Accountability with voucher threats, responses, and the test-taking population: Regression discontinuity evidence from Florida

Abstract

Florida's 1999 A-plus program was a consequential accountability program that embedded vouchers in an accountability regime. Under Florida rules, scores of students in several special education (ESE) and limited English proficient (LEP) categories were not included in the computation of school grades. One might expect these rules to induce F schools (that faced stigma and threat of vouchers) to strategically classify their weaker students into these excluded categories. The interplay of theseruleswiththoseoftheMcKayprogramfordisabled students, however, created an interesting divergence of incentives as far as classifications into excluded LEP and ESE were concerned. Because classifying students into ESE made them eligible for McKay vouchers that were funded by public school revenue, the McKay program acted as a strong disincentive to such classification. Using a regression discontinuity strategy, I investigate whether the differences in incentives led the F schools to exhibit different behaviors as far as classifications into excluded ESE and LEP were concerned. Indeed, I find robust evidence in favor of classification into excluded LEP in high-stakes grade 4 and entry grade 3. In contrast, I do not find evidence of classification into excluded ESE. © 2013 Federal Reserve Bank of New York.