It’s less common in the region I live in, but school closures don’t just happen during the pandemic. A few kids get sick, absentee rates rise, and soon the local school board decides to close a class for a week. It’s meant as a health measure, but what does it do to learning?

A team of Japanese researchers — Masato Oikawa, Ryuichi Tanaka, Shun-ichiro Bessho, Akira Kawamura, and Haruko Noguchi — decided to find out. Their new paper in the Journal of The Japanese and International Economies looks at class closures caused not by COVID-19, but by the seasonal flu. That might sound mundane, but it’s precisely what makes their work valuable: by studying shorter, unplanned interruptions in schooling before 2020, they can isolate the real educational effects of losing a few days of class time — without all the chaos of a pandemic.

Using detailed administrative data from a large city near Tokyo, the authors tracked thousands of primary and middle school students between 2015 and 2017. They compared test results from classes that had been temporarily closed due to influenza outbreaks with those that hadn’t, controlling for prior performance, school characteristics, and local factors.

Their conclusion? Lost learning time hurts — but not everyone equally.
Math scores dropped among children from low-income households, especially among boys and those already struggling academically. The average decline was modest (roughly 0.05 standard deviations), but for the most disadvantaged students, the effect was 10 times larger than for their more affluent peers. Language results, by contrast, were mostly unaffected. Interestingly, the study also found that high-quality teachers could mitigate the impact of closures on disadvantaged students.

In other words, a short flu-related shutdown might not seem like much, but for some students it leaves a measurable scar.

Why this matters beyond Japan

The study echoes what we saw during the COVID-19 crisis: when learning time disappears, inequality widens. But unlike the pandemic, these “mini-closures” offer a clearer natural experiment. Everyone faced the same shock, within the same system — only some classes happened to close.

This makes the findings more than just a Japanese curiosity. They highlight a general principle: instructional time matters, and its loss compounds existing disadvantages. The problem isn’t just fewer hours in the classroom, but also behavioural changes that may follow — disrupted routines, less structured time, and the absence of teacher guidance that keeps learning on track.

A note of nuance

Of course, not every closure is equally harmful. The effects were concentrated in primary school mathematics, not in older grades or language learning. That suggests that the damage is greatest where skills are most cumulative and fragile. And while the average effect is small, minor differences accumulate over the years, especially for those who already start behind.

Still, the authors avoid alarmism. They show that high-quality teaching can buffer shocks: high-performing teachers reduced or even eliminated learning loss for their vulnerable students. That’s an encouraging message: disruption doesn’t automatically mean decline.

A broader reminder

It’s tempting to think that brief interruptions in schooling are harmless, or that children will naturally “catch up.” This study — quiet, careful, and grounded in real-world data — shows that’s not always the case. Even a few missed days can deepen inequality if no support follows.

And yet, it also reminds us that the best antidote isn’t panic or endless remediation, but what education research keeps pointing to: investing in strong, well-supported teachers.

Abstract ot the study:

This paper examines the impact of class closures on the academic achievement of primary and middle school students, with a particular focus on heterogeneous effects related to their household socioeconomic backgrounds. Using administrative data from students in a Japanese city within the Tokyo Metropolitan Area, we analyze the effects of class closures due to influenza epidemics on students’ language and mathematics test scores. Our findings indicate that class closures adversely affect the mathematics test scores of economically disadvantaged students. The magnitude of these negative effects on disadvantaged students varies by subject, grade level, gender, the timing of closures, and students’ prior academic achievement. In particular, male students from economically disadvantaged households are more susceptible to class closures, and those with lower prior academic achievement experience more severe adverse effects. These deleterious effects among economically disadvantaged male students appear to be driven not only by reductions in in-school instructional time but also by behavioral changes that may diminish their learning capacity. Furthermore, we find that high-quality teachers can mitigate the negative impact of class closures on economically disadvantaged students. These results highlight the importance of public programs designed to safeguard student learning environments against such adverse disruptions.

Image: https://www.bbc.com/news/uk-england-derbyshire-42305871