Poster #26 - Uneven Fiscal Impacts of Charter Schools: National Evidence

• In Event: Saturday Poster Luncheon

Abstract

Since the first charter school opened in Minnesota over 30 years ago, the sector has expanded to nearly eight thousand schools serving 3.7 million students (NCES, 2024). Advocates argue that charter schools improve education by expanding choice, promoting innovation, and encouraging competition. However, concerns about the fiscal impact of charter schools on traditional public schools and districts persist, and the evidence is inconclusive. Some studies document negative fiscal impacts and others find neutral or positive impacts. Much of this prior work focuses on single states or districts, which may explain why findings across studies are inconsistent. Namely, differences in impacts may reflect differences in state policy, particularly around charter school funding.

In this paper, we begin to explore this possibility. Using a national sample, we examine how charter school expansion is related to changes in district per-pupil expenditures (PPE) and assess how this relationship varies across states. This provides preliminary evidence on how policy context may shape the fiscal impacts of charter schools.

We use data from the National Longitudinal School Database (NLSD), which combines several publicly available datasets, including the Common Core of Data, Civil Rights Data Collection, Private School Survey. These datasets include student characteristics and district-level fiscal data. The advantage of using the NLSD is that charter schools are linked with their geographic host districts, and steps are taken to confirm schools’ charter status, which reduces measurement error in key variables. Our analysis spans from 2008-09 to 2019-20, the last year available in the NLSD and focuses on 20 states with consistent methods of reporting of charter expenditures.To examine the relationship between charter school expansion and district expenditures, we estimate a series of models linking measures of PPE (total, current, variable) to district charter share, controlling for district characteristics that are likely correlated with both charter school share and education costs. These include the log of total enrollment (TPS+charter) as a measure of scale, racial composition, the share of students who are English Language learners, the share of students with disabilities, the share of students in poverty, the share of students enrolled in grades 9-12, and labor-market specific time trends. All models are estimated using district fixed effects, which are identified by within-district variation in charter shares and capture the PPE change associated with changes in the charter share over time. Costs are adjusted using the Extended Comparable Wage Index. 

We find significant variation in the relationship between charter share and PPE. There is a statistically significant negative relationship between charter share and PPE in six states, a statistically insignificant relationship in 12 states, and a statistically significant positive relationship in two states (New York and Ohio). In future work, we plan to examine how these relationships vary by district characteristics, including urbanicity, racial composition, and poverty. Aside from demonstrating the potential importance of state policy variations and impacts, this analysis will lay the foundation for future work examining how specific aspects of state charter school funding policy may shape variation in fiscal impacts.

Authors

• Sarah A. Cordes, Temple University
  Presenting Author

• Danielle Duffy Duffy-Chipman, University of Kentucky
  Non-Presenting Co-Author