Rigorous studies have been conducted on various school voucher programs – most notably those in Milwaukee, the District of Columbus, and Florida – but this study by CATO’s Matthew Carr is the first of its kind to study Ohio’s EdChoice Scholarship program. Specifically, it examines whether traditional public schools are spurred to improve in the face of a threat of losing students to private schools. It does not examine whether vouchers are effective for students who use them but rather investigates a key school choice theory: whether competition “creates incentives for systemic improvements.”
To test this, Carr collected achievement data from the Ohio Department of Education on EdChoice-eligible schools over three distinct time periods beginning in 2006 (note, eligibility changed multiple times), which creates a unique research design in that there are three “treatment” periods enabling analyses of whether each school changed its own behavior in response to the voucher threat. (In contrast, other studies have compared fundamentally different types of schools, eligible v. non-eligible schools.)
The study measured school improvement by looking at fourth- and sixth-grade reading and math scores; and the percent of students scoring at various levels (limited, proficient, advanced) to gauge the extent to which schools under threat focused on “bubble students” (those just above and below the proficiency cut-off and upon whom a school’s rating depends most heavily). It also controlled for factors such as school quality (rating on A-F scale), and percent of students who are white, disadvantaged, and/or disabled. Unique to this study (and impressive) is that Carr manages to tease out the “scarlet letter” effect, i.e., did schools improve not because of the voucher threat but rather because of the stigma associated with receiving a highly publicized poor rating from the state?
Several significant findings emerged. The voucher threat was correlated with a achievement gains in fourth grade reading (the equivalent of 2200 extra students reaching proficiency) and this did not result primarily from stigma. Second, performance among bubble students didn’t shift much; rather, students in the lowest and highest performing categories made gains. Carr theorizes that voucher-threatened schools may have focused on those students most likely to exit, and calls for more research in this area.
The Impact of
Ohio’s EdChoice on Traditional Public School Performance
CATO Institute
Matthew Carr
August 2011