The Disaster of School Closures Should Have Been Foreseen

The Atlantic

17-04-2025

Of the many mistakes made in the COVID era, none were as glaring as prolonged school closures. The damages go beyond loss of learning, a dire consequence in its own right: Millions of families, both children and parents, still carry the scars of stress, depression, and isolation.
The closures began at a time of understandable panic, but that was only the beginning of the story. On February 25, 2020, Nancy Messonnier, the director of the CDC's National Center for Immunization and Respiratory Diseases, led a press conference to address the developing coronavirus crisis. Messonnier warned the public that, without vaccines, non-pharmaceutical interventions—things like business closures or social-distancing guidelines—would be the most important tools in the country's response. 'What is appropriate for one community seeing local transmission won't necessarily be appropriate for a community where no local transmission has occurred,' she said. The school closures that would be implemented the following month—and that endured through the end of the school year in nearly all of the roughly 13,800 school districts in the United States, in regions that had wildly different infection levels—showed this directive was not followed.
At the time of the initial closures, in mid-March, COVID was spreading quickly, but large areas in the U.S. were absent any known cases. Still, to the extent that a planned response to influenza was an appropriate universal pandemic guide, these closures were aligned with the CDC's most recent update to its pandemic playbook, released in 2017. According to that document, an initial two-week closure of schools would be sufficient to fulfill a first objective of buying authorities time to assess the severity of the pandemic. Given the news being reported of care rationing in northern-Italian hospitals, following this plan was not unreasonable—and, as part of broader stay-at-home orders, it may have had some effect on disease transmission.
'Italy spooked us,' Jennifer Nuzzo, director of the Pandemic Center at the Brown University School of Public Health, told me. 'We did not want to be Italy. The governors all saw China and Italy lock down and decided to follow their example.'
Indeed, had China not locked down, perhaps the rest of the world might not have done so either. China, governed by an authoritarian regime that rules the country with tremendous top-down power, does not share the same attitude toward personal liberties as Western democracies. And neither the CDC's pandemic playbook nor the pre-2020 consensus within the public-health field favored a lockdown of society of the breadth that we would experience. On a countrywide scale—from both an epidemiological perspective and a human-rights perspective—closing all nonessential business, closing all schools, prohibiting most social interactions and nonessential travel, and so on, was not considered feasible or wise. Because we initially lacked the ability to test, Nuzzo said, shutting schools, along with other facets of society, did make sense at first. The problem, in Nuzzo's mind, was not closing down in March; it was that there was no plan beyond that. By 'no plan,' Nuzzo was referring to two interrelated problems: all the potential harms of closures, and the challenge of unwinding interventions after they'd been implemented.
In Messonnier's press conference, she mentioned the CDC's 2017 pandemic report directly, said that school closures were part of the plan, and recognized that they were likely to be associated with unwanted consequences such as missed work and loss of income. 'I understand this whole situation may seem overwhelming and that disruption to everyday life may be severe,' she advised. 'You should think about what you would do for child care if schools or day cares close.' There was no mention of how the government might aid families during school closures, or, for example, about what a single parent with a job as a cashier in a grocery store and a 4-year-old at home was supposed to do. Rather, in just one line amid a lengthy speech, people were told to simply 'think about' it. To government officials and many others at the time, this was a regrettable but entirely reasonable approach—a presumed temporary loss of wages and child-care issues were lower-order concerns compared with the coming onslaught of a pandemic.
Yet what was positioned as a secondary issue—a mere abstraction, warranting just a brief mention—led to catastrophic consequences for millions of children, and their families. A year later, my kids, along with tens of millions of other students, were still trudging through remote learning, either as their exclusive form of schooling or through so-called hybrid schedules during which they could attend classes only part time. (Meanwhile, bars, restaurants, and all manner of other businesses had long since reopened, as had many private schools.) Teachers in much of the country had been prioritized for vaccines—making them eligible for protection before some other, more vulnerable populations—yet schools in half the country still weren't open full time, and in many places weren't open at all.
While federal public-health officials made recommendations regarding schools, the actual closures were carried out at the state and local levels, in response to misplaced public fears and aggressive campaigning by teachers' unions. Randi Weingarten, the high-profile head of the American Federation of Teachers, said in a February 8, 2021, New York Times article that she hoped things would be 'as normal as possible' by the following fall. Class-action lawsuits in multiple states had been filed on behalf of children with special needs on the claim that the conditions of IDEA—a federal law that requires certain services (such as physical and occupational therapy, supplemental aids and equipment, etc.) for children with disabilities—were not being met in remote-learning models.
Opinion pieces with titles such as 'Remote School Is a Nightmare. Few in Power Care,' had been appearing in major news outlets since the previous summer. Working parents, especially mothers, were dropping out of the workforce in staggering numbers because of child-care obligations during the pandemic. An analysis by the U.S. Chamber of Commerce found that nearly 60 percent of parents who had left the workforce had done so for this reason. The psychic toll on parents and children was never—and can never be—calculated. It won't show up in statistics, but it was real for millions of families. And millions of children, especially those without resources for tutors or parents to oversee them during the day, were losing ground with their academics. Worse, they were suffering from isolation, frustration, and, for an increasing number of them, depression from spending their days alone in front of an electronic screen. Untold numbers of other children became 'lost,' having dropped out of school entirely. Those in power who advocated for school closures were not adequately prepared for these consequences, which were still pervasive a full year into the pandemic.
But they should have been.
The damaging effects of school interruptions were not unforeseen. They were explicitly warned about in the academic literature. Exhibit A is a 2006 paper called 'Disease Mitigation Measures in the Control of Pandemic Influenza,' in the journal Biosecurity and Bioterrorism, written by Nuzzo; her mentor (and global eminence on disease-outbreak policies), D. A. Henderson; and two others. 'There is simply too little experience to predict how a 21st century population would respond, for example, to the closure of all schools for periods of many weeks to months,' these authors wrote. 'Disease mitigation measures, however well intentioned, have potential social, economic, and political consequences that need to be fully considered by political leaders as well as health officials. Closing schools is an example.'
The authors went on to warn that closures would force some parents to stay home from work, and they worried about certain segments of society being forced to bear an unfair share of the burden from transmission-control policies. They wrote:
No model, no matter how accurate its epidemiologic assumptions, can illuminate or predict the secondary and tertiary effects of particular disease mitigation measures … If particular measures are applied for many weeks or months, the long-term or cumulative second- and third-order effects could be devastating.
Nearly a decade and a half before the pandemic—in a stark rebuke to the approach championed by the CDC, White House Coronavirus Response Task Force coordinator Deborah Birx, and other powers that be—the paper had called out the major harms that would come to afflict many families in our country as a result of school closures. Yet, from the spring of 2020, health officials who directed our pandemic response ignored many of the consequences they must have known to expect. Or, at the very least, they failed to provide adequate information about them to the public. The officials had opened a bottle of medicine while disregarding the skull and crossbones on the warning label.
And the portents were not just in Henderson's and Nuzzo's paper. A 2011 paper by researchers from Georgetown assessed the decision making behind—and the consequences from—several hundred brief school closures enacted during the 2009 H1N1 pandemic. The authors noted that the child-care costs to families were substantial, and that hardships from closures were inequitable. 'Officials considering closure must weigh not only the total amount of disruption but also the extent to which social costs will be disproportionately borne by certain segments of society,' the authors wrote. Even the CDC playbooks themselves warn of some of these issues. Both the 2007 original and the updated 2017 report cautioned that school closures could lead to the secondary consequence of missed work and loss of income for parents who needed to stay home to take care of their school-age children. This effect, the latter report noted, would be most harmful for lower-income families, who were also hit hardest by COVID in the first place. With prescience, and comic understatement, the authors noted that school closures would be among the 'most controversial' elements of the plan.
Meanwhile, the second of Nuzzo's points—that unwinding interventions is often incredibly difficult, and there must be a plan on how to do so—was also a well-established phenomenon. Just as public-health experts are biased toward intervention, they, along with the public, are also biased toward keeping interventions in place. This is a known phenomenon within the literature of implementation science, a field of study focused on methods to promote the adoption of evidence-based practices in medicine and public health. Westyn Branch-Elliman, an infectious-diseases physician at UCLA School of Medicine with an expertise in implementation science, told me that de-implementation is generally much harder than implementation. 'People tend to err on the side of intervening, and there is often considerable anxiety in removing something you believe has provided safety,' she said.
There also is a sense of inertia and leaving well enough alone. It's not unlike legislation—oftentimes repealing a law, even an unpopular one, poses bigger challenges than whatever barriers existed to getting it passed. Although the initial school closures may have been justifiable (even if off-script in many locations), there was no plan on when and how to reopen. Officials repeated a refrain that schools should open when it was 'safe.' But 'safe' was either pegged to unreachable or arbitrary benchmarks or, more often, not defined. This meant there would be limited recourse against a public that had been led to believe this intervention was a net benefit, even long after evidence showed otherwise. The lack of an exit plan—or an 'off-ramp,' as many health professionals would later term it—would prove disastrous for tens of millions of children in locations where social and political pressures prevented a reversal of the closures.
Without sufficient acknowledgment of the harms of school closures, or adequate planning for unwinding this intervention, officials showed that their decisions to close were simply reactive rather than carefully considered. The decision makers set a radical project in motion with no plan on how to stop it. In effect, officials steered a car off the road, threw a cinder block on the accelerator, then jumped out of the vehicle with passengers still in the back. No one was in the front or even knew how to unstick the pedal.