Skip to main content
Statement banner
Policy Insights

How Do School Closures Affect Student Learning? It’s Worse Than You Think

By Per Engzell, Arun Frey & Mark Verhagen (Leverhulme Centre for Demographic Science, University of Oxford)

Per Engzell, Arun Frey & Mark Verhagen of the Leverhulme Centre for Demographic Science, Oxford University find that during the first COVID-19 lockdown in the Netherlands, students learned less than during a normal year. These results highlight the costs of keeping students out of schools and of the difficulties in compensating for these costs.
How Do School Closures Affect Student Learning? It’s Worse Than You Think
Source: sommersby

The impact of school closures on student learning has been a topic of heated debate since they first became a tool in the battle against the COVID-19 pandemic. During the first pandemic wave in spring 2020, most countries instituted school closures in some form – to the point that 95% of the world’s student body was affected. In contrast, going into a second lockdown, countries like Germany, France and England have opted to close workplaces, restaurants and bars, while letting schools remain open. This policy is grounded in the concern that keeping students from learning will have irreparable consequences down the line, unlike the immediate damage to jobs and the economy that can at least in principle be kept at bay through furlough schemes and stimulus packets. But is there evidence for this view? Data from the Netherlands answer that question with a resounding “yes”.

In a recent working paper, we provide new evidence on learning loss in the spring of 2020. Until recently, most knowledge about the effects of school closures was based on natural experiments where schools closed for other reasons: extreme weather, teacher strikes, or skill stagnation that occurred as a result of students being out of school during the summer. However, the COVID-19 pandemic presents a singular event that must be studied in its own right. We do so with unique data from the Netherlands, where students’ performance is tracked throughout their school career. Dutch primary schools closed for eight weeks beginning on March 16 and reopened on May 11. We can observe the same students as they were tested through nationally standardised tests shortly before and after lockdown. We then compare the learning progress during this period to that of students in the three prior years.

The Netherlands is an exceptional country and more of a best case scenario, as they were unusually well prepared to handle the pandemic. For example, data from Eurostat show that the Netherlands leads the world in broadband access: In 2019, more than 90% enjoyed broadband even among the poorest quartile of households. These figures are similar if we look at the presence of a computer at home among PISA respondents. Adding to this advantage, the policy response has been swift: Already in March 2020, the Ministry of Education devoted 2.5 million euros to purchase online learning devices for students in need, and this scheme was extended with another 3.8 million in June. According to PISA, parental support in the Netherlands also ranks close to the top and is distributed relatively equally across family backgrounds.

Despite having such a strong infrastructure, our study found robust evidence that students are still learning less in lockdown than in a normal year. The lost progress amounts to about a fifth of a school year, almost exactly the same period that schools remained closed (assuming a regular school year is ten months). In other words, the data suggest that nearly every hour out of school was an hour of lost student learning. These losses are not distributed equally. Students from low-educated homes were disproportionately affected, with a 50% larger drop in performance than their more advantaged peers. We could also confirm that most of the drop in performance reflects the actual knowledge learned as opposed to testing performance as such. Performance on tests not related to the curriculum also declined, but to a much lesser extent than in core subjects such as Maths, Reading and Spelling.

As large parts of the world are heading into a second wave of the pandemic, it is vital to know how school closures impact students' progress, and consider the disproportionate harms to students from disadvantaged homes. With new school closures being discussed as a way to combat the pandemic, our study gives crucial input for decision-makers. The results are sobering; even in the “best-case scenario” of a short lockdown and good infrastructure for remote learning, Dutch students learned little or nothing from home. These results highlight the urgency of meeting students’ educational needs – and of measures to compensate for the progress already lost. Responding to the pandemic involves inevitable trade-offs and there are no easy choices to make. Nonetheless, among all interventions on the policy menu, school closures should remain close to the bottom of the list. The costs of keeping students out of schools are much more difficult to make up for than damage to the economy, and may be here to stay with us for a generation to come.


The study 'Learning Inequality During the Covid-19 Pandemic' is authored by Per Engzell, Arun Frey, and Mark Verhagen and can be found here:

The version is a working paper and thus not yet peer reviewed.

Data on contextual factors in the Netherlands cited above come from the following sources: