Back to School. Freedom at Last.

The end of summer marks the beginning of freedom. You’re not really supposed to say this, but school isn’t just about educating kids; it’s about watching them. Working parents know the dread of scrambling for sitters or rearranging their own schedule when school is unexpectedly out of session. One 2016 study estimated that paying out of pocket for child care during periods when schools are closed during the standard workday would cost $6,600 annually.

A recent National Bureau of Economic Research (NBER) working paper indicates that school closures may even drive parents to drink. The researchers found that in places with longer-than-average school closures during the coronavirus pandemic, antidepressant use by mothers skyrocketed. When schools gradually began to reopen, new antidepressant prescriptions returned to pre-pandemic levels. The researchers also looked at alcohol purchases from grocery stores and found increased alcohol purchases in counties with longer-than-average pandemic school closures, which suggests that people—the authors could not disaggregate the trend by gender—were self-medicating because of the stress of school closures.

[Read: America’s child-care equilibrium has shattered]

The authors of the paper, which is titled “School Closures and Parental Mental Health,” conclude that “the school system plays an important role in maintaining population mental health outcomes and in helping families cope with stress.” Simply put, schools aren’t just for education, one of the authors, the Indiana University economist Sumedha Gupta, told me. “At the end of the day, schools play an important role for families,” she said. “The educational piece is one [part], but … schools provide counselors, therapies for children who have learning disabilities. So it is day care, plus all these other things.”

By now regret over long school closures during the coronavirus pandemic has filtered through academic and policy circles. Study after study has revealed the learning loss and behavioral difficulties that students experienced during remote learning. But shutting down schools wasn’t just bad because of learning loss, because schools aren’t just for kids. The problem isn’t merely what school closures have done to kids; it’s also what they’ve done to parents.

During the pandemic, some teachers bristled at the presumption that educating the next generation of citizens wasn’t their only role. Social media abounded with sentiments like those expressed in a July 2020 blog post titled “School Is Not Childcare,” whose author, an educator, took note of complaints that closed schools could not respond to children’s hunger or protect them from harm at home. “But what does that really have to do with education? Is that really the role of the school system?” the author asks. No, she concludes, “that is the job of the community, the government, and the people.”

The working paper by Gupta and her Indiana University colleagues Dario Salcedo and Kosali Simon tells a distressing story about what happens when this care disappears: Mothers suffer. Prior research shows that time spent on child care skyrocketed for women during the pandemic. “What really hit home with me was this mothers-versus-fathers effect,” Gupta said, adding, “Women are still bearing the brunt of these things.”

Strikingly, the paper found “no statistically significant effects of school closures on fathers’ antidepressant demand; all effects are among mothers.” The increased demand for antidepressants was concentrated among Black and Asian mothers, and increased alcohol consumption was concentrated in counties with higher Black and Asian populations.

One important question is: How do we know these effects were driven by school closures and not by the countless other stressors of a global pandemic? After all, people were hospitalized; more than 1 million Americans died; kids were scared; parents were losing their jobs. How could the researchers know why mothers started demanding SSRIs?

[Read: The biggest disruption on the history of American education]

The answer is: They can’t be totally sure. The Johns Hopkins economist Nicholas Papageorge, who was not involved with the new paper but spoke approvingly of it, told me that school closures might have been correlated with other public-health closures, including those involving bars, parks, and other spaces where people get away from domestic stresses. In which case, mothers could be driven to drink at least in part because of other changes happening in their community that correlate with school closures.

However, as Papageorge noted, the recent findings are consistent with other research on the knock-on effects of school closures. A 2021 study detected an increase in parental stress, intimate-partner violence, child abuse and neglect, and parental substance abuse during the pandemic. A systematic evidence review in 2022 noted deteriorating mental health among parents during school closures.

But perhaps the most persuasive evidence in the NBER paper’s conclusions is the distributional changes that the researchers detected. The new paper indicates that women with elementary-age children suffered the most, consistent with expectations that young children require the most time and attention. Further, the researchers found no effect for fathers, which is consistent with research about who was likely to be most burdened with child care.

One clear upshot of the new research is that extending school would be helpful to mothers and to the broader community. Fortunately, year-round school is also good for kids. Philadelphia’s mayor is piloting a year-round school option in 25 schools starting this year, citing a long-studied phenomenon called “the summer slide”—which refers to the “losses in learning, health and well-being, college and career opportunity, and support needed to break cycles of intergenerational poverty” that happen in the summer, as one paper put it. Schools elsewhere should be considering similar interventions.

I am very sympathetic to the argument that we expect too much from schools and teachers without providing them adequate support. But the problem isn’t viewing schools as also serving these broader functions; the problem is not equipping them with the resources to adequately do so. And perhaps this is the hardest pill to swallow: Even without adequate resources, schools still play the role of keeping an eye on kids, freeing mothers to work, reducing crime, lowering hunger, and watching for signs of abuse. Whether they are closed for Presidents’ Day or because of a teacher strike or a pandemic, the cost isn’t just a few points on a standardized test. It may well be the stability of an entire community.