differences that compares schools run by state and local authorities.
Unit of observation: schools. Source of the data: Prova Brasil and School Census. Grade level: fifth grade. Period: 2005, 2007 (pre-treatment) and 2009 (post-treatment). Dependent variables: proficiency of students in Portuguese, proficiency of students in Math, standardized performance in Portuguese and Math, percentage of students with insufficient performance in Portuguese and percentage of students with insufficient performance in Math. Independent variables: the vector of control variables was chosen based on a Lasso regression, which selects the variables that better predict the variation in the proficiency score. The selected variables are: if the school has science and computer lab, sport court, library and access to the internet; instruction hours per day; students per class; approval rates; GDP per capita of the municipality where the school is located; and socioeconomic characteristics of fifth graders. The vector of socioeconomic variables includes the percentage of mothers with a high school diploma; the percentage of students that already repeated or dropped out school; the percentage of whites and girls; the percentage of students that already work, that previously studied in a private school and that have a computer at home; the percentage of students whose parent's incentive them to study, to do the homework, to read, to not miss classes, and that talk about what happens in the school.
If you want to have access to the dataset (.dta), please write to Vivian Amorim: vamorim@worldbank.org or vivianamorim5@gmail.com