What happened in Louisiana? The first experimental evaluations in the world to find negative effects of a private school choice program on student achievement were of the Louisiana Scholarship Program. Education policy scholars have been debating what went wrong in the bayou for over three years now. Some researchers have argued that Louisiana’s onerous regulations deterred the best private schools from participating in the program. Others instead have claimed that the program would have done even worse without the quality‐enhancing regulations in place. Of course, both theories are plausible. But which theory is supported by evidence?
Our just‐released study is the first experimental evaluation of the effects of various regulations on the willingness of private school leaders to participate in voucher programs.
We randomly assigned a hypothetical voucher program participation offer to 3,080 private school leaders in Florida in 2018. Specifically, we randomly assigned one of three different regulations — or no additional government regulation at all — to each of these private school leaders and asked them whether they would participate in the voucher program the next year. Because regulations increase the costs of program participation, we expected each regulation to reduce the likelihood that private school leaders would be certain to participate the following year.
Relative to no regulations, our most conservative statistical models find that open‐enrollment mandates reduce the likelihood that private schools are certain to participate in a choice program by about 17 percentage points, or 70 percent. State standardized testing requirements reduce the likelihood that private schools are certain to participate by 11 percentage points, or 44 percent. We find no evidence to suggest that the prohibition of copayment affects anticipated program participation overall. Our main results suggest that private school leaders are most concerned about maintaining their specialized missions.
It appears that regulations tend to reduce private school participation in voucher programs overall. But that finding alone doesn’t explain why the Louisiana experiment found negative effects on student test scores. Although the Louisiana program includes all three regulations in our study, those requirements could have prevented a lot of low‐quality private schools from participating in the program. But do regulations increase average quality levels of participating private schools? Or are they more likely to deter higher‐quality private schools from participating in school choice programs?
Because we had access to two proxies for private school quality — tuition levels and enrollment trends — our analysis is able to examine this important question. The clearest result is that higher‐quality schools — as measured by tuition — are more likely to be deterred by the regulation that mandates that all schools take the voucher amount as full payment. This result is intuitive, as it is much more costly for a school with tuition of $20,000 to take a $6,000 voucher as full‐payment than for a school with tuition of $10,000 to do so.
In addition, our model with all controls finds that a $1,000 increase in tuition is associated with a 1.4 percentage point larger negative effect of a state standardized testing mandate on intended program participation, and a 10‐percentage point increase in enrollment growth from 2014 to 2016 is associated with a 2‐percentage point larger negative effect of the open‐enrollment regulation on intended program participation.
Importantly, because random assignment leads to equivalence in expectation across treatment and control groups on both measurable and unmeasurable factors, each of our estimates of the impact of regulatory requirements on the expressed willingness of private school principals to participate in a private school choice program is causal.
Of course, one study certainly does not settle the voucher regulation debate. But our results suggest that costly regulations tend to reduce the quantity and quality of private schools that elect to participate in school choice programs. In other words, estimates from existing private school choice program evaluations could be lower‐bounds of the true effects of private schooling in general. And decreasing voucher regulations could improve the efficacy of these types of programs.