By the spring of 2022, when American schools had largely returned to in-person instruction, the test score data began arriving. The first NAEP results post-pandemic showed the largest drop in reading and math scores in the history of the assessment. Children had lost, on average, the equivalent of half a school year of learning in math and a quarter year in reading. The losses were not evenly distributed. Children in high-poverty schools lost roughly twice as much as children in low-poverty schools. Black and Hispanic students lost more than white students. The students who had the fewest resources for remote learning lost the most, exactly as predicted in the spring of 2020 by the researchers who said this would happen.
Three years on, the recovery has been partial and uneven. Some districts are nearly back to baseline. Many are not. Some students never returned to school at all—chronic absenteeism roughly doubled and has stayed elevated. The federal government spent about $190 billion in emergency school aid (ESSER funds), the largest one-time investment in K-12 education in American history. The money is now spent or expiring, and the question of whether it bought a generation of educational recovery or whether it largely papered over a damage that will persist into adulthood remains unsettled.
The honest answer is that we are running an enormous natural experiment in real time, and the early results are not encouraging. Tutoring works when it is high-dose (three sessions a week) and high-quality, but few districts implemented it at scale because the labor supply was not there. Summer school works at the margin but is voluntary and reaches mostly the students who need it least. Extended school day works when teachers are not exhausted, which most are. The interventions that the research said would help were the ones districts found hardest to operationalize. The interventions that were operationally easier—digital learning platforms, teacher coaching, generic professional development—are the ones whose effects on actual student learning are smallest.
Law 4 is planning, and Law 5 is revision. The pandemic was not a planning failure in the sense that no one could have anticipated it; the failure was that the existing plan—decentralized school districts making independent choices in a national emergency—was the wrong instrument for a coordinated response. Some districts reopened in fall 2020 with adequate safety measures. Some did not reopen for over a year. The variance was not driven by epidemiology; it was driven by local politics, union dynamics, parental composition, and the inertia of the system. Children's outcomes followed accordingly.
Emily Oster's analysis of the period is essential because she was among the few public voices saying, in real time, that the in-school transmission data did not justify keeping schools closed as long as many districts kept them closed. She was attacked for this position. The data, with hindsight, largely supported her. The longer school closures lasted, the worse the academic and mental health outcomes became, and the small benefit to community transmission that closures provided once vaccines were available was, in most analyses, swamped by the cost. This is not a partisan claim; it is what the cohort outcomes show.
The collective question for the next decade is whether the educational system can metabolize what happened. The lost learning is real and durable—students who fall behind in math compound their disadvantage in later math courses, and reading deficits affect every subject. The mental health damage is real and durable—anxiety, depression, and chronic absenteeism rose sharply and have not returned to baseline. The teacher workforce was burned by the experience and has not recovered; many of the most experienced teachers retired early. The students entering kindergarten in 2020 and 2021 missed crucial early socialization and are now entering elementary school visibly behind the cohorts ahead of them in social and self-regulation skills.
What is next, then, is not a return. It is a long, expensive, deliberately structured catch-up that the political system is already losing patience for. ESSER funds are gone. Public attention has moved on. The states that committed to long-term high-dose tutoring are still doing it; most are not. The cohort of children who were five to fifteen during the pandemic will carry the academic and emotional residue of that period into their working lives, and the population-level economic cost—projected by Eric Hanushek and others at trillions of dollars in lifetime earnings—will materialize slowly over the next four decades.
A society that wants to absorb this honestly has to do two things at once. It has to fund and execute a sustained recovery program for the affected cohort, at a scale and duration that matches the damage. And it has to revise its planning instruments so that the next time a national shock hits the school system—and there will be a next time—the response is not seventy thousand uncoordinated local decisions made by exhausted boards. Neither task is happening at the scale it requires. The cohort is being asked, in effect, to absorb the loss. They will, because they have no choice. But the loss will show up in their lives.