Poster #120 - Understanding the Long-Term Dynamics of Special Education Expenditures in Massachusetts

• In Event: Friday Poster Reception

Abstract

Motivation

Approximately 15% of students enrolled in public schools – about 7.5 million students – are identified as students with disabilities and thus are eligible to receive special education services. Special education costs amount to approximately 14 billion dollars of investment through the federal government (Dragoo, 2018). Additionally, recent research indicates that there is significant state-by-state variation how special education costs are distributed across federal, state, and local levels (Kaput & Schiess, 2024).

This study, focused on Massachusetts, makes a unique contribution to the special education finance research field by providing new insights into the long-term dynamics of special education expenditures. In our analysis, we focus on the Chapter 70 and Circuit Breaker programs, which are, respectively, the foundation budget and a reimbursement program for excess special education costs.

We answer two primary research questions:

1.        How have the components of special education expenditures changed over time?

2.        How do districts respond to the Chapter 70 / Circuit Breaker programs to determine special education expenditures?

Method

Our data is a combination of publicly available and restricted-use data from Massachusetts from 2013-2023, which includes data on student enrollment as well as education finance expenditures and reimbursements across all public schools in the state. In our analysis, we leverage OLS regression to understand which district and school factors contribute to changes in special education expenditures over time. We use a model that predicts year-over-year changes in special education expenditures as a function of district and school factors, including student enrollment characteristics (e.g., disability classification, race/ethnicity), use of public collaboratives, and private school outplacement. This model is: 

𝐸𝑥𝑝𝑒𝑛𝑑𝑖𝑡𝑢𝑟𝑒𝑠𝑇𝑦𝑝𝑒𝑑𝑡 = 𝛼0 + 𝜋𝑡 + 𝛾𝑑 + 𝜖𝑑𝑡

Where ExpendituresType are the various special education expenditure types, 𝜋𝑡 is a variable for each year, and 𝛾𝑑 represents a vector of district characteristics, as described above. This model is also expanded to include interactions with district characteristics and further parameterize which type of special education expenditure drives overall costs. 

Findings

Some emergent patterns in our results point to important areas of further investigation. For example, we find that smaller districts who enroll a lower proportion of students with disabilities relative to their total enrollment in 2013 are consistently more likely to submit a reimbursement claim through the Circuit Breaker program, but that trend has declined within that group over time. We also find a wider distribution of per pupil costs (i.e., eligible expenses divided by students claimed) in Circuit Breaker claims over time. Additionally, we find that districts claim higher eligible expense amounts when they claim more students in Circuit Breaker as a proportion of their total enrollment of students with disabilities.

Implications and Future Directions

Overall, these findings indicate that districts may indeed engage in different special education service provision behaviors based on their engagement with state funding mechanisms. This has notable implications for how policymakers and practitioners may move forward in covering special education costs in the face of rising enrollment of students with disabilities, rising costs, and potentially fewer federal dollars.

Author

• Christopher Cleveland, Brown University
  Presenting Author