Scientists published a groundbreaking study on school re-openings. Major outlets fumbled the stats.

A new study published in Epidemiology provides evidence that returning to in-person learning during the COVID-19 pandemic benefited children’s mental health. While some mainstream reporting has misinterpreted these results as an absolute drop in cases, the research analyzes insurance claims from California to show that school reopenings instead helped slow the rising rates of mental health diagnoses and related healthcare costs among youth.

The onset of the COVID-19 pandemic forced a massive shift in the daily lives of families across the United States. By late March 2020, almost all public schools had closed their physical doors to prevent the spread of the virus. Students transitioned to remote learning formats that fundamentally altered their educational and social routines. Public health experts expressed immediate concern regarding the potential long-term effects of this isolation on children.

Schools provide more than just academic instruction to young people. They offer socialization, structure, and access to mental health resources. The shift to remote learning created a natural environment to test how the mode of instruction impacts well-being. Previous studies documented general declines in youth mental health during the pandemic. However, there was less clarity regarding how much of that decline was specifically attributable to school closures versus the pandemic in general.

Some earlier research relied on parent surveys or small samples, which can introduce bias. Other studies looked at emergency department visits but lacked a comprehensive view of ongoing mental health treatment. This study aimed to isolate the impact of returning to the classroom using objective medical data. The researchers sought to understand if the reopening of schools correlated with a change in the trajectory of mental health diagnoses.

“During the pandemic, school closures were a major disruption to children’s routines, but early research focused primarily on academic outcomes. While there was anecdotal evidence that closures were linked to worsening mental health, the causal link had not been rigorously explored,” said study author Pelin Ozluk, a real world evidence expert at UCB (Union Chimique Belge), who conducted the study along with researchers from Harvard School of Public Health and Elevance Health, Inc.

Free daily newsletter

The research team utilized a quasi-experimental design to analyze the relationship between school reopenings and mental health outcomes. They focused on a specific period between March 2020 and June 2021 across 24 counties in California. This allowed the investigators to compare outcomes between districts that had returned to in-person instruction and those that remained remote.

“The staggered school reopenings in California created conditions that mimicked a natural experiment: while most schools closed at the same time, they reopened at very different times,” Ozluk explained. “This allowed us to follow the same child over time—nearly 200,000 children in total—and compare their outcomes before and after their specific school reopened, effectively isolating the impact of school from other confounding factors.”

The study drew data from a large database of medical claims from the second-largest private health insurer in the state. The final sample consisted of 185,735 children between the ages of 5 and 18. These children had continuous medical insurance eligibility throughout the study period. The researchers linked this health data with administrative information from the California Department of Education to determine the reopening status of each child’s district.

The researchers employed a statistical method known as difference-in-differences analysis. This approach controls for general trends that affect everyone, such as the overall progression of the pandemic or the rollout of vaccines. By focusing on the differences that arose specifically when schools reopened at different times, the team could isolate the effect of in-person learning. They tracked monthly prevalence of diagnoses and prescription fills.

The findings indicate that school reopenings were associated with a distinct improvement in mental health outcomes. Specifically, the analysis showed a 1.2 percentage point drop in the monthly probability of a child receiving a mental health diagnosis after their school reopened. This represents a relative reduction of approximately 43 percent compared to the pre-reopening baseline. The study suggests that the physical school environment plays a protective role.

“The key takeaway is that in-person school is not just for academics; it is an essential public health intervention,” Ozluk told PsyPost. “While a 1.2 percentage-point improvement may sound small as an absolute number, it represents a 43 percent improvement relative to the baseline of children seeking care at that time. In the context of a statewide population, this represents thousands of children who avoided a mental health crisis because they returned to the classroom.”

In addition to diagnoses, the researchers looked at healthcare expenditures. The researchers found a 10.6 percent decrease in spending related to mental health services following the return to classrooms. This reduction included costs for both medical visits and prescription medications. This financial metric provides another layer of evidence regarding the demand for mental health services.

The researchers examined specific conditions including anxiety, depression, and attention-deficit/hyperactivity disorder. The data showed that the largest improvements occurred in diagnoses for anxiety and depression. The effects on attention-deficit/hyperactivity disorder were less pronounced and appeared later in the timeline. The reduction in diagnoses typically became evident about six months after schools reopened.

The impact of returning to school was not uniform across all groups. The results suggests that the benefits were strongest among girls. Girls saw a significant decrease in both diagnoses and related costs after returning to in-person learning. The researchers theorize this may be due to gender differences in socialization. Girls often rely more on social conversation and peer relationships, which were severely restricted during closures.

The researchers also found that the effects were larger for adolescents aged 14 to 18 compared to younger children. This supports the idea that older children, who are more socially oriented toward their peers, suffered more from the isolation of remote learning. The reopening of high schools provided a restoration of their primary social network. Younger children may have been more buffered by their immediate family environment.

It is important to clarify a key nuance regarding these statistics. The overall prevalence of mental health diagnoses rose from 2.8 percent during the shutdowns to 3.5 percent during the reopening period. The study demonstrates that reopenings acted as a brake on this worsening trend. The return to school slowed the rate of diagnosis compared to what would have occurred had schools remained closed.

Some media outlets previously misinterpreted these findings to suggest that mental health diagnoses dropped below pre-pandemic levels. This is not accurate. The reduction is relative to a counterfactual scenario where schools remained closed. The data shows that while the mental health crisis continued, open schools significantly mitigated its severity.

“A major misinterpretation (notably in the New York Times) implied that mental health diagnoses ‘dropped’ by 43% after schools opened,” Ozluk said. “This is not accurate. In reality, mental health conditions worsened across the board during the pandemic, rising from a 2.8% prevalence during shutdowns to 3.5% during the reopening period. ”

“The improvement we report (the 1.2 percentage point and 11% decrease) refers to the effects specifically associated with the policy of reopening. Our findings do not suggest that cases dropped back to pre-pandemic levels; rather, they show that school reopenings acted as a ‘brake’ on a worsening crisis.”

“Furthermore, because we used a difference-in-differences design to follow the same children over time, statewide trends like vaccine rollouts or a general easing of COVID-19 fears were already ‘teased out’ of the data,” Ozluk continued. “This nuance was not properly addressed in previous reporting, leading some experts to falsely attribute the results to vaccines rather than the schools themselves.”

As with all research, there are also limitations to consider. The data came from a commercial insurer, meaning the sample overrepresents children from higher-income households. These families likely had more resources to mitigate the effects of the pandemic than lower-income families. Consequently, the study may underestimate the true benefit of school reopenings for vulnerable populations who rely more heavily on school-based services.

“Since the publication of this research, I have moved into the pharmaceutical industry, so I will not be working primarily on children’s issues moving forward,” Ozluk said. “However, I believe future research must focus on lower-income households. A major caveat of our current study is that the sample consisted of privately insured children. The impact on children in lower-income families—who often rely more heavily on school for socialization and mental health resources—was likely even more profound.”

The study, “Effect of School Reopenings on Children’s Mental Health during COVID-19: Quasi-Experimental Evidence from California,” was authored by Pelin Ozluk, Jeff Romine, Gosia Sylwestrzak, and Rita Hamad.