In a press conference last week, organizers advocating for the Arkansas Educational Rights Amendment presented damning evidence on academic performance in private schools receiving vouchers. They argued that the research record offers a compelling argument that private schools that receive taxpayer funds should be held to the same standards as public schools.
Josh Cowen, professor of education policy at Michigan State University and a prominent researcher on vouchers and other education policies, gave a guest presentation on that research. If taxpayer money is going to offer funding for private schools, Cowen said, citizens should demand “transparency, accountability and public information — as much as possible on the same scale as what public schools have to provide.”
And the evidence on student performance only highlights the need for rigorous accountability, Cowen said. The most robust research on voucher programs, dating back to 2010, has produced “horrific academic results,” he said.
The For AR Kids ballot question committee, a coalition of local progressive groups, is currently collecting signatures to put a measure before voters in November that would “require identical academic standards and identical standards for accreditation, including assessments of students and schools based on such standards, for any school that receives State or local funds.”
The ballot initiative comes partly in response to Arkansas LEARNS, the education overhaul backed by Gov. Sarah Huckabee Sanders and passed by the Republican supermajority in the Legislature last year. Among the law’s most controversial provisions was the creation of a voucher program to help families cover the cost of private tuition. Under LEARNS, private schools are given wide latitude to police themselves and do not have to follow the same testing, accreditation, reporting, or other accountability measures that public schools must. LEARNS began this year and will be phased in until all K-12 students in the state are eligible to apply starting with the 2025-26 school year.
In addition to equalizing the standards, the ambitious For AR Kids proposal would require the state to offer universal (voluntary) pre-school for all 3-and-4-year-olds, as well as afterschool and summer programs; require services to meet the needs of children whose families make less than 200 percent of the federal poverty level and students with disabilities; and explicitly codify the adequacy and equity standards imposed by the Lake View court decisions.
In order for the proposal — which would amend the state constitution — to make it onto the 2024 ballot, supporters must collect at least 90,704 signatures from at least 50 counties. (Arkansas has 75 counties total.) Ballot committee leaders said they believe they’re past the halfway point in collecting signatures and confident that they will meet the county threshold, but the group has not offered more specific numbers.
In addition to being a nationally prominent researcher on education reform efforts, Cowen has frequently partnered with districts and state agencies, often with a focus on voucher evaluation, for more than two decades.
“One of the reasons I got vocal about this, it’s not just as an issue-area expert, but my entire career has been devoted to this idea that evidence should inform public policy,” he said. The voucher push, which has been popping up in GOP-controlled states across the country, has been the opposite, he said. All of the evidence was “as one-sided as social science gets,” he said. Vouchers aren’t just sucking money out of public schools or programs with a proven track record of success, he argued, they’re making education outcomes worse. The evidence is “pretty dreadful,” he said, but that evidence is “manifestly not informing public policy.”
Research finds vouchers lead to declining student performance
Cowen pointed to four independent, statewide evaluations for voucher programs dating back to 2010 — the only such evaluations available in the research record for fully scaled-up programs. The results are overwhelming, clear and startlingly bad.
“If that’s not enough of an argument for why we need oversight, I don’t know what is,” Cowen said. “Speaking as a researcher, we don’t get straightforward patterns very often in the research community — we have one here.”
Cowen said that these results were extreme and highly unusual. Some education policy initiatives fail because they don’t move the needle much, Cowen said, but vouchers are an outlier in that they actually appear to make student outcomes get significantly worse. “You can’t really find too many examples of education reforms that had negative impacts,” he said.
The only comparison for “the absolute academic declines that were seen in these statewide systems,” according to Cowen, is the learning loss associated with natural disasters or the recent COVID 19 pandemic. The chart above demonstrates the scale of the effect for students who switch from public schools to private schools via vouchers.
Very early on, small pilot programs showed promise, but once vouchers were scaled up to serve larger programs — such as the statewide, universal program in Arkansas — the results tanked. This challenge is scaling up is a frequent problem in education reform efforts, Cowen said. The results have been consistently atrocious dating back at least to 2010, he added.
The main problem, Cowen said, is that while there are a small number of excellent private schools, there are many more of what he calls “subprime” schools. These are private schools that simply aren’t equipped to meet reasonable academic standards. Some may be in financial distress. Some have plenty of space because most families don’t want to send their kids there. Some may pop up and only survive via the largesse of voucher cash. Kids most in need of help are much more likely to wind up at these low-performing schools, because their families may be priced out of better schools with tuitions higher than the voucher amount. Or they may simply not be let in to those better schools, which can pick and choose who they admit, unlike public schools that have to take all comers.
“The reason [for the learning loss] is simple,” Cowen said. “The schools were just not very good. There’s not enough high-quality private schools to go around is what it comes down to. And they don’t take anybody.”
Vouchers are controversial for lots of other reasons. For one thing, many question whether it’s an appropriate use of taxpayer money if 70% of the kids never went to public school in the first place (that’s where the figure typically settles in all of the statewide voucher programs, Cowen said) — amounting to a welfare check for private school families. But this is surely the most devastating critique: They don’t help students frustrated with their public school wind up at better schools. Instead, they funnel them to bad schools. They learn less, and perform worse. It’s hard to imagine a more counterproductive policy.
Transparency on test results
Since the federal No Child Left Behind law was passed 23 years ago, we’ve become accustomed to annual reporting on how schools are performing in terms of student test scores. A parent can quickly ascertain whether a given school is doing well on that particular metric, and the state gives out publicly available grades for each school based on their performance.
If that’s the standard for public schools, Cowen argued, parents should have access to the same information for private schools receiving public funding for vouchers.
In fact, Cowen was part of a research team that found evidence in Milwaukee that private schools receiving vouchers actually saw their test results improve once they started having to publicly report them by school. Working on behalf of the state, Cowen and his team tracked a representative sample of around three thousand voucher kids over a period of five years. In the fifth year of the program, there was a policy change: Private schools receiving vouchers had to start publicly reporting test scores.
This set up a natural experiment, since the researchers had access to the test scores of the kids they were following (redacted by name to protect their privacy). What kind of impact would it have on those students’ performance if these private schools had to report test results just like public schools? The results were dramatic: “Lo and behold, achievement for these children jumped up quite significantly when [the schools] knew they were going to be on the hook for the reporting.”
These results sharply contradict a pro-voucher talking point that asking private schools to abide by the same responsibilities as public schools on student assessment would amount to over-regulation.
“It’s not conclusive, but it strongly suggests that what you need at minimum is to set public and private schools on the same performance-reporting standards,” Cowen said. “And then this helps kids and families who are in either sector make comparisons.”
Apples to apples
The testing component of LEARNS, touted as a way to enforce accountability, falls short. While private schools receiving LEARNS vouchers are required to do some sort of (vaguely described) testing, they do not have to take the same tests public schools are required to take. This sets up an asymmetry that makes it impossible to compare student performance at public schools versus private schools.
Cowen suggested that this was by design. This bifurcated testing system has emerged as a disturbing pattern in statewide voucher programs in recent years. In the early days of implementing vouchers, he said, voucher advocates were eager to get comparable test results because they were certain the vouchers would lead to improvement. When it turned out that student performance was getting worse when students switched to voucher schools, advocates kept right on pushing vouchers but suddenly preferred to keep test results secret or sabotage efforts to do apples-to-apples comparisons with public schools.
The value of oversight
Private schools can’t simply be trusted to police themselves, Cowen argued. We expect that issues like discrimination, financial malfeasance or safety in other areas of the private sector require government regulation and a watchdog function. Why would private schools be any different? Discrimination has been an ongoing issue with private schools receiving vouchers, he said, as they generally have little to no restrictions on their admissions policies. In Arkansas, for example, we uncovered details about a private school that was being promoted by the state as a participant in LEARNS but explicitly discriminated against LGBT students and families.
Another consistent issue, Cowen said, is that parents typically are not made aware of their rights, particularly parents of students with disabilities.
“You really need robust systems of information that are out there,” Cowen said. “You can’t just give parents a coupon, call it a Freedom Account and expect everything to go well for these students, because when it comes to these voucher systems, it’s not really about school choice at all. It’s the school’s choice. And we see this most devastatingly for kids with disabilities.”
Crowding out other priorities
“All of this stuff costs more and more money,” Cowen said of voucher programs, which have blown past budget projections in states like Arizona, with jaw-dropping price tags in states like Florida and Ohio.
“They start to crowd out things like just sustaining public school spending at the cost of living and deal with other costs,” he said. They also systematically reduce public school funding because some kids leave the public schools, leaving less funding to follow them. Because of various fixed costs, that leaves schools in a budget lurch even if their student populations are a bit lower.
“Heating, cooling, teacher salaries, transportation, capital expenditures — all of that has to stay with the schools as the kids leave,” Cowen said. “If you lose 15 or 20 kids from a particular school district, it’s not a lot in the grand scheme of things. But those 15 to 20 kids can really devastate the public school’s bottom line.”
Not to mention that spending money on vouchers leaves less money to spend on the interventions that actually work, Cowen said. “We do know unambiguously that over the last 10 years, as the evidence has accumulated against voucher systems, it has also accumulated in favor of deep and sustained investment in public schools,” he said. “It doesn’t sound very clever. … But the evidence is really clear at this point that these types of investments — especially in early childhood — work. And side items like meals for kids, transportation for kids, tutoring programs, and all of these things that we do have really good evidence for, aren’t available when you start expanding these voucher programs. They become opportunity costs.”
In this way, the Arkansas Educational Rights Amendment addresses the issue from both angles: It holds private schools accountable if they’re going to receive vouchers and it also forces the Legislature’s hand in prioritizing more effective interventions.