Search and Filter

Browse institutions A–Z

When Students Leave: Predicting Public School Closures

Abstract: Persistent enrollment losses have renewed questions about district finances and school facilities. I link national enrollment, demographic, and district finance data to examine where enrollment decline is associated with elevated closure risk and how well that risk can be identified in advance. The share of school-age children in traditional public schools fell from 83.0 percent in 2019\u201320 to 77.8 percent in 2024\u201325. The share in private schools or unaccounted for in enrollment data rose from 9.4 to 13.3 percent. Most districts, including most with steep enrollment losses, maintained real spending per pupil. Nonetheless, 730 districts serving 2.3 million students combined enrollment losses of at least 10 percent with federal revenue losses of at least 2 percent of revenue. In six annual out-of-sample tests, the 10 percent of schools with the highest predicted closure risk accounted for 53\u201363 percent of subsequent closures. Applied to traditional public schools open in 2024\u201325, the one-year model predicts about 550 closures in 2025\u201326. The three-year model predicts about 1,700 closures by 2027\u201328. Risk is concentrated among small schools with declining enrollment and in districts with recent closures. These include districts that maintained spending per pupil. The findings support targeted facility planning, transition support, and monitoring.

Research Categories

Keywords
Public School Enrollment Decline, School Closures, Demographic Change, School Finance, Educational Resource Allocation, Education Policy
Education level

Citation

Document Object Identifier (DOI)
10.26300/aq91-g957
EdWorkingPaper suggested citation:
Goulas, Sofoklis. (). When Students Leave: Predicting Public School Closures. (EdWorkingPaper: -1648). Retrieved from Annenberg Institute at Brown University: https://doi.org/10.26300/aq91-g957

Machine-readable bibliographic record: RIS, BibTeX