School Vouchers From Milton Friedman to Arizona's Universal Plan

In 1955 a University of Chicago economist named Milton Friedman contributed an essay called "The Role of Government in Education" to an academic collection on economics and the public interest. It was not a campaign document. Friedman was not yet famous, and the essay was the kind of thing read mostly by other economists. His argument was tidy: society has a good reason to pay for children's schooling, because an educated citizenry benefits everyone, but no particular reason to run the schools. Give parents a voucher worth a set amount per child, let them spend it at any approved school, and competition would do the rest. Friedman added a footnote of sorts that history would not let him escape. He acknowledged that in the South, where states were at that moment searching for ways around the Supreme Court's ruling in Brown v. Board of Education, vouchers could be used to fund segregated private schools, and he said he found that prospect objectionable while still preferring his plan to forced segregation or forced integration. The idea and its first real-world use arrived in the same decade, and they did not look alike.

In short: Vouchers began as a market theory, were first used at scale to fund white flight from desegregation, were reborn in Milwaukee as an anti-poverty tool, and have now become universal programs open to wealthy families too. Arizona's 2022 law is the clearest test of that last version. The argument for choice deserves a hearing, but universal programs mostly pay families for choices they were already making, and the cost lands on public budgets.

The first vouchers were for leaving integrated schools

The earliest large-scale use of public money for private tuition in the United States came not from economists but from segregationists. After Brown, Virginia's leaders adopted a policy they called massive resistance, and part of it was "tuition grants" that let white families take state money to private academies. In Prince Edward County, officials closed the public schools entirely in 1959 rather than integrate them and kept them closed until 1964. White students went to a private academy supported by tuition grants and private donations. Most Black students in the county simply went without formal schooling for years, unless their families sent them away to relatives or found makeshift classes.

Other Southern states passed similar grant programs. Federal courts struck many of them down during the 1960s as transparent attempts to preserve segregation. That history is why many Black civil rights leaders treated the word "voucher" with suspicion for a generation, and why the debate over public money for private schools still carries a weight in some communities that people elsewhere don't always understand.

Milwaukee, and a Democrat named Polly Williams

The idea came back in a form Friedman might have recognized, sponsored by a politician his admirers did not expect. In 1990 Wisconsin created the Milwaukee Parental Choice Program, the first modern voucher program in the country, limited to low-income families in the city. Its champion in the legislature was Annette "Polly" Williams, a Black Democratic state representative from Milwaukee's north side, a former welfare recipient and a onetime state chair of Jesse Jackson's presidential campaign. She worked with Tommy Thompson, the Republican governor, to pass it.

Williams' case was not about markets. It was about power. The families she represented, she argued, had no exit from schools that were failing their children, while richer families could move to the suburbs or pay tuition. A voucher was a way to give poor parents the leverage that money already gave everyone else. She also insisted that the program stay limited to poor families, and later broke with conservative allies who wanted to expand it to everyone.

Cleveland followed with its own program, and in 2002 the Supreme Court decided Zelman v. Simmons-Harris, upholding Cleveland's vouchers by a vote of five to four even though most participating students attended religious schools. Because the money reached the schools through parents' independent choices, the majority held, the program did not violate the Establishment Clause. That decision cleared the constitutional path that the voucher movement has walked ever since.

Friedman's design, and what got left out

It is worth going back to the 1955 essay to see what Friedman actually proposed, because both his admirers and his critics tend to compress it.

Friedman separated two questions that most people run together: who should pay for schooling, and who should provide it. On the first, he was not a libertarian purist. He accepted that a stable democracy requires a minimum of shared literacy and knowledge, and that this benefits everyone, not just the child being educated. That "neighborhood effect," in his phrase, justified public funding. On the second question, he saw no reason the government should own and run the schools any more than it owns and runs grocery stores. A voucher would separate the two: public money, private provision, parental choice.

He also acknowledged that his plan had problems. He worried about rural areas where there might be only one school, a natural monopoly where choice would mean little. He accepted that the government could set minimum standards for schools that took vouchers, like the standards it set for restaurants. And, as noted, he confronted the segregation question in the setting of 1955, conceding that his plan could be used to preserve segregation, which he said he deplored.

What the essay barely addressed is the question that dominates today's debate: what happens to the public schools that remain, and to the children who stay in them. Friedman imagined a market in which schools competed on equal terms. He did not spend much time on the fixed costs that districts carry, on children with disabilities whom private schools might not serve, or on families who lacked the time, information or transportation to make the market work. Seventy years of experience have filled in what the theory left blank.

Arizona builds an account for everyone

Arizona has long been a laboratory for school choice, with tax-credit scholarship programs dating to the late 1990s and one of the country's largest charter sectors. In 2011 the state created Empowerment Scholarship Accounts, a newer design than the classic voucher. Instead of a coupon redeemable at a school, families receive a portion of the state's per-student funding in an account they can spend on private tuition, tutoring, therapies, curriculum, and other approved expenses. At first the accounts were for students with disabilities. Over the next decade lawmakers added other groups: children in low-rated schools, children of military families, children living on tribal land.

In 2022 the legislature removed the eligibility limits entirely, and Governor Doug Ducey signed the law that made every school-age child in Arizona eligible, regardless of income. Supporters called it the most expansive choice program in the country. Opponents, led by a grassroots group called Save Our Schools Arizona, tried to put the expansion to a public referendum; the petition drive fell short of the required valid signatures, and the law took effect.

What happened next is the part worth studying. Applications poured in, and a large share came from families whose children were already enrolled in private school or being homeschooled, meaning the state was now paying for education it had not been paying for before. Program costs climbed well past the legislature's early projections and became one of the most contested lines in the state budget. Reporting by Arizona news outlets documented purchases that raised eyebrows and a review system that struggled to keep up. Katie Hobbs, a Democrat elected governor in 2022, called for limits and more oversight; the Republican-led legislature refused to roll the program back.

Since then, other states have gone universal or nearly so, including Iowa, Utah and Florida in 2023 and Texas in 2025. The question Arizona answered first, what happens to a state budget when a choice program open to everyone meets families already paying private tuition, is now a national one.

What changed between Milwaukee and Phoenix

Put the two programs side by side and the shift in the movement becomes plain. Milwaukee's program began small, for poor families in one city, with caps on enrollment and a sponsor who saw it as a remedy for a specific injustice. Arizona's account program is open to every child in the state, with no income test, and pays for a wide menu of expenses beyond tuition.

That change was deliberate. Over the 2010s, many choice advocates concluded that means-tested programs were politically fragile: they served a constituency with little clout, and opponents could cut them in a bad budget year. Universal eligibility, they argued, would create a broad base of middle-class families who would defend the program the way retirees defend Social Security. The education savings account, which lets families buy services rather than only a school seat, fit a broader vision of education as a set of purchases assembled by each family.

The strategy has worked politically in a number of states. Whether it serves the poor children in whose name vouchers were first sold is a different question, and it is the one the rest of this post tries to answer.

The best case for universal choice

Choice supporters make an argument that deserves to be taken seriously. Parents, they say, should be able to find the school that fits their child: a smaller school for an anxious kid, a program built around a disability, a religious school that shares the family's values. Wealthy families have always had that freedom by buying a house in the right district or paying tuition. Means-tested programs, in this view, stigmatize and invite political attacks; universal programs build a broad coalition that protects the policy. And public school systems that face real competition, they argue, will finally have reason to improve.

Some of that lands. A child with autism whose family found a specialized program through an education account is not an abstraction, and families who use these programs report high satisfaction. Districts that have had to compete for enrollment sometimes have responded with new programs.

But the argument for universal accounts runs into three problems that the Arizona experience makes concrete.

It is a transfer to people who need it least. When a significant share of new money goes to families who were already paying private tuition, the policy works as a subsidy for an existing choice. Families with more income are more likely to be in a position to use these accounts, because private tuition often exceeds the account value and someone has to cover the gap, along with transportation. A program sold as freeing poor children from bad schools ends up, in large part, reimbursing families who were never trapped.

Public schools' costs don't shrink as fast as their revenue. When a few students leave a school, the building still needs heat, the bus still runs its route, and the third-grade class still needs a teacher. Districts in rural areas, where there may be no private school within driving distance, gain nothing from choice and still share the budget squeeze.

Accountability is thin. Public schools must test students and publish results by group. Most universal account programs ask much less of the private schools and vendors they fund. Parents can judge for themselves, supporters reply, and that is a fair point about individual families. It is a weak answer when taxpayers want to know whether public money is buying learning.

The inequality question underneath

The deepest disagreement is about what public education is for. If schooling is a private good, a service a family buys for its own child, then a voucher is just a more efficient way to buy it. If schooling is also a public good, the place where children of different backgrounds learn alongside one another and where a community invests in its common future, then a policy that breaks it into individual accounts loses something no market can replace.

Progressives do not have to pretend that public schools are serving every child well to make that case. Many aren't. The honest progressive answer is that the solution is to fund and fix the schools most children will still attend, not to spend scarce dollars subsidizing a minority who leave, many of whom would have left anyway. There is also a fair self-criticism to make: when public systems ignore parents who are desperate for something different, those parents become the strongest voices for vouchers. Polly Williams was a Democrat. Her constituents were not wrong to be angry.

What you can do

Voucher and education account programs are written and funded in state legislatures, often in the budget, where they are hardest to see.

• Ask your state legislators whether your state has a voucher, account or tax-credit scholarship program, what it costs this year, and how much of that goes to students who were already in private school. Their contact information and latest votes are at the representatives lookup.
• Follow reporting on choice programs, school funding fights and the federal role on the education topic page.
• Read about who pays and who benefits from tax and spending choices like these on the topic page on inequality and taxing the rich.
• Find parent coalitions and public education advocates in your state, including groups that organize around budgets, in the organizations directory.
• Several voucher fights have been decided at the ballot box. Make sure you are registered and know your state's deadlines with the voting guide.