Photo: Christopher Lee/The New York Times/Redux
Last week, Texas Republicans advanced a measure that would create a massive new school voucher program. Should it become law, parents could use $1 billion in public funds to subsidize private-school tuition and the costs of homeschooling. It would not matter if the private school is religious or secular. (A homeschooling parent could be a Flat Earther and it wouldn’t matter, either.) As the New York Times put it, “The bill was championed by an ascendant wing of the Republican Party, closely allied with President Trump and important conservative donors, including Betsy DeVos, Mr. Trump’s wealthy former education secretary, and Jeff Yass, a billionaire financier from Pennsylvania and a Republican megadonor.” Money is winning the day — but what about kids?
They might not fare well, if research is any guide. In The Privateers: How Billionaires Created a Culture War and Sold School Vouchers, Josh Cowen traces the history of voucher programs in the U.S. and argues that the evidence doesn’t support their expansion. Cowen, a professor of education policy at Michigan State University, writes that he once supported school vouchers until the evidence began to show major declines in academic outcomes, especially among disadvantaged students. Louisiana’s program is a prime example. One paper showed that school vouchers “caused unprecedented large, negative impacts on student achievement,” Cowen writes, deficits that continued into the program’s second year. Another paper, produced by a team at the Massachusetts Institute of Technology, showed “similarly large negative impacts.” Cowen says voucher advocates are pushing school choice in spite of poor results because it entrenches the culture war. Vouchers are no longer about education attainment, if indeed they ever were. They’re part of a much broader assault on the very notion of the public good.
This interview has been edited for length and clarity.
As I understood it, the pro-voucher argument was originally twofold: They would improve academic achievement and create better opportunities for disadvantaged kids. What does the research tell us about the veracity of those claims?
From the academic-achievement standpoint, the bigger and the more recent the voucher system is, the worse the results have gotten. And that was true right up until the eve of the pandemic. Since then, we haven’t really been able to collect any data because all of the new expansions don’t include any data-collection requirements that would allow us to ask that question in a rigorous way. I say this in the book, but the reason for that is that the results were so bad they just stopped collecting data.
The smaller targeted programs that really were means tested in the ’90s and the early aughts didn’t have the negative results on student academics that we saw over the last decade. They maybe weren’t the greatest thing ever, but they certainly weren’t causing student achievement to drop.
Now we’re rolling back social services, and health care, and retirement security — there are all these other things going on that really, in my view, gets vouchers back to historical origins and what they were meant to be, which is a disinvestment in the public sector.
As you note in your book, there is research that gets used to make a case for vouchers. Can you walk me through what that evidence says and what the flaws might be?
The way that this works with these guys — and this was true even before the data got really bad on test scores — is they asked, What were the legislative or evaluation outcomes? And then they would buffer those results with a series of non-required, non-mandated reports on other things.
They would say, “Well, parent satisfaction is higher.” Which seemed true in some surveys. There was one study that I singled out because I found it particularly distasteful, but they said vouchers improved character because there were fewer out-of-wedlock births. So they’ve started to move the goalposts to a whole lot of nonacademic outcomes. First there were no voucher impacts on achievement, nothing to write home about, and then increasingly, these horrific academic declines over the past decade.
You still kept seeing all these reports coming out. Basically, the argument being “Well, test scores just aren’t a good way to evaluate what private schools and vouchers can do.” Apparently they’re good enough to evaluate public schools in their view, but not private schools.
If we’re thinking about test scores as one measure of academic achievement, and using that to draw a conclusion about school vouchers, how conclusive is the evidence against school vouchers at this point?
It’s as conclusive as social science gets on test scores. And I really can’t overstate the magnitude of some of these results. So the other side will say, “Well, Josh, the parents are happier.” Well, they’re happier if you believe the survey research. I don’t want to give you the impression that I’m discounting parent satisfaction as a metric for how we should consider school success.
But if that’s true, we need to talk about how parents feel about their public schools too, especially the schools they are in, which is always higher than their view of public schools more generally.
These were some of the largest academic declines in the history of education research. Take a kid in Louisiana, for example, who is started at the median, or 50th percentile, of the state exam. So if they started on the exam with roughly half the state students scoring better, and half the state students scoring below them, the Louisiana voucher program effects — translated into percentiles — are the equivalent of pushing that kid down from the 50th percentile to the 34th percentile the following year. Meaning now after using the voucher, that child is below 66 percent of students on the exam, down from below only 50 percent of kids the year before, and ahead of only 33 percent of them. Those are massive declines. In recent years, only COVID-19 and Katrina did that much to student learning. And occasionally virtual charter schools.
You just don’t see stuff like that in social science or in education research. And that became true as the programs have gotten larger. And instead of the early period of time on these early demonstration trials that I talk about in the book, where you had a lot of school recruitment, a lot of these conservative philanthropists would incentivize the higher-end private schools to take a handful of voucher kids to then study.
In Cleveland, for example, the Walton Foundation literally built two schools for voucher kids to use and then gave the data to the researchers to study those schools. Think about that as maybe a medicine that performed well in the laboratory, and then when handed out or prescribed to the general population had very harmful effects. You would immediately stop giving out the medicine to the general population. You wouldn’t say, “It performed okay in a few trial studies.” You’d stop the medication.
The through line here is this claim that public schools, or public services more generally, were so inefficient that they could no longer be trusted. That’s basically the nutshell case that the other side makes. But as science goes, we don’t get results like this. This is what I say in the book and what I say in a lot of public speaking. You actually have to go to what COVID-19 itself did to academic results, or to what Hurricane Katrina did to test scores in Louisiana 20 years ago.
Can we look at the evidence that exists and draw some sort of conclusion about why vouchers don’t improve academic achievement or test scores? I’m thinking, for example, of creationism and how that might influence a child’s ability to go major in a STEM subject.
That’s one big thing right there: These schools often aren’t interested in these academic subjects. So you’ll hear the other side on this question say, “Well, these private schools just don’t have the curriculum aligned with the state standardized tests that they’re using.” I mean, curriculum alignment is a real thing, and individual parents couldn’t use a national normed exam and compare it to a state standardized exam to see how their kid was doing in a voucher school relative to how they were doing in public school. But curriculum alignment itself really can’t explain the masses and size of the declines. The real issue is two things. The first is what you’ve just raised. It’s tough to teach calculus when you’re teaching creationism.
I’m a Christian man. I practice my faith very rigorously, especially, over the last few weeks, during the Lenten season. No one’s beating up on religious schools here, but I don’t think it’s a good use of taxpayer dollars.
More concretely, though, how many times have you heard this idea that vouchers are needed to save academic learning? I mean, every time the NAEP comes out, or the COVID-19 results come out, they say, “Oh, this is why we need vouchers.” Well, on that metric, the actual metric of standardized tests, they do really poorly, at least for the kids who are in public schools. And that gets back to the second explanation, which is that three-quarters of voucher users are already in private school. The most accurate way to think of this is just a taxpayer subsidy for garden-variety interest groups.
What that means is that there are lots of high-end private schools out there. Historically, they’re not the ones that take kids who transfer from public to private school, at least not in huge volumes. The good elite providers are already full. They’re the ones that have waiting lists; they’re the ones whose tuition vastly exceeds what the voucher covers.
No one is saying that the kids who are already there for years, and all of a sudden get taxpayer dollars as a new subsidy, that they’re necessarily getting a bad education. What I am saying is that the kids who leave public schools to transfer to vouchers tend to be able to get in only to some of these what I call subprime providers. So it’s not just that they may be teaching creationism instead of calculus. It’s actually really hard to run a school. And some of these schools just aren’t any good. Forty percent of private schools in the state of Wisconsin, where I did research, that have ever taken a voucher have closed.
How much of the voucher push is about disciplining or undermining teachers’ unions and organized labor?
I think it’s a big part of it. It’s the through line, as I say in the book, between these three different sets of billionaires. But really, I only talk mostly about two of the sets in the book, because the latest flavor is brand-new, and that’s the tech-bro billionaires. Elon Musk is getting into this game, and Jeff Yass, the billionaire, who really spends a lot in Texas. Then you have the Betsy DeVos world, where they want to advance God’s kingdom using vouchers, and churches need to become centers of community again, and vouchers will help us do that. Again, I take issue with that, as a religious man, but that’s how they see it. And the Koch people just couldn’t care less about Christian schools. What links all three of these things together is a deep hostility to labor unions. And public education still is one of the largest proportionate areas where labor is still strong.
I tend to associate school vouchers with the right wing, especially with the Trump administration. I don’t think that’s an isolated perspective. So I was curious to know if support for school vouchers was ever bipartisan.
Not like charter schools. There are some constituency groups that have been more, I would say, willing to consider vouchers. These tend to be communities who have felt like public schools haven’t always lived up to their promise. Some Black pastors, for example, who might actually see vouchers’ religious component as a positive thing.
In general, I think public schools do great, and the data shows that, but only when they’re well invested in. For example, take students with disabilities. While there’s a legal responsibility to provide those kids with an education that does not exist in the voucher sector, maybe the public schools fall short in the way that parents expected. The problem with vouchers as a solution to any of that is that it makes all of those problems far worse, because the private schools really do get to deselect anybody they want.
It’s why I called vouchers education’s version of predatory lending, because you do see the other side, and these do come from right-wing messaging, going into some of these communities and saying, “We’ve got a deal for you.” But those are exactly the kids hurt most by vouchers, we have to explain. And so two things can be true at once: that we need to do better for all of these families, and reinvest in their children, and in their needs as parents, and that vouchers are not the solution and in fact make all of these problems worse.
Do we know how likely it is for voucher schools to be religious?
The answer is that voucher schools are disproportionately religious. We also have some anecdotal evidence from these places I write about in the book, where often you get churches that maybe used to run a school 40 years ago and decide to restart a school once vouchers come to town.
And this is why you’ve got Betsy DeVos’s 501(c)4, American Federations of Children, and other organizations very much leaning into that. “Let’s save our Catholic schools using vouchers.” Some studies I note in the book are from an economist based at Notre Dame that showed that vouchers become the dominant source of revenue for churches. And the reason for that is these churches start to set up schools or restart schools as a result.
Is that what people mean when they mention “pop-up schools”?
There are some pop-up schools, especially in today’s voucher market, that are like these micro schools, which are a little different than just schools with one teacher and 12 kids. But they’re tiny private schools, as the name implies. Those may not be religiously affiliated.
But often you see double-wides appear on church grounds. You see a lot of that happening again; churches are reactivating their school buildings. And I’ve been to a bunch of private schools that are literally run out of church basements. I mean, that’s what a pop-up school often is.
Yeah, I went to a Christian school that didn’t have any windows.
At least you could find it. You’ve got schools in North Carolina, for example, that NPR reporters couldn’t even find, just some entities charging the state for tuition at a school that didn’t seem to physically exist. That’s an extreme example of the deregulated version of these things. Most states are a little bit in the middle of that. Wisconsin has thrown voucher schools out for not having their financial books together. And then you’ve got the far end of that, which would be like a North Carolina or an Arizona where it’s just anything goes.
As someone who spent some time being homeschooled, I’ve been curious about this recent push to subsidize homeschooling. Is that relatively recent, or was it always part of the idea?
It’s very recent. Let me stipulate that the homeschool movement is very divided on this question. It was really conservative homeschool networks in Texas and Tennessee that helped push this process out further than it would have. They really opposed vouchers in those two states, which were among the Republican coalition against these bills.
If you think of the religious homeschool movement as a big piece of this, then the big players in Republican legal and political circles — and the Alliance Defending Freedom most of all — have been pushing, over the last 15 years or so, to get more public funding for homeschoolers, even to the point of trying to originate casework litigation strategies that would require all states to fund homeschool parents at the same rate they’re funding public-school parents on religious grounds.
Obviously, your book is about school vouchers, but I did want to mention St. Isidore, which is a Catholic virtual charter school in Oklahoma. It’s a new development. Does it bear any relationship to the school voucher program? Are we talking about the same forces and the same people involved?
It’s all part of the same movement. The argument was that a private entity should be able to run a public school. Even if they’re religious, they should be able to run a religious public school.
You have to tell that part of the story alongside the expansion of universal vouchers to everybody, and alongside this push for the Ten Commandments in classrooms in many states now, as well as students who are allowed to leave public schools for part of the day to get credit for Bible studies. There’s this very big effort to really push religion much more into the public-school sector. And the St. Isidore case would complete that circle. If you think about vouchers as basically public funding going to these private religious schools, but they’re not. No one’s arguing they’re being publicly run, but they are private entities. They actually act as something like third-party contractors with states to provide K-12 schools.
And the fact that the parents take the dollars and then spend them at the private school is a distinction without a difference. But the voucher lobby seems to think it’s this big deal that the parents touch the dollars first. They don’t actually even physically often do that. Like in Iowa, for example, the state basically holds the money and asks for it until a parent can show they got into a private school. And then the state releases the dollars to the private school in the parent’s name. But these are just different ways to accomplish the same thing, which is to have publicly funded religious education. St. Isidore takes it a step further, or they’re purporting to take it a step further, where they really would be a religious public school as opposed to a private school that does take public dollars.
Are there other developments that you think are of particular concern that we should be following?
I’ll just go back to the same point that I’ve tried to hit a couple of times, which is that it’s really important to see this walk back from public education as part of this larger push to disinvest from the public good, to disinvest from health care, the Affordable Care Act, to disinvest from Social Security in the long run.
I think people need to understand that we’re in a time when we need to have serious conversations about whether we’re going to renew, and rebuild, and reinvest in these public services, or whether this is going to be a country where everyone’s on their own for these steps.
