Chapter X

The Distribution of Income

A central element in the development of a collectivist sentiment in this century, at least in Western countries, has been a belief in equality of income as a social goal and a willingness to use the arm of the state to promote it. Two very different questions must be asked in evaluating this egalitarian sentiment and the egalitarian measures it has produced. The first is normative and ethical: what is the justification for state intervention to promote equality? The second is positive and scientific: what has been the effect of the measures actually taken ?

The ethical principle that would directly justify the distribution of income in a free market society is, "To each according to what he and the instruments he owns produces." The operation of even this principle implicitly depends on state action. Property rights are matters of law and social convention. As we have seen, their definition and enforcement is one of the primary functions of the state. The final distribution of income and wealth under the full operation of this principle may well depend markedly on the rules of property, adopted.

What is the relation between this principle and another that seems ethically appealing, namely, equality of treatment? In part, the two principles are not contradictory. Payment in accordance with product may be necessary to achieve true equality of treatment. Given individuals whom we are prepared to regard as alike in ability and initial resources, if some have a greater taste for leisure and others for marketable goods, inequality of return through the market is necessary to achieve equality of total return or equality of treatment. One man may prefer a routine job with much time off for basking in the sun to a more exacting job paying a higher salary; another man may prefer the opposite. If both were paid equally in money, their incomes in a more fundamental sense would be unequal. Similarly, equal treatment requires that an individual be paid more for a dirty, unattractive job than for a pleasant rewarding one. Much observed inequality is of this kind. Differences of money income offset differences in other characteristics of the occupation or trade. In the jargon of economists, they are "equalizing differences" required to make the whole of the "net advantages," pecuniary and non-pecuniary, the same.

Another kind of inequality arising through the operation of the market is also required, in a somewhat more subtle sense, to produce equality of treatment, or to put it differently to satisfy men's tastes. It can be illustrated most simply by a lottery. Consider a group of individuals who initially have equal endowments and who all agree voluntarily to enter a lottery with very unequal prizes. The resultant inequality of income is surely required to permit the individuals in question to make the most of their initial equality. Redistribution of the income after the event is equivalent to denying them the opportunity to enter the lottery. This case is far more important in practice than would appear by taking the notion of a "lottery" literally. Individuals choose occupations, investments, and the like partly in accordance with their taste for uncertainty. The girl who tries to become a movie actress rather than a civil servant is deliberately choosing to enter a lottery, so is the individual who invests in penny uranium stocks rather than government bonds. Insurance is a way of expressing a taste for certainty. Even these examples do not indicate fully the extent to which actual inequality may be the result of arrangements designed to satisfy men's tastes. The very arrangements for paying and hiring people are affected by such preferences. If all potential movie actresses had a great dislike of uncertainty, there would tend to develop "cooperatives" of movie actresses, the members of which agreed in advance to share income receipts more or less evenly, thereby in effect providing themselves insurance through the pooling of risks. If such a preference were widespread, large diversified corporations combining risky and non-risky ventures would become the rule. The wild-cat oil prospector, the private proprietorship, the small partnership, would all become rare.

Indeed, this is one way to interpret governmental measures to redistribute income through progressive taxes and the like. It can be argued that for one reason or another, costs of administration perhaps, the market cannot produce the range of lotteries or the kind of lottery desired by the members of the community, and that progressive taxation is, as it were, a government enterprise to do so. I have no doubt that this view contains an element of truth. At the same time, it can hardly justify present taxation, if only because the taxes are imposed after it is already largely known who have drawn the prizes and who the blanks in the lottery of life, and the taxes are voted mostly by those who think they have drawn the blanks. One might, along these lines, justify one generation's voting the tax schedules to be applied to an as yet unborn generation. Any such procedure would, I conjecture, yield income tax schedules much less highly graduated than present schedules are, at least on paper.

Though much of the inequality of income produced by payment in accordance with product reflects "equalizing" differences or the satisfaction of men's tastes for uncertainty, a large part reflects initial differences in endowment, both of human capacities and of property. This is the part that raises the really difficult ethical issue.

It is widely argued that it is essential to distinguish between inequality in personal endowments and in property, and between inequalities arising from inherited wealth and from acquired wealth. Inequality resulting from differences in personal capacities, or from differences in wealth accumulated by the individual in question, are considered appropriate, or at least not so clearly inappropriate as differences resulting from inherited wealth.

This distinction is untenable. Is there any greater ethical justification for the high returns to the individual who inherits from his parents a peculiar voice for which there is a great demand than for the high returns to the individual who inherits property? The sons of Russian commissars surely have a higher expectation of income -- perhaps also of liquidation -- than the sons of peasants. Is this any more or less justifiable than the higher income expectation of the son of an American millionaire ? We can look at this same question in another way. A parent who has wealth that he wishes to pass on to his child can do so in different ways. He can use a given sum of money to finance his child's training as, say, a certified public accountant, or to set him up in business, or to set up a trust fund yielding him a property income. In any of these cases, the child will have a higher income than he otherwise would. But in the first case, his income will be regarded as coming from human capacities; in the second, from profits; in the third, from inherited wealth. Is there any basis for distinguishing among these categories of receipts on ethical grounds? Finally, it seems illogical to say that a man is entitled to what he has produced by personal capacities or to the produce of the wealth he has accumulated, but that he is not entitled to pass any wealth on to his children; to say that a man may use his income for riotous living but may not give it to his heirs. Surely, the latter is one way to use what he has produced.

The fact that these arguments against the so-called capitalist ethic are invalid does not of course demonstrate that the capitalist ethic is an acceptable one. I find it difficult to justify either accepting or rejecting it, or to justify any alternative principle. I am led to the view that it cannot in and of itself be regarded as an ethical principle; that it must be regarded as instrumental or a corollary of some other principle such as freedom. --

Some hypothetical examples may illustrate the fundamental difficulty. Suppose there are four Robinson Crusoes, independently marooned on four islands in the same neighborhood. One happened to land on a large and fruitful island which enables him to live easily and well. The others happened to land on tiny and rather barren islands from which they can barely scratch a living. One day, they discover the existence of one another. Of course, it would be generous of the Crusoe on the large island if he invited the others to join him and share its wealth. But suppose he does not. Would the other three be justified in joining forces and compelling him to share his wealth with them? Many a reader will be tempted to say yes. But before yielding to this temptation, consider precisely the same situation in different guise. Suppose you and three friends are walking along the street and you happen to spy and retrieve a $20 bill on the pavement. It would be generous of you, of course, if you were to divide it equally with them, or at least blow them to a drink. But suppose you do not. Would the other three be justified in joining forces and compelling you to share the $20 equally with them? I suspect most readers will be tempted to say no. And on further reflection, they may even conclude that the generous course of action is not itself clearly the "right" one. Are we prepared to urge on ourselves or our fellows that any person whose wealth exceeds the average of all persons in the world should immediately dispose of the excess by distributing it equally to all the rest of the world's inhabitants ? We may admire and praise such action when undertaken by a few. But a universal "potlatch" would make a civilized world impossible.

In any event, two wrongs do not make a right. The unwillingness of the rich Robinson Crusoe or the lucky finder of the $20 bill to share his wealth does not justify the use of coercion by the others. Can we justify being judges in our own case, deciding on our own when we are entitled to use force to extract what we regard as our due from others? Or what we regard as not their due ? Most differences of status or position or wealth can be regarded as the product of chance at a far enough remove. The man who is hard working and thrifty is to be regarded as "deserving"; yet these qualities owe much to the genes he was fortunate (or unfortunate?) enough to inherit.

Despite the lip service that we all pay to "merit" as compared to "chance," we are generally much readier to accept inequalities arising from chance than those clearly attributable to merit. The college professor whose colleague wins a sweepstake will envy him but is unlikely to bear him any malice or to feel unjustly treated. Let the colleague receive a trivial raise that makes his salary higher than the professor's own, and the professor is far more likely to feel aggrieved. After all, the goddess of chance, as of justice, is blind. The salary raise was a deliberate judgment of relative merit.

The operative function of payment in accordance with product in a market society is not primarily distributive, but allocative. As was pointed out in chapter i, the central principle of a market economy is co-operation through voluntary exchange. Individuals co-operate with others because they can in this way satisfy their own wants more effectively. But unless an individual receives the whole of what he adds to the product, he will enter into exchanges on the basis of what he can receive rather than what he can produce. Exchanges will not take place that would have been mutually beneficial if each party received what he contributed to the aggregate product. Payment in accordance with product is therefore necessary in order that resources be used most effectively, at least under a system depending on voluntary co-operation. Given sufficient knowledge, it might be that compulsion could be substituted for the incentive of reward, though I doubt that it could. One can shuffle inanimate objects around; one can compel individuals to be at certain places at certain times; but one can hardly compel individuals to put forward their best efforts. Put another way, the substitution of compulsion for co-operation changes the amount of resources available.

Though the essential function of payment in accordance with product in a market society is to enable resources to be allocated efficiently without compulsion, it is unlikely to be tolerated unless it is also regarded as yielding distributive justice. No society can be stable unless there is a basic core of value judgments that are unthinkingly accepted by the great bulk of its members. Some key institutions must be accepted as "absolutes," not simply as instrumental. I believe that payment in accordance with product has been, and, in large measure, still is, one of these accepted value judgments or institutions.

One can demonstrate this by examining the grounds on which the internal opponents of the capitalist system have attacked the distribution of income resulting from it. It is a distinguishing feature of the core of central values of a society that it is accepted alike by its members, whether they regard themselves as proponents or as opponents of the system of organization of the society. Even the severest internal critics of capitalism have implicitly accepted payment in accordance with product as ethically fair.

The most far-reaching criticism has come from the Marxists. Marx argued that labor was exploited. Why ? Because labor produced the whole of the product but got only part of it; the rest is Marx's "surplus value." Even if the statements of fact implicit in this assertion were accepted, the value judgment follows only if one accepts the capitalist ethic. Labor is "exploited" only if labor is entitled to what it produces. If one accepts instead the socialist premise, "to each according to his need- from each according to his ability" -- whatever that may mean -- it is necessary to compare what labor produces, not with what it gets but with its "ability", and to compare what labor gets, not with what it produces but with its "need."

Of course, the Marxist argument is invalid on other grounds as well. There is, first, the confusion between the total product of all co-operating resources and the amount added to product -- in the economist's jargon, marginal product. Even more striking, there is an unstated change in the meaning of "labor" in passing from the premise to the conclusion. Marx recognized the role of capital in producing the product but regarded capital as embodied labor. Hence, written out in full, the premises of the Marxist syllogism would run: "Present and past labor produce the whole of the product. Present labor gets only part of the product." The logical conclusion is presumably "Past labor is exploited," and the inference for action is that past labor should get more of the product, though it is by no means clear how, unless it be in elegant tombstones.

The achievement of allocation of resources without compulsion is the major instrumental role in die market place of distribution in accordance with product. But it is not the only instrumental role of the resulting inequality. We have noted in chapter i the role that inequality plays in providing independent foci of power to offset the centralization of political power, as well as the role that it plays in promoting civil freedom by providing "patrons" to finance the dissemination of unpopular or simply novel ideas. In addition, in the economic sphere, it provides "patrons" to finance experimentation and the development of new products -- to buy the first experimental automobiles and television sets, let alone impressionist paintings. Finally, it enables distribution to occur impersonally without the need for "authority" -- a special facet of the general role of the market in effecting co-operation and co-ordination without coercion.

A capitalist system involving payment in accordance with product can be, and in practice is, characterized by considerable inequality of income and wealth. This fact infrequently misinterpreted to mean that capitalism and free enterprise produce wider inequality than alternative systems and, as a corollary, that the extension and development of capitalism has meant increased inequality. This misinterpretation is fostered by the misleading character of most published figures on the distribution of income, in particular their failure to distinguish short-run from long-run inequality. Let us look at some of the broader facts about the distribution of income.

One of the most striking facts which runs counter to many people's expectation has to do with the sources of income. The more capitalistic a country is, the smaller the fraction of income paid for the use of what is generally regarded as capital, and the larger the fraction paid for human services. In underdeveloped countries like India, Egypt, and so on, something like half of total income is property income. In the United States, roughly one-fifth is property income. And in other advanced capitalist countries, the proportion is not very different. Of course, these countries have much more capital than the primitive countries but they are even richer in the productive capacity of their residents; hence, the larger income from property is a smaller fraction of the total. The great achievement of capitalism has not been the accumulation of property, it has been the opportunities it has offered to men and women to extend and develop and improve their capacities. Yet the enemies of capitalism are fond of castigating it as materialist, and its friends all too often apologize for capitalism's materialism as a necessary cost of progress.

Another striking fact, contrary to popular conception, is that capitalism leads to less inequality than alternative systems of organization and that the development of capitalism has greatly lessened the extent of inequality. Comparisons over space and time alike confirm this view. There is surely drastically less inequality in Western capitalist societies like the Scandinavian countries, France, Britain, and the United States, than in a status society like India or a backward country like Egypt. Comparison with communist countries like Russia is more difficult because of paucity and unreliability of evidence. But if inequality is measured by differences in levels of living between the privileged and other classes, such inequality may well be decidedly less in capitalist than in communist countries. Among the Western countries alone, inequality appears to be less, in any meaningful sense, the more highly capitalist the country is: less in Britain than in France, less in the United States than in Britain -- though these comparisons are rendered difficult by the problem of allowing for the intrinsic heterogeneity of populations; for a fair comparison, for example, one should perhaps compare the United States, not with the United Kingdom alone but with the United Kingdom plus the West Indies plus its African possessions.

With respect to changes over time, the economic progress achieved in the capitalist societies has been accompanied by a drastic diminution in inequality. As late as 1848, John Stuart Mill could write, "Hitherto [1848] it is questionable if all the mechanical inventions yet made have lightened the day's toil of any human being. They have enabled a greater population to live the same life of drudgery and imprisonment, and an increased number of manufacturers and others to make fortunes. They have increased the comforts of the middle classes. But they have not yet begun to effect those great changes in human destiny, which it is in their nature and in their futurity to accomplish." * This statement was probably not correct even for Mill's day, but certainly no one could write this today about the advanced capitalist countries. It is still true about the rest of the world.

The chief characteristic of progress and development over the past century is that it has freed the masses from backbreaking toil and has made available to them products and services that were formerly the monopoly of the upper classes, without in any corresponding way expanding the products and services available to the wealthy. Medicine aside, the advances in technology have for the most part simply made available to the masses of the people luxuries that were always available in one form or another to the truly wealthy. Modern plumbing, central heating, automobiles, television, radio, to cite just a few examples, provide conveniences to the masses equivalent to those that the wealthy could always get by the use of servants, entertainers, and so on.

Detailed statistical evidence on these phenomena, in the form of meaningful and comparable distributions of income, is hard to come by, though such studies as have been made confirm the broad conclusions just outlined. Such statistical data, however, can be extremely misleading. They cannot segregate differences in income that are equalizing from those that are not. For example, the short working life of a baseball player means that the annual income during his active years must be much higher than in alternative pursuits open to him to make it equally attractive financially. But such a difference affects the figures in exactly the same way as any other difference in income. The income unit for which the figures are given is also of great importance. A distribution for individual income recipients always shows very much greater apparent inequality than a distribution for family units: many of the individuals are housewives working part-time or receiving a small amount of property income, or other family members in a similar position. Is the distribution that is relevant for families one in which the families are classified by total family income ? Or by income per person? Or per equivalent unit? This is no mere quibble. I believe that the changing distribution of families by number of children is the most important single factor that has reduced inequality of levels of living in this country during the past half century. It has been far more important than graduated inheritance and income taxes. The really low levels of living were the joint product of relatively low family incomes and relatively large numbers of children. The average number of children has declined and, even more important, this decline has been accompanied and largely produced by a virtual elimination of the very large family. As a result, families now tend to differ much less with respect to number of children. Yet this change would not be reflected in a distribution of families by the size of total family income.

A major problem in interpreting evidence on the distribution of income is the need to distinguish two basically different kinds of inequality; temporary, short-run differences in income, and differences in long-run income status. Consider two societies that have the same distribution of annual income. In one there is great mobility and change so that the position of particular families in the income hierarchy varies widely from year to year. In the other, there is great rigidity so that each family stays in the same position year after year. Clearly, in any meaningful sense, the second would be the more unequal society. The one kind of inequality is a sign of dynamic change, social mobility, equality of opportunity; the other, of a status society. The confusion of these two kinds of inequality is particularly important, precisely because competitive free-enterprise capitalism tends to substitute the one for the other. Non-capitalist societies tend to have wider inequality than capitalist, even as measured by annual income; in addition, inequality in them tends to be permanent, whereas capitalism undermines status and introduces social mobility.

The methods that governments have used most widely to alter the distribution of income have been graduated income and inheritance taxation. Before considering their desirability, it is worth asking whether they have succeeded in their aim.

No conclusive answer can be given to this question with our present knowledge. The judgment that follows is a personal, though I hope not utterly uninformed, opinion, stated, for sake of brevity, more dogmatically than the nature of the evidence justifies. My impression is that these tax measures have had a relatively minor, though not negligible, effect in die direction of narrowing the differences between the average position of groups of families classified by some statistical measures of income. However, they have also introduced essentially arbitrary inequalities of comparable magnitude between persons within such income classes. As a result, it is by no means clear whether the net effect in terms of the basic objective of equality of treatment or equality of outcome has been to increase or decrease equality.

The tax rates are on paper both high and highly graduated. But their effect has been dissipated in two different ways. First, part of their effect has been simply to make the pre-tax distribution more unequal. This is the usual incidence effect of taxation. By discouraging entry into activities highly taxed -- in this case activities with large risk and non-pecuniary disadvantages -- they raise returns in those activities. Second, they have stimulated both legislative and other provisions to evade the tax -- so-called "loopholes" in the law such as percentage depletion, exemption of interest on state and municipal bonds, specially favorable treatment of capital gains, expense accounts, other indirect ways of payment, conversion of ordinary income to capital gains, and so on in bewildering number and kind. The effect has been to make the actual rates imposed far lower than the nominal rates and, perhaps more important, to make the incidence of die taxes capricious and unequal. People at the same economic level pay very different taxes depending on the accident of die source of their income and the opportunities they have to evade the tax. If present rates were made fully effective, the effect on incentives and the like might well be so serious as to cause a radical loss in the productivity of the society. Tax avoidance may therefore have been essential for economic well-being. If so, die gain has been bought at the cost of a great waste of resources, and of the introduction of widespread inequity. A much lower set of nominal rates, plus a more comprehensive base through more equal taxation of all sources of income could be both more progressive in average incidence, more equitable in detail, and less wasteful of resources.

This judgment that the personal income tax has been arbitrary in its impact and of limited effectiveness in reducing inequality is widely shared by students of the subject, including many who strongly favor the use of graduated taxation to reduce inequality. They too urge that the top bracket rates be drastically reduced and the base broadened.

A further factor that has reduced the impact of the graduated tax structure on inequality of income and wealth is that these taxes are much less taxes on being wealthy than on becoming wealthy. While they limit the use of die income from existing wealth, they impede even more strikingly -- so far as they are effective -- the accumulation of wealth. The taxation of the income from the wealth does nothing to reduce the wealth itself, it simply reduces the level of consumption and additions to wealth that the owners can support. The tax measures give an incentive to avoid risk and to embody existing wealth in relatively stable forms, which reduces the likelihood that existing accumulations of wealth will be dissipated. On the other side, the major route to new accumulations is through large current incomes of which a large fraction is saved and invested in risky activities, some of which will yield high returns. If die income tax were effective, it would close this route. In consequence, its effect would be to protect existing holders of wealth from the competition of newcomers. In practice, this effect is largely dissipated by the avoidance devices already referred to. It is notable how large a fraction of the new accumulations have been in oil, where the percentage depletion allowances provide a particularly easy route to the receipt of tax-free income.

In judging the desirability of graduated income taxation it seems to me important to distinguish two problems, even though the distinction cannot be precise in application: first, the raising of funds to finance the expenses of those governmental activities it is decided to undertake (including perhaps measures to eliminate poverty discussed in chapter xii); second, the imposition of taxes for redistributive purposes alone. The former might well call for some measure of graduation, both on grounds of assessing costs in accordance with benefits and on grounds of social standards of equity. But the present high nominal rates on top brackets of income and inheritance can hardly be justified on this ground -- if only because their yield is so low.

I find it hard, as a liberal, to see any justification for graduated taxation solely to redistribute income. This seems a clear case of using coercion to take from some in order to give to others and thus to conflict head-on with individual freedom.

All things considered, the personal income tax structure that seems to me best is a flat-rate tax on income above an exemption, with income defined very broadly and deductions allowed only for strictly defined expenses of earning income. As already suggested in chapter v, I would combine this program with the abolition of the corporate income tax, and with the requirement that corporations be required to attribute their income to stockholders, and that stockholders be required to include such sums on their tax returns. The most important other desirable changes are the elimination of percentage depletion on oil and other raw materials, the elimination of tax exemption of interest on state and local securities, the elimination of special treatment of capital gains, the co-ordination of income, estate, and gift taxes, and the elimination of numerous deductions now allowed.

An exemption, it seems to me, can be a justified degree of graduation (see further discussion in chapter xii). It is very different for 90 per cent of the population to vote taxes on themselves and an exemption for 10 per cent than for 90 per cent to vote punitive taxes on the other 10 per cent -- which is in effect what has been done in the United States. A proportional flat-rate-tax would involve higher absolute payments by persons with higher incomes for governmental services, which is not clearly inappropriate on grounds of benefits conferred. Yet it would avoid a situation where any large numbers could vote to impose on others taxes that did not also affect their own tax burden.

The proposal to substitute a flat-rate income tax for the present graduated rate structure will strike many a reader as a radical proposal. And so it is in terms of concept. For this very reason, it cannot be too strongly emphasized that it is not radical in terms of revenue yield, redistribution of income, or any other relevant criterion. Our present income tax rates range from 20 per cent to 91 per cent, with the rate reaching 50 per cent on the excess of taxable incomes over $18,000 for single taxpayers or $36,000 for married taxpayers filing joint returns. Yet a flat rate of 23% per cent on taxable income as presently reported and presently defined, that is, above present exemptions and after all presently allowable deductions, would yield as much revenue as the present highly graduated rate.2 In fact, such a flat rate, even with no change whatsoever in other features of the law, would yield a higher revenue because a larger amount of taxable income would be reported for three reasons: there would be less incentive than now to adopt legal but costly schemes that reduce the amount of taxable income reported (so-called tax avoidance); there would be less incentive to fail to report income that legally should be reported (tax evasion); the removal of the disincentive effects of the present structure of rates would produce a more efficient use of present resources and a higher income.

If the yield of the present highly graduated rates is so low, so also must be their redistributive effects. This does not mean that they do no harm. On the contrary. The yield is so low partly because some of the most competent men in the country devote their energies to devising ways to keep it so low; and because many other men shape their activities with one eye on tax effects. All this is sheer waste. And what do we get for it ? At most, a feeling of satisfaction on the part of some that the state is redistributing income. And even this feeling is founded on ignorance of the actual effects of the graduated tax structure, and would surely evaporate if the facts were known.

To return to the distribution of income, there is a clear justification for social action of a very different kind than taxation to affect the distribution of income. Much of the actual inequality derives from imperfections of die market. Many of these have themselves been created by government action or could be removed by government action. There is every reason to adjust the rules of the game so as to eliminate these sources of inequality. For example, special monopoly privileges granted by government, tariffs, and other legal enactments benefiting particular groups, are a source of inequality. The removal of these, the liberal will welcome. The extension and widening of educational opportunities has been a major factor tending to reduce inequalities. Measures such as these have the operational virtue that they strike at the sources of inequality rather than simply alleviating the symptoms.

The distribution of income is still another area in which government has been doing more harm by one set of measures than it has been able to undo by others. It is another example of the justification of government intervention in terms of alleged defects of the private enterprise system when many of the phenomena of which champions of big government complain are themselves the creation of government, big and small.

1 Principles of Political Economy (Ashley edition; London: Longmans, Green & Co., 1909), p. 751.

2 This point is so important that it may be worth giving the figures and calculations. The latest year for which figures are available as this is written is the taxable year 1959 in U. S. Internal Revenue Service, Statistics of Income for 1959. For that year: Aggregate taxable income reported on

Individual tax returns......................................................................	$166,540 million
Income Tax before tax credit .........................................................	39,092 million
Income tax after tax credit..............................................................	38,645 million

A flat rate tax of 23½ per cent on the aggregate taxable income would have yielded (.235) X $166,540 million = $39,137 million.

If we assume the same tax credit, the final yield would have been about the same as that actually attained.