American students just posted their worst math and reading scores in a quarter-century, according to the latest OECD PISA results. The rational response would be to intensify diagnostic efforts. The Trump administration did the opposite — it dismantled the primary agency capable of explaining what's going wrong. The National Center for Education Statistics, a federal statistical agency responsible for The Nation's Report Card and dozens of longitudinal studies tracking student outcomes, was reduced from roughly 100 employees to four during the initial DOGE cuts. It has since recovered to 12 staffers. Nearly $900m in Department of Education contracts were canceled, and most of NCES's ongoing data collections appear inactive or reduced in scope, according to an American Statistical Association tracker. The most consequential casualty is the Early Childhood Longitudinal Study, a $47m survey following 15,000 first-graders — the first cohort born after Covid. Researchers had designed new questions to capture pandemic effects on childcare, hybrid instruction, and developmental trajectories. The contract was canceled mid-collection, meaning the data on how Covid reshaped early learning will never exist. Since 1998, the ECLS has been cited in nearly 3,000 academic papers informing decisions on Head Start, preschool programs, and full-day kindergarten. Other longitudinal studies were killed in their final years. The Beginning Postsecondary Students study, tracking 30,000 college entrants from 2019 through 2025, was terminated near completion. The High School Longitudinal Study of 2009, following 23,000 ninth graders into their early 30s, suffered the same fate. In each case, the bulk of investment had already been spent on sample design, recruitment, and questionnaire development. The government also owed several million-dollar termination fees to contractors — meaning the cuts didn't even save much money. Litigation has forced some reinstatements. The National Postsecondary Student Aid Survey and the Baccalaureate and Beyond study were restored after lawsuits from research organizations including the Institute for Higher Education Policy. But the longitudinal studies that require continuous tracking of specific cohorts cannot be restarted — those children have aged out of the study window, and the developmental data is permanently lost. The administration frames this as reform. Education department spokesperson Savannah Newhouse said NCES and IES "were in need of reform — and that work is well underway." But what was terminated wasn't bureaucratic overhead. It was the measurement infrastructure that allows policymakers to distinguish which interventions work from which don't. Former NCES commissioner Peggy Carr, fired after nearly four decades, and former statistician Carolyn Grim both describe the cuts as destroying the feedback loop between policy and outcomes. The structural problem is stark: the US is now making education policy without the instruments to evaluate it. Billions flow annually into early childhood programs, school funding formulas, and college financial aid based on data these studies generated. Without successors, those programs will continue on autopilot — funded by habit rather than evidence — while test scores continue their decline with no diagnostic capacity to understand why.
