Riddle of the year: How is a public school like the U.S. Post Office? Answer: It’s inefficient, it costs more each year than the last, it is a perpetual subject of complaint about which nothing is ever done. It is, in short, a typical government monopoly.
The Post Office is a legal monopoly; no one else can carry first-class mail for profit. The public school is a monopoly by virtue of the money it receives from state and local governments. In order to compete with it, an unsubsidized private school must be, not merely better, but so much better that its customers are willing to forgo their share of that money.
There is a simple solution: Subsidize schooling instead of schools. Give each student a tuition voucher redeemable by any qualified school, public, private, or parochial.
The value of the voucher would be the state’s per capita expenditure on schooling. Public school systems would have to support themselves on the money brought, in the form of vouchers, by their pupils. Private and parochial schools could, if they chose, supplement the vouchers with additional tuition, charitable donations, or church monies.
The school system would then be open to real competition. An educational entrepreneur who found some way of providing a better education at a lower cost would make money and expand his operations; his competitors, public as well as private, would have to improve or shut down.
Such an entrepreneur would have the best possible incentive to find good teachers and pay them what they are worth. Many different teaching methods would be tried. Those that failed would disappear; those that succeeded would be copied.
The state would have to determine what was a qualified school in order to guarantee that the vouchers were spent on schooling. Some supporters of private education fear that this power would be used to control schools that are now independent. For that reason they either oppose all subsidies to private schools or prefer tax rebates.
The trouble with tax rebates is that they are useless to the poor, who, since they get the worst education of all from the public schools, would be the greatest beneficiaries of a competitive system. If rebates are used, they should be combined with a system of direct vouchers for parents whose tax payments are less than the amount of the rebate.
Even with rebates, the state (or the federal government) decides what qualifies as an educational expense. Even if there is no subsidy at all, there are still compulsory schooling laws; the state decides what is or is not a school. A state which wishes to control its private schools can do so now.
The best solution to this problem would be for any state instituting a voucher system to include, as part of the initial legislation, the provision that any institution can qualify as a school on the basis of the performance of its graduates on objective examinations. In New York, for instance, the law might state that any school would be recognized if the average performance of its graduating class on the Regents exam was higher than the performance of the graduating classes of the bottom third of the state’s public schools. A new school could operate provisionally, accumulating vouchers until its first class graduated. A school dealing with retarded or otherwise disadvantaged children could petition the state for special recognition if it was unable to satisfy the usual criterion. Such legislation would be sufficient to prevent parents from setting up fake schools in order to transfer the voucher money to their own pockets. At the same time, it would make it almost impossible for the state to control either the method or content of private schooling.
The state could force schools to teach certain ideas by putting them on the exam but could not prevent them from teaching others and would have no control over how they were taught. A teacher who disagreed with the orthodox position could always tell his class that ‘this is what the examiners want you to write on the test. On the other hand, what I think is true is…’
A voucher system with such precautions would not only prevent the state from controlling the pupils now in private schools, it would also greatly reduce the state’s power over students who are now in public schools.
According to a Gallup poll some years ago, 30 percent of parents would send their children to private schools if they were free and 29 percent would send them to parochial schools. Parochial schools and many private schools already charge less than the amount the state spends on public schools, so with vouchers they could be free. Private schools that currently cost more than that could charge much lower tuition than they now do and entrepreneurs could and would take the opportunity to set up new schools funded entirely by vouchers. Thus the number of pupils in schools run by the state would be cut in half.
For those people who view the power of government to make sure that everyone learns the same things in the same way as desirable, this is a disadvantage of the voucher system. For those of us who prefer a free and diverse educational system, it is an advantage.
Some critics argue that the use of vouchers by parochial school pupils would violate the separation of church and state. When I first wrote this chapter around 1970, it was unclear what position the courts would take on whether it was constitutional to use vouchers for religious schools; when I revised it in 1988 it was still unclear, but the odds that the Supreme Court would rule them constitutional had improved somewhat, at least in the judgment of a law professor friend of mine who specialized in church-state issues. When the case finally reached the court in 2002, he turned out to be correct.
However the courts rule, is the voucher system in fact an illegitimate subsidy of religion? No. The state is subsidizing parents in the purchase of schooling for their children; they can buy that schooling wherever they wish. For them to use the subsidy to buy schooling from a parochial school is no more a state subsidy of religion than for a welfare client to buy his food at a church bazaar. Of course, the parochial school hopes to achieve its end of teaching religion at the same time that it provides the state what it is paying for: education in secular subjects. Similarly, the church hopes to use its profits from the bazaar to finance religious projects.
One argument sometimes made against a voucher system is that it would subsidize the rich and impoverish the public school system by transferring money to upper-class parents who currently send their children to private prep schools.
Unfortunately for this argument, only about half of 1 percent  of all pupils in the U.S. go to private, nonreligious schools (about 250,000). The great majority of pupils in nonpublic schools (about 5.5 million) are in church-related schools, and their parents are frequently poorer, not richer, than the average of the community.
In states where Roman Catholics make up a large percentage of the population the voucher system would substantially increase the state’s educational costs, since the state would have to provide vouchers for many children now in parochial schools. But without some form of state aid, the parochial schools may well close down,  in which case the state will have to pay for their pupils anyway. Furthermore, these are precisely the states where it is now difficult to get money for public schools, since parents whose children are not in the public schools are notoriously unfriendly to new taxes for schooling.
A related objection to the voucher system is that it would increase educational inequalities. Presently, it is said, all children, rich and poor, go to the same public schools. Under a voucher system, poor parents would send their children to public schools or to private schools that subsisted only on vouchers, while richer parents could supplement the vouchers with additional tuition payments and so put their children in better schools.
But under our present system, the school a child goes to is determined by where he lives and where he lives is largely determined by the income of his parents. Under the voucher plan a ghetto parent who was deeply concerned about his child’s education might be able to scrape up a thousand dollars a year or get a small scholarship, add that to the value of his voucher, and so send the child to a good private school. Under the present system he has the choice of either paying $5,000 a year for a good private school or buying a $200,000 house in a suburb with a good school system.
Thus the voucher system, although it does not eliminate class distinctions in education, blurs them. Today a small elite goes to private prep schools, middle-class children go to moderately good suburban schools and the inner city poor get schools that are often little more than custodial institutions.
Under a voucher system the motivated middle-class parent could afford the differential between the cost of a public school education and a good prep school. Low-income parents who felt that they were being short-changed in the schooling provided to their children would have the option of setting up their own schools, perhaps along the lines of the Harlem Street Academies, or persuading someone to set up private schools for them and financing them with vouchers.
The voucher plan, like other free market mechanisms, provides the ultimate form of decentralization and does so in a way that protects the rights of even small minorities. If 60 percent of one school district’s population wants one kind of school, the other 40 percent can take their vouchers and set up their own school. If a local minority is too small to support a school of its own it can pool its resources with similar groups elsewhere.
When I originally wrote this chapter I predicted that a voucher plan would be adopted in some state sometime in the next few years. I was wrong. There were several attempts to introduce such plans, but they were bitterly and successfully opposed by the educational bureaucracy and the teachers’ unions.
That is no reason to give up. It took a long time to get the country into its present situation and it will take a long time to get it out. While attempts to get the government out of the schooling business have so far been unsuccessful, both the ideology of government control and the public’s support for the public school system have been growing gradually weaker. I am not willing to make any more predictions, but I can still hope.
For years we have been told that all the public school system needs is more money. For years we have watched its per pupil spending rise, with little visible effect on quality. It is time to try something new.
[Since this chapter was written for the first edition and revised for the second, there has been considerable progress. No state has yet adopted a full fledged voucher program giving all students the option of a voucher for the full cost of public schooling. But a considerable number have adopted programs giving a significant number of students the option of a voucher for a substantial fraction of that cost.]
 This figure and those following are from the Statistical Abstract of the United States: 1967.
 New York Times, (1967) September 22, page 32; September 4, page 44; (1969) June 16, page 1.