Schools and the transmission of Sars-Cov-2: Evidence from Italy

This paper studies the effect on the spread of Sars-Cov-2 in Italy of schools’ re-openings and closures. Exploiting different re-opening dates across regions after the summer break of 2020, I show that early-opening regions experienced more cases in the 40 days following school re-openings compared with late-opening ones. However, there is great uncertainty around the estimates, and this suggests a wide dispersion in the effects of school re-openings on Sars-Cov-2 transmission. I also study the effect of school closures in Campania, one of the biggest regions in Southern Italy. Using a synthetic control approach, I show that school closures are associated with lower numbers of cases relative to the counterfactual group, particularly in younger age groups. In contrast, I find no significant effects on older age groups, which are more likely to require hospitalization. Finally, by exploiting survey data, I provide descriptive evidence on the increased incidence rate among teachers and students relative to the general population, following school re-openings. •I study the effect of school openings on the spread of Sars-Cov-2 in Italy, exploiting staggered re-openings across regions.•Early-opening regions experienced more cases in the 40 days following school re-openings.•I also study the effect of school closures in Campania, one of the biggest regions in Southern Italy.•School closures are associated with lower case numbers relative to the counterfactual synthetic group.•The effects are driven by younger age groups, while the response in older age groups is muted.•Incidence within schools is larger where classrooms are more crowded and school buildings are older.