U.S. Learning Rebound Underway According to Education Scorecard

Math achievement continues to rebound. Reading gains are starting to happen. And “districts on the rise” are proving that academic improvement post pandemic is possible everywhere. These are just a few of the positives highlighted in the latest Education Scorecard, released today by the Center for Education Policy Research (CEPR) and researchers at Stanford and Dartmouth. “The recovery of U.S. education has begun,” says Professor Tom Kane, one of the study’s authors and founding director of CEPR. 

Now in its fourth year, the report — which dropped the word “recovery” from its title — annually tracks district-level learning loss and the pace of post-pandemic recovery. The Scorecard uses data from the Stanford Education Data Archive, which links state test results for roughly 35 million grade 3–8 students to a common national scale to track district-level changes in achievement across the country. This year’s report, From Learning Recession to Learning Recovery: Understanding the Sources of U.S. K-12 Improvement, draws on data from the 2024–2025 school year.

Despite the positives, there are still challenges impacting learning recovery, the report notes. High student absenteeism continues to slow progress, with 23% of students chronically absent in 2024–2025, down from the post-pandemic peak, but still higher than the 15% before the pandemic. And middle-income school districts — those with 30–70% of students receiving federal lunch subsidies — received little federal relief aid during the pandemic and remain far behind 2019 levels of achievement.

Recently, Kane sat down to talk about the report’s biggest discoveries, U-shaped recovery, and why there’s a hunger for this kind of education information.

One of the big Scorecard changes this year is that you dropped the word "recovery." Why?
When we started this in 2022, it was at a moment where we knew there were achievement losses. We didn't know how big they were. We were just hoping to support a national conversation about what to do to speed the recovery. But over the first three years, we recognized that there had been a great need for a timely reporting of district-level results. We dropped “recovery” because we hope to continue it into the future. It's not just about pandemic recovery; it's more the education scorecard, an annual reporting on district-level, local results.

Other than dropping that word, what's the biggest change since last year’s report?
The biggest change is that this year we’re identifying 108 communities around the country — what we’re calling “districts on the rise” — that have seen unusual improvements in math and reading. We're doing that to shine a spotlight on successful local leaders and to remind everybody that improvement is possible. We shouldn't assume that we have to live with continuing losses. And a lot of these places that have seen improvement are not on the usual list of places flagged as having education leadership. For instance, Baltimore; Detroit; Compton; and Birmingham are not places we typically celebrate and yet they have all improved more than 0.3 grade equivalence in math and reading relative to similar districts in their state.

The report mentions that since 2022, recovery has been U shaped, with middle-income districts at the bottom of the U. Did you expect to discover this?
That was surprising. What we found was the highest-income and the lowest-income districts nationally have seen the most recovery since 2022. The main factor driving that was the federal pandemic relief money, which was highly targeted at the highest poverty districts. For instance, here in Massachusetts, Springfield Public Schools received about $10,000 per student in federal pandemic relief, but Framingham Public Schools, which is not a very high-income district, received about $2,500 per student.

All districts didn’t receive the same amount?
Here’s something that isn’t well known — just how targeted the federal dollars were. Back in the spring of 2021, while we were still in the middle of school shutdowns, Congress was trying to get a lot of money out the door quickly to school districts and said, “Okay, how are we going to decide how to target these dollars?” They settled on using the existing federal Title I formula, but remember, schools were still closed and that was before anybody knew what the magnitude of losses for students would be. The dollars were not targeted based on academic achievement losses. They were targeted based on this formula that was really meant for the highest-poverty districts in the country, not for middle-income or higher-income communities. I don't think most people realize just how targeted the Title I program is and just how targeted the federal pandemic relief was. That means that communities like Framingham often lost considerable ground academically but did not receive much federal pandemic relief. In this year’s study, we’re suggesting that states use some of their school improvement dollars to help those middle-poverty districts recover.

Math rebounded almost immediately after the pandemic ended compared with reading. Is this surprising?
I'm not surprised that the rebound was larger in math. We've long known that math achievement is more responsive to in-school interventions and students learn math in school, but reading it's a mixture of in-school and out-of-school interventions. What was surprising was that reading continued to decline since 2022. Based on past research, we would have expected a smaller recovery in reading but not a continued decline.

What does this tell you? 
The per-year loss in reading in the years leading up to the pandemic and in the two years after the pandemic were the same as during the pandemic, which implies it was something else going on besides the pandemic. The leading hypothesis is social media and whether it's distracting students in school or just displacing reading time outside of school. That's why I think one of our top priorities ought to be investigating ways to improve reading achievement even in the presence of social media.

The study noted that reading, though, is starting to recover, at least in some states that implemented science of reading curriculum. But why only some?
None of the states which had avoided science of reading policies as of January 2024 improved in reading between 2022 and 2025. That includes Massachusetts, California, New Jersey, and Illinois, but all states that did improve were implementing many of the elements of science of reading. I don't want to overstate it though, because there were a lot of states that were doing lots of science of reading interventions that did not improve like Florida, Nebraska, and Arizona. This tells me that there are a bunch of different elements to science of reading and we need to learn which elements are the most effective and the most important. For instance, how important is having a literacy coach in every elementary school? What happens to student achievement gains in the years before and after the literacy coaches arrive? Is that having a positive effect?

Student absenteeism also continues to be a sticky problem for academic gains post pandemic. Why?
The percentage of students chronically absent essentially doubled during the pandemic. There's been a slow drift down, but chronic absenteeism rates have leveled off at a level much higher than they were in 2019, before the pandemic. We actually don't have good evidence on what’s causing this, but family routines have not snapped back to pre-pandemic levels, and it doesn't look like they're snapping back anytime soon. What's frustrating is that school districts are paying for teachers’ salaries and paying for buses whether or not children are present, and when a student is absent, it's not just affecting the absent student; it's affecting their peers because when absent students return to the classroom, teachers have to reteach material and it slows down learning for everybody. We can't allow a higher level of student absenteeism to be acceptable.

Since the Scorecard launched in 2022, the work has been mentioned more than 5,000 times in the media. What does this say about our thirst for this kind of information?
It tells me that there is hunger for timely, comparable data on local achievement trends, not at the state level, not at the national level, but local district achievement trends where the rubber really hits the road.  And that's what the Education Scorecard is providing. It's providing comparable measures of achievement changes at the local level going back to 2009.

Why did you say in the report that "recovery of U.S. education has begun"? 
There definitely are signs of leadership in states making big investments and literacy improvement in local districts that are achieving impressive gains. We have to remind people that prior to 2013 the United States was making dramatic progress in math and modest progress in reading. There was more than a two-grade equivalent improvement in math achievement between 1990 and 2013. So in some ways, the greatest misconception that is holding us back is the sense that K–12 is a hopeless quagmire. Part of what we're trying to do with this Scorecard is show that that's not true. There are states and districts that are making substantial progress now and the country was making substantial progress prior to 2013. Yes, we have had struggles since 2013, but that was the exception, not the rule.