Econlib

The Library

Other Sites

 Front Page arrow Titles (by Subject) arrow E. G. WEST, The Uneasy Case for State Education - New Individualist Review

Return to Title Page for New Individualist Review

Search this Title:

Also in the Library:

Subject Area: Economics
Subject Area: Political Theory
Subject Area: History
Order this book from Liberty Fund

E. G. WEST, The Uneasy Case for State Education - Ralph Raico, New Individualist Review [1961]

Edition used:

New Individualist Review, editor-in-chief Ralph Raico, introduction by Milton Friedman (Indianapolis: Liberty Fund, 1981).

About Liberty Fund:

Liberty Fund, Inc. is a private, educational foundation established to encourage the study of the ideal of a society of free and responsible individuals.


The Uneasy Case for State Education

E. G. WEST*

This article has been adapted by the author from his recent book Education and the State, published in London by the Institute of Economic Affairs.

THE CASE FOR substantial government intervention in primary education is normally examined by political economists in the context of two principles: “state protection” and “neighborhood effects.” According to the first, if it is generally agreed that the state exists basically to give protection to its members, it must have special obligations to children since they are least able to protect themselves on their own. According to the second principle, the way I choose to act often has serious spillover effects upon my neighbors; and similarly the actions of my neighbors, taken individually, can substantially affect me. The choice of any one individual to educate, or not to educate, his children is shown to be a particular case in point. Even if all individuals choose voluntarily to purchase education they may still “underinvest” from the point of view of society as a whole. This being so there is again said to be a strong presumption in favor of government intervention.

The first part of this article attempts to examine the “protection principle” in more than usual detail. The second part will raise questions which are not normally raised about the empirical foundation for common assumptions about the “neighborhood effects” of education.

WHATEVER THE MEANING of laissez faire, the most ardent of its nineteenth century English supporters rarely argued that it should operate outside the boundaries of a proper legal framework. Within this framework they were prepared to make many exceptions to their general principle of freedom of contract and their special reservation for children was a prominent example of this attitude. However much they disliked overinterfering governments, they believed that some element of governing was necessary. But what was the true duty of government? The answer of one of the classical economists, Nassau Senior, disciple of Bentham and friend of John Stuart Mill, was quite clear:

I detest paternal despotisms which try to supply their subjects with the self-regarding virtues, to make men by law sober, or frugal, or orthodox. I hold that the main, almost sole, duty of Government is to give protection. Protection to all, to children as well as to adults, to those who cannot protect themselves as well as those who can.1

Senior went on to remind his readers that children were more defenseless than others. In view of this the state had extra protective obligations towards them and in particular there was a strong presumption in favor of state intervention in education. This view, which seems to have been readily accepted by liberal economists ever since,2 will now be examined in more detail in order to gather its precise implications.

A MOMENT’S REFLECTION will show that it is much easier to state the “protection of minors” principle than to draw practical policy conclusions from it. If it is agreed that the state should be responsible for seeing that children are protected, the question arises: whom should it appoint to carry out this duty in practice? The first obvious point to clear up before deciding this issue is for the state to ascertain from its members how important a role they want the family to play. If they aim at giving the family a central place, then the question of protecting infants cannot be settled in isolation of this policy, for it establishes a presumption in favor of delegating the duty of child protection first and foremost to the parents and only withdrawing this arrangement when special circumstances require it. For these reasons the following remarks of John Stuart Mill, which seem to have had widespread influence on this subject, were too evasive and too legalistic:

In this case [education] the foundation of the laissez-faire principle breaks down entirely. The person most interested is not the best judge of the matter, nor a competent judge at all. Insane persons are everywhere regarded as proper objects of the care of the state. In the case of children and young persons, it is common to say, that though they cannot judge for themselves, they have their parents or other relatives to judge for them. But this removes the question into a different category; making it no longer a question whether the government should interfere with individuals in the direction of their own conduct and interests, but whether it should leave absolutely in their power the conduct and interests of somebody else.3

Few would disagree with Mill’s general sentiments. If we do not tolerate cruelty to animals still less should we allow the possibility of continued cruelty to children by granting absolute power over them to any single person. But Mill was mistaken in assuming that the common argument, that parents and relatives could judge for their children, was a claim for absolute power. What most people envisage is something in the nature of a fiduciary power, to be removed in cases where abuse can be shown.

Whatever the true interpretation of his statement, Mill’s anxiety to put this question into a different category did not make it any less important or less urgently in need of an answer: for the state is not a disembodied entity; it has to work through individuals to whom it prescribes certain powers. Now the function of supervising a child is such a personal and delicate matter that it is most important to visualize it in the form of a competition for influence between one individual, the parent, who by nature is closer to the child and therefore has better opportunity for gaining knowledge of its best interests, and another individual, appointed by the state, who has the advantage of some presumed expertise in protecting children. When the problem is expressed in this way the question of absolute power is beside the point. Certainly the difficulty is not one which can be solved easily by any universal or dogmatic ruling. Accordingly, the following analysis is not intended to stake an unconditional claim for any of the parties concerned, but is merely an attempt to examine the issues objectively. Insofar as I am critical of the ability of the “protection” argument to justify universal state schooling, it must be emphasized that my criticism is not to the prejudice of the “neighborhood effects” argument for this system.

BEFORE WE CAN CHOOSE those individuals deemed to be best able to protect a child, we have to solve the even more difficult task of defining the danger against which we are trying to protect it. Even state action against physical cruelty is not always simple to administer since the criterion is rarely a matter of unanimous agreement. But at least where physical injury has resulted in permanent damage to an individual, then the evidence is usually so obvious that the case for state protection against further assault is clear enough. In education, on the other hand, the argument is for protection not against physical injury but against ignorance. Here it is often more difficult to see in what respect any faculties can be said to be permanently injured, either in effect or in intention. When, for instance, the head of a temporarily distressed family sees his five- and six-year-old children remaining ignorant of reading and writing, we cannot say that their faculties are injured in the same sense as in the case of physical brutality. In this instance, the faculties of learning need not in any way be permanently damaged; it is quite possible for them to remain intact to be used later. If the state does decide to intervene in such cases, therefore, it cannot be on the grounds of the same sort of protection as that directed against physical aggression of any kind; rather is it likely to be based on the widely accepted principle of government relief of poverty. This initial clarification seems necessary if we wish to avoid using the term “protection” with its more usual connotation.

But there is a much more stubborn difficulty. When we now speak of protection against ignorance, we have to ask: ignorance of what? A person may be most ignorant of one thing but quite expert at another. Too hasty attempts to prescribe learning priorities can lead to results which not only endanger spontaneity and individuality but which can involve fundamental contradictions in any society which professes to encourage and support the ideal of liberty. Confusion on this subject often arises from a dogmatic insistence that the relevant ignorance is necessarily ignorance of what is taught in schools between statutorily prescribed ages. Schooling is only one instrument in the removal of ignorance; if other means are being used, the need for protection may well be superfluous. There are additional sources of learning in real life: the parent, the family, its friends, the church, books, television, radio, newspapers, correspondence courses, etc., “on the job training,” and personal experience. J. S. Mill himself can be quoted in this respect:

Even if the government could comprehend within itself, in each department, all the most eminent intellectual capacity and active talent of the nation, it would not be the less desirable that the conduct of a large portion of the affairs of the society should be left in the hands of the persons immediately interested in them. The business of life is an essential part of the practical education of a people; without which, book and school instruction, thought most necessary and salutary, does not suffice to qualify them for conduct, and for the adaptation of means to ends. Instruction is only one of the desiderata of mental improvement; another, almost as indispensable, is a vigorous exercise of the active energies; labor, contrivance, judgment, self-control: and the natural stimulus to these is the difficulties of life.4

The best means by which individuals are likely to “protect” themselves or their children from “ignorance” should therefore be open to constant comparative appraisal. That a parent, for instance, wishes to take his child away from school at an early age does not necessarily signify that he is negligent. Insofar as the school has become less efficient than other means of education, the parent himself may be acting from motives of protection and be making the same kind of shrewd comparative assessment that he makes before transferring his buying from one source of his child’s food or clothing supply to another. Again, to assume that the education given in a school is always and under all circumstances to be preferred to alternative types of education is probably to assume also that all schools and home environments are homogeneous. This is by no means self-evident—especially in a changing society. It is interesting a recall that J. S. Mill himself was deliberately kept at home throughout his childhood by a father who was strenuously motivated by protective impulses. His father, James Mill, indeed kept his son away from school, “lest the habit of work should be broken and a taste for idleness acquired.”5 It is quite true that James Mill has been the subject of severe criticism from subsequent state educationists on the ground that he was too forceful a task-master. The question is, however, whether a state school could have produced a “better” John Stuart Mill. Such questions, of course, give rise to all sorts of speculations, but no apology is offered for asking them, since, as I hope to show, they lie at the root of the problem I am discussing.

THE “PROTECTION OF minors” argument has in the past been used to support pressure not merely to educate a minority of neglected children but also to establish universal schooling whereby every child is provided for by state schools. There are two major difficulties in the way of our accepting such reasoning, the first political and the second economic. On the political difficulty, it must first be observed that in order to justify a vast and comprehensive system comprising thousands of new state schools one must establish that such provision is needed to fill an obviously widespread deficiency and that the majority of parents and relatives are either negligent or ignorant. Now if this is the contention, it must imply either widespread schizophrenia or self-abnegation, for it envisages an electorate which virtually condemns most parents and relatives for being ignorant or negligent about their children when that same electorate consists to a large extent of the parents and relatives themselves. Otherwise, the question immediately arises of why, if such ignorance and negligence is so serious, should we presume that it will not equally express itself at the ballot box and with equally “unfortunate” results when the parents and relatives choose their representatives?

The extent of this presumed schizophrenia could best be checked by a survey among parents to ask their intentions if given hypothetical refunds of indirect and direct taxes in the form either of money vouchers spendable only on education in lieu of “free” state education or of income tax allowances. The only statistical survey that has attempted to elicit answers to this question suggests that negligence would be far from typical.6 Alternatively we can make an historical investigation to discover parental behavior before the inception of state education, bearing in mind that such state intervention had to be supported by increased tax revenues which might otherwise have been spent voluntarily on education. I think it can be shown from historical evidence that nineteenth century parental behavior was much more responsible than is commonly supposed.7

The second major difficulty in the way of accepting the “protection” argument to justify a universal state school system is an economic one. Nowhere does it seem to have been shown why other forms of state intervention could not achieve the intended result of state protection more effectively and with less cost than the present system of state schools (which amounts virtually to a system of nationalized schools). Much as John Stuart Mill wanted the protection of children, even he did not in the end prescribe compulsory state schools, nor even compulsory private schooling, but only compulsory education. Accordingly he held that the state should be interested not merely in the number of years of schooling but in checking the results of education whatever its sources, and he contended that an examination system was all that was necessary. If a young person failed to achieve a certain standard, then extra education would be prescribed at the parents’ expense. Another sanction which Mill also entertained was to make the right to vote conditional on some minimum degree of education.

Under Mill’s scheme, if it were operating today, it is conceivable that some children would attempt to attain the necessary standards by much more dependence on parental instruction,8 correspondence courses, evening classes, local libraries, etc. These in turn would be measured against particular services offered by the private schools whose relative efficiency would be measured by parents and their children in terms of the size of their classes, for instance, or the qualifications of their staff and the personal attention they gave the children.9 There are examples of this kind of minimum state intervention in other spheres. Thus although the state insists on the acquisition of a minimum competence in driving before allowing persons to take their vehicles on the roads, it has so far found it unnecessary to prescribe the particular way in which persons should acquire the knowledge and the skill, or to nationalize the driving schools and supply training “free” by raising taxes on all. Again, protection against the supply of adulterated food to children (or to anybody else) is effected simply by a system of inspection, reinforced by regulations, breaches of which are punishable by the law.

THE CASE OF FOOD IS interesting. Protection of a child against starvation or malnutrition is in the same category of importance as protection against ignorance. It is difficult to envisage, however, that any government, in its anxiety to see that children have minimum standards of food and clothing, would pass laws for compulsory and universal eating, or that it should entertain measures which lead to increased taxes in order to provide children’s food “free” at local government kitchens or stores. It is still more difficult to imagine that most people would unquestioningly accept this system, especially where it had developed to the stage that for “administrative reasons” parents were allocated those stores which happened to be nearest their homes; or that any complaint or special desire to change their pre-selected stores should be dealt with by special and quasi-judicial inquiry after a formal appointment with the local “Child Food Officer” or, failing this, by pressure upon their respective representatives on the local “Child Food Board” or upon their Congressman. Yet strange as such hypothetical measures may appear when applied to the provision of food and clothing, they are nevertheless typical of English and American state education as it has evolved by historical accident or administrative expediency.

Presumably it is recognized that the ability in a free market to change one’s food market when it threatens to become, or has become, inefficient is an effective instrument whereby parents can protect their children from inferior service in a prompt and effective manner. If this is so, then one should expect that the same arguments of protection would in this respect point in the direction not of a state school system where it is normally difficult to change one’s “supplier” but in the direction of a free market where it is not. In this sense one must question John Stuart Mill’s assertion that in the case of education the principal of laissez faire breaks down entirely; for if by laissez faire he meant (as he seems to have meant) a free market system, then our reasoning suggests on the contrary that this is a technique which with some qualifications is admirably suited to protection of all kinds and not least the “protection of minors.”

So much for the analysis of the basic issues in the “protection” thesis. The conclusion is that if there is a logical case for a universal state school system, as distinct from marginal intervention to meet special cases, we must look elsewhere than the protection principle.

THE ABSENCE OF A justification for state intervention on the grounds of the “protection principle,” brings us to the second line of reasoning, which is widely believed to give still stronger support for such intervention. As previously indicated, this belongs to what is known as the “neighborhood effects” argument; a brief general account of its present place in economics may be helpful prior to the particular application of it to education.10

Roughly speaking, the “neighborhood effects” argument stems from the common observation that “no man is an island.” Many of his actions intentionally or unintentionally affect other people. Where these spillover effects are very pronounced, and do not show any signs of ever being organized or brought under control by the market, one’s normal reaction is to explore the possibility of government intervention. The most obvious instance of the resort to government is seen in the establishment of a state system of law and order to curb and to make socially accountable individual acts of aggression. Here, indeed, we have the basic raison d’être of the state in the first place. Beyond personally aggressive private actions, however, there are other particular “neighborhood effects” which are also identified and often listed in a descending order of seriousness or scope. Several of these are also commonly claimed to warrant state intervention. The most frequently quoted example in economics is the firm whose factory chimneys offend the neighborhood with smoke and thereby cause people outside the factory to spend extra amounts on laundries, bronchitis cures, etc. Another alleged instance is the injury to “local amenities” caused by the “unfortunate” construction of a new house by a private speculative builder. Again there is the individual motorist who parks his car to the detriment of other road users. Campaigns to make socially accountable those responsible for road traffic and aircraft noises or exhaust fumes also belong in the same category of “neighborhood effects.” All of them are cited as cases where the costs taken into account by the individual in the market are unlikely to include any element of what are called social costs.

One must observe already that although most people react to such situations by readily calling upon government to put these things right, they do not immediately see how complicated is their request. For one thing, if they examine more carefully the above examples of social costs they will see that they are not exclusively a consequence of private action. The noise from publicly-operated airways, buses, and trains, for instance, has its ultimate sanction not in private action but in public legislation, that is, in government itself. Nationalized chimneys give off smoke no less than private ones. Public gasworks can spoil amenities while offensive smells can come from publicly operated sewage works.

A further complication is that social costs can be negative as well as positive; that is, some spillover effects may be unintentionally beneficial to the neighborhood. Thus a farmer who drains his own land may improve that of the neighboring farm, even though he cannot charge for benefits rendered Still more complex are the many instances where both positive and negative social costs are produced by the same agent. An example of this is where a new industrial plant gives off smoke (detrimentally) but also reduces unemployment in the same neighborhood (beneficially). Again the owner of the plant could be the central or local government as well as a private company.

It seems to have been widely believed at one time that the moment that one had pointed out a privately originating “neighborhood effect,” such as that of the smoke-laden atmosphere, actual intervention by the state was adequately justified. This, however, is not so. The identification of a “neighborhood effect” is only a necessary but not a sufficient condition for intervention. There are many serious offsetting considerations, the most important being that the task of measuring the chain reaction of costs and benefits is often insuperable. The administrative costs of intervention alone may be so high as to exceed the net benefits which such action sought to secure, even if they could be measured.11 Furthermore, it is likely that a particular mode of state intervention to meet a privately originating “neighborhood effect” may itself incur a second (publicly originating) “neighborhood effect” which has still more serious consequences than the first. For instance, suppose that the government responds to the smoke pollution of a factory by placing a special local tax upon the offending firm. The resulting increase in relative costs of production may so discourage the expansion of the same firm as to encourage it to invest its surplus funds elsewhere. The result may be that the firm’s decision is now the cause of another “neighborhood effect,” not smoke pollution this time but the more serious problem of unemployment.12

How then does the “neighborhood effects” analysis apply to education? Political economists usually have two particular instances in mind. The first is expressed in their contention that the social benefits of education are not confined to the “educatee” but spread to society as a whole, most noticeably in the form of reduced crime and more “social cohesion.” This can be expressed negatively: The private actions of an uneducated person may have unfortunate consequences for others in society. The idea seems to be, for instance, that just as the government can do something about “anti-social” smoke (e.g., by taxation) it can do something about “anti-social” conduct (e.g., by education). The second general example is the idea that education is an investment whose benefits also spill over to the economic advantage of society as a whole. There is nothing in this argument so far to justify government provision of schools. If the government were satisfied that neighborhood effects were substantial, it could increase the quantity of education by a mixture of subsidies, vouchers, and compulsory legislation. But what is the evidence for neighborhood effects? Are economists’ assumptions always well grounded?

CONSIDER FIRST THE familiar proposition that state provision of education will successfully meet the “neighborhood effect” of crime. It will be helpful first to give a few quotations to illustrate the widespread influence of the idea both in the nineteenth century and today.

In 1847 T. B. Macaulay proclaimed in Parliament:

I say that all are agreed that it is the sacred duty of every government to take effectual measures for securing the persons and property of the community; and that the government which neglects that duty is unfit for its situation. This being once admitted, I ask, can it be denied that the education of the common people is the most effectual means of protecting persons and property?13

In the previous year W. T. Thornton had typically expressed the prescription of the Utilitarians:

No one now denies that proper schools for the lower orders of people ought to be founded and maintained at the cost of the state. The expense no doubt would be considerable, but it would scarcely be so great as that already incurred for prisons, hulks, and convict ships; and it is certainly better economy to spend money in training up people to conduct themselves properly, than in punishing them for their misdeeds.14

Such Utilitarian calculation of the crime-reducing potentialities of formal education is still in evidence among responsible authorities today. Thus, referring to the “neighborhood effects” of education, the Robbins Report on Higher Education (1963) said:

There are, of course, also important social and political benefits of education which accrue to the populace as a whole—a better informed electorate, more culturally alive neighborhoods, a healthier and less crime-prone population, and so on. What is not always recognized is that these social and political consequences may in turn have significant economic effects—the efficiency with which goods are exchanged is obviously enhanced by general literacy, to the extent that education reduces crime (even if only by keeping children off the streets during the day) the country can shift resources that would have had to be used for the police function to other ends, and so on.15

The Robbins Report makes this statement before freely admitting that the evidence is still not sufficient to make it anything more than an inspired hunch. The Crowther Report of 1959 and the Newsom Report of 1963, referring to the education of persons between fifteen and eighteen years (which include the most crime-prone ages), although much more hesitant on the matter, nevertheless, as I shall show in more detail below, favored still more education in the current twentieth century campaign against delinquency and crime.

IN THE LIGHT OF THE “neighborhood effects” argument examined in the preceding paragraphs, what can be made of its application to education? It is important to remember the complications; not only must the “neighborhood effect” be first identified in a meaningful way, but also the possible side-effects of the proposed government intervention itself should all be examined before such intervention is fully justified. How far then, first of all, has the particular “neighborhood effect” relationship between education and crime been reasonably established in practice? In other words, what evidence have we to show that the belief in state education as a general insurance against crime is anything more than dogma?

In answering such questions the early economists were inclined to rely upon crude statistics. Because the latter were presented as showing an inverse relationship between crime and education (usually measured by the degrees of schooling) the general inference was that ignorance (or the deprival of schooling) was a major cause of crime. To the extent that poverty was connected with ignorance and undesirable habits, it too was thought to be an important contributory factor. In view of this kind of reasoning it would be intriguing to know the reaction of these early commentators to present-day statistical evidence, for this shows that crime has increased at the same time as state education has been growing. Certainly this does not deny that crime could have grown equally or even more in the absence of state education; but scientific objectivity demands that all things should be suspect, especially where there is a positive correlation. One can at least speculate, judging by their weakness for such crude statistical inferences on this subject, that the early economists would be tempted to point to the possibility that our state education, as distinct from their nineteenth century parochial education, was a predisposing cause of crime!

Today indeed there is at least a growing scepticism about the potentialities of state education as a crime reducer. Thus The Times Educational Supplement in 1963 declared:

It is strange that as education spreads and poverty decreases, juvenile crime should steadily rise.16

Similarly, the idea that poverty is a major cause of crime is not so confidently held. Indeed some social scientists conclude from the evidence that the idea has no firm basis. Lady Wootton writes:

. . . it is a conclusion which would, I think, have surprised our grandfathers. The converse was implicit—and sometimes explicit—in the thought of not so many generations ago; as it is implicit also in the thought of those who express disappointment that the coming of a “welfare state,” which they believe (though mistakenly) to have banished poverty, has not also greatly reduced the criminal statistics.17

The most energetic nineteenth century advocates of state education would no doubt have been also perplexed by the fact that the same welfare state which has failed to reduce crime is one which also includes an extensive education system financed by public funds of unprecedented magnitude.

Today, of course, we would claim to be much more aware of the complexity of crime and its causes. Certainly much more sophisticated reasoning surrounds the subject and it has been established in particular that a proportion of convicted persons is suffering from mental disorders which, it is claimed, need psychiatric treatment. Many other possibly conducive factors are still being investigated and these include divorce, broken homes, the persistence of crime in some families, poor church attendance, mothers’ employment outside the home, health, and type of employment. Meanwhile an ordinary person may be forgiven the thought that highly organized crimes today call for such a degree of skill and intelligence that education can only serve to be a complementary rather than a competitive factor. The cleverer the criminal the more effective the crime. But it is interesting to observe that so far English social scientists have not even yet reported any very clear correlation between education and crime. Thus in 1958 Lord Pakenham published the results of research into the causes of crime (financed by the Nuffield Foundation), which included the following observation:

I do not think, however, that the distinguished experts, including the representatives of the National Union of Teachers who gave evidence before us, would claim that, up to the present, much progress has been made in connecting education and crime.18

Insofar as people today still press for more education to reduce crime they usually mean a lengthening of the school life, i.e., a rise in the school-leaving age. In this connection we touch upon a piece of evidence which English educationists have found particularly perplexing. The Crowther Committee (1959) discovered the fact that the last year of compulsory education was also the heaviest year for juvenile delinquency and that the tendency to crime during school years was reversed when a boy went to work. Not only was this a long-standing phenomenon but also when in 1947 the school-leaving age was raised from fourteen to fifteen the most troublesome age group moved up from the thirteen-year-olds to the fourteen-year-olds.19

HOW SHOULD AN economist treat such information within the general framework of the “neighborhood effect” analysis? It seems reasonable at least for him to conclude that the popular belief, as quoted for instance from the Robbins Report that state education makes the public less crime prone, is unsupported by the available evidence. Beyond this he could argue, but with less certainty, that the evidence showed a prima facie relationship in the opposite direction, i.e., that state education involved adverse external effects and aggravated or even helped to cause the prevailing trend towards increased criminal behavior. Certainly one could not object to the tentative conclusion that if any further official action were to take place it should first concentrate on a proper investigation of this question and that in the meantime there was to be a presumption against any further increases in the duration of compulsory schooling on this account.

It therefore comes as a surprise to find the Crowther Report concluding that there was nothing in the current state of affairs

. . . to make any thoughtful person doubt the value of being at school; indeed, the delinquency may arise, not because boys are at school, but because they are not at school enough.20

As is well known in Great Britain, the Report argued for the raising of the compulsory school-leaving age. In doing so it referred to the beneficial external effects on society of presumed increased economic growth,21 but at the same time it apparently refused to consider the possibility of any adverse external effects. Yet the Crowther policy of first raising the school-leaving age despite the evidence about delinquency and then trying to justify the measure after the event by offering hopes of future improvements in schooling seems to start at the wrong end. Indeed, such proposals seem to substitute dogma for reason and to betray the attitude that come what may the schools should not yield. In this attitude they probably reflect the limitations of state-sponsored committees which inevitably comprise many members and witnesses such as local education officers, state school headmasters, and the heads of teacher-training establishments who have a direct interest and belief in the expansion of state education itself. Such committees seem to welcome the rationality implicit in the “neighborhood effect” argument when it suits them, but are too ready to discard it the moment it becomes inconvenient.

The Crowther Committee contended that delinquency among older pupils probably arose because a boy had more time on his hands to get into mischief when at school compared with when he was at work. The Newsom Report22 (on the education between the ages of thirteen and sixteen of pupils of average or below average ability) pointed out the difficulty of getting enough staff and resources to keep the boys occupied sufficiently. Accordingly this Report recommended that: “The school programme in the final year ought to be deliberately outgoing—an initiation into the adult world of work and leisure.”23 Again, undaunted by the evidence, the Report also recommended the raising of the school-leaving age. Once more the innocent observer must be allowed to question why, if the rate of delinquency declines when boys go out to work, should the commencement of this work be delayed still further in favor of schemes for simulating work in school? Why do we accept indiscriminately arguments about the need to protect young people from the “uncertain pressures of adult life” as long as possible and neglect the possibility that the pressures of school life may be in some cases the crucial ones? Such thoughts are not so revolutionary when it is remembered that, in these days, going out to work does not necessarily mean the end of formal education, since technical colleges and day-release schemes are now typically provided for the young worker’s continued instruction. Those who show an overweening concern about the welfare of school-leavers do not seem to have shown why their proposals for extra protection could not be implemented, for instance, by schemes for appointing special supervisors, which would have the additional merit of being far less expensive than schooling. This is not of course to argue that all expenditure on further schooling is wrong. Where it is appropriate, and this may apply to most people, it is to be welcomed; but to apply a measure to all on the grounds that it is suitable to some is to sacrifice prudence to mere legislative expediency.

The reduction of crime is only one of many kinds of social benefit which society is supposed to expect from education; but it has been selected here for first attention because of its prominence in traditional reasoning and because it is a good illustration of the facility with which unchallenged and unverified theories can become assimilated in the folklore of educational debate. In addition to the errors of fact which are involved in it, such thinking also suffers through a lack of conceptual clarity. What do we mean when we say that education reduces crime? Are we thinking of education in the wide sense or do we mean only formal schooling? If the former, what kind of education have we in mind? If the latter, do we mean only state schooling or do we include non-state schooling?

The “reduction of crime” is only one of the propositions which economists normally associate with “neighborhood effects.” Space does not allow a review of all of them, so the remainder of this article will confine itself to one more example of unsupported assumptions in this field: the assumption that widespread literacy could not have been achieved or maintained today without the predominant role of government.24

MANY ECONOMISTS ARE anxious to remind us that without widespread literacy it would be difficult to maintain a market economy as well as a political democracy. What, then, is to be feared? Was the populace not making itself literate before the government stepped in?

To what extent does nineteenth century history indicate that the English people, for instance, were in need of governmental help on this account? The evidence shows indeed that the majority of people in the first half of the nineteenth century did become literate (in the technical sense) largely by their own efforts.25 Moreover, if the government played any role at all in this sphere it was one of saboteur!

As long ago as the first few years of the nineteenth century it was a subject for government complaint that the ordinary people had become literate; for the government feared that too many people were developing the “wrong” uses of literacy by belonging to secret “corresponding societies” and by reading seditious pamphlets. In 1803, for example, Thomas Malthus echoed the government’s fears by asserting the probability that: “The circulation of Paine’s Rights of Man . . . has done great mischief among the lower and middle classes of this country.”26 Far from subsidizing literacy, the early nineteenth century English government placed severe taxes on paper in order to discourage the exercise of the public’s reading and writing abilities. Yet despite this obstacle, by the time government came around to subsidizing on a tiny scale in the 1830’s, between two-thirds and three-quarters of the people (according to one modern specialist27 ) were already literate. Even then the subsidies were financed from a taxation system which burdened the poor more than the rich.28

Those who argue that the system, not of government subsidy, but of government provision in the form of nationalized schools, was the key to the expansion of literacy have an even more difficult task. Using one of the generally accepted indexes of nineteenth century literacy, the figures showing the number of persons signing the marriage register with marks, it is argued that in 1870 when nationalized schools (called board schools) were introduced, 20 per cent of the men and 27 per cent of the women were signing their names with a mark. Much of the subsequent improvement in literacy, so the argument continues, must therefore be attributed to the 1870 legislation. One must observe, however, first that even if such figures are accepted they indicate that most people were already literate without the help of the nationalized schools. In other words, this evidence does not demonstrate the need for universal provision of these schools. Second, that judging by the growth rate in literacy prior to 1870 it is not clear that at least as good an improvement would not have occurred without the 1870 Act. Third, the men signing the marriage register in 1870, having an average age of 28 years, had on the average left school seventeen years before. If most people learn to read and write at school, an assumption which seems to be implied in this sort of reasoning, the 1870 marriage registers in England reflect the schooling of the early 1850’s. A more appropriate figure to test the literacy rate of school leavers around 1870 is that of the 1891 Census. This shows that only 6.4 per cent of the men and 7.3 per cent of the women were signing the registers with a mark. These men must have left school on an average around 1874. They could therefore have barely benefitted from the new state schools since the building program had scarcely got underway at this time. Finally, there is no evidence that the state school system, which had become universal by the time of the Second World War, has been a complete success. Even by 1948 there were still in England and Wales 5 per cent of fourteen-year-old school-leavers officially classified as nearly or completely illiterate.

Here then we have the paradox of a public managing to educate itself into literary competence from personal motives and private resources, despite the obstacle of an institution called government which eventually begins to claim most of the credit for the educational success. The notion held by many people that had it not been for the state they or at least most of their neighbors would never have become educated is a striking monument to the belief of the Victorian legal theorist Dicey, that people’s opinions and convictions eventually become conditioned by the legislated institutions they make themselves.

[* ] E. G. West has been lecturing at the University of Newcastle-on-Tyne, and will take up a Readership in Economics at the University of Kent at Canterbury, England, in 1966. He is currently engaged in a research fellowship at the University of Chicago.

[1 ]Suggestions on Popular Education (London: J. Murray, 1861), p. 6.

[2 ] See, for instance, Milton Friedman, Capitalism and Freedom (Chicago: University of Chicago Press, 1962), p. 86.

[3 ]Principles of Political Economy (London: Longmans, 1915), p. 957.

[4 ]Ibid., p. 948.

[5 ] J. S. Mill, Autobiography (New York: Holt, 1873), p. 36.

[6 ] See Choice in Welfare, First Report (London: Institute of Economic Affairs, 1963), Part III.

[7 ] See E. G. West, Education and the State (London: Institute of Economic Affairs, 1965), chaps. ix, x, and xi. The English government did not make education compulsory in the nineteenth century until most parents were already sending their children to school voluntarily. The same can be said of the State governments in the United States.

[8 ] In England and Wales there are 100,000 qualified women teachers at home compared with 160,000 teaching in schools. It is obvious that there are thousands of parents who are qualified, even in this formal sense, to teach in their own homes. Under Mill’s system it is doubtful whether all this educational capital would be as underused as it is today. Since under his scheme the state would not have involved itself in the heavy taxation now needed to finance “free” schools, average incomes after taxes would be much higher. These added incomes wou’d, for instance, enable many married teachers quietly to buy the help of auxiliaries in the form of domestic help and/or labor-saving devices to allow them time to teach their own five- or six-year-olds at home. Today it is just conceivable that the government could allow tax rebates for this purpose, but the administrative and political obstacles are formidable.

[9 ] Bearing in mind that everybody would be more able to afford tuition-charging schools to the extent that they would be asked to pay less indirect taxation (which nobody now escapes) than a state “free” school system makes necessary.

[10 ] For the application of the neighborhood effects argument to education in England, see the Report of the Committee on Higher Education (London: HMSO, 1963), I, chap. i, and Appendix IV, Part III. See also 15 to 18 Report of the Central Advisory Council for Education [Crowther Report] (London: HMSO, 1963), chap. vi. For further modern critical treatment of the general argument, see J. M. Buchanan, “Politics, Policy and the Pigovian Margins,” Economica, XXIX (Feb. 1962), 28; R. H. Coase, “The Problem of Social Cost,” Journal of Law and Economics, III (Oct. 1960), 1-44: and R. Turvey, “On Divergence between Social Cost and Private Cost,” Economica, XXX (Aug. 1963), 309-13.

[11 ] Expressed more formally: the persistence of such “externalities” as smoking chimneys may be quite consistent with optimal allocation of resources since the costs of using the market and/or government agencies to correct them may be larger than the potential gain.

[12 ] Another aspect requiring deeper analysis: the larger the influx of new residents to the area, the more harmful the effects of the smoke and the higher the tax necessary. But this increasing tax is itself an increasing harm to the firm, thus encouraging it to move or expand elsewhere. Since each new resident is not aware of the harm he is imposing on the firm and its employees, he is causing an adverse neighborhood effect; and one which is not counteracted by government intervention, but indeed has its origins in such intervention. This is the argument of R. H. Coase, loc. cit.

[13 ] House of Commons, Hansard, April 19, 1847.

[14 ] W. T. Thornton, Over Population and Its Remedy (London: Longman, 1846), p. 379.

[15 ]Report of the Committee on Higher Education (London: HMSO, 1963), Cmnd. 2154. Appendix IV, Part III, Para. 54. Italics supplied.

[16 ] September 1963. Mr. William Singer, President of the Ulster Teachers’ Union, told his annual conference on April 21, 1965: “There is a growing body of opinion which believes that our educational system must bear its share of responsibility for many of the problems of behavior, which show themselves in juvenile delinquency and vandalism.”

[17 ] Barbara Wootton, Social Science and Social Pathology (London: Allen and Unwin, 1959), p. 80.

[18 ]Causes of Crime (London: Weidenfeld and Nicholson, 1958).

[19 ] Crowther Report, op. cit., Para. 63.

[20 ]Ibid.

[21 ] This proposition is examined in West, op. cit., chap. vii.

[22 ]Half Our Future (London: HMSO for the Ministry of Education, 1963).

[23 ]Ibid., p. 79.

[24 ] Further detailed investigation of assumptions concerning other allegedly favorable spillover effects from education, for instance, upon economic growth, equality of opportunity, and social cohesion, will be found in West, op. cit.

[25 ] See West, op. cit., chap. ix.

[26 ] T. R. Malthus, Essay on Population (London: Dutton Everyman, 1958), p. 190.

[27 ] R. K. Webb, “The Victorian Reading Public,” in From Dickens to Hardy (London: Pelican Books, 1963).

[28 ] Moreover, the effects of the subsidies to schools were probably more than offset, in the early years at least, by the continuation of the “taxes on knowledge,” i.e., the enormous taxes on paper, newspapers, and pamphlets which were not removed until the 1850’s and 1860’s. Here we have a good instance of the way in which government action itself (this time the way in which taxation was levied) can result in socially adverse spillovers or “neighborhood effects” (i.e., social costs). If it were seeking the positive neighborhood effects of education, it would have been more practical for the government first to have attended to the negative neighborhood effects for which it was responsible; that is, it should have abolished the “taxes on knowledge” before considering subsidizing the pursuit of it. To subsidize and tax the same activity is illogical and costly.