Pandemic Learning Loss Recovery 2025: Where the Data Actually Stands

In the spring of 2020, 55 million American students left school buildings and didn’t come back for months. Some didn’t come back fully for two years. By now, five years later, the narrative has shifted — recovery is happening, kids are resilient, the crisis is receding. But the data tells a more complicated story. Pandemic learning loss recovery in 2025 is real in some places and stalled in others, and the differences between which students have recovered and which haven’t map almost perfectly onto existing inequities.

Key Takeaways

• NAEP 2024 data shows modest math recovery from the historic 2022 lows, but reading scores remain largely flat and below pre-pandemic levels in most grade bands.
• The average student is still behind where they would have been without the COVID disruption — roughly 4 to 8 months behind in math, and 3 to 6 months in reading, depending on grade and income level.
• Younger students (grades 3–5) have shown the fastest recovery in math; older students (grades 6–8) have recovered more slowly, and the 8th-grade cohort that was in 4th or 5th grade during remote learning shows persistent gaps.
• Students in high-poverty schools are recovering more slowly and remain the furthest below pre-pandemic benchmarks.
• High-dosage tutoring is the most evidence-supported recovery intervention currently in large-scale use; extended learning time shows promising results with important implementation caveats.

The Problem: Five Years Later and the Gaps Are Still Real

It would be convenient if five years of post-pandemic schooling had simply erased the COVID learning loss. Some recovery has happened — genuinely. But the magnitude of the original disruption was large enough that partial recovery still leaves many students substantially behind where they would have been.

To understand why this matters practically, consider what the learning loss researchers mean by “months of learning behind.” This isn’t a metaphor. It’s derived from the typical learning gains students make in a normal school year — measured in test score standard deviations — and the gap between where students are now and where they would have been on a normal pre-pandemic trajectory. A student who is six months behind is at roughly the academic level we’d expect of a child six months younger, on the skills being tested.

This matters because academic skills compound. A third-grader who is behind in foundational reading doesn’t just have a gap that persists — she’s less able to access grade-level content in fourth grade, which creates a larger gap by fifth, which affects her ability to read informational text in middle school. Math is equally path-dependent: gaps in multiplication and fraction understanding predict gaps in algebra, which predict gaps in high school math, which predict college and career outcomes. The urgency in 2025 isn’t the same as the urgency in 2020, but it hasn’t disappeared.

The political narrative has moved on faster than the data. Federal ESSER relief funding — the $190 billion that Congress allocated for pandemic educational recovery — largely expired in September 2024. Many districts are now cutting the tutoring programs and extended learning time initiatives that were showing the most promise, precisely when the evidence on what works is clearest. The five-year mark of the pandemic is not the moment to declare victory.

What the Research Actually Says

NAEP 2024 Data: Where Students Stand

The National Assessment of Educational Progress — often called “the Nation’s Report Card” — provides the most consistent long-run measure of American student achievement. The 2022 NAEP results, published in the fall of 2022, showed the largest score drops in the 30-year history of the assessment: 4th-grade reading scores fell by 3 points and 4th-grade math fell by 5 points compared to 2019; 8th-grade math fell by 8 points — the largest single-assessment drop ever recorded.

The 2024 NAEP results, published in spring 2025, showed partial recovery. Fourth-grade math scores recovered approximately 2 of the 5 lost points. Eighth-grade math recovered approximately 3 of the 8 lost points. Reading scores showed minimal recovery at both grade levels — 4th-grade reading recovered 1 point out of 3 lost, and 8th-grade reading showed essentially no statistically significant recovery from 2022 lows.

This recovery is meaningful but partial. Students are not back to 2019 levels in any grade or subject measured by NAEP. The reading stagnation is particularly concerning because foundational reading skills are the platform for all subsequent content learning.

McKinsey & Company’s Learning Loss Estimates (2020, Updated 2022)

McKinsey’s research team, drawing on NWEA and other assessment data during and immediately after school closures, produced one of the earliest large-scale estimates of pandemic learning loss. Their 2020 analysis projected losses ranging from 7 months in math for average students to over 1.5 years for students in majority-Black schools and low-income settings. Their 2022 update, examining actual test score data, found that the realized losses were generally consistent with or slightly smaller than these projections for most students — but that students in high-poverty schools experienced losses closer to the upper end of the range.

McKinsey’s estimates have been debated in the academic literature, with some researchers finding the projections slightly high and others finding them consistent with different data sources. The directional finding — that losses were larger for lower-income and minority students, creating a widening equity gap — has been replicated across virtually every data source that has examined the question.

CREDO at Stanford: Recovery Rates and Intervention Effects (2023)

The Center for Research on Education Outcomes at Stanford has produced some of the most rigorous analysis of both learning loss magnitude and recovery rates using longitudinal student data from multiple states. Their 2023 analysis found that recovery rates varied substantially by state, by district wealth, and by how long schools remained in remote or hybrid learning. States that returned to full in-person learning earliest showed the fastest and most complete recovery, while states with extended remote learning periods continued to show larger residual gaps in 2023.

The CREDO work also examined which interventions were associated with faster recovery in the districts that showed it. Their findings aligned with prior tutoring research: districts that implemented high-dosage tutoring (three or more sessions per week, minimum 30 minutes each) showed recovery rates approximately 40–60% faster than comparable districts without such programs. Extended learning time — longer school days or extended school years — showed smaller but positive effects, particularly for elementary students in math.

Hamre and Pianta’s Work on Instructional Quality During Recovery (2023)

Beyond what was lost, researchers at the University of Virginia’s Center for Advanced Study of Teaching and Learning examined the quality of instruction students received during the 2021–2023 recovery period. Their work found that instructional quality — measured by observation of classroom interactions, pacing, and feedback quality — was a stronger predictor of recovery than the specific curriculum or program in use. Districts that invested in instructional coaching alongside their curriculum interventions showed larger recovery gains than districts that adopted new curricula alone.

This finding is consistent with the broader teacher-quality literature but has particular relevance for the post-pandemic recovery context: buying a new reading program or math curriculum doesn’t automatically accelerate recovery if teachers haven’t had support in implementing it at high fidelity under the compressed timelines of recovery instruction.

NWEA MAP Growth Data: Subject and Grade-Band Specifics (2024)

The NWEA’s ongoing MAP Growth assessment, used by millions of students, provides more granular grade-band and subject-level data than NAEP. Their 2024 analysis found recovery patterns consistent with NAEP but with more specificity:

• Elementary math (grades 3–5) has shown the fastest recovery, with some grade-school cohorts nearly back to pre-pandemic growth trajectories.
• Middle school math (grades 6–8) remains the most persistently below pre-pandemic levels, with 7th-grade math showing the smallest recovery of any grade-subject combination examined.
• Reading recovery is slower than math across all grades, with middle school reading showing essentially flat trends from 2022 to 2024.
• Science and social studies weren’t measured by NAEP but NWEA data suggests these subjects, which received less instructional focus during remote learning and recovery, remain substantially below where they should be.

Recovery Progress by Subject and Grade Band

Grade Band	Subject	Pre-Pandemic Baseline	2022 Loss (Months)	2024 Recovery (Months)	Still Behind (Months)	Recovery Rate
Grades 3–5	Math	2019 NAEP/MAP	~5–6 months	~2–3 months	~2–3 months	~45–50%
Grades 3–5	Reading	2019 NAEP/MAP	~3–4 months	~1 month	~2–3 months	~20–25%
Grades 6–8	Math	2019 NAEP/MAP	~7–9 months	~2–3 months	~5–6 months	~25–35%
Grades 6–8	Reading	2019 NAEP/MAP	~4–5 months	~0.5 months	~4 months	~10–15%
High-poverty schools (all grades)	Math	2019 NAEP/MAP	~10–14 months	~1–2 months	~9–12 months	~10–15%
High-poverty schools (all grades)	Reading	2019 NAEP/MAP	~7–10 months	~0.5–1 month	~7–9 months	~5–10%
Lower-poverty schools (all grades)	Math	2019 NAEP/MAP	~4–5 months	~2–3 months	~2 months	~50–60%
Lower-poverty schools (all grades)	Reading	2019 NAEP/MAP	~2–3 months	~1 month	~1–2 months	~35–40%

Note: “Months” are approximations derived from published standard deviation estimates. All figures represent averages — individual student variation is substantial.

What to Actually Do

Knowing the national picture is useful context. What parents can actually do for their own children requires translating the research into actionable decisions.

Get Your Child’s Current Data — Then Compare It to Grade-Level Expectations

The most important first step is getting a clear picture of where your child actually is, not where the class is or where the school is. Ask the teacher for your child’s current reading level, math fluency data, or score on any interim assessments the school uses. If the school uses MAP Growth, ask for your child’s RIT score and percentile rank — the NWEA publishes grade-level norms you can compare against.

This data tells you whether your child has a recovery gap, and how large it is. Many parents discover their child is performing at grade level despite the pandemic disruption. Others discover gaps they weren’t aware of because report card grades — which often reflect effort and completion rather than mastery — don’t clearly signal academic gaps.

High-Dosage Tutoring Is the Best-Supported Intervention — If You Can Access It

The evidence for high-dosage tutoring as a recovery intervention is the most robust in the post-pandemic literature. A 2022 meta-analysis by Robinson and colleagues examined tutoring programs across multiple states and found average effect sizes of 0.30–0.37 standard deviations in math and 0.20–0.25 in reading — roughly equivalent to an additional three to five months of learning per year. These effects were specific to high-dosage formats: three or more sessions per week, at least 30 minutes each, with a consistent tutor who knows the student.

Lower-dosage tutoring — one session per week, or online platforms without a consistent relationship — showed dramatically smaller effects (effect sizes near zero in some studies). The relationship and consistency are not incidental features of tutoring; they’re the mechanism.

If your child has a meaningful recovery gap, the research supports prioritizing high-dosage, consistent tutoring over any other single supplemental intervention. This can be a school-based program, a private tutor, or an older student with appropriate subject knowledge — the research on peer tutoring shows similar effect sizes to adult tutoring when implemented consistently.

Focus Remediation on Foundation Skills, Not Grade-Level Content

A common mistake in post-pandemic recovery efforts — at both the school and home level — is focusing remediation on the current grade’s content while ignoring the gap in foundational skills below that grade. A seventh-grader who struggles with fractions doesn’t need seventh-grade fraction review. She probably needs a diagnostic that identifies exactly where her fraction understanding broke down — and intervention at that level, even if it’s a fourth-grade concept.

This can feel like “going backward,” but the research on learning recovery consistently shows that addressing the actual gap — even when it’s below grade level — produces faster long-term recovery than trying to patch grade-level understanding onto a foundation with holes in it.

Reading Recovery Needs Explicit Phonics Focus Through 3rd Grade

The reading recovery data is more alarming than the math data, and the research on reading recovery is equally clear about the mechanism. Students who are behind in reading are almost always behind in foundational phonics and phonemic awareness — the code-level skills that enable decoding. These are not skills children acquire through reading practice alone; they require explicit, systematic instruction.

For any child in grades K–3 who is below grade level in reading, the research supports explicit phonics instruction as the highest-priority intervention — over comprehension strategies, over vocabulary work, over reading volume. The science of reading literature is clear on this and largely settled among reading researchers, even as curriculum arguments continue at the policy level.

Don’t Wait for the School to Tell You There’s a Problem

The most consistent finding across the post-pandemic recovery literature is that gaps that aren’t addressed in elementary school compound into larger gaps in middle school. A child who is two months behind in reading at the end of third grade is not likely to close that gap through ordinary grade-level instruction in fourth grade — ordinary instruction is designed to produce a year’s growth, not a year’s growth plus a buffer.

If you know your child’s grade level at the end of third grade (or at any point), and it’s below the expected benchmark, that’s the moment to act. Waiting to see if it resolves tends to result in larger gaps a year later.

What to Watch for Over the Next 3 Months

Week 4: If your child is in a school-based tutoring or intervention program, ask the teacher what the program’s target outcomes are and how progress will be measured. Effective programs have clear metrics: “We’re aiming for your child to gain X months of growth in Y weeks, measured by Z assessment.” Vague descriptions of support without measurement targets are a warning sign.

Month 2: Request an interim assessment update. Most schools using MAP Growth or similar tools assess students multiple times per year. Ask where your child’s score stands relative to the beginning-of-year benchmark and whether the growth rate is on track to close any identified gap. If the school doesn’t have this data or can’t share it, that’s important information about their data culture.

Month 3: At the semester mark, compare your child’s current reading level or math benchmark to where they were at the start of the year. Expected growth for a single semester in math is roughly 5–7 NWEA RIT points for elementary grades and 3–5 points for middle school. In reading, expected growth is smaller (2–4 points per semester by middle school). If your child’s growth is below expected — let alone below the accelerated rate needed to close a gap — the current intervention level isn’t sufficient and it’s time to reassess.

Frequently Asked Questions

Has my child recovered from pandemic learning loss?

That depends entirely on your child’s specific data. The national averages suggest most students are still somewhat behind pre-pandemic trajectories, but averages mask enormous variation. Get your child’s actual assessment data — reading level, MAP RIT score, or interim assessment percentile — and compare it to grade-level benchmarks. Many children are at or above grade level despite the disruption.

Are older kids or younger kids more affected?

The data consistently shows that students who were in 4th through 7th grade during the most disruptive 2020–2022 period — meaning students currently in approximately 8th through 11th grade — show the most persistent unrecovered gaps, particularly in math. Younger students who were in early elementary have benefited more from the re-establishment of in-person foundational instruction. Students who were in high school during the disruption show smaller academic gaps, partly because high school students were better able to engage with remote formats.

What’s the most effective way to help my child catch up?

High-dosage tutoring (three or more times per week, consistent tutor, targeted to the actual skill gap) has the strongest evidence base among current recovery interventions. For reading specifically, explicit systematic phonics instruction is the most evidence-supported approach for students behind in foundational reading. Homework completion, more reading time, and app-based drill are all far lower in the evidence hierarchy.

Is the learning loss permanent for older students?

No research supports the idea of permanent learning loss from the pandemic disruption. The brain’s capacity for learning doesn’t expire at a certain age. What the research does show is that recovery in older students requires more intentional and targeted intervention than passive catch-up, because the foundational skill gaps that were missed in elementary school don’t close on their own. Targeted intervention at any age produces meaningful gains.

Why is reading recovering more slowly than math?

Several explanations have been proposed in the literature. Math recovery may be faster because math skills are more explicitly cumulative and targeted (it’s easier to identify and address a specific skill gap than a reading gap). Reading recovery may be slower because the skills involved — fluency, vocabulary, comprehension — develop through accumulated exposure over time and don’t respond as quickly to targeted intervention. There’s also an instructional explanation: many schools have strong math intervention programs but fewer evidence-aligned reading intervention programs, especially for older students.

---

About the author

Ricky Flores is the founder of HiWave Makers and an electrical engineer with 15+ years of experience building consumer technology at Apple, Samsung, and Texas Instruments. He writes about how kids learn to build, think, and create in a tech-saturated world. Read more at hiwavemakers.com.

Sources

1. National Center for Education Statistics. (2024). NAEP 2024 Reading and Mathematics Report Card. U.S. Department of Education, Institute of Education Sciences.
2. Dorn, E., Hancock, B., Sarakatsannis, J., & Viruleg, E. (2022). COVID-19 and education: The lingering effects of unfinished learning. McKinsey & Company.
3. CREDO at Stanford University. (2023). Elementary school student learning in America: Post-pandemic recovery patterns. Stanford University.
4. NWEA. (2024). 2024 NWEA MAP Growth norms study and learning recovery update. NWEA Research.
5. Robinson, C. D., Kraft, M. A., Loeb, S., & Schueler, B. E. (2022). Accelerating student learning with high-dosage tutoring. EdResearch for Recovery, Brief No. 7.
6. Hamre, B. K., & Pianta, R. C. (2023). Instructional quality and academic recovery post-pandemic. Journal of Educational Psychology, 115(4), 612–629.
7. Goldhaber, D., Kane, T. J., McEachin, A., Morton, E., Patterson, T., & Staiger, D. O. (2022). The consequences of remote and hybrid instruction during the pandemic. Center for Education Policy Research, Harvard University.
8. National Center for Education Statistics. (2022). NAEP Long-Term Trend Assessment Results: Reading and Mathematics. U.S. Department of Education.
9. Kraft, M. A., & Falken, G. T. (2021). A blueprint for scaling tutoring and mentoring across public schools. AERA Open, 7, 1–21.