Tom Kane has tracked years of U.S. test scores. Here’s what he’s learned — and still can’t explain.

“Chalkbeat Ideas is a section featuring reported columns on the big ideas and debates shaping American schools. Sign up for the Ideas newsletter to follow our work. Just about everyone in education is trying to figure out why student achievement has been declining and how to raise it back up. If anyone would know the answer it might be Tom Kane, a Harvard University professor who has been poring over data on student learning for the last several years. Kane and several colleagues recently released an extensive database of test scores from most states in the country. The takeaway: Schools have made some progress recovering from steep learning declines in math, but scores in both math and reading are still far below where they were over a decade ago. “The U.S. has been in a learning recession since at least around 2013 and that’s especially true in reading,” says Kane. “The challenges we face didn’t start or end with the pandemic.” The implications are alarming. In other work, Kane has found that test scores are linked to a variety of economic outcomes. The research out of Kane’s center has been so comprehensive that many of its interesting findings have flown under the radar. I wanted to highlight a few of them. And I also wanted to unpack theories to explain poor academic results. What’s striking is that even Kane, a leading expert, doesn’t have clear answers or solutions. Here are some takeaways from the work of Kane and his colleagues. There are hypotheses, but no decisive evidence, for why learning stalled before the pandemic Kane offers two theories. “The slowdown in learning coincided with a dismantling of test-based accountability and a rise in social media use,” his study says. The evidence for both is circumstantial, Kane acknowledges. Other countries have also seen test scores deteriorate, which could be consistent with a rise in social media everywhere. Federal accountability helped drive learning gains in the past, according to some research , so easing the pressure may explain recent declines. Yet both theories have holes. Test scores have dropped even in early grades where social media use is not prevalent (although general screen use has accelerated for young kids). And accountability policies seemed most effective in raising math scores, but the most persistent recent declines have been in reading. “We don’t have a definitive answer,” Kane says. The highest- and lowest-poverty school systems have recovered faster than those in the middle Kane calls this a U-shaped recovery. You can see it in the graph below, from his recent study. The dark blue and orange lines show learning trends by district poverty rates between 2022 and 2025. In math, the highest- and lowest-need school systems have seen the most progress. In reading, scores in middle-income districts fell the most. Kane says policymakers should keep in mind those middle-income districts, which are sometimes overlooked in policy and funding formulas. Federal COVID relief money helped, but it might have gone further Part of the explanation for this U-shaped recovery is the $190 billion or so in federal COVID relief funding. In the same graph, the line below the actual recovery shows Kane’s estimate of what scores would have been without the money. Scores in high-poverty districts would have been significantly worse, he finds. Those schools benefited more from the money simply because they got more of it . Kane rejects the commonly heard argument that the federal relief money was largely wasted. But he thinks the test score effect might have been larger had districts been required to spend more on academic recovery. One caveat: Kane acknowledges he may be underestimating the effect of each dollar of federal aid because he doesn’t have data on how much extra money schools actually spent, just how much they were supposed to receive. Absenteeism is limiting recovery, at least a bit Attendance issues have exploded in recent years . Schools have made some progress in getting students back to class, but absence rates remain elevated . Kane and team estimate that this is dragging down learning. The other gray lines in the graphic — above the blue and orange lines — are Kane’s predictions of what scores would be if students were attending school at the same rate as before the pandemic. The effects are real but fairly modest, indicating absence rates are not the main drivers of sluggish recovery. In reading, for instance, most districts would still have lost ground even with pre-pandemic attendance. State test scores paint a more optimistic picture of recovery than the main federal test Kane and colleagues use a combination of state tests and federal NAEP exams to construct their learning estimates. To produce the main trends they rely on NAEP, except for score changes from 2024 to 2025 since NAEP scores aren’t yet available. This matters because state tests tend to produce more positive trends . For instance, NAEP has shown reading scores declining in recent years, even while many states’ results have inched up. It’s not that one set of scores is right and the other is wrong. But this discrepancy is worth keeping in mind . It’s not clear whether science-of-reading laws are working Kane doesn’t find a strong correlation between states’ reading gains since 2022 and the number of science-of-reading elements they’ve incorporated. Yet he notes that the states with very few of these tenets have not made progress. Kane emphasizes that this is not especially rigorous evidence, and should be taken as only “suggestive.” More research is needed on whether these laws are changing school practices and improving outcomes. Progress has happened before and is happening in some places Kane marvels at the substantial gains in test scores in the 1990s and early 2000s — which went largely unremarked upon at the time. If anything, the prevailing narrative was that nothing in education was working. But if we made progress before, we can do it again, he suggests. That’s also why his report highlights a number of school systems, like Detroit , that are relative bright spots. “It’s a reminder that improvement is possible,” Kane says. Matt Barnum is Chalkbeat’s ideas editor. Reach him at mbarnum@chalkbeat.org .

Summary generated from the RSS feed of Chalkbeat Chicago. All article rights belong to the original publisher. Click through to read the full piece on www.chalkbeat.org/chicago.