Effects of two school closings on learning gains in Dutch primary education: Differences between students and schools

In this study we analyse the inequality in average learning progress of ~325k unique primary school students in the Netherlands of almost 1700 schools (~25% of all schools) during the first year of the COVID-19 crisis. We study the effect of both the first school closure period in spring 2020, which lasted 6 weeks, as well as the effect of going back to school after the summer holiday which was then again followed by a school closure for another 6 weeks between Dec 2020 and Feb 2021. We analyse learning progress in reading comprehension, spelling and arithmetic, and show that large learning losses occurred during the first school closure period. Furthermore, despite the second school closure, during the first semester of the new school year most students and schools had larger learning progress than usual and so the learning losses after one year COVID-19-crisis are a lot smaller than after the first school closure. However, there are large difference between different types of students and different types of schools. We present differentiated results by grade level,gender, student socio-economic status (SES), student home and family situation, as well as bydifferent school characteristics. The results show that the learning loss in the first period is thelargest for reading comprehension, and that learning losses are also still largest after one year. Furthermore, the results show that learning losses are largest for students from low socioeconomic background, student with an unstable or more difficult home (divorced parents, or many siblings), and for schools with many low-SES students. Several robustness checks confirm these results