Comparing the Impacts of Universal vs. Means-Tested Vouchers on Student Achievement and Absenteeism in Indiana

• In Event: Universal Private School Choice: Implementation, Organizational Responses, Student Outcomes, and Fiscal Effects

Abstract

Over the past five years, private school voucher and other choice programs have expanded rapidly from targeted interventions to near-universal or universal policies. Despite this shift, rigorous evidence on student outcomes under expanded eligibility remains limited, particularly in statewide programs where participant composition and policy design have evolved over time (see Wolf & Zimmer, 2025). Evaluations of statewide voucher programs such as Indiana’s Choice Scholarship Program (ICSP) (Berends & Waddington, 2018) and similar programs in Louisiana (Abdulkadiroglu et al., 2018; Mills & Wolf, 2017) and Ohio (Figlio & Karbownik, 2016) found negative effects on student achievement—especially in mathematics—for students switching from public to private schools. However, these findings reflect earlier, more targeted program designs and may not generalize to current policy contexts characterized by broader eligibility and shifting participant populations. 

This study examines the longitudinal impacts of Indiana’s voucher program as it has transitioned from a means-tested program to near-universal eligibility. We ask: (1) What is the impact of receiving a voucher and enrolling in a private school on students’ math and English/language arts (ELA) achievement and chronic absenteeism? (2) How do these impacts vary across policy phases? and (3) How do effects differ across student and school subgroups? 

We use thirteen years of longitudinal, student-level administrative data already acquired from the Indiana Department of Education spanning the full history of the ICSP (2011–12 through 2024–25, excluding 2019–20). These data include comparable achievement measures for public and private school students in grades 3–8, along with detailed demographic and enrollment records. Indiana’s requirement that voucher-participating private schools administer the same statewide assessments as public schools provides a unique opportunity for consistent cross-sector comparisons. Outcomes include standardized math and ELA achievement and chronic absenteeism. 

Given the absence of random assignment, we employ complementary quasi-experimental strategies. We first construct a matched comparison sample using exact matching on grade, cohort, geography, and voucher eligibility, combined with propensity score weighting that accounts for other student pre-treatment characteristics. We also estimate a more restrictive “switcher” design comparing students who move from public to private schools using vouchers to similar peers who remain in public schools. Across both approaches, we estimate effects within a difference-in-differences framework adjusting for prior achievement and attendance and including grade-by-year fixed effects. We model program phases corresponding to shifts in eligibility rules and prior schooling requirements and examine heterogeneity across student demographics and school contexts. Robustness checks include alternative matching specifications, falsification tests, and sensitivity analyses. 

Preliminary findings suggest that negative achievement effects were concentrated in early program phases, with largely null achievement effects as the program expanded. These patterns coincide with an increase in voucher users who were previously enrolled in private schools. By leveraging a rare longitudinal dataset spanning the evolution of a statewide voucher program, this study provides updated, policy-relevant evidence on how effects may change as programs move toward universal eligibility.

Authors

• Mark Berends, University of Notre Dame
  Non-Presenting Co-Author

• Joseph J Ferrare, University of Washington Bothell
  Non-Presenting Co-Author

• Joseph Waddington, University of Notre Dame
  Non-Presenting Co-Author

• Ron W Zimmer, University of Kentucky
  Presenting Author