The headline number
Of the 74,397 US public schools open with at least 200 students in 2017-18 that were still operating in 2024-25, 785 lost more than half their enrollment in those seven years. Another 9,617 lost more than a quarter of their students, and roughly 31,564 (42% of the universe tracked) shrank by at least 10%.
These are not closures. These are schools that remain open, with their buildings, staff, and budgets, but with classrooms much emptier than they were five years ago.
What's driving it
Three forces overlap. The first is structural. US birth rates have declined every year since 2007, and the babies of the late-2000s have aged into K-12 with smaller cohorts behind them. The second is post-COVID disenrollment, which permanently shifted families to homeschool, private, and charter alternatives in 2020-2021 and never fully reverted. The third is school choice expansion: charter and voucher programs that pull students out of neighborhood schools without closing them.
Schools that lose half their enrollment in seven years rarely recover. Per-pupil funding formulas mean each lost student takes operating budget with them. Once the school crosses below the viability threshold for its district, consolidation or closure follows.
Where the steepest declines cluster
Steep declines concentrate in two geographies. Legacy urban districts like Detroit, Cleveland, St. Louis, and Baltimore, where neighborhood schools compete against charters and where the underlying population has been shrinking for decades. And remote rural districts where the loss of a single major employer ripples through the local school enrollment within two years.
What it means for parents and policymakers
A school that has lost a third or more of its enrollment looks different than one that's stable. Class sizes shrink, electives get cut, sports teams fold, and the building feels half-empty. For families considering a school, the trajectory matters as much as the snapshot.
Methodology
Source: NCES Common Core of Data, public school universe files for 2017-18 and 2024-25. Sample restricted to schools that appear in both years with at least 200 students in 2017-18, to exclude noise from very small schools where percentage changes are dominated by a handful of students. Closed schools (those absent from the 2024-25 file) are not included in this study. See our separate analysis of school closures.