This paper studies the effect of longer school days — induced by voluntary all-day programs in German primary schools — on school performance. We combine data from the National Educational Panel Study covering 5771 primary school students with municipality-level information on all-day school investments. Facing the challenge of selection into all-day school programs, we instrument all-day school expansion with construction subsidies from a large federal investment project. Results imply that all-day programs lead to improvements in language and math skills as measured by teacher assessments and to a higher probability of being recommended for the academic track after primary school. The heterogeneity analysis reveals that boys benefit more than girls from all-day programs in terms of the assessment by their math teacher. Furthermore, there is a significant negative effect on non-native speakers' math and German test scores.
all-day school skill development educational inequality