School choice skeptics often raise concerns about the students who are “left behind” in traditional public schools when statewide school vouchers are enacted.
What will become of those kids whose parents, some claim, weren’t savvy enough, sufficiently motivated, or simply didn’t have the time or resources to apply for a private school voucher? And what about those students who barely missed the eligibility cut-off—won’t their public schools suffer as their eligible classmates exit for nearby private schools, taking valuable financial resources with them?
Policymakers across the country are right to raise concerns about the systemic effects of choice-based reforms. Thankfully, researchers can help answer some of those questions by studying existing programs.
I’ve reviewed the available research on this topic and conducted a new study of my own to determine how private school competition affects public schools after the establishment of a statewide voucher program. The following sums up what the education industry knows so far.
There’s a hefty literature on how public schools respond to competition from statewide school vouchers or their close-cousin, tax-credit scholarship programs. In my review of more than 1,000 titles, I uncovered 21 studies on the competitive effects of school vouchers specifically. Twenty of those studies found neutral to positive results and one found exclusively null results, but none found negative effects on public schools.
I found increases in private school competition are associated with neutral to positive changes in public school test scores.
There are limits to what that body of literature can tell us, however. Sure, we know a lot about the direct and indirect effects of highly targeted programs, concentrated within a single city, but how much do we really know about the current generation of voucher programs? After all, the characteristics of more recently established voucher programs have changed. We are more likely to see statewide programs now, with more generous income caps determining student eligibility for participation.
That’s why I undertook a competitive effects study of two of the three largest means-tested statewide voucher programs currently in operation in the U.S.
Indiana’s Choice Scholarship Program (CSP) and the Louisiana Scholarship Program (LSP) started distributing vouchers in the 2011-12 and 2012-13 school years, respectively. Although they have grown considerably since then, both programs served between approximately 4,000 and 5,000 students in their first year. Examining public school outcomes in these programs’ first years of operation makes it easier to separate out the pure competitive response from compositional changes to schools brought about by changing peers and any potentially significant changes to schools’ financial resources brought about by enrollment shifts.
Imitating the geocoded competition measures used by Figlio and Hart in their study of the Florida Tax Credit Scholarship Program, I created distance, density, diversity, and concentration measures of private school competition to test if the introduction of CSP and LSP voucher programs were related to changes in public school test scores. This allowed me to compare public schools facing more (and more varied) private school competition to public schools facing fewer (and more homogenous) private school competition.
- Distance – Did student test scores increase more in public schools that were located geographically close to a private competitor, compared to public schools that were located further from any private competitors?
- Density – Did student test scores increase more in public schools with a higher count of private competitors within a five-mile radius?
- Diversity – Did student test scores increase more in public schools where there were lots of different types of private schools (eg. Non-sectarian, Catholic, Lutheran, Baptist) within a five-mile radius?
- Concentration – Did student test scores increase more in public schools where the private school market share was distributed among many different types of private schools, as opposed to being in the hands of a single type of private school (eg. If all the private schools were affiliated with the same church)?
I found increases in private school competition are associated with neutral to positive changes in public school test scores. The positive impacts are generally small but appear to be largest in the lower-performing public schools, using measures of competition that include concentration and diversity of private school competitors; they also tended to occur in Louisiana rather than in Indiana.
Overall, these studies demonstrate that private school choice programs have null to modest, statistically significant, positive impacts on the academic achievement outcomes of non-participants who remain in their assigned public schools.
The full study is available as a working paper on the website of the Program on Education Policy and Governance: http://www.hks.harvard.edu/pepg/PDF/Papers/PEPG14_05_Egalite.pdf