The Friedman Foundation for Educational Choice

Advancing Milton and Rose D. Friedman's Vision of School Choice for All Children

Selling School like Groceries: The Voucher Idea

The New York Times Magazine | September 23, 1975

A fable may dramatize the true source of the nation’s present discontent with our public schools: Suppose that, 50 or 75 years ago the U.S. had adopted the same institutional arrangements for the distribution of food as it did adopt for elementary and secondary schools. Suppose, that is, that the retail provision of groceries had been nationalized, that food was paid for by taxes and distributed by government-run stores. Each family would be assigned to a store, as it is now assigned to a school, on the basis of its location. It would be entitled to receive, without direct payment, a collection of foods, as its children are entitled to receive a collection of classes. It would be able to choose among foods, as its children choose among subjects. Presumably, this would be done by giving each family some number of ration points and assigning point prices to various foods. Private grocery stores would be permitted (just as private schools are), but persons shopping in them would be taxed for the support of the public stores just the same.

Can there be any doubt what retail food distribution would be like today if this system had been in effect? Would there be supermarkets and chain stores? Would the shelves be loaded with new and improved convenience products? Would stores be using every device of human ingenuity to attract and retain customers?

Suppose that under such a system you were unhappy with your local grocery. You could not simply go to a different store unless you were able and willing to pay twice for your groceries, once in taxes and again in cash. No, you would have to work through political channels to change the elected or appointed Grocery Board, or the Mayor, or the Governor, or the President. Obviously this would be a cumbrous, inefficient process. And suppose you had different ideas from your neighbors about the kind of service you wanted? What then? You would have to find a neighborhood of likeminded people to which you could move.

Of course the well-to-do would escape all this by patronizing the few luxury establishments that would arise to cater to them. They would willingly pay twice for their food, just as they now pay twice for the schooling of their children. (I do not blame them. It is right and proper that parents should deny themselves in order to purchase the best products they can for their children’s bodies or minds. I blame only those well-meaning persons who, while sending their own children to private schools, self-righteously lecture the “lower classes” about their responsibility to put up with government-supplied pabulum in the “public interest”).

Consider the producer rather than the consumer. The supermarket is a modern invention that has contributed enormously to the well-being of the masses. What would the inventor have done -- if he had existed at all -- in the hypothetical world of government grocery stores? In the actual world, all he had to do to try out his idea was to use his own capital, or persuade a few people to venture some capital, and set up shop. In the hypothetical world, he would have to launch a successful political campaign to persuade a local grocery board, an entrenched civil service and harried legislators that his idea was worth trying. Obviously, innovations would come primarily from the few private stores serving the well-to-do.

Ask yourself what activities in the U.S. have participated least in the technological revolution of the past century -- is there any doubt that schooling, mail service and legislative activity heed the list? Grocery stores would be on the list, too, if they had been government-owned.

Schooling is not groceries. Yet the many and important differences do not invalidate the comparison. The delivery of mails is not the same as the delivery of schooling, yet both are inefficient and technologically backward for the same reason: They are conducted mostly by government agencies enjoying an effective monopoly. The delivery of groceries is not the same as the production of hi-fi equipment. Yet both are highly efficient and technologically progressive for the same reason: They are conducted mostly by private enterprises operating in a competitive market.

The same contrast applies to schooling itself. As Adam Smith wrote nearly 200 years ago, “Those parts of education, it is to be observed, for the teaching of which there are no public institutions, are generally the best taught.” That is equally true today: Music or dance, secretarial skills, automobile driving, airplane piloting, technical skills, all are taught best when they are taught privately. Try talking French with someone who studied it in public school. Then with a Berlitz graduate.

As the fable suggests, the way to achieve real reform in schooling is to give competition and free enterprise greater scope, to make available to children of low- and middle-income parents, particularly those living in slums, a range of choice in schooling comparable to that which the children of upper-income parents have long enjoyed. How can the market be used to organize schooling more effectively? The most radical answer is to put schooling precisely on a par with food: Eliminate compulsory schooling, government operation of and government financing of schools except for financial assistance to the indigent. The market would then have full rein.

A slightly less radical answer is to put schooling on a par with smallpox vaccination or getting a driver’s license: Require every person to have at least a minimum level of schooling as he or she is required to be vaccinated or to take a driving test before he or she is granted a license -- but let the schooling be obtained privately and at the parents expense again, except for the indigent.

These solutions have much to be said for them as ways to further both freedom and equal opportunity. However, they are clearly outside the range of political feasibility today. Accordingly, I should discuss a more modest reform -- one that would retain compulsory schooling, government financing and government operation, while preparing the way for the gradual replacement of public schools by private schools.

The City of New York now spends about $1,500 per year for every child enrolled at public elementary and secondary schools. Parents who send their child to a private school therefore save the City about $1,500. But they get no benefit from doing so. The key reform would be for the City to give such parents a voucher for $1,500 to pay for schooling their child (and for no other purpose). This would not relieve them of the burden of taxes; it would simply give parents a choice of the form in which they take the schooling that the city has obligated itself to provide.

To widen still further the range of choice, parents could be permitted to use the vouchers not only in private schools but also in other public schools; and not only in schools in their own district, city or state, but in any school anywhere that is willing to accept their child. This would involve giving every parent a voucher and requiring or permitting public schools to finance themselves by charging tuition. The public schools would then have to compete both with one another and with private schools.

Today the only widely available alternative to a local public school is a parochial school. The reason is that only churches have been in a position to subsidize schooling on a large scale and only subsidized schooling can compete with “free” schooling. (Try selling a product that someone else is giving away!) The voucher plan would produce a much wider range of alternatives. In the first place, choice among public schools themselves would be enormously increased. The size of a public school would be set by the number of customers it attracted, not by politically defined geographical boundaries. Parents who organized nonprofit schools, as a few pioneers have, would be assured of funds to pay the costs. Voluntary organizations -- ranging from vegetarians to Boy Scouts to the Y.M.C.A. -- could set up schools and try to attract customers. And, most importantly new sorts of private schools would arise to tap the vast new market -- perhaps Mom and Pop schools like Mom-and-Pop grocery stores, perhaps also highly capitalized chain schools, like supermarkets.

But why require parents to spend the voucher in a single school? Why not offer divisible vouchers? Let part be spent for the core school and the rest for mathematics lessons, music lessons or vocational training purchased from another source. One does not buy all one’s groceries at a single store. Why should one buy all of a child’s schooling at a single school?

Let us examine in detail some problems with the voucher plan and some objections that have been raised to it:

(1) The church-state issue.

Parents could use their vouchers to pay tuition at parochial school and would this violate the First Amendment? Whether it does or not is it desirable to adopt a policy that might strengthen the role of religious institutions in schooling?

On June 25, 1973 the Supreme Court struck down by a 6-to-3 majority laws in New York and Pennsylvania that provided for reimbursing parents for part of the tuition paid to nonpublic elementary and secondary schools. The majority held that the laws “have the impermissible effect of advancing religion.” The minority (in three separate dissents) urged that “government aid to individuals generally stands on an entirely different footing from direct aid to religious institutions (Chief Justice Burger); that “preserving the secular functions of these schools is the overriding consequence of these laws” (Justice White); and, that the Court has failed to “distinguish between a new exercise of power within constitutional limits and an exercise of legislative power which transgresses those limits” (Justice Renquist).

To this non-lawyer, the tuition-reimbursement plans appear to be at least kissing cousins of the voucher plan that I have outlined. However, as I read the decisions, two differences might lead the Court to rule favorably on a full-fledged voucher plan: (1)The voucher plan would apply to all parents, not simply to those with children in nonpublic schools; and, (2) it would grant the same sum to all parents not, as in the particular tuition-reimbursement plans struck down, a sum much smaller than the per-pupil cost (a point that Justice Powell referred to explicitly in the majority decision).

Whatever the fate of a full-fledged voucher plan, it seems clear that the Court would accept a plan that excluded church-connected schools but applied to all other private and public schools. Such a restricted plan would be far superior to the present system and might not be inferior to a wholly unrestricted plan. Schools now connected with churches could qualify by subdividing themselves into two parts: a secular part reorganized as an independent school eligible for vouchers, and a religious part reorganized as an after-school or Sunday activity paid for directly by parents or church funds.

The constitutional issue will have to be settled by the courts. But it is worth emphasizing Justice Burger’s point that the vouchers would go to parents, not to schools. Under the G.I. Bill, veterans are free to attend Catholic or other religious colleges, and, so far as I know, no First-Amendment issue has ever been raised. Recipients of Social Security and welfare payments are free to contribute to churches from their government subsidies, with no First Amendment question being asked.

Indeed, I believe that the penalty now imposed on parents who do not send their children to public schools produces a real violation of the spirit of the First Amendment, whatever lawyers and judges may decide about the letter. The penalty abridges the religious freedom of parents who do not accept the liberal humanistic religion of the public schools, yet, because of the penalty, are impelled to send their children to public schools.

In practice, the voucher plan might well reduce the role of parochial schools by eliminating their privileged position as the only effective alternative to public schools for most people. In the first instance, parochial schools would benefit, but soon they would encounter far greater competition than they do today. However, I hasten to add that my advocacy of the plan in no way hinges on whether this conjecture is correct.

(2) Financial cost.

A second objection to the voucher plan is that it would raise the total cost of schooling -- because of the cost of paying for children who now go to parochial and other private schools. This is a “problem” only if one neglects the present discrimination against parents who send their children to nonpublic schools; universal vouchers would end the inequity of using tax funds to school some children but not others, and current laws impose the responsibility on the state to school all children, not just children now in public schools. Moreover, there are two off-setting considerations: First, growing financial difficulties are forcing many nonpublic schools to close, which also raises governmental costs. Second, under a voucher plan, parents who now send their children to non-public schools might be more favorable to higher public expenditures for schooling.

There would, however, be a real problem of finding enough money to begin with. One way to meet it is to make the amount of the voucher less than current expenditures per public-school child. Take present total expenditures on schooling, divide by the number of eligible children and let the resulting sum -- or some amount between that sum and present expenditures per public-school child -- be the amount of the voucher. If $1,500 is now spent per public-school child in New York, $1,300 spent in a competitive school would provide a far higher quality of schooling. Witness the drastically lower cost per child in parochial schools. The fact that elite, luxury schools charge high tuition is no counter-argument, any more than the $7.25 charged at the “21” Club for the 21 Burger means that McDonald’s cannot sell a hamburger profitably for 25 cents.

The net effect of a compromise plan, making the voucher somewhat less than the current average cost, would be: (a) to require public schools to economize somewhat; (b) to give parents of children who now attend a public school the alternatives of keeping them there or transferring them to any other school, public or private, at no financial cost if the school’s tuition is $1,300 or less, or at a cost equal to the excess of the school’s tuition over $1,300; (c) to enable parents of children who now attend a parochial school to keep them there if they wish, relieved of the tuition they have been paying (almost invariably less than $1,300), or to take advantage of any of the other alternatives available to parents of public school children; (d) to relieve parents of children now attending elite private schools of $1,300 of the annual cost of schooling their children, as a partial offset to the school tax payments they will continue to have to make.

(3) The possibility of fraud.

How can one assure that the voucher is spent for schooling not diverted to other family expenses? The answer is that the voucher would have to be spent in an approved school or teaching establishment. True, this does mean some government regulation of the schools, but of course private schools are regulated to an extent now, to assure that attendance at them satisfies compulsory schooling requirements. Compared to current regulation of public schools, the government requirements in a voucher plan would be a mere trifle.

A less obvious problem is the kind of schooling for which the vouchers may be used. The major justification for both compulsory schooling and government financing of schooling is the so-called “neighborhood effect” --i.e., the assertion that schooling benefits not only the children and their parents but also the rest of us by promoting a stable and democratic society. But do all kinds of schooling contribute to responsible citizenship? Where should the line be drawn?

The voucher plan does not create this problem. It simply makes it more visible. Much that is taught today in public schools cannot readily be justified as conferring benefits on the community at large. I regard it as a virtue of the voucher plan that it forces us to face this issue rather than evade it.

(4) The racial issue.

Voucher plans were adopted for a time by a number of Southern states as a device to avoid integration. They were ruled unconstitutional. Discrimination under such a plan can be easily prevented by permitting vouchers to be used only in schools that do not discriminate. However, a more difficult problem has troubled some students of vouchers. This is the possibility that a voucher plan might increase racial and class separation in schools, exacerbate racial conflict and foster an increasingly segregated and hierarchical society.

I believe that it would have precisely the opposite effect, that nothing could do more to moderate racial conflict and to promote a society in which black and white cooperate in joint objectives, while respecting each other’s separate rights and interests. Much objection to forced integration reflects not racism but more or less well-founded fear about the physical well-being of children and the quality of their schooling. Integration has been most successful, when it has been a matter of choice not coercion.

Violence of the kind that has been rising apace in public schools is possible only because the victims are compelled to attend the schools that they do. Give them effective freedom to choose and students -- black and white, poor and rich -- will desert in droves any school that cannot maintain order. Let schools specialize, as private schools would, and the pull of common interest will overcome the pull of color, leading, I believe to far more rapid integration than is now in process -- in fact, not on paper.

The voucher idea would completely eliminate the busing issue. Busing would occur, and might indeed be increased, but it would be wholly voluntary -- just as the busing of children to music and dance classes is today.

I have long been puzzled that black leaders have not been the most vigorous proponents of the voucher plan. Their constituents would benefit from it most; it would give them power over the schooling of their children, eliminate the domination of both the citywide politicians and, even more important, of the entrenched bureaucracy. Black leaders themselves frequently send their children to private schools. Why do they not help others to do the same? My tentative answer is that vouchers would also free the black man on the street from domination by his leaders, who correctly see that control over local schooling is a powerful political lever.*

(5) The economic-class issue.

The question that has perhaps divided students of vouchers more than any other is their likely effect on social and economic class structure. Some have argued that the great value of the public school has been as a melting pot in which rich and poor, native and foreign-born, black and white have learned to live together. This image had much validity for small communities and still does. But it has almost none for large cities and their suburbs. In them, the public school has fostered residential stratification, by tying the kind and the cost of schooling to residential location. Most of the country’s outstanding public schools are in high-income enclaves, Scarsdale or Lake Forest or Beverly Hills. Such schools are better regarded as private tax shelters than as public schools. If they were, in the strict sense, private, their cost would not be deductible in computing federal income tax. But the cost is deductible as local taxes because the high-cost and high-quality school is nominally public.

Elementary schools would probably still be largely local under a voucher plan. But even they might be less homogeneous than they are now, because of the indirect effect of the voucher plan in making residential areas more heterogeneous. And secondary schools almost surely would be less stratified. Schools defined by common interests, one stressing, say, the arts; another, the sciences; another, foreign languages, would perforce attract socially and economically more heterogeneous clienteles from a wide variety of residential areas.

One feature of the voucher plan that has aroused particular concern is the provision that parents may “add on” to the voucher; that is, if the voucher is for, say, $1,300, a parent could add another $500 to it and send his child to a school charging $1,800 tuition. Some fear that the result might be even wider differences in school expenditures per child than now exist, because low-income parents would not add to the amount of the voucher while middle-income and upper-income parents would supplement it extensively.

This possibility has particularly worried Christopher Jencks and his associates at the Center for the Study of Public Policy in Cambridge, Mass., who conducted an O.E.O.-financed study of vouchers. In their 1970 report. “Education Vouchers,” they assert, on the basis of the most casual empiricism, that “an unregulated market would redistribute resources away from the poor and toward the rich.” Being confirmed egalitarians, they respond by proposing first that the voucher be larger for children from low-income families than for others; second, that voucher schools be required (among other things) to “(a) accept a voucher as full payment of tuition; (b) accept any applicant so long as it had vacant places; and, (c) if it had more applicants than places, fill at least half of these places by picking applicants randomly.”

I have great sympathy for the proposed compensatory voucher. For a variety of reasons, costs of schooling are greater in slum areas than elsewhere, so vouchers of equal dollar amount would not purchase equal schooling. In addition, I share the motives of those who believe that taxpayers should be willing to help the children of the poor. Moreover, with few exceptions, governmental expenditures benefit disproportionately middle- and upper-income groups, so compensatory vouchers would only help to redress the balance As a realist, however, I believe that it is a mistake to recommend a compensatory voucher. In the first place, the political facts that account for the present bias in government spending will also pervert compensatory vouchers. In the second place, equal per-child vouchers, while falling short of the ideal, would be a great improvement over what now exists.

The proposed restrictions on voucher schools are not justified in principle from the point of view even of confirmed egalitarians like Jencks, let alone of persons like myself who regard freedom as the primary social goal, though welcoming as a desirable by-product the tendency for a free society to foster equality of both opportunity and outcome. Equality surely should refer to the whole of the income or wealth of families. Can the egalitarian say, it is all right for the well-to-do to spend any income that the tax collector leaves them on riotous living, but they must be penalized (by being denied a voucher) if they try to spend more than the publicly specified sum on schooling their children?

The very poor would benefit the most from the voucher plan. For the first time, poor parents would have a real opportunity to do something about their children’s schooling. Social reformers, and educational reformers in particular, often self-righteously take for granted that the poor have no interest in their children or no competence to choose for them. This, I believe, is a gratuitous insult. The poor have had limited opportunity to choose. But U.S. history has amply demonstrated that given the opportunity, they will often sacrifice greatly, disinterestedly and wisely for their children’s welfare. Even the poorest are capable of scraping up a few dollars to improve the quality of their children’s schooling, although they could not replace the whole of the present cost of public schooling.

In the middle classes, there is now much private expenditure on schooling, but most of it goes for music, dancing, golf and similar skills that supplement the public school. The voucher plan would enable people to spend their own money as well as the voucher on what they regard as most valuable.

The net effect, I believe, would be a larger total sum spent on schooling, with that auto, if anything, more evenly divided. But this, too, like Jencks’ opposite conclusion, is a judgment not a documented finding. Fortunately, we do not need to rely on casual empiricism. The current experiment in the Alum Rock School district in San Jose, Calif. goes even further than Jencks by restricting the use of the vouchers to public schools. A proposed experiment in New Hampshire is designed to go in the other direction -- toward an unrestricted voucher -- though unfortunately the recent Supreme Court decisions have induced New Hampshire to exclude parochial schools. The results of these experiments may provide some solid evidence by which to judge the likely outcome of unrestricted vouchers.**

(6) Doubt about new schools.

Is this not all a pipe dream? Private schools now are almost all either parochial schools or elite academics. Will the effect of the voucher plan not simply be to subsidize these, while leaving the bulk of the slum dwellers in inferior public schools? What reason is there to suppose that alternatives will really arise?

My grocery fable perhaps suggests the answer. The absence of alternatives when there is no market does not mean that none would arise when there is one. Today, cities, states and the Federal Government spend about $50 billion a year on elementary and secondary schools. That sum is about a third larger than the total spent annually in restaurants and bars for food and liquor. The smaller sum surely provides an ample variety of restaurants and bars for people in every class and place. The larger sum, or even a minor fraction of it would equally provide an ample variety of schools. It would offer a vast market that would attract a host of entrants, both from the public schools and from other occupations. In the course of giving occasional talks on the voucher plan, I have been enormously impressed by the number of persons who have come to me afterward and said something like: “I have always wanted to teach [or run a school] but I couldn’t stand the educational bureaucracy, red tape and general ossification of the public schools. Under your plan, I’d love to try my hand at starting a school.”

(7) The impact on public schools.

It is essential to separate the rhetoric of the professional public-school bureaucracy from the real problems that would be raised. The National Education Association and the American Federation of Teachers claim that vouchers would “destroy the public-school system” which has been the foundation and cornerstone of our democracy. Claims like this are never accompanied by any evidence that the public school system is, in fact, under current conditions achieving the great results claimed for it -- whatever may have been true in earlier times. Nor is it ever made clear why if the public school system is performing so magnificently, it need fear competition from nonpublic, competitive schools.

The real problem arises from the defects, not the virtues of the public schools. In small, closely knit communities where public schools, particularly elementary schools, are now reasonably or highly satisfactory, I doubt that even the most comprehensive voucher plan would have much effect. The public schools would remain dominant, perhaps improved by the threat of potential competition. But elsewhere, and particularly in the urban ghettos where the public schools are failing so dramatically, most parents would doubtless try to send their children to other schools.

Difficulties would arise from a possible exodus of quality pupils from some public school. The parents who are most concerned about their children's education are likely to be promptest in transferring them, and even if their children are no smarter than those who remain, they will be highly motivated to learn and will have favorable home backgrounds. Such a sorting-out process goes on now, but, it can be argued, the voucher plan will greatly accelerate it, so that many public schools will be left with “the dregs,” society’s rejects, and will become, by virtue of the well documented effect of the students on a school, even poorer in quality than now.

As the private market took over, the quality at all schools would rise so much that even the worst, while it might be relatively lower in the scale, would be better in absolute quality. And many of today’s “rejects” are rejects only because the schools are so poor. As store-front academies and similar institutions have demonstrated, many “rejects” perform admirably in a school that evokes their enthusiasm instead of their antipathy.

Nonetheless, it is possible that, at least for a time, some children who remain in public schools will get even poorer schooling than they do now. This raises the moral dilemma that we are all familiar with: Are we justified in imposing poorer quality schooling on some children to leaven the schooling of others? It is easy to answer “yes” for other people’s children, almost impossible to be so saintly or so diabolical -- I do not know which -- as to say “yes” for one’s own children.

I am comforted in my own negative answer by three considerations. First, the possibility is purely hypothetical; it has neither been demonstrated by experiment nor rendered probable by any persuasive indirect factual evidence. Second, I am convinced that at worst the phenomenon would be temporary. Third, the children who would benefit most are from the very same social and economic groups as those who, it is feared, would be harmed.

There is no doubt what the key obstacle is to the introduction of market competition into schooling: the perceived self-interest of the educational bureaucracy. The role of this interest group is nothing new, as demonstrated by a fascinating study by British economist Edward West on the development of compulsory and government advanced schooling in New York State (Journal of Law and Economics, October, 1967). West demonstrated that teachers and public officials who wanted higher pay and more job-security spearheaded the pressure that led to full assumption of financing by the government, which came with the Free School Act of 1867, and compulsory schooling, which came later. On the present situation, let me quote an eloquent if bitter judgment by Kenneth B. Clark:

“It does not seem likely that the changes necessary for increased efficiency of our urban public schools will come about because they should.... What is most important in understanding the ability of the educational establishment to resist change is the fact that public school systems are protected public monopolies with only minimal competition from private and parochial schools. Few critics of the American urban public schools -- even severe ones such as myself -- dare to question the givens of the present organization of public education.... Nor dare the critics question the relevance of the criteria and standards for selecting superintendents, principals and teachers, or the relevance of all of these to the objectives of public education -- providing a literate and informed public to carry an the business of democracy -- and to the goal of providing human beings with social sensitivity and dignity and create respect for the humanity of others.

“A monopoly need not genuinely concern itself with these matters. As long as local school systems can be assured of state aid and increasing federal aid without the accountability which inevitably comes with aggressive competition, it would be sentimental wishful thinking to expect any significant increase in the efficiency of our public schools. If there are no alternatives to the present system -- short of present private and parochial schools, which are approaching their limit at expansion -- then the possibilities of improvement in public education are limited."

Let me give the last word to my great master. Before there was a United States of America, Adam Smith wrote:

“Were there no public institutions of education, no system, no science would be taught for which there was not some demand; or which the circumstances of the times did not render it either necessary or convenient or at least fashionable to learn.... Were there no public institutions for education, a gentleman...could not come into the world completely ignorant of everything which is the common subject of conversation among gentleman and men of the world....

“The public can facilitate the acquisition [of the most essential parts of education] by establishing in every parish or district a little school, where children may be taught for a reward so moderate, that even a common laborer may afford it; the master being partly, but not wholly paid by the public; because, if he was wholly, or even principally paid by it, he would soon learn to neglect his business.”

Reprinted with the permission.

Return to the Friedmans on School Choice