The pandemic-era learning loss and what's next

The 2022 NAEP results showed math scores down 5 points for fourth graders and 8 points for eighth graders compared to 2019—the largest declines ever recorded. Reading scores dropped 3 points across both grades. By 2024, eighth-grade math had recovered slightly but eighth-grade reading had actually dropped further, reaching levels not seen since the early 1990s. The pattern is uneven across states and districts; some places have recovered fully, others have moved backward. The aggregate is a system that, three years after returning to in-person learning, has not undone the loss. The students who were in middle school during the closures are entering high school below grade level in numbers that the system does not know how to absorb.

Remote learning was not equally remote. A child with parents who could work from home, in a house with reliable internet and a quiet space and an adult who could supervise schoolwork, experienced school closures as a difficult inconvenience. A child whose parents had to leave the house to work, who shared a phone for schoolwork with siblings, who did not have an adult home during the school day, experienced school closures as the disappearance of school. The gap widened because the same external shock landed differently on differently resourced families. Robin Lake and her colleagues documented this in granular detail through the Center on Reinventing Public Education tracking. The schools that served poorer students were also more likely to stay remote longer, which compounded the effect.

By the 2020-2021 school year, public school districts in the United States made very different choices about reopening. Some reopened in person in August 2020. Some stayed remote through June 2021. Some operated in hybrid models that changed monthly. The variance was correlated with state political composition, district urbanity, and union strength, not with epidemiological conditions. Districts in Florida and Texas reopened earlier; districts in California and the Northeast stayed remote longer. The academic outcome data tracks this variance: longer closures, worse outcomes, particularly for students who were already behind. The conclusion that schools should have reopened sooner is now broadly accepted; in 2020 and early 2021 it was a politically dangerous position to hold.

Chronic absenteeism—a student missing 10 percent or more of school days—roughly doubled during and after the pandemic. In some districts it tripled. Three years on, it has come down somewhat but remains far above pre-pandemic levels. The students who are chronically absent are disproportionately the students who fell furthest behind, which produces a vicious cycle: they are behind, school is harder and less satisfying, so they attend less, so they fall further behind. Absenteeism is the leading indicator of dropout. The cohort of students who became chronic absentees during the pandemic and have not returned to regular attendance represents the most concentrated long-term cost of the period.

The federal ESSER program disbursed about $190 billion to school districts in three tranches. Districts had unusual flexibility in spending. Some invested in evidence-based interventions—high-dose tutoring, extended learning time, mental health staff. Some spent on facilities, technology, or operational costs that displaced what would otherwise have been local spending. The variance in how districts spent reflects the same fragmentation as the closure decisions. The Center on Reinventing Public Education and Eric Hanushek's research suggest that the gains from ESSER spending have been small relative to the loss, and that districts that concentrated on evidence-based academic interventions got more learning recovery per dollar than districts that did not. The funds are now expiring, and the recovery work, where it has begun, will have to continue on local budgets that cannot afford it.

The single intervention with the strongest evidence for learning recovery is high-dose tutoring: small groups (one to four students), several times per week, with a trained tutor following a structured curriculum, integrated into the school day. The effect sizes are large and consistent across multiple randomized trials. The implementation problem is labor: a serious tutoring program requires hiring large numbers of tutors at reasonable wages, training them, supervising them, and integrating them into school schedules. Most districts that attempted this at scale ran into hiring constraints. The intervention works; the operational infrastructure to deliver it does not exist at the scale the recovery requires.

Academic loss and mental health loss are not separable. Adolescent depression and anxiety rose sharply during the pandemic, particularly among girls and particularly in communities with longer closures and more social isolation. CDC data show suicidal ideation rates among teen girls reaching historic highs in 2021. Three years on, these indicators have improved somewhat but remain elevated. A student who is depressed does not learn well, which is part of why the academic recovery has stalled. Schools have responded by hiring counselors—often using ESSER funds—but the demand exceeds the supply by orders of magnitude, and the funding for these positions is expiring. The mental health response was real but is not sustained.

The children who entered kindergarten in 2020 and 2021 are now in second and third grade. Teachers report that this cohort entered with visibly weaker social and self-regulation skills—the developmental work of preschool and kindergarten interrupted by closures, masks, and isolated home environments. Reading instruction in the early grades depends heavily on social interaction, phonemic instruction in groups, and the kind of teacher-child relationship that remote learning could not produce. This cohort is likely to carry early-reading deficits for years, because foundational reading skills compound. The early-childhood loss may turn out to be the most consequential of all the pandemic-era effects.

The pandemic accelerated retirements among experienced teachers, increased mid-career exits, and made the profession less attractive to new entrants. Education school enrollments, already declining, dropped further. The teachers who stayed are exhausted, dealing with classrooms full of students with academic and behavioral needs that exceed their training, and increasingly leaving the profession themselves. The recovery from the academic loss depends on the teacher workforce being able to deliver it. The teacher workforce is, in many districts, less capable now than it was in 2019. Richard Ingersoll's tracking of teacher turnover shows the deterioration continuing well past the immediate pandemic period.

Public attention to pandemic-era learning loss has dropped sharply. School board races have moved on to other concerns. Federal interest is largely exhausted with the end of ESSER. State legislatures are facing fiscal pressure as one-time pandemic federal funds disappear. The recovery is at the moment when it most needs sustained political commitment, and the political system has moved on. This is a common pattern in disaster recovery—the acute attention fades while the long recovery is still underway—but the consequence in education is that the cohort of affected students will graduate into adult life with deficits that the system stopped trying to fix.

A serious plan would commit to a decade of sustained recovery investment, not a three-year emergency response. It would mandate high-dose tutoring for students more than a grade level behind, funded at the state and federal level rather than relying on local capacity. It would extend the school year or school day for the affected cohort, with teacher compensation that reflects the additional work. It would expand mental health staffing as a permanent feature of schools, not a pandemic add-on. It would invest in early childhood education for the cohorts entering school now, on the recognition that the early-years gap is the most reparable. None of these are politically easy. All of them are cheaper than the lifetime earnings loss the cohort will otherwise carry.

Law 5 is revision. The school system that entered the pandemic had a particular planning structure: highly decentralized, locally controlled, with limited federal coordination capacity. That structure produced the variance in closure decisions and the variance in recovery responses. A revision would not centralize K-12 education—the American political tradition will not allow it and probably should not—but it would build coordination capacity for shocks: a national framework for school continuity during emergencies, evidence-based default protocols that local districts can adopt, faster federal response infrastructure for the next pandemic or climate disaster. The absence of this capacity in 2020 was the deeper failure underneath all the specific bad calls. Building it now, while the memory is still fresh, is the work the political system is not currently doing.

1. Oster, Emily. The Family Firm: A Data-Driven Guide to Better Decision Making in the Early School Years. New York: Penguin Press, 2021. 2. Lake, Robin, and Travis Pillow. "The Alarming State of the American Student in 2022." Center on Reinventing Public Education, Arizona State University, 2022. 3. Hanushek, Eric A., and Ludger Woessmann. "The Economic Impacts of Learning Losses." OECD Education Working Papers, No. 225. Paris: OECD Publishing, 2020. 4. Goldhaber, Dan, Thomas J. Kane, Andrew McEachin, Emily Morton, Tyler Patterson, and Douglas O. Staiger. "The Educational Consequences of Remote and Hybrid Instruction during the Pandemic." NBER Working Paper 30010, 2022. 5. Reardon, Sean F., Demetra Kalogrides, and Erin M. Fahle. Education Recovery Scorecard. Stanford University and Harvard University, 2024. 6. National Center for Education Statistics. NAEP 2022 Long-Term Trend Assessment Results. Washington, DC: U.S. Department of Education, 2022. 7. Kraft, Matthew A., and Grace Falken. "A Blueprint for Scaling Tutoring across Public Schools." EdWorkingPaper 21-335. Annenberg Institute, Brown University, 2021. 8. Dee, Thomas S. "Higher Chronic Absenteeism Threatens Academic Recovery from the COVID-19 Pandemic." Proceedings of the National Academy of Sciences 121, no. 3 (2024): e2312249121. 9. Ingersoll, Richard M., Elizabeth Merrill, Daniel Stuckey, and Gregory Collins. "Seven Trends: The Transformation of the Teaching Force." CPRE Research Report #RR 2018-2, University of Pennsylvania, 2018. 10. Oster, Emily. "COVID-19, School Closures, and Outcomes." Journal of Economic Perspectives 37, no. 4 (2023): 51–70. 11. Fahle, Erin M., Thomas J. Kane, Sean F. Reardon, and Douglas O. Staiger. "The First Year of Pandemic Recovery: A District-Level Analysis." Education Recovery Scorecard, 2023. 12. Lake, Robin. Rethinking Education: How America Can Reset Its K-12 System. Cambridge, MA: Harvard Education Press, 2024.