How a School Voucher Works, From the State Check to the Classroom Door

In the folklore of the school-choice fight, a voucher involves a suitcase of cash. It does not. The reality is duller and, for that reason, far more persuasive. A voucher is a payment rule: a few paragraphs of state law, a form, a check, and a school that either takes the child or does not. There is no suitcase. There is paperwork, and paperwork is where the whole argument lives.

I want to walk the money from the front door of the statehouse to the front door of the classroom, because once you have seen it move, most of the scary talk falls away and the real questions stand up where you can see them. Then we will go back in time to find out where the idea came from, which is a lot older than the people shouting about it on television.

First, understand the money that already exists

Every state pays for public schools out of a stew of three ingredients. Local property taxes cover a large share in most places. The state adds its own money, usually through a formula that tries to even out the differences between rich and poor districts. And Washington contributes a thinner slice, much of it tied to particular programs such as aid for low-income schools or for students with disabilities. The exact mix differs from state to state, and your state education agency publishes the numbers for your district if you want the truth about yours.

Divide the district's total spending by the number of students and you get the figure everyone calls per-pupil spending. That number is the starting point for nearly every school choice debate, so keep it in your head. A voucher law answers one question about it: if a family chooses not to use the district school, how much of the public money set aside for educating that child goes with the child, and where does it go?

Notice how modest that question is. A voucher does not create new education spending out of thin air. It changes who decides which school receives a payment the state was already prepared to make for that child.

The voucher in six steps

Here is the mechanism, in order, as it works in a typical program. Your state's statute will vary on the details, which is why I say typical.

1. The legislature writes the rules. A voucher begins as a bill. The bill says who is eligible (all children, or low-income families, or children in failing schools, or children with disabilities), how much the voucher is worth, which schools may participate and what they must do in return. Most of the political fights happen here.
2. The state sets the amount. The law ties the voucher to something. It might be a flat dollar figure. It might be a percentage of the state's per-pupil aid, or of the district's per-pupil spending. It is usually capped at the school's actual tuition, so the state is never paying more than the school charges.
3. The family applies. A parent fills out an application with the state agency or a designated administrator, proves eligibility (usually residency and, in income-tested programs, household income), and names a participating school.
4. The school admits the child. This step is the one critics tend to skip. The voucher is not an admission ticket. The school still decides whom it can serve, subject to whatever the statute requires about lotteries, nondiscrimination and space. A voucher gives a family a payment. It does not give a family a seat.
5. The state pays. In many programs the payment goes to the school directly. In others the check is made out to the parent and the parent endorses it over to the school, a design choice that matters legally, as we will see in a moment. Either way the money moves in installments, typically tied to enrollment confirmation.
6. The state checks. Participating schools usually have to meet some conditions: a fiscal audit, background checks, a test or reporting requirement, sometimes a minimum operating history. The depth of this step is the single biggest difference from program to program, and the single most useful thing to read in your own state's statute.

That is the whole machine. Six steps, and not one of them requires a suitcase.

A worked example, clearly hypothetical

Numbers help, so let us invent a state. Call it Prairie State, and understand that every figure below is made up for illustration. Do not go looking for it on a map.

Prairie State's average district spends $11,000 per pupil in total. Of that, say, $5,000 comes from the state, $5,000 from local property taxes and $1,000 from federal programs. The legislature passes a voucher for low-income families worth $7,000 a year or the school's tuition, whichever is lower.

Two families use it.

• The first family picks a neighborhood school that charges $6,400 in tuition. The voucher pays $6,400, because that is less than $7,000, and the family owes nothing, assuming the program requires the school to accept the voucher as payment in full.
• The second family picks a school that charges $9,500. The voucher pays the $7,000 cap. The family covers the other $2,500, since $9,500 minus $7,000 is $2,500. Spread over ten monthly installments, that is $250 a month, which the family has to find or raise from a scholarship fund or relatives.

Now look at the proportions. The voucher of $7,000 is about 64 percent of the district's $11,000 (7,000 divided by 11,000 is 0.636). The first family's school is operating on 58 percent of what the district spends (6,400 divided by 11,000 is 0.58). Whether the child does better or worse for that is an empirical question, and we will come to the honest evidence in a minute. But it tells you something that the typical voucher is worth substantially less than the typical per-pupil figure, and that is on purpose. The sticker price of a district school is not what a family gets to carry out the door.

Later in this post I will raise the objection that the money leaving the district hurts the children who stay. There is a full treatment of that argument, with arithmetic, in Vouchers Drain the Public Schools, and the Answer Is in the Arithmetic, and I would encourage you to read it before you argue with anybody.

Why some programs write the check to the parent

There is a reason you will hear about checks endorsed by parents. For decades the legal fear was the Establishment Clause. If the state wrote a check straight to a religious school, a lawyer could argue the state was funding religion. If the state wrote a check to the family, and the family decided where to take it, the argument got harder. The legal question became whether the program was a true private choice, a program neutral toward religion in which the money reaches a religious school only because individual families picked it.

The Supreme Court answered that question in Zelman v. Simmons-Harris in 2002, in a ruling about Cleveland that I will get to below. For a full run of the cases that shaped the law, from the 1940s through the 2020s, read From Everson to Carson, the Court Cases That Opened the Door to Choice. The point for now is that the payment design in step 5 is not a quirk. It is a choice made by lawyers with a courtroom in mind.

Where the idea came from: Milton Friedman's 1955 essay

If you want the intellectual birth certificate, it is an essay by the economist Milton Friedman called "The Role of Government in Education," published in 1955. Friedman made a distinction that is still the cleanest way to think about the whole subject. Paying for something and running something are two separate jobs. A government might have good reasons to pay for the schooling of every child, he argued, since a stable democracy depends on an educated citizenry. That does not mean the government has to own and operate the schools.

His proposal was to have the government give parents vouchers, redeemable for a specified sum per child per year at an approved school of the parents' choice, with government's role limited to setting minimum standards. Parents could add their own money to go to a costlier school. Schools, public and private, would compete for the children. He pointed to the GI Bill as a working model: the government paid for veterans' college education, the veteran chose the school, and nobody in Washington ran the lecture halls.

You can agree or disagree with every sentence. But notice the structure of the argument. It is not "public schools are bad." It is "the thing we want is educated children, the thing the state does badly is monopoly, and the remedy is to separate the financing from the delivery." That is why the idea has found friends in places Friedman never visited. If your complaint is with a government monopoly, you do not need to hate teachers to want an exit.

Friedman, incidentally, was not the first to notice the principle. Some towns in northern New England had already been paying tuition for children to attend schools elsewhere for generations, because they were too small to run a high school of their own. Maine's tuition arrangement dates back to the 1870s. The family in a tiny town picked the school, the town paid the bill, and nobody called it a revolution. It was simply how a small town educated its teenagers.

Milwaukee, 1990: the first modern program

The idea sat on the academic shelf for about three and a half decades. Then Wisconsin's legislature acted. The Milwaukee Parental Choice Program was enacted in 1989 and began operating in the 1990 school year, and it was the first modern publicly funded voucher program aimed at the children of a big city.

Two details deserve attention because they cut against the way the story gets told. First, the program began small and tightly targeted. It was limited to low-income Milwaukee families and to a modest number of participating students, and at the beginning it was open only to nonreligious private schools. Second, its leading champion was not a Republican think-tank boffin but a Democratic state representative from Milwaukee, Annie Polly Williams, who worked with Republican Governor Tommy Thompson to get it through. She was frustrated by how the city's schools were serving the children of her inner-city district and decided that the parents closest to the problem should have an exit.

That is worth remembering the next time somebody tells you school choice is a rich man's scheme. In its first modern incarnation it was a low-income program, pushed by a Democratic legislator from the inner city. Try fitting that into the cartoon.

In 1995 Wisconsin expanded the program to allow religious schools to participate, and the state's supreme court upheld the expanded program in Jackson v. Benson in 1998. That decision was an important proof of concept and pointed the way to the national ruling that followed.

Cleveland and the national test

Ohio launched its own program in 1995, the Cleveland Scholarship and Tutoring Program, aimed at a school district that was in such trouble that a federal judge had placed it under state control. Families could use the aid to attend participating private schools, and the aid was sent to the school selected by the parents, with the parents endorsing the check. The program also offered tutoring aid for children who stayed in public schools.

Taxpayers sued, arguing that the program violated the First Amendment because most of the private schools that took part were religious. In 2002, in Zelman v. Simmons-Harris, the Supreme Court ruled 5 to 4 that the program was constitutional. Chief Justice Rehnquist wrote for the majority that the program was neutral toward religion and offered genuine choice, since families could choose among religious private schools, nonreligious private schools, charter schools, magnet schools or the public school with tutoring. The fact that most participants ended up in religious schools did not change the analysis, because the state did not steer them there. The families chose.

That ruling is the legal foundation under almost every voucher program in America today. It did not require any state to adopt a program. It said only that the federal Constitution does not forbid one. What happens after that is up to the states, and the states have their own constitutions, a subject I take up in The Blaine Amendments and the Old Bigotry Still Sitting in State Constitutions.

The strongest objections, taken one at a time

I promised fairness, so here are the three best arguments against the voucher as it is usually designed, and what I think of them.

One: the voucher is too small to cover tuition at many schools. True, and often by design, as the Prairie State example showed. The answer is that the voucher does not have to buy the best school in town to be an improvement. It has to buy a better fit for the child than the one the family was assigned by their street address. For many families, a school at $6,400 is perfectly reachable. And many private schools price their tuition at what the market will bear because they have never before had a customer base with a public subsidy attached.

Two: private schools may not be accountable. This is the objection with the most meat on it, and the honest response is that it depends on the statute. Some programs require state testing and publish the results. Some require only a basic financial audit. Some require accreditation. If you want accountability, write it into the law, and demand it from your legislator. I would argue the most powerful accountability in the world is a parent who can take the child and the money elsewhere. But a good law does not rely on only one kind.

Three: it takes money from public schools. This is the argument I think deserves the most careful answer, because the answer is not "no." The honest answer is: it depends on how the funding formula treats a departing child, how much of the district's spending is fixed in the short run, and whether the program also draws in children who were already in private school. Those are real variables. They are covered, with numbers, in the companion post I mentioned above. What I would say here is that the debate should be about the formula, not about slogans.

What to look for in your own state

Before you form an opinion about vouchers, spend twenty minutes with the statute. Look up your state legislature's website, search the education title for the words "scholarship," "voucher," or "education savings account," and read four things: who is eligible, how much is the benefit, what must a participating school do, and who audits the money. If your state has no program, read the bills that have been introduced. The text will tell you more than a month of cable news.

If you want to see how this mechanism compares with the other tools in the school choice family, such as education savings accounts and tax-credit scholarships, my comparison table in Vouchers, ESAs, Tax Credits and Charters Sorted Out in One Table lays them out side by side. And if you are curious why some arguments about choice seem to float away from the facts, Seven School Choice Myths Your Neighbor Believes and What the Record Shows takes the popular ones apart.

The bottom line

• A voucher is a payment rule, not a mystery: a statute sets eligibility and the amount, the family picks a school, the school admits the child, and the state pays and audits. Read your state's version before you argue about it.
• The idea is far older than today's fights: Milton Friedman laid out the principle in 1955, New England towns had paid tuition for generations, and Milwaukee turned it into law in 1990 with a Democrat as its champion.
• The Supreme Court settled the federal question in Zelman in 2002. What remains is state law, state constitutions and the design of the formula, which means your state legislator is the person to call.
• Ask your representative three things: who is eligible, what must a participating school report, and who checks the books. A program that cannot answer those is not ready.
• A family that can choose a school is not a threat to the neighborhood. It is the neighborhood, doing what neighborhoods are supposed to do, which is decide for themselves.