Powered by 
It’s National School Choice Week (Jan. 25–31), an event first announced in 2011, and the organizers proclaim that school choice is now “a mainstream expectation for American families.” They offer their own survey indicating broad acceptance of the ideas animating school choice. Michael T. Hartney, the Bruni Family Fellow at the Hoover Institution, sees a more complicated picture.
Has the campaign for school choice been won?
Since it’s Super Bowl season, a football analogy seems apt. The school choice movement has made real progress moving the ball down the field over the past quarter century, but this is no time for spiking the ball in the end zone. And the reason is straightforward: there’s a big disconnect between the abstract, low-intensity support captured by public opinion polls—who’s against more “choice” in the abstract?—and what happens once specific choice policies are contested in real-world politics. Ballot initiative campaigns are a case in point. They don’t just inform voters—they mobilize opposition, elevate concerns about tradeoffs, and make local context and perceived self-interest especially salient.
These insights arise out of some new research I have undertaken at Hoover. In that work, my colleagues and I examined voting patterns in three statewide school choice referenda on the November 2024 ballot—in Kentucky, Nebraska, and Colorado. Despite years of polling showing broad support for school choice, all three measures were overwhelmingly rejected by voters.
Importantly, we found that support for school choice is highly contextual and self-interested. Voters who live in communities with low-performing public schools and in places where there are many private schools are supportive, but voters who live elsewhere and do not stand to benefit, or who cannot see how they would benefit, are more easily persuaded to vote against choice. And 2024 was no outlier. Since the 1970s, every single school choice referendum has failed, most by 2-to-1 margins. Again, it’s not that this poor showing reflects a rejection of the core idea of school choice. Rather, it demonstrates how voters respond once specific choice proposals are debated in real campaign settings.
In a previous interview you said that “the decentralized nature of American education is a cautionary tale for reformers who have grand ambitions about a single policy reform.” Is that an obstacle to universal school choice?
I do think decentralization is a real obstacle—but it’s an obstacle to scaling a single model, not to expanding choice writ large. In a country with roughly thirteen thousand school districts, governed by fifty different state systems, the idea that one choice program is going to scale everywhere just isn’t realistic. One basic constraint is supply. The private school sector isn’t equally robust across the country, and it’s not as if passing a voucher or ESA law suddenly creates a dense ecosystem of high-quality private schools overnight.
Building new schools is hard. It takes time, capital, personnel, and community buy-in. So even under the most expansive choice regimes, a large share of students will continue to be educated in traditional public schools. That’s simply a fact. Which is why the question of scale needs to be reframed.
The goal shouldn’t be to scale up a single model everywhere. It should be to ensure that every child has access to a high-quality option, regardless of ZIP code. In some places, that might mean charter schools. In others, it might be a strong parochial sector, or a well-designed magnet system. The idea that the same solution will work in every community ignores both local context and the structure of American education—and isn’t a particularly useful way to think about how choice can be a lever to improve our K–12 system.
Is statewide open enrollment moving in the right direction or has progress stalled? What stands in the way of wider acceptance?
When people talk about scaling school choice, open enrollment is often the most obvious place to start. Even voters with a strong attachment to public schools, and politicians who may be skeptical of vouchers, tend to be more comfortable with this form of public school choice because it works within the existing system.
But that comfort only goes so far. There are real obstacles to making open enrollment work at scale. One of the biggest comes from housing markets. Economists have long shown that the quality of local public schools is capitalized into home values. Homeowners therefore have strong incentives to protect the schools attached to where they live, because those schools are tied directly to the value of their largest asset. When open enrollment raises the possibility that students from outside the community might enroll, opposition often forms quickly, even among people who support choice in principle.
It is no accident that in states like Ohio, the districts that decline to participate in open enrollment are often affluent suburban districts surrounded by large, low-income urban systems. These are districts with high test scores and high home values. They have little interest in opening their doors to students from neighboring districts, especially when those students are perceived as bringing academic or fiscal risk.
There are also other barriers. In some states, choosing another public school across district lines requires approval from both the sending district and the receiving district. Districts that are losing students are also losing revenue. They have little incentive to make it easier for families to leave, and they can use procedural rules to block transfers.
As No Child Left Behind’s disappointing experience with public school choice showed, giving parents a formal “right” to exit a failing school means very little if there aren’t enough high-quality options within reach.
What are these fast-growing microschools, and what are their advantages and disadvantages? Does this kind of radical decentralization make it more difficult to measure accountability and performance? Or does it help that they’re outside the direct control of school boards?
Microschools are small-group learning arrangements in which children from a handful of families are educated together. Their growth accelerated during the pandemic, and current estimates suggest that roughly one million to two million students are now enrolled in these environments. The appeal is flexibility and customization. For many families, microschools were a lifeline during COVID, especially when children were shut out of in-person instruction at their zoned public schools for a full academic year or more.
The downside, as you note, is accountability. Advocates respond that traditional public schools are hardly models of performance oversight. Indeed, there is a certain irony in the fact that microschools expanded precisely when district systems proved slow and rigid in restoring high-quality in-person instruction. Still, there are reasonable concerns, especially when education savings accounts are used to fund microschools. I tend to agree with the sentiment voiced by former Hoover visiting fellow Mike Petrilli, who recently argued, “It’s nearly impossible to link à la carte [educational] services to whether students are actually learning. Some argue that accountability to parents is sufficient, but when taxpayer dollars are involved, that cannot be the only answer. Public education is a collective investment, because all of us benefit when all children—not just our own—are well educated and able to navigate adult life.”
How solid is the evidence that charter schools actually spur improvement among the traditional schools?
The short answer is that the evidence is reasonably solid, though more modest than advocates on either side sometimes suggest. As Ohio State University political scientist Vladimir Kogan explains in his excellent new book, No Adult Left Behind, the most consistent effects of choice schools are often not on the students who attend them, but on the traditional public schools that are forced to compete. A large body of research finds that when traditional public schools face credible competition, they tend to improve, at least in the short run. These competition effects are generally positive, but not dramatic. Schools respond to the risk of losing students by trying to retain families, sometimes by strengthening academic programs and sometimes by improving other features parents care about.
But evidence like this, by itself, has never been enough to settle the political fight over school choice. As Howard Fuller, one of the central figures behind the early voucher movement, has put it bluntly, opponents of choice aren’t waiting to be persuaded by your latest research study: “We’re going to have to fight these people. And you can’t do it with data. You can’t do it with graphs. You can’t do it with ‘oh my God, if y’all just understood how wonderful we are.’ They don’t care how wonderful you are—they’re coming at you because of politics.”
But the fact that evidence doesn’t settle the politics raises a natural question: does that mean there isn’t mounting evidence of competitive effects? Hardly. Research by Hoover-affiliated economists David Figlio and Krzysztof Karbownik, for example, finds that private school choice programs can generate positive effects on the achievement of nearby public school students. And ongoing work by Rochester economist and visiting Hoover fellow John D. Singleton extends this evidence to the charter school arena. Focusing on charter expansion in Texas, Singleton shows that public schools that are exposed to nearby charter entry experience meaningful gains in student achievement. Notably, this occurs before the charter schools actually open, suggesting that it is the competitive threat itself that drives improvement!
Speaking of incentives, what’s the best pitch to teachers—and their unions—that might overcome their misgivings about universal school choice?
Teachers’ unions are responding to straightforward incentives. Their leadership depends on maintaining membership, resources, and bargaining power. Because private and charter schools are less unionized, school choice is always going to make union leaders recoil.
Where things get more nuanced is when you move from the unions as organizations to individual educators. Many teachers strongly support their unions, but they also care deeply about their day-to-day professional environment. Research consistently shows that one of the most important factors for teacher satisfaction and retention is having a supportive principal. That often matters as much as, or more than, pay. Educators also care deeply about mission fit, whether that means a school’s instructional philosophy, its student population, or the broader goals it is trying to pursue.
School choice can appeal to teachers on these dimensions because it expands the range of working environments available to them. Some educators are drawn to magnet schools with a clear mission. Others prefer charter schools built around a particular instructional model or school culture that promotes discipline. Some even choose private schools, despite lower pay, because they are drawn to a school’s emphasis on character formation.
So, while unions as organizations are likely to remain skeptical, school choice can still resonate with many teachers. Recent polling underscores that individual educators are not reflexively anti-choice. For example, survey data from EdChoice shows that a majority of teachers who report being union members express support for education savings accounts. Likewise, the most recent EdNext poll that surveyed teachers (2021) found that roughly half supported school vouchers.
As we approach the 250th anniversary of the American founding, what’s the best way to make sure students everywhere participate in an accurate, uplifting study of their country’s values?
We can start by injecting some choice-like competition into the civics arena!
Beginning in 2030, states will have the option to participate in the NAEP Civics exam in a way that yields comparable state-level results (which have been available for math and reading since 2003). Congress could even “require” this by linking state participation on NAEP civics to federal K–12 funding. That could provide an incentive to address the fact that only a minority of states require dedicated civics instruction and even fewer administer high-quality assessments that make results both transparent and comparable. The limited data we do have already tell us something troubling: roughly one in three students fail to demonstrate even basic proficiency—a low bar that amounts to merely being able to describe the basic structure and function of American government.
Having all states participate on the NAEP civics exam would allow policymakers to see where they stand, learn from peers that are doing better, and apply pressure to improve civics instruction. During the No Child Left Behind era we saw this dynamic work in reading and math, where cross-state comparisons helped spur real policy change. There’s no reason civics should be exempt from that same logic.
Michael T. Hartney is the Bruni Family Fellow at the Hoover Institution and participates in Hoover’s Working Group on Civics and American Citizenship. He is an associate professor at Boston College and a non-resident senior fellow at the Manhattan Institute for Policy Research. His scholarly expertise is in American politics and public policy with a focus on state and local government, interest groups, and K–12 education policy.