The fourth annual Education Scorecard — a joint project of Stanford University’s Educational Opportunity Project and Harvard’s Center for Education Policy Research — released new data today confirming what many educators already suspected: the pandemic learning decline was not the beginning of a crisis. It was the acceleration of one that was already underway. Researchers describe it as a “learning recession” that predates COVID-19 by years, making the recovery more complicated than simply returning to a pre-pandemic baseline that was itself insufficient.
The new data carries two distinct findings that shouldn’t be collapsed into one narrative. Most states are finally showing measurable gains in math — a signal that targeted interventions and instructional shifts are working in at least one discipline. Reading is a different story. Most states have not made comparable progress, and the gap between math recovery and reading recovery is widening in ways that will compound over time. A child who can’t read proficiently by third grade faces a statistical headwind that follows them through every subsequent grade level, every standardized test, and eventually into the labor market.
The states that have made reading gains share one variable that stands out in the data: they changed how reading is taught at the legislative level, moving toward structured literacy approaches grounded in phonics and decoding rather than the whole-language and balanced literacy models that dominated classrooms for decades. The science of reading has been settled for years. The policy has been slower to follow. The Scorecard suggests that where legislatures finally moved, outcomes moved with them. The schools recovering fastest aren’t the ones that waited for the crisis to pass. They’re the ones that changed something specific — and the data is now clear enough to say what that something was.