Does CEP Reduce School Funding? Evidence from Michigan

• In Event: Beyond “Money Matters”: How Policy Design Shapes the Distribution and Effectiveness of School Spending

Abstract

This study examines how changes in poverty measurement under the Community Eligibility Provision (CEP) and the expansion of Medicaid Direct Certification (Medicaid-DC) reshape school funding. CEP, a federal nutrition program that allows high-poverty schools to provide universal free meals, eliminates household income applications and relies instead on administrative certification to identify economically disadvantaged students. While this design reduces administrative burden and expands access to meals, it can also mechanically reduce administratively measured poverty when not all eligible students are captured through direct certification. Despite widespread concern that such changes may affect funding systems that rely on these measures, there is very little empirical evidence on whether CEP alters school-level resources.

Using Michigan as a case study, I leverage the staggered rollout of CEP and the subsequent expansion of Medicaid Direct Certification in 2018 to examine how changes in measured poverty translate into school funding. I find that CEP adoption reduced administratively measured poverty without changing underlying student need, leading to an average decline of nearly 10 percent in total operational spending. These reductions are concentrated almost entirely in added-needs instructional spending, the category most directly tied to poverty-based state aid, and are largest among schools with lower levels of poverty at baseline. A variety of diagnostic checks indicate that these declines are largely mechanical, reflecting the removal of application-based certification rather than shifts in student composition or increased food service costs. Using a recent meta analysis on the effects of school spending, the average decline in total operational spending of $861 per pupil implies a reduction of roughly 0.027 standard deviations in test scores and 2.4 percentage points in college-going, with nearly identical estimates when focusing on added-needs instructional spending.

I then show that Medicaid-DC causally increased added-needs spending by correcting this mismeasurement, raising spending by approximately 14 percent in early-adopting CEP schools and four percent in schools adopting CEP in 2018. However, by 2019, many CEP schools had partially recovered, but the majority continued to spend less than their pre-CEP levels, despite substantial increases in the number of students identified as economically disadvantaged after Medicaid-DC.

These findings highlight a previously underexamined mechanism through which policies that alter administrative measurement can generate unintended fiscal consequences. More broadly, they show that correcting poverty measurement is necessary but insufficient to ensure equitable school funding when aid formulas are constrained by fixed budgets.

Author

• Brittany Nicole Vasquez, University of Pennsylvania
  Presenting Author