Milton Friedman's 1955 Voucher Essay and Its Seventy-Year Road

Imagine a country that decides food is so important that the government must feed everyone, and so it builds a government grocery store in every neighborhood, assigns each family to the one nearest its home, and makes the stores free. Now imagine someone suggesting, quietly, that families might get a food card and pick their own grocer. The reply writes itself: "You want to destroy the grocery system!" No. He wants to keep the food and fire the monopoly.

That, in a paragraph, is the essay an economist at the University of Chicago published in 1955. It is a short piece with a long shadow, and it deserves a careful reading, because the school choice movement is routinely described as a recent scheme cooked up by lobbyists. It is nothing of the kind. It is a seventy-year-old argument with a clear author, a clear logic and a record you can check.

The claim in one sentence

Milton Friedman's essay, "The Role of Government in Education," appeared in 1955 in a volume called Economics and the Public Interest, edited by Robert A. Solo and published by Rutgers University Press. He reworked it as a chapter of Capitalism and Freedom in 1962, which is how most readers have met it since.

The claim is this: paying for schooling and running schools are two different jobs, and nothing in the first requires the government to do the second. Governments had fused the two. They collected taxes, built the buildings, hired the staff, set the calendar and assigned the children. Friedman asked why a society that wanted every child educated needed to do any of that rather than simply fund the child.

Let us define the terms the way he did, because the argument turns on them.

• Financing means that government pays, from general revenue, for a minimum level of schooling.
• Administration means that government employees operate the schools.
• A voucher is the device that separates the two: a government-funded certificate of a specified value that the parent may spend on schooling at any approved school, public or private.

His proposal, in plain words, was that government would require a minimum level of schooling, would pay for it by giving parents a voucher redeemable for a stated maximum sum per child per year at approved institutions, and would otherwise step back. Parents could add their own money. Approved schools could be run for profit or by nonprofit institutions. The state's role would resemble what it does when it inspects restaurants: set a floor, check the floor, and leave the cooking to the cooks.

Why he thought government should pay at all

Here is where the usual caricature collapses. Friedman was not an abolitionist about public funding of schooling. He argued for it, and he rested the argument on what he called neighborhood effects.

The idea is old and respectable in economics. Some transactions affect people who are not party to them. If you educate your child, the benefit is not only yours. A stable and democratic society, he wrote, is impossible without widespread acceptance of some common set of values and without a minimum degree of literacy and knowledge. Education contributes to both. So the child's schooling pays a dividend to strangers, and strangers should therefore share the cost.

Notice what that argument justifies and what it does not.

1. It justifies public financing of general education, the reading, arithmetic, civics and common knowledge on which a free republic depends.
2. It does not justify government operation of the schools. A dividend paid to neighbors tells you who should chip in. It says nothing about who should own the building.
3. It does not justify subsidizing specialized vocational training. Friedman treated that as an investment in a person's own earning power, one the person captures personally, which makes it closer to any other investment in human capital than to a public good.

That last distinction is worth a moment. Many of the people who invoke Friedman today as a patron saint of "education spending" have never noticed that he drew a line around it.

The steelman: what the defenders of the district school say

A fair post states the opposing case as its defenders would. The case for the traditional district school runs like this.

First, the common school is the engine of a common citizenship. Children of different backgrounds sit in the same rooms and learn to live as fellow citizens. Second, public accountability matters: a school run by elected boards answers to voters, while a private school answers to a customer base and a board of its own. Third, choice can be a polite word for sorting, as families with information and means move first. Fourth, a voucher may drain money from the schools that remain, leaving the children who stay behind with less.

Those are serious worries, and they were raised in 1955 as they are now. The honest answers, in order:

1. Common citizenship is a result, not a building. Friedman's own neighborhood-effects argument is the case for the common core of knowledge and values. His proposal kept a minimum content requirement for approved schools for exactly that reason. The question is whether a single government-run institution is the only way to deliver it, and the historical record of private and religious schooling in America says it is not. The Founding generation did not think of schooling as a government monopoly at all; the Northwest Ordinance of 1787 said that "Religion, morality, and knowledge, being necessary to good government and the happiness of mankind, schools and the means of education shall forever be encouraged," and the verb there is encouraged, not operated. For more on that clause see Religion, Morality and Knowledge, the Northwest Ordinance's School Clause.
2. Accountability cuts both ways. A family that can leave holds a power no school-board election can match, because it works on Monday morning rather than every second November. Friedman's point is that exit is a form of accountability, and one that does not need a majority.
3. Sorting already exists, and it is done by real estate. Under assignment by address, the price of a house is the price of admission to a school. A voucher sorts nobody by income that the housing market has not already sorted. Your ZIP Code Picks Your Child's School, the Case for Open Enrollment makes that argument in detail.
4. The money follows the child, and so does the obligation. The remaining objection, stranded costs, is a real transitional problem. It is a problem of arithmetic and budgeting, not of principle, and it is the standard objection to ending any monopoly.

The road: from an essay to a movement

An essay does not change a country by itself. It needs people who take it up. Here is the road, with the dates I am sure of.

1955. The essay appears.

1962. Capitalism and Freedom carries it, as chapter VI, to a general audience.

1980. Milton Friedman and his wife and collaborator, Rose Friedman, publish Free to Choose, the book that accompanied their television series, with a chapter on schools that returns to the argument.

1990. Wisconsin launches the Milwaukee Parental Choice Program, the first modern, publicly funded voucher program for low-income children in a major American city. It was driven locally by a Democratic state legislator from Milwaukee, Polly Williams, whose story is told in Polly Williams, the Milwaukee Democrat Who Started the Voucher Movement. The point matters: the first modern voucher program was a project of poor families in an American city, not a project of an economics seminar.

1996. The Friedmans create a foundation to promote school choice, the Milton and Rose D. Friedman Foundation. It later changed its name to EdChoice.

2002. In Zelman v. Simmons-Harris, the Supreme Court held that Ohio's Cleveland voucher program did not violate the Establishment Clause, because aid reached religious schools only through the independent choices of parents. That ruling removed the main federal constitutional obstacle.

2011. Arizona creates the first education savings account program, a design that goes further than a voucher.

Seventy years from the essay, the idea runs through vouchers, tax-credit scholarships (see Tax-Credit Scholarships, the Choice Program Built on Private Donations), charter schools in a different form, and the education savings accounts discussed in Arizona Made Every Child Eligible for an ESA, How the Idea Spread So Fast. Each is a variation on the one distinction he drew: the state may pay without owning.

Where today's accounts go beyond what he proposed

Honesty requires saying what Friedman did not propose, because a conservative movement that borrows an authority should not stretch him.

A voucher buys schooling. An education savings account buys an education. Under the 1955 proposal, the certificate was redeemable at an approved school. Under an ESA, the parent receives a deposit in a restricted account and spends it on a menu: tuition, tutors, curriculum, therapies, testing fees, sometimes the costs of homeschooling. That is a larger step toward the parent as the buyer of services, and Friedman never wrote that particular design.

He wanted a floor of content, and many programs have a thinner one. His proposal had the government require a minimum amount of common content in approved schools, to be verified the way a restaurant is inspected. Some programs today require annual testing or nationally norm-referenced exams; some require almost nothing. Friedman's neighborhood-effects argument is exactly what a legislator should read before deciding how thin the floor can be.

He did not decide the question of regulation and strings. Anyone who has watched private schools accept public money, and then public rules, understands the worry that a voucher can become a leash. That concern, that the check comes with a clipboard, is a legitimate conservative objection to a program of any design, and a state that adopts choice should write its guardrails with that in mind.

He was more radical than most programs. Friedman, in the end, wanted a system in which most schools were private and competing. The programs that exist are, by his lights, modest: a limited number of seats, an income cap in many states, a dollar value below per-pupil spending in the district school. If you think vouchers are a revolution, the man who invented the idea would have called them a rounding error.

The numbered argument

So let me do what Friedman would have done and set it out in premises.

1. A free society needs an educated citizenry, and the benefit of an educated child reaches beyond the family. That justifies public financing.
2. Public financing does not entail public operation. The two functions are separable, and the Founding-era record, including the Northwest Ordinance, treats the state's role as encouraging schools, not owning them.
3. A monopoly provider, public or private, serves its producers before its customers. That is not an insult to teachers, who are good people working inside a structure, and many of whom would flourish in a competitive one. It is an observation about incentives.
4. When families may leave, schools must keep them. When families cannot leave without paying twice, the only families who can leave are the wealthy.
5. Therefore the fair way to protect the poor family's chance is to put the money in the family's hands.

I think that argument survives contact with the evidence of seven decades. It does not require you to believe that every private school is good, or that every public school is failing. It requires you to believe that the family, not the district, is the right unit of choice.

What the left should have to answer

I will give the other side one more courtesy, which is a question worth asking in return.

If the justification for a government-run system is that it equalizes opportunity, why does the wealthiest family in any metro area already have a choice of schools, by buying a house or a tuition contract, while the poorest has none? If the answer is "because the district school is good enough for everyone," then the district school should have no fear of competition. If the answer is "because choice drains the district," then the district school is admitting that families would choose to leave.

A system that is confident of its quality does not need to lock the door. Notice, too, the unspoken premise in so much of the coverage: that a child assigned to a school is the system's child, and the money assigned to the child is the system's money. The text of the 1955 essay turns that premise upside down. The child belongs to the family, and the money, held in trust, follows.

What the Founders would have recognized

The Founders did not have school districts as we do. They had a mix of town schools, church schools, tutors, academies and parents at the kitchen table, and they trusted that mix. Their documents speak of encouraging schools and the means of education, and of the diffusion of knowledge. Nothing in the Constitution's text mentions education at all, which is why the Tenth Amendment reserves it to the states and to the people.

What Friedman added in 1955 was a mechanism for a mass society to do what the early republic did by habit: let the people who love the child choose the teacher. That is not a break with the American tradition. It is a restoration, carried out with the tools of a twentieth-century economist.

The case in brief

1. Milton Friedman's 1955 essay separated financing schooling from operating schools, and that separation is the foundation of every modern choice program.
2. His case for public financing rested on neighborhood effects, and it stopped short of public operation and short of subsidizing vocational training.
3. The standard objections, common citizenship, accountability, sorting and stranded costs, are real and have answers, and the sorting that already exists runs through real estate.
4. Today's education savings accounts go beyond what he proposed, and conservatives should say so honestly while defending the principle.
5. The family, not the district, is the right unit of choice.

What you can do: look up your own state's choice law on your legislature's website, read who is eligible and what the program requires of approved schools, and tell your state representative in a short letter which guardrails you think it should have.