Throughout the first quarter of the 21st century, schools built up capacity to accommodate growing student populations, adding classroom space, teacher positions, and the many resources that go into educating a child. The Covid-19 pandemic accelerated the build-up through the injection of giant sums of federal money to supplement state and local budgets. In 2025, more students graduated from high school than in any other year in American history. Today, however, the population of young people is sharply declining, and the trend shows no signs of slowing.
Districts around the country are seeing what happens to schools funded by per-pupil formulas when there are simply fewer pupils to go around. It’s a concern for teachers unions, with leaders desperately trying to strong-arm blue-state governors to keep existing budgets intact. It’s a concern for rural communities, who have seen this play out before. And it’s a concern for families who feel their children are pawns in an opaque and disruptive process that has the largest impact on communities already struggling.
It’s very easy to find emotional reasons to argue against closing underenrolled schools; it’s trickier to argue against the cold, hard fact that keeping empty schools running costs a lot of money while leaving students with less access to high-quality programming, safe facilities, and the non-academic resources that can help put struggling students on a better path.
This background makes a new analysis from Stanford’s Francis Pearman compelling. Using a synthetic difference-in-differences model, Pearman examined 796 California districts between 2011 and 2019 to see whether closure decisions led to balanced budgets. The sample design is important: Pearman looked at school district finances after the Great Recession but before the pandemic. He excluded districts with school closures underway at the beginning of the study period because the analysis requires a pre-treatment baseline, plus dropped single-school districts and any districts with missing data. These choices allow for a sophisticated statistical analysis but potentially exclude some districts facing circumstances similar to those making closure decisions today.
Pearman found that the closures in California reduced the number of students within a district by an average of 287 students, and reports that closures “left districts’ funding deficits and probability of achieving a balanced budget unchanged.” The null result held for districts in financial distress and for financially stable systems.
Why? Spending went down by about $447 per pupil, with a standard error of $388—nearly as large as the estimate—but revenue fell at effectively the same rate. California’s Local Control Funding Formula (LCFF) creates a large degree of state allocation variance based on local student counts, as state funding (about 55 percent of total funding at the start of the study period) drops when students leave the district with no hold-harmless cushion. Plus, closing schools had no effect on the total number of teachers and staff despite shuttered buildings and a drop in enrolled students. Fixed costs like required salary and benefits payments did not change following the closures.
Rather than truly right-sizing schools, it appears the districts just moved around adults.
With over three-fourths of all public school expenditures going to personnel costs and the average public school teacher salary in California during the 2024–25 school year reaching six figures, district leaders have limited tools to adjust their budgets. California has the second-strongest teachers union in the country according to Fordham’s latest analysis, working hard to protect its members’ jobs and raise salaries even when those costs extend beyond what districts can afford. Beyond union pressure, the regulatory environment in California limits the ability to repurpose closed buildings to optimize revenue—a point Pearman considers in discussing how the state’s surplus land policies constrain local districts.