Vouchers, responsees, and the test-taking population: Regression discontinuity evidence from Florida

While there is a rich literature that investigates whether accountability regimes induce schools to manipulate their test-taking population by strategically excluding weaker students, no study thus far investigates whether voucher programs induce schools to engage in similar strategic behavior. This paper analyzes a Florida program that embedded vouchers in an accountability regime. Specifically, it investigates whether the threat of vouchers and the stigma associated with the Florida program induced schools to strategically manipulate their test-taking population. Under Florida rules, scores of students in several special-education and limited-English-proficient (LEP) categories were not included in the computation of school grades. Did this rule induce the threatened schools to reclassify some of their weaker students into these excluded categories so as to remove them from the effective test-taking pool? Using a regression discontinuity strategy, I find evidence in favor of strategic reclassification into the excluded LEP category in high-stakes grade 4 and entry-grade 3. In contrast, I find no evidence that the program led to reclassification into excluded special-education categories, which is consistent with the substantial costs of classifying into special-education categories during this period. These findings have important policy implications.