Skip to main content

New Data Show How the Pandemic Affected Learning Across Whole Communities

Most comprehensive picture yet of COVID's unequal impacts — and the urgency to expand learning opportunities

Today, The Education Recovery Scorecard, a collaboration with researchers at the Center for Education Policy Research at Harvard University (CEPR) and Stanford University’s Educational Opportunity Project, released 12 new state reports and a research brief to provide the most comprehensive picture yet of how the pandemic affected student learning. Building on their previous work, their findings reveal how school closures and local conditions exacerbated inequality between communities — and the urgent need for school leaders to expand recovery efforts now.

The research team reviewed data from 8,000 communities in 40 states and Washington, D.C., including 2022 NAEP scores and Spring 2022 assessments, COVID death rates, voting rates, and trust in government, patterns of social activity, and survey data from Facebook/Meta on family activities and mental health during the pandemic.

>> Read an op-ed by researchers Tom Kane and Sean Reardon in the New York Times.

They found that where children lived during the pandemic mattered more to their academic progress than their family background, income, or internet speed.  Moreover, after studying instances where test scores rose or fell in the decade before the pandemic, the researchers found that the impacts lingered for years.  

“Children have resumed learning, but largely at the same pace as before the pandemic. There’s no hurrying up teaching fractions or the Pythagorean theorem,” said CEPR faculty director Thomas Kane. “The hardest hit communities — like Richmond, Virginia, St. Louis, Missouri, and New Haven, Connecticut, where students fell behind by more than 1.5 years in math — have to teach 150 percent of a typical year’s worth of material for three years in a row — just to catch up. That is simply not going to happen without a major increase in instructional time.  Any district that lost more than a year of learning should be required to revisit their recovery plans and add instructional time — summer school, extended school year, tutoring, etc. — so that students are made whole. ”

“It’s not readily visible to parents when their children have fallen behind earlier cohorts, but the data from 7,800 school districts show clearly that this is the case,” said Sean Reardon, professor of poverty and inequality, Stanford Graduate School of Education. “The educational impacts of the pandemic were not only historically large, but were disproportionately visited on communities with many low-income and minority students. Our research shows that schools were far from the only cause of decreased learning — the pandemic affected children through many ways — but they are the institution best suited to remedy the unequal impacts of the pandemic.”

The new research includes:

  • A research brief that offers insights into why students in some communities fared worse than others. 
  • An update to the Education Recovery Scorecard, including data from 12 additional states whose 2022 scores were not available in October. The project now includes a district-level view of the pandemic’s effects in 40 states (plus D.C.). 
  • A new interactive map that highlights examples of inequity between neighboring school districts. 

Among the key findings:

  • Within the typical school district, the declines in test scores were similar for all groups of students, rich and poor, white, Black, Hispanic. And the extent to which schools were closed appears to have had the same effect on all students in a community, regardless of income or race. 
  • Test scores declined more in places where the COVID death rate was higher, in communities where adults reported feeling more depression and anxiety during the pandemic, and where daily routines of families were most significantly restricted. This is true even in places where schools closed only very briefly at the start of the pandemic. 
  • Test score declines were smaller in communities with high voting rates and high Census response rates — indicators of what sociologists call “institutional trust.” Moreover, remote learning was less harmful in such places. Living in a community where more people trusted the government appears to have been an asset to children during the pandemic.
  • The average U.S. public school student in grades 3-8 lost the equivalent of a half year of learning in math and a quarter of a year in reading.

The researchers also looked at data from the decade prior to the pandemic to see how students bounced back after significant learning loss due to disruption in their schooling. The evidence shows that schools do not naturally bounce back: Affected students recovered 20–30% of the lost ground in the first year, but then made no further recovery in the subsequent three to four years.   

“Schools were not the sole cause of achievement losses,” Kane said. “Nor will they be the sole solution. As enticing as it might be to get back to normal, doing so will just leave the devastating increase in inequality caused by the pandemic in place. We must create learning opportunities for students outside of the normal school calendar, by adding academic content to summer camps and after-school programs and adding an optional 13th year of schooling.”

The Education Recovery Scorecard is supported by funds from Citadel founder and CEO Kenneth C. Griffin, Carnegie Corporation of New York, and the Walton Family Foundation.


The latest research, perspectives, and highlights from the Harvard Graduate School of Education

Related Articles