4. Private and Public Expenditures: A Reappraisal

4 Private and Public Expenditures: A Reappraisal ERNEST VAN DEN HAAG In The Affluent Society) John Kenneth Galbraith per suasively refurbishes the arguments in favor of widening the public and narrowing the private sector of our economy. Let me examine some of his major ideas.l I Galbraith argues foremost that there is an "imbalance" between the public and the private sector. Economists have overlooked that "imbalance" because their obsolete "conven tional wisdom" leads them to focus on "scarcity," whereas we have "affluence." Galbraith's argument here rests on his confusion or equivoca tion between the technical meaning of "scarcity"-i.e., need for allocation-and the colloquial meaning-i.e., insufficiency, or poverty. If the two meanings of "scarcity" are separated, his argument is seen to be without merit. Economists are aware that our affluence has increased (col loquial meaning); but affluence does not make rational allocation obsolete (technical meaning). Allocation, the subject matter of economics-the problem: how can we make the best use of our resources?-becomes "obsolete" only with the millennium. For, 51 52 The New Argument in Economics though we can and do produce a great deal in temporal society, we cannot produce enough to satisfy all desires; thus it remains rational to allocate-i.e., to choose between alternative satisfac tions, and to economize-i.e., to satisfy desires with the least expenditure of resources.

Indeed, Galbraith himself advocates forcible reallocation of resources from the private to the public sector, thus implying that the scarcity, which makes allocation (choice) necessary, and which he is at pains to deny, is still with us. In his equivocation between the technical and colloquial meanings of "scarcity," Galbraith is not altogether original. He follows the "conven tional wisdom" of millenarians, utopians, and Marxians. How ever, the chiliastic sects realized that there is scarcity in temporal society; and the Marxists realized that there is scarcity in non Communist society. The discovery that scarcity is "obsolete" in our society is original with Galbraith, but no improvement on "conventional wisdo'm." (The supposed disappearance of scarcity leads Galbraith to conclude that we need no longer emphasize productivity-a consistent conclusion, as dangerous as it is wrong.) Galbraith's "imbalance" itself turns out to be a rhetorical device disguised as an argument. No indication is given-let alone substantiated-as to where a "balance" could be found, or by what means one might locate it. Hence the word "imbalance"

does no more than denote Galbraith's dissatisfaction with the present allocation of resources between the public and the private sector; his conclusion merely restates that premise; and his reasoning but affirms what it is supposed to prove, namely, that the public sector should be expanded. (Galbraith suggests, often with striking illustrations, that spe cific public services are lagging. He takes it for granted that the lag is caused by insufficient financing. The term "imbalance" helps to evade analysis of the distribution and of the effectiveness of the use of money within the public sector-the actual problems causing most of the "deficiencies" Galbraith mentions. More money is not likely to improve matters and may make them worse. Reorganization is called for.) Private and Public Expenditures: A Reappraisal 53 Economis ts took the idea of balance from mechanics to describe a state without an endogeneous tendency to change (equilibrium). Galbraith gives it a laudatory sense-balance becomes an unspecified but' ideal distribution between public and private sectors-and proceeds to use the descriptive term prescriptively. By this device, he retains the authority the term derived from its objective reference, though he has surreptitiously cut it adrift. This public relations stratagem has worked well; in defense of reason, I suggest that from now on, unless an author tells us how he determines "imbalance" in the economy, we ought to grant only that it exists in his mind.

Colin Clark has tried to prove that we suffer from the op posite imbalance, that the public sector is overexpanded when ever it uses (or, better, attempts to use) more than 25 per cent of the national income. When that happens, he thinks, the ostensive welfare goals are defeated, and the economy generally suffers. At least Clark tells us how to locate his imbalance and its effects. He tries to prove a meaningful proposition. I do not think he succeeds, but Galbraith does not even try-and for good reason: he does not have a genuine proposition to which evidence could be relevant. In Galbraith's defense, one may argue that some public expenditures are directly complementary to private ones. With out roads, cars are of no use. But even here, "imbalance" is misleading. The proper quantitative relationship between road and car expenditures has not been worked out with any precision, and it may be impossible to do so in any way that would permit us to speak of a general "balance."

Here Galbraith follows convention, but would have done better to abandon it: we usually treat public expenditures as a dependent variable and the private expenditures to be comple mented by them as an independent one. Yet this. ~s certainly wrong. Both must be treated as dependent variables. They are functionally interdependent. The problem is not: how much road space is needed for automobile traffic growing at a given rate? but: how much automobile traffic should be supported, induced or balked, in view of public costs and various alter54 The New Argument in Economics native expenditures and means of transportation? Roads invite automobile traffic as much as they accommodate it. (The unde veloped countries illustrate very well that automobile traffic is as much a function of road-building as vice versa.) Other factors-taxation, public transport, etc.-also influence automo bile traffic. When roads are built with public funds, surely we must base decisions on the comparative desirability of foster ing and accommodating automobile traffic, and not exclusively on the desires of actual and prospective automobile users. (In the case of toll roads, this problem differs-although toll roads do involve costs in addition to those paid for by users.) To treat public expenditure as a dependent variable com plementing independent private expenditures is about as reason able as it would be to say, "Since many people want to drive at a speed of 100 miles per hour, injuring more pedestrians than we now can take care of, we must multiply hospitals and cemeteries." Should we not instead, or as well, discourage speed and other hazards, and perhaps automobile traffic?

Whenever possible, the cost of the public complement to private expenditure should be defrayed by the most direct and main beneficiaries. A fuller application of this ancient rule of fiscal equity might, apart from its intrinsic merit, greatly reduce public expenditures: the private expenditures that are now subsidized by complementary public ones might be reduced if the complementary public costs are borne by the direct private beneficiaries through taxes.2 Though easily abused, the attempt to determine theoretically complementarities between specific public and private expendi tures can be fruitful when the relationship between the inter dependent variables is treated as independent, and both variables are regarded as dependent. I doubt, however, that the attempt to find a general balance between public and private expenditure is fruitful. I can conceive of situations where 70 per cent of the national income might well be spent publicly even in peacetime and of others where ten per cent is too much. So many variables are involved-e.g., size and distribution of the national income, type of expenditure, tax structure-that I do not think a general Private and Public Expenditures: A Re'appraisal 55 rule can do justice to concrete situations unless it be interpreted as a warning to pay heed to the undesired and possibly self defeating effects .of very high general levels of taxation. Clark's attempt to weigh diminishing returns of taxation certainly has this merit.

II Galbraith next offers a value judgment: ·consumers spend too much on trivialities. I share his judgment; but he miscon ceives the problem and proposes irrelevant and indefensible solutions, likely to make matters worse. Galbraith argues as though the problem were simply that there are not so good (i.e., private) and good (i.e., public) expendi tures; and he calls on all right-thinking citizens to make sure that the "good guys" (government) get more money and the "bad guys" (consumers) less. This will do for a Western; but is it economic analysis? It replaces the all-too-real problem: What values are we to live by, or how can people be free and wise? with a pseudo problem: How can we keep people right thinking and buying? and then offers a pseudo solution: by letting the government spend more of their income for them. Thoughtful men have always agreed that consumers prefer trivialities and vulgarities to the satisfaction of their real needs; but they have never agreed on the. real needs that ought to be satisfied, or on how to make consumers do what is good for them. This is not a new problem arising from aflluence and advertising, as Galbraith suggests. Nor is Galbraith's. solution new. It ignores what makes the problem problematic: neither an objective nor an agreed-upon standard by which purchases could be judged more or less trivial in general is available. (I am making a factual statement: such a standard is not available.

Nothing is implied about its existence; and should it exist, about its nature, applicability or imposition.) Since Plato, those who in the past proposed what Galbraith proposes-let the government decide (and spend) more, con sumers less-were more consistent than he: they did not believe 56 The New Argument in Economics in individual freedom; and they felt that right values can be objectively ascertained and collectively imposed. If Galbraith held these views, his theory would be consistent, however unac ceptable to me. But though they are implied in much of The Affluent Society) I think he would repudiate such views when stated explicitly. One of the major defects of that book is that it disguises inchoate and chaotic philosophical ideas as economic analysis. Trivialities and vulgarities now constitute a greater proportion of our total output than in the past, not because taste has de teriorated-it probably was always bad-but because it is satisfied more often. Affluence changes the ability to satisfy taste rather than the taste: whereas in the past only the rich had the privilege to indulge their taste, now the poor can too. I do not see why transferring expenditures to a government no wiser than the voters who elect it and less able to satisfy individual tastes would solve this problem. It would probably replace private folly with official silliness. Si monumentum quaeris) circumspice. And it would make harder the satisfaction of the minority tastes (in cluding Galbraith's), which are usually more interesting than those of the majority.

I see one advantage of democracy in making reasonably cer tain, not that the government is better than the voters in general, but that it is not much worse; without democracy, this last happens quite easily. Still, in a democracy, if the government is likely to be not worse than the average voter, it is also likely to be worse than some voters. Therefore, the transfer of the power to decide from the individual to even a democratic govern ment is justifiable only where the object is so indivisible, or indiscriminate, as to make collective decisions imperative. The argument in favor of expansion of the public sector need not be based on the demerits of private expenditure alone. It can be based also on the merits of public expenditure. But these can never be general. Each expenditure must be considered on its own merits: Will benefits exceed costs by more than the benefits of alternative expenditures, public or private? How are benefits and costs distributed? The admitted triviality of many Private and Public Expenditures: A Reappraisal 57 private expenditures and the "imbalance" -whatever it means -are utterly irrelevant unless the general superiority of public expenditures can be shown. By pointing to the triviality of private expenditures-which each consumer is likely to admit in regard to the expenditures of all other consumers-Galbraith suggests that the government would restrain those other con sumers, and that, anyway, it could not do worse. But it always can and usually does. (These two fallacies are perennial asser tions in the appeals for dictatorship: [1] followers of would-be dictators usually think, that matters can only improve; and [2] that the dictator would frustrate the [trivial] wishes of others, but not their own [un trivial] ones.) III Though/reproaching consumers for their silliness, Gal brai th does not hold them responsible for it: according to the principles of ritualistic liberalism, the people always have their heart (or is it their stomach?) in the right place; when they err, it is because they have misread their heart's prompting, owing to some wicked seducer. This theory is popular all around: it enables people to eat their cake and profess that they did not really want it; and it enables the theorist to have his cake (people are good) and eat it (people act badly). In the past, the seducers were devils, capitalists, or Jews. Madison Avenue is fast replacing them in the folklore of our society. Yet recent myths are less con venient than the ancient ones. Jews and capitalists suffer from them, whereas in the past the devils were the main sufferers.

Let me note that Galbraith's myth, though structurally analogous to those mentioned, if more urbane, is functionally quite differ ent: Galbraith does not urge liquidation of Jews, capitalists, or advertising men. His theoretical mountain gives birth to a quite modest, practical mouse: a general sales tax. But then mice can be quite voracious, and they multiply fast. (The general sales tax may have technical merits as such. But Galbraith wants it because he believes it to be the best means to achieve his basic end of increasing the general level of taxation, of 58 The New Argument in Economics decreasing private and increasing public expenditure. I object to this end, rather than to a sales as compared to an income tax.) Galbraith argues that consumers no longer satisfy endogeneous desires, but exogeneous ones "contrived" by advertising. Indus try thus first produces the desires it then satisfies, whereas before The Affluent Society) the needs satisfied arose independently.

Now to the extent to which we grant the truth and undesirability of this "dependence effect," it might argue against advertising. Surprisingly, Galbraith seems to think of it as an argument for reducing the purchasing power of consumers and increasing that of the government by means of a general sales tax. His not entirely explicit train of thought seems to be: consumers' purchases are trivial; they satisfy a demand "contrived" by advertising; therefore we should transfer purchasing power from consumers to the government; public expenditures will be less contrived and less trivial. There is no logical connection among the various parts of this argument. Why should the government not do worse than consumers? Why are public expenditures less (rather than differently) con trived than private ones? Is the political process that spontane ous? Why is "contrived," i.e., influenced demand better or worse (more or less trivial) than uncontrived, i.e., uninfluenced (if that is conceivable) demand? Is the "contrived" demand for education, books, and soap worse than uncontrived spontaneous dirt and ·ignorance? One may spontaneously desire trivial things -as any child knows and does. Culture is contrivance, i.e., social influence, and there is no society without it. There is no less contrivance in primitive, or for that matter, in Soviet society than in ours. The problem is not that there is influence ("contrivance"), but its quality, source and direction, above all, whether influence is monopolized or whether it comes from many competing sources. The high taxes Galbraith proposes obviously would not affect these problems except by reducing the power of individuals and adding to the power of the government. Influence would not be eliminated; it would be centralized .

. Galbraith's argument is irrelevant then. Is it true? Are con sumers hapless victims of Madison Avenue? One would expect Private and Public Expenditures: A Reappraisal 59 that they bought Ford's Edsel car in droves and not small foreign cars·; or that the political party that spends most on advertising always wins. Presidents Thomas Dewey and Richard Nixon, and the Republican Congressional majority would testify to these beliefs were they true. They are not. But believers are unshaken by evidence. Advertising is only one influence among others in political or purchasing decisions and is frequently offset by competing advertising. Yet advertising men as well as their opponents (for different reasons) cherish the belief that advertising is always the decisive influence. Advertising agencies want to impress prospective clients with the importance of what they may do for them. The motivation of opponents is more complex: though their faith is as strong. 50me are looking for scapegoats; others are mildly paranoid; still others want to convince people that they lose nothing in getting government "protection" from their own use of freedom, since advertising mysteriously has deprived them of that freedom anyway.

Many consumers spend their money in ways many other consumers disapprove of. The first group of consumers assumes that the second cannot possibly have freely decided to spend its money as it does. The disapproved expenditure pattern is taken for evidence of their lack of freedom, just as apologists for dictatorship have always presented the election of a less than perfect democratic government, or of one they disapproved of, as evidence of a lack of political freedom. If you do not "really" have it, what can you lose in giving it up? The moral connection between sumptuary and political free dom is anything but tenuous: if consumers are incapable of freely choosing among advertised products, why should they be deemed capable of freely choosing among propagandizing political parties? What is freedom if not the right to choose among competing influences? If these influences are considered coercion, or if people are deemed too incompetent to choose among them, freedom can never be more than sham, and we might just as well install a dictatorship. It would not follow that dictatorship will lead to better results; but there would be no reason to expect 60 The New Argument in Economics worse ones. We would give up something we never had, or were never able to use competently. But I like freedom because it allows me to make and judge my own choices, however foolish.

We might well oppose and seek to restrain seduction. But we cannot (unless it be practiced on the legally incompetent) treat it as though it were rape, precisely because it gives choice (and therewith part of the responsibility) to all parties con cerned: it requires consent. Democracy does so no less than the "dependence effect": with freedom there always are com peting influences. And advertisements for detergents no more hypnotize or coerce than advertisements for political candidates. If people are competent as citizens to choose among candidates, why should they not be competent as consumers to choose among goods? (I find it hard to believe that those who ~onfuse influence with coercion can be sincere, but I may overestimate both their intelligence and their malevolence.) Has the influence of advertising increased? The proportion of the population affected certainly has. Not, however, because Madison Avenue has discovered new tricks-there seems to have been little basic progress in propaganda techniques since Roman times-but because literacy and other communications media have spread; and because many people who did not in the past, do now have enough disposable income to follow fashions in purchasing. However, even the fixed purchasing patterns of the past were socially influenced. The peasant's dress, food, and hous ing were no more individual creations than the wigs of noble men, or their extravagant garments. Nor were these things less trivial than the things advertised in Vogue today. Versailles, a medieval cathedral, or Vanderbilt's yacht were neither more of a necessity or more spontaneous than our tail-finned cars, or a copy of The Affluent Society-though perhaps of more lasting value.

The actual change that has occurred is not any new "depend ence effect," but a general change from a pre-industrial, tradition directed to an industrial, consumer-directed society with its concomitant mass culture. 3 This means that social influences are Private and Public Expenditures: A .Reappraisal ()l more changeable, that they come more often from helow, that people are more attuned to them (more outer-directed). Rather than coming mainly from the nobility, the king, and the church, or having ~een internalized, influences currently come from a greater variety of sources; the advertising industry is among them, though pertIaps it transmits more than it creates. Characteristically, advertising agencies are not in the per manent service of a class, ideology, or church, but of whoever hires their services. Mostly private firms try to increase or maintain their profits by doing so. But firms are as interested in producing what people want as they are in making people want what they produce. The former is more profitable and more certain of success. In this sense, advertising probably is less engaged in '''contriving'' needs than the church was, or than Harvard University is when it raises funds. Indeed, the bad taste Mr. Galbraith deplores is indulged in largely because of the absence of the ancient contriving agencies-church and court. We are now catering to the taste of the masses, whereas before it could be ignored. There is no evidence to indicate that people would buy fewer or better things without advertising.

They might,buy slightly different things from slightly different people. IV The public expenditures Galbraith advocates to replace private expenditures-more public works, education, welfare services~are neither new nor supported by new arguments. This part of The Affluent Society amounts to tiresome exhortation. Yet a discus.sion of the specifics of public expenditures might have been fihitful. I am convinced that we need public services not now ulldertaken, and that many public services presently offered at high cost to taxpayers are "contrived": they become necessary because the government prevents private industry from rendering them or makes it excessively costly to do so, or finally, because they, offset noxious government activities elsewhere. A classical ilhistration is the case of farm subsidies.

62 The New Argument in Economics Through price support the government raises farm prices, i.e., the cost of living, mainly the price of food, but also of cotton, tobacco, etc. Taxpayers, via the government, spend more than $5 billion annually for the privilege of buying food at prices higher than those of a free market. This money is spent to buy supplies that would depress the market price, or to pay farmers for not producing them. No way to dispose of most of the supplies purchased has been found; they are stored at huge additional cost until they spoil. Some are sold at a loss abroad or given away. The subsidy, of course, perpetuates the misal location of resources between agriculture and the rest of the economy that it is supposed to correct, and also within agricul ture. More is produced of what is needed less, and less of what is needed more, ad infinitum} and subsidies are required ad infinitum to keep it that way.

There is no respectable argument known to economists for these subsidies. The argument usually heard is that without them the income of farmers would be lower than that of workers, or lower than it was in some past period. But most of the subsidy does not go to low-income farmers; their farms are so small and unproductive that higher prices or payments for noncultivation of their property add little to their income. Most of the subsidy goes to big-scale producers of unneeded farm products. (50 per cent of our farmers produce less than 10 per cent of the agricul tural output.) If we wanted to help low-income farmers, we could do so without raising food prices and fostering misallocation, and without subsidizing farmers whose income far exceeds the in come of the taxpayers who must pay the subsidy. Without chang ing the whole complex structure of present legislation-a task that should not be neglected, but that requires time-the govern ment could do so simply by purchasing only from farmers whose net income from all sources in any given year does not exceed, say, $7,000. (Surely farmers with higher incomes need no subsidies.) Any amount paid to· farmers whose income exceeds $7,000 must be returned-and the government can return what it purchased from them. In time the government might Private and Public Expenditures: A Reappraisal 63 limit .itself to purchases from progressively older farmers. This ultimately would solve the problem; only those would remain or go into farming who can make an income that satisfies them by selling their crops at unsubsidized market prices. This, of course, would correct the allocation of resources and make further subsidies unnecessary.

I have outlined elsewhere how this proposal could be carried out and what the effects would be.4 I have received heartening letters from economists. But nothing else. The mood of our times is such that if a problem can be perpetuated and made worse by high subsidies and brought nearer solution by low ones, we prefer high subsidies. I deliberately speak of the mood of the times. For in terms of their economic interests, the farmers who would not lose under the proposal-let alone consumers-constitute an overwhelming majority. But our gen eral tendency is to make individual efforts less and less de pendent on the value placed by the market on one's products or services, and more and more on political considerations: on the political power exercised by the selling group, or the power that politicians can gain by catering to it. Above all, the nature of the subsidy is carefully hidden. We speak of price supports when we make relief payments to some farmers who need them (and who should be trained for a different occupation rather than kept farming and receiving relief) and to many who do not, By now the farm program has become a classical case. Some newer but equally dubious programs seem more complicated.

For instance, the government feels it necessary to subsidize low and middle-rent apartment-house building in many cities. It is contended that low-and middle-income families, although more affluent than ever before, cannot afford the rent for unsubsidized housing. But if their income is too small, we should subsidize them, not housing. Why a subsidy in kind, which deprives them of choice? Is it feared that they would make a choice that ap pears wrong to the government? What evidence is there to show that they are· wrong and the government right? As soon as we turn from vacuous generalizations about the affluent private and the starved public sector of our economy 64 The New Argument in Economics to an actual scrutiny of public expenditures, we find that, though it is true that some needed services are not performed, or not performed well, it is also true that many unneeded services are performed at immense cost. Often they serve as pretexts to perform still other services or pay subsidies to offset the effect they have. Increasing public expenditure is unlikely to remedy this state of affairs; it will make it worse.

v Is additional public expenditure on education necessary, sufficient, or at least favorable to additional learning? The im pression that the more we spend for education, the more educa tion we get, is widely held but sadly mistaken. It derives, of course, from individual experience in purchasing items of elastic supply and easily measured quality. Unfortunately, education does not have these characteristics. The supply both of educable students and of capable educa tors is limited, particularly in the short run. Given a minimum that we have long exceeded, the size of expenditure for education plays a comparatively small part in determining whether this supply is productively utilized and increased in the long run. Since neither teaching and research ability nor educability is measurable except in crude and arbitrary probability terms, immense amounts of money can be easily wasted without anybody's being the wiser. (Note, particularly, that most meas urements tend to include, or altogether to rest on, the inter pretative scheme they are supposed to validate.) The available resources, including money, are used badly now. Simple increases of total expenditure are more likely to hinder the necessary changes than to foster them.

Much of the clamor for more expenditure on "education" stems from a confusion. The solution of the world's problems by philosophy, religion, or psychology certainly is more impor tant than chewing gum. But people spend more money on gum than on philosophers, ministers, or psychologists. Food is more important than sartorial elegance, but many people spend more Private and Public Expenditures: A Reappraisal 65 on clothing. Health is more important than beer, but it does not follow that more should, or could, usefully be spent on medicine than on beer. The point is simple, but the confusion ubiquitous: how often are we told that more is spent on something less essential than on educationas though this were evidence that too little is spent on education! Yet expenditure need not and cannot be proportional to importance, moral or material. It must pay for the cost of whatever can be produced and used of the important thing or service. Any expenditure beyond that is waste-no matter how much more money be spent on less important things.

The returns on educational expenditure have been diminish ing. To increase educational expenditure now is like giving sugar to diabetics who crave it no less than educational lobbies crave money, and for the same reason: improper utilization. Unfortunately, the lobbies are supported by many well-meaning people who have been assured that the money will result in better nutrition. But the trouble is that the patient cannot actually absorb it. Education is weak, .but not undernourished. It is weak partly from obesity, partly from the faulty diet, and, above all, from lack of exercise. More money may assuage our guilt feelings, but it will do nothing else. Let us look at some of the specifics. The "classroom shortage" is often cited as evidence for too little expenditure. At present about 70,000 new classrooms are built per year. At the usual 28 pupils per classroom, the need (compounded of the backlog owing to overcrowding and deteri oration, and increased enrollment) is for about 60,000 new classrooms per year. 5 There are, of course, local shortages and surpluses-owing both to shifts of population and to local mis takes-but no general shortage and certainly no needfor·an over all federally financed building program.

This calculation does not take into account three obvious measures which, singly or in combination, could relieve any fore seeable shortage if it were to occur, and which might well make it possible to reduce the present high rate of school build ing. They are: lengthening the school year by one month; teach66 The New Argument in Economics ing on Saturday; double shifts. (By teaching on Saturday, and/or teaching a month longer per year, schools could make up any loss of time through double shifts.) The five-day week most adults enjoy was the result of in creased productivity and of a decreased need for production. But the need for education has not decreased, nor has its productivity increased. The long summer vacations (which adults are not usually granted) derive from the original need for sum mer child labor in an agricultural economy. This need no longer exists. And reduction of vacation time by one month leaves plenty for students and teachers. Finally, why should schools not be kept open, say, from 8:00 a.m. to 6:00 p.m.? By proper utilization of gymnastic and other special facilities, this span, with double shifts, could amount to six hours of schooling per day per child. If desired, this arrangement might also make it easier to dispense with expensive and elaborate cafeterias. The total, or the yearly school work of students or teachers, need not be increased. Utilization of school buildings by greater numbers of children and teachers is thus possible without re quiring additional classrooms. The children and teachers who attend on Saturday may take off on other days.

A second reason frequently offered for greater public expendi ture on education is "the teacher shortage" and the "need to raise teachers' salaries." There was a teacher shortage in the early 1950's owing not so much to "underpayment" as to sharp increases in enrollment. Since that time, the increase in the number of teachers has regularly exceeded the increase in the number of pupils, so that the National Education Association itself admits that by 1962 enough new fully certified teachers will be available to fill all the new positions and those left by teachers who have quit. This certainly does not indicate that salaries must be raised to get more teachers. Teachers themselves, though they complain, apparently feel that they could not earn as much as gratifyingly in other occupa tions. About nine per cent of all classroom teachers quit their jobs every year. One-third of these do so to get married; it is unlikely that a higher salary would prevent this. Another twenty Private and Public Expenditures: A Reappraisal 67 per cent just change jobs within the teaching profession. Fifteen per cent retire for old age or disability. Only about thirteen per cent of those who leave quit teaching for other occupations.

This is about one per cent of all classroom teachers. The figure does not indicate economic pressure to leave. An unknown number of those who have quit return; and we do not know how many among this one per cent of the total take other jobs because they pay better, and how many prefer them for non financial reasons. Moreover, every year the number of those entering teaching from other gainful employments exceeds the number of those leaving for nonteaching employments. Something of the nature of the shortage is revealed by the New York Times) reporting, under the front-page headline (9/18/53) "200 New City Teachers Missing," that 200 newly appointed teachers had failed to show up because of low salaries. On 9/22/53, the same New York Times reported-this time on page 33 and without an editorial deploring the financial plight of teachers-"2,OOO Flock to Get City Teachers' Jobs." Accord ing to the Daily News) there were 3,OOO-and the jobs were those very 200 underpaid jobs.6 (As do all newspapers, the Times selects news and its prominence according to reader interest and expectations. Readers sensitized to a "teacher shortage" and "underpayment" are likely to be more interested in the first rather than in the second story. The Times does not seem to be worse than other papers, possibly better, certainly more dignified and comprehensive. How useful or ethical this news selection is, lies beyond my scope here to discuss.) This is not to say that there is no shortage; there is in some fields, just as in others there is a surplus. Nor can we conclude that teachers' pay is satisfactory. Some teachers are underpaid as others are overpaid. But the comparative income of teachers has improved. In 1909, the yearly salary of teachers was about the same as that of workers in manufacturing. Today, teachers make about $500 more than workers in manufacturing (including only the total time worked on the average: fifty weeks for workers, forty for teachers). Teachers' salaries also have risen more than those of government workers, dentists, and lawyers.

68 The Nero Argument in Economics All these data depend on the periods selected for comparison. But the trend suggested is unmistakable. Of course, the educa tional requirements-or, better, the degree requirements-which teachers must fulfill have risen. Sixteen to seventeen years of formal education are now required for all teachers in almost all states. (Abroad, twelve years is the norm for elementary school teachers.) Would higher salaries attract more able teachers? In, occupa tions such as art, politics, literature, scholarship, judicial admin istration, and teaching, the average quality of the personnel attracted depends on the income offered only in the sense that a comparative minimum must be exceeded (even this is largely conjectural). There is no evidence that the quality is improved by raising incomes, particularly average incomes, beyond this minimum. The attractiveness of these occupations depends on the inherent satisfaction of pursuing them, and on such matters as prestige and intraoccupational advancement opportunity.

The interoccupational mobility between white-collar and blue collar occupations is small. The student who considers teaching asa career compares it only with other white-collar occupations. No possible salary scale is likely to attract those whose major ambition is to become rich; they will go into business or an independent profession. On the other hand, the girl who even tually wants to raise a family but who has some intellectual interests as well, and who likes children in general, may find teaching attractive. Whether she becomes a teacher will not be greatly influenced by salary-provided her income compares favorably with white-collar occupations as easily open to her and not much less attractive. Exceptional teachers, who are certainly needed, must be attracted by exceptional advancement possibilities in prestige and money "vithin the occupation, though these cannot be expected to compare with those in business and the professions. Increases in the average salary will not attract these teachers: they are not distinctive and cannot be great enough. An across-the-board raise will have no positive effects whatever in intensifying the attraction teaching holds for those actually talented for it, or in fostering a better utilizaPrivate and Public Expenditures: A Reappraisal 69 tion of the talent already available in the schools. To make teaching attractive for the talented, we must raise the standards of teacher education. We need notprolong the years of study, but we can make them more demanding and less silly by eliminating "education" courses and substituting subject-matter courses. The idea that more money is an all-powerful remedy is abetted by generosity and by the strong guilt feelings Americans seem to have toward their young: people realize that all is not well with education. And yet there is an unwillingness among teachers and parents to scrutinize the difficulties that have brought about the present state of. affairs.

If we look at the matter in comparative terms-and there are no absolute standards-we find that our expenditure on educa tion is extraordinarily high, and that our troubles are clearly related to the way these expenditures and, indeed, all our educa tional resources have been utilized. Less may be more here. The proportion of the national income spent by the United States on public education compares favorably with that of other advanced nations,7 and the per capita expenditure is·higher than that of any other country. As a percentage of our expenditures on consumption, expenditures on public education have risen~ contrary to Galbraith's unsupported surmise-from 3.1 per cent in 1929 to 4.3 per cent in 1956; this despite the immense increase in consumption expenditure. Per capita expenditure for public education (calculated in constant dollars) has risen steadily in the United States; so has expenditure for private education; so has the percentage of the national income spent on public educa tion and that spent on private education. Average expenditure per pupil per year, in constant dollars} has increased by more than 600 per cent since 1900. We spend, in real terms, six times as much for each pupil than we did sixty years ago-hardly a sign of financial neglect. On the contrary, one may suspect that the cost of education has risen by more than is justified by its accomplishments.

In terms of formal schooling, our expenditures have assured our pre-eminence. Thus, in the United States, 66.2 per cent of the 15-to 19-year-olds are in school, versus 17 per cent in 70 The N ewA rgument in Economics Europe and 48.6 per cent in the Soviet Union; and 12 per cent of the 20-to 24-year-olds, versus 3.7 per cent in Europe and 8.2 per cent in the Soviet Union. The number of school days per year has increased as well (though abroad it often remains greater). The size of classes, i.e., the teacher-student ratio has been consistently declining. (Incidentally, there are mountains of research but no conclusive evidence linking class size to learning rates; after kindergarten it seems to matter little.) Our school buildings and our auxiliary personnel and apparatus are far more opulent and costly than those of any other nation. We are still advancing fast, perhaps too fast, in statistical terms. According to theU nited States Office of Education, of 1,000 fifth-grade pupils in 1924, 612 were in high school four years later, whereas of 1,000 fifth-grade pupils in 1952, 904 were in high school in 1960 (nearly 50 per cent more); of the first 1,000, 30 per cent graduated; of the second, 60 per cent (i.e., 100 per cent more). And, whereas only 118 out of 1,000 fifth-grade pupils· in 1924 entered college (in the depression year 1932), 319 out of 1,000 in the fifth grade entered college in 1960. The difference leaves· no doubt about the enormous increase of our expense on education, and about the statistical effect it had: an increased proportion of the population is subjected to increased amounts of schooling. Doubts start when we consider the return on this investment.

We spent, and are spending, a great deal of money and, in purely statistical terms, the results are excellent. Our trouble is that students do not learn much in school, and, in particular, not enough of what they need to learn. It is difficult to see how still better buildings, more highly paid teachers, or greater ex~ penditures would solve a problem that has been created largely by emphasis on total expenditures and neglect of their use, of education.

Private and Public Expenditures: A Reappraisal 71 NOTES 1. Most of Galbraith's arguments, were they correct, would not be neces sary, and altogether they are not sufficient, for the conclusions he draws. (1 have discussed The Affluent Society as a whole in Commentary, Sep tember, 1960, and January, 1961.) 2. The general sales tax proposed by Galbraith in place of the excise taxes here suggested would have the opposite effect. 3. See Ralph Ross and Ernest van den Haag, The Fabric of Society} passim} particularly chap. XV. 4. See "A Modest Farm Proposal," The New Leader} February 27, 1961. 5. The most reasonable analysis of the incredibly confused issue is found in Roger A. Freeman's School Needs in the Decade Ahead (Washington, D.C.: The Institute for Social Science Research, 1958; 3rd ed., 1960). His projections have since been abundantly confirmed. 6. Quoted by Freeman, Ope cit.} p. 67.

7. This proportion is mentioned here because frequently quoted, and usually wrongly. It is not very significant, because (a) with a high na tional income, a smaller percentage is needed; (b) demographic factors distort the picture; (c) much depends on how much education is pub licly financed. Much depends finally on how "education" is defined and how expenditure is calculated.