Should we increase instruction time in low achieving schools? Evidence from Southern Italy

•We study the effect of lengthen instruction time on mathematics and language scores.•Identification relies on a difference-in-differences assumption and we investigate both average and quantile effects.•We find a positive effect for mathematics but not for language.•Effects for mathematics are driven by the best students in schools with low socio-economic background.•In the least problematic schools, more time spent at school on activities in language has negative effects on mathematics.

This paper investigates the short term effects of a large scale intervention, funded by the European Social Fund, which provides additional instruction time to selected classes of lower secondary schools in Southern Italy. Selection is addressed using institutional rules that regulate class formation: first year students are divided into groups distinguished by letters, they remain in the same group across grades at the school, and the composition of teachers assigned to groups is stable over time. Using a difference-in-differences strategy, we consider consecutive cohorts of first year students enrolled in the same group. We compare participating groups to non-participating groups within the same school, as well as to groups in non-participating schools. We find that the intervention raised scores in mathematics for students from the least advantaged backgrounds. We also find that targeting the best students with extra activities in language comes at the cost of lowering performance in mathematics. We go beyond average effects, finding that the positive effect for mathematics is driven by larger effects for the best students.