Progressive Taxation Reconsidered by F. A. HAYEK

AMONG the measures of economic policy which are gradually transforming our society and producing far-reaching results which few people yet clearly grasp, few are as firmly established and as widely accepted as the redistribution of income by progressive taxation. Though it is a comparatively recent feature and one which only in the course of the last generation has assumed the proportion of a major factor in social change, there has been until quite recently very little re-examination of its effects. It is accepted as right and desirable even by most people who are anxious to preserve a free market economy, and to most of them it indeed appears as the main hope of establishing within such a system the greater degree of economic justice or equality for which they yearn. So firmly has the opinion that progressive taxation is both innocuous and desirable been established that even those who were alarmed by some of its visible effects seem to feel that any critical examination of the principle as such would be a futile waste of effort and that anyone who undertook it would thereby mark himself as an unpractical doctrinaire. Quite lately, however, a change in this attitude is noticeable. After a long period in which there was practically no questioning of the principle as such and the discussions on the whole merely repeated the old arguments, there is a new critical attitude noticeable in the occasional references to the problem; and there have already appeared some notable major contributions to the discussion.1 There is, however, still much need for a systematic re-examination of the whole complex of problems raised by progressive taxation of the kind which is now actually practiced. This can not be attempted in a single article and what the following paragraphs will undertake is merely to sketch a few considerations which do not yet seem to have received the attention which they deserve.
The main reason why the whole subject requires reconsideration is that the gradual increase in the rates of taxation over the past fifty years has, in its cumulative effect, made the problem different in kind and not merely in degree. With scales of progression approaching and even exceeding ninety per cent of income their significance is of an altogether different nature from what it was when the upper limits were in the region of ten or at most fifteen per cent. This seemed to be the extreme figures which had to be seriously considered when, around the beginning of this century, the whole issue was for the last time thoroughly discussed. It was then still possible to treat the whole issue as if it were a problem of allocating a given tax burden among the various classes of society; and though it did raise important issues of principle if the comparatively wealthy were made to contribute a few per cent more of their income, no important economic effects were expected from this. To suggest at that time that progression might ever be carried to the figures it has now reached would have been treated by its advocates as a malicious travesty of the principle showing a disreputable contempt for the wisdom of democracy.
With the change in scale has come a general recognition of the fact that the only ground on which progressive taxation could be rationally justified was a desire to change the distribution of incomes2 and that this could not be based on any scientific argument but had to be recognized as frankly a political decision, an attempt to impose upon society a pattern of distribution determined by political choice. All the ingenious theories of just taxation which had been developed in the early days of the discussion and which can still be found in the textbooks on public finance3 have lost their relevance in view of the no longer disputed fact that present policy is guided almost exclusively by the desire to produce an all-round reduction of income inequalities.
There is only one among these older theories which needs some brief consideration because it is still often asserted that it provides something like a scientific foundation for policy. This is the use of the conception of decreasing marginal utility in support of a proportionally greater taxation of the larger incomes. In spite of its abstract character it has had great influence in making scientifically respectable what originally had been frankly based on arbitrary postulates. How important it seemed at the time may be gauged by such statements as that of the late Lord Stamp who in 1929 wrote that “it was not until the marginal theory was thoroughly worked out on its psychological side that progressive taxation obtained a really secure basis in principle.”4 Yet I do not believe it is overstating the case to say that modern developments within the realm of utility analysis itself have left no justification whatever for this use of marginal utility. Not only does it fall with the abandonment of interpersonal comparisons of utility—a conclusion which seems to me inescapable notwithstanding the ever-recurring objection that individually most of us have definite views about whether a particular need of A is greater than a certain need of B. But the fact that we may have views about this, of course, does not prove that if these views differ there is any objective basis for deciding between them; and this is the question which has been at issue and which must undoubtedly be answered in the negative. But what is more, it is exceedingly doubtful whether even the conception of decreasing marginal utility as such, applied to income as a whole, has any clear meaning if we count as income all the benefits derived by a person from his disposal over his resources. The recognition that utility has definite meaning only as a relative concept, i.e., that we can say only that a given object is more, equally, or less useful than some particular other object, and that it is meaningless to speak of the utility of a thing in isolation, implies that in order that we should be able to speak of the utility, or the marginal utility, of income we have to define income so as to leave out of it something which can serve as a standard of comparison. We can meaningfully speak of the utility of income in terms of effort or of some other such magnitude, say, leisure. But if we were seriously to follow up the consequences of the contention that the utility of income in terms of effort is decreasing, this would lead to very curious results in our context: it would in effect mean that as a person’s income grows, the incentive in terms of additional income required to induce the same marginal effort would increase. This might lead to an argument in favor of degressive taxation but certainly not to one for progressive taxation. It is scarcely worth while to follow this line of thought further. In retrospect we must probably say that the whole episode of introducing utility analysis into this discussion was a regrettable mistake (in which some of the most distinguished economists of the time shared) and that the sooner we can undo the effects produced by the quasi-scientific sanction which economic theory gave to a dangerous instrument of policy, the better it will be.
For what follows we shall take it for granted that today the only grounds on which progressive taxation can be defended is the desire for a more equal distribution of income. This it attempts to achieve mainly by flattening the top of the income pyramid. It differs from other measures of more specific controls of income distribution in that it does not directly manipulate the income of specific groups but, as it were, alters the scale of incomes which can be earned. How far it succeeds in this, i.e., how far its effects are not counteracted by an adjustment of gross incomes and the burden thus partially shifted, is a question we will not consider here. It has recently been attempted to show, by an ingenious argument of the “Keynesian”5 type, that so far as the aggregate amount of profits is concerned, the attempt to reduce them by taxation cannot succeed. This argument is based on rather special assumptions (especially the assumption that the volume of investment can, for the purpose of this argument, be treated as fixed) and I doubt whether under actually existing conditions there is any validity in the contention. At any rate, there seems to me little doubt possible that in actual fact progressive taxation does succeed in greatly reducing the net incomes in the higher brackets compared with what they would otherwise be, and the further discussion will proceed on the assumption that this is the case. Progressive taxation is, of course, not the only method by which a redistribution of incomes can be brought about. It would be possible to effect a considerable amount of redistribution under a system of proportional taxation. To achieve this it would merely be necessary to devote a substantial part of tax revenue to finance services which benefit mainly the relatively poor—or to subsidize them directly. Yet there are several limitations to the extent to which this could be carried. Not only is it doubtful how far the people in the lower income classes would be willing to have their freely spendable income reduced by taxation in return for services offered free. It is also particularly difficult to conceive how in this manner the differentials in the higher income classes could be substantially altered. There might well be brought about in this manner a considerable transfer of incomes from the rich as a class to the poor as a class; but it would not bring about that flattening of the top of the income pyramid which is the characteristic effect of progressive taxation. For the comparatively well-to-do it would presumably mean that they would all be taxed proportionately on their whole incomes and that the differences in the services they received would be negligible. It is in this class, however, that the changes in the income structure resulting from progressive taxation are most significant. The consequences for progress, for the allocation of resources, the effect on incentives, on social mobility, and on investment, operate mainly through the effect on this group (which, in the most advanced countries today includes, of course, many of the highly skilled manual workers). Whatever may be the possible developments in the future, for the present at any rate it seems beyond question that progressive taxation is the main tool available for effecting a redistribution of income and that without it the scope of such a policy would be very limited.
A distinction must, of course, be drawn between the progressive character of a particular tax, such as the income tax, and the progressive character of the burden which the tax system as a whole imposes upon incomes. It is well-known that the heavier incidence of indirect taxation on the lower incomes may make the effect of the tax system as a whole regressive in the lower brackets, even though the income tax is progressive, and that, on the other hand, a progressive income tax may be used to make the tax burden as a whole proportional to incomes by compensating for the degressive effects of indirect taxation. The argument for a progressive income tax which does no more than this is probably very strong and it seems to us the only valid argument in favor of a progressive tax, but, be it noted, only in favor of progressive scales for one particular tax, and not in favor of a progressive character of the tax system as a whole. The significance of this argument is today, however, some-what diminished because it seems probable that the regressive character of taxation in the lowest income brackets is largely compensated for by the redistributive effects of government expenditure.
It is, however, still worth while to look a little more closely at the information we have on the distribution of the tax burden between the different income classes, since it throws an interesting light on the alleged inevitability of relieving the lowest incomes from the burden of taxation. The most detailed investigation of this kind known to me concerns the situation in Great Britain,6 but similar studies for other countries, especially the United States,7 suggest that the main results of the British investigation reveal a situation which also prevails elsewhere. It will be useful to reproduce here some of the results of that investigation. According to it the total burden of taxation on different fully “earned” incomes of a family with two children were as follows during the last pre-war year (1937/38) for which these figures were worked out:8



 Income£

 Per cent takenin taxation

100  
. . . . . . . . . . . . . . . .
18

150  
. . . . . . . . . . . . . . . .
16

200  
. . . . . . . . . . . . . . . .
 15

250  
. . . . . . . . . . . . . . . .
 14

300  
. . . . . . . . . . . . . . . .
 12

350  
. . . . . . . . . . . . . . . .
 11

500  
. . . . . . . . . . . . . . . .
 14

1,000  
. . . . . . . . . . . . . . . .
19

2,000  
. . . . . . . . . . . . . . . .
24

2,500  
. . . . . . . . . . . . . . . .
25

5,000  
. . . . . . . . . . . . . . . .
33

10,000  
. . . . . . . . . . . . . . . .
41

20,000  
. . . . . . . . . . . . . . . .
50

50,000  
. . . . . . . . . . . . . . . .
58



It will be noticed that the lowest rate of taxation occurs at an income of £350; other data given in the same work suggest that it may actually have been as high as £500 and that this situation had prevailed during the preceding twenty years, while during the first two decades of the century the income with the lowest taxation had gradually risen from £150 and was again somewhat reduced by the severe taxation of the Second World War.
In our immediate context these figures are significant in two respects. In the first instance they show that the argument that progressive taxation is inevitable because the poorest must be relieved from bearing a proportional share of the tax burden is, so far as the effects of the tax system as a whole is concerned, just humbug. It may be questioned whether any tax system has been ever able to dispense with the individually small but so very numerous contributions from the smallest incomes. At any rate, this has not been the situation since progressive taxation has become an important feature and is not the position today. We have already conceded that the regressive character of indirect taxation may be a valid argument for compensating for it by making the income tax progressive. But in view of the actual practice of democratic countries in modern times the necessity of exempting the poorest from the tax burden can hardly be advanced as an argument for making the tax structure as a whole progressive. (Since this argument has usually been coupled with the contention that the prohibitive costs of a direct taxation of small incomes made the exemption a practical necessity and in consequence also a progressive structure of the income tax inevitable, it may be mentioned that the techniques for levying small contributions developed in connection with social insurance, etc., have deprived also this argument of most of its validity.)
The second interesting point arises if one compares these figures about the relative tax burden in the different income classes with the proportional numbers of taxpayers in each class, or, what amounts to much the same thing, their relative strength in the electorate. If the figures given above were plotted in a diagram together with a curve representing the relative frequency of the taxpayers in each class, it would be found that the two curves were approximately mirror images: this means that it was not the poorest but the most numerous and therefore politically most powerful classes which were left off relatively lightly, while not only those above them but also those below them were burdened more heavily—approximately in proportion to their smaller political strength. I am not suggesting, of course, that this is the deliberate result of a diabolic policy; it seems to be rather the unforeseen but almost inevitable result of the democratic process when it is not guided by at least the desire to apply the same uniform principle to all. Once it is admitted that a majority has a right to impose upon minorities burdens of a kind which the majority does not bear itself, there is little reason that this will be used only for discrimination against the rich.
There is one more consideration which ought to be kept in mind in this connection. There is clearly little justification for specially favoring those lower middle incomes which we have seen to be the actual gainers under the prevailing tax structures. But if we consider only that part of the scale of progression where it again exceeds the tax burden imposed on the very lowest incomes, it becomes clear that the part of revenue which depends on the progressive character of the tax system as a whole is negligible. If we remember that, e.g., in Great Britain, according to the latest information, “only about 10½ per cent of the total income over £.155 a year lies in the slice above £ 1000”9 while in the United States the sum of all incomes of $10,000 and more amounted in 1952 to only 17 per cent of the total of “adjusted gross income,”10 it becomes clear how relatively small the financial yield from the progressive taxation of these incomes is. It is almost certainly considerably smaller than the additional revenue which would be obtained if the lower middle groups just mentioned were taxed as heavily as the poorest.
This is important because it in effect disposes of the supposed fiscal necessity of making the tax system as a whole progressive. It just is not the case that the sums actually raised could not be raised without resort to progression; they could in fact, probably, be raised without increasing the burden on the very poorest at all and by merely bringing up the proportional burden on those lower middle groups to that actually borne by the poorest.
It seems that the conclusion we must draw from this is that rates of taxation in the upper part of the progressive scale have very little to do with the benefit the resulting redistribution of income confers on the lower income classes or the relief in the tax burden they actually obtain. They must be regarded as purely punitive rates, as an expression of the dislike of the majority of the idea that anybody should enjoy the command of such large incomes. It is in this region, however, where marginal tax rates rise in Great Britain and the United States more or less rapidly from the neighborhood of 20 per cent (for a married couple with two children) to 90 per cent or over, that the effect of progressive taxation is so very important. The percentage of the population directly affected by it is comparatively small; but it is probably the section of the population which in a free society predominantly decides on how efficiently the resources will be used. It will be for this reason that in what follows we shall be concerned mainly with the effect of progressive taxation on this group.
Before we go on to examine some of the specific effects of this kind of taxation we will pause for a moment to consider how it has come about that we have arrived at a scale of progression leading up to rates which a generation ago would have been regarded as thoroughly unreasonable. We have already been able to eliminate real financial necessity as an explanation—though this does not exclude the possibility that mistaken beliefs about the extent to which the burden might be shifted to the rich may not have had a determining influence. Indeed, it seems more than likely that the illusion that by means of progressive taxation the cost of additional expenditure can be raised from the rich has made such expenditure much more attractive and that as a result even the poor now have to give up a larger proportion of their income than they would have consented to do.
Another factor which has operated in a similar direction was, of course, inflation. It is now well understood how a general rise in money incomes tends to lift everybody into a higher tax bracket even though his real income may have remained the same. In this manner many members of the majority must have found themselves unexpectedly the victims of discriminatory rates for which they had readily voted in the belief that it would never affect them. This particular effect of progressive taxation is often represented as a special merit of the system because it tends to make inflation (and deflation) produced by unbalanced budgets to some extent self-correcting. If the source of inflation is a budget deficit, the tax revenue will tend to rise proportionately more than incomes and thus to close the gap; and if a budget surplus has produced deflation the resulting fall of incomes will soon bring an even greater reduction of revenue and wipe out the surplus. I doubt, however, whether with the ever-present bias in favor of inflation which at this time seems particularly strong, this is really an advantage. Even without this consideration the needs of government finance have been in the past the main source of recurrent inflations and only the knowledge that an inflation, once started, was difficult to stop has acted in some measure as a deterrent. With a tax system under which inflation produces a more than proportional increase in revenue by way of a disguised increase in taxes which needs no vote of the legislature, this device may become almost irresistibly tempting.
These special factors, however, had done no more than speed up further a process which is practically inevitable once the principle of progressive taxation has been accepted. From the very beginning it has been one of the main arguments against it that once the principle is adopted there is no stopping on the road to steeper and steeper scales. As early as the sixteenth century, as Professor Seligman pointed out, Guicciardini had argued that “it lies in the nature of things that the beginnings are slight, but unless great care is taken, the rates will multiply rapidly and finally reach a point that no one could have foreseen.”11 The nineteenth century literature, particularly in its discussions of democracy, is full of such warnings. The best-known statement of the fears is probably that of J. R. McCulloch: “The reasons that made the step taken in the first instance, backed as they are sure to be by agitation and clamor, will impel you forwards. Having once given way, having said that a man with 500 £ a year shall pay 5 per cent, another with 1000 £ 10 per cent, and another with 2000 £ 20 per cent, on what pretence or principle can you stop in your ascending scale? Why not take 50 per cent from the man of 2000 £ a year, and confiscate all the higher classes of income before you tax the lower? In such matters the maxim of obsta principiis should be firmly adhered to by every prudent and honest statesman. Graduation is not an evil to be paltered with. Adopt it and you will effectually paralyze industry and check accumulation; . . . The moment you abandon . . . the cardinal principle of exacting from all individuals alike the same proportion of their income of their property, you are at sea without rudder or compass, and there is no amount of injustice or folly you may not commit.”12
The question why these pessimistic prognostications of the opponents of progressive taxation have come to be confirmed and not the confidence of its supporters that it would be used in moderation raises a problem of much wider application than merely in our field. It is the problem why it is apparently necessary, in social no less than in private action, to abstain altogether from certain kinds of measures if we want to avoid consequences which would follow if we applied the principle underlying them as a general rule. The problem is very similar to that why in individual ethics, when a kind of action is held to be bad because bad consequences frequently spring from it, it is still held to be bad if in the particular instance no such bad consequences seem to follow. Yet while, on the whole, we still accept in private ethics the need of such hard and fast rules which prohibit certain classes of actions irrespective of whether we can see that they will have immediate bad effects, similar rules applied to social action are generally regarded as superstitions which should not be allowed to interfere with our freedom to experiment. Yet, if we want to avoid altogether undesired results of what we are doing, strict adherence to general rules, even in instances where their justification is not readily seen, is probably even more important here than in individual behavior. This might not be true if social organization was ever designed as a whole and if in designing it we could judge each individual feature in relation to all others. But a social structure is never really the result of design, not even in what is called a planned society. It rather results from the application to particular and partial decisions of general conceptions or ideals ruling that society. The arguments underlying these principles of action cannot and are not re-examined in every individual instance; the mere fact that a principle has been applied in other instances becomes the main ground for it being applied again. But the cumulative effect of it being applied separately in many different instances or in many successive decisions will, of course, be very different from what on any of these occasions has been foreseen. Though as an isolated measure, and applied to a limited degree, action of the kind may seem innocuous enough and any possible barm that could follow from it negligible compared with the importance of the immediate object, the joint effects of many measures of this type may be exceedingly harmful.
To any one who views the social process realistically it can offer little reassurance to be told that a principle which, if carried very far, is admittedly dangerous, will have only beneficial effects if used in moderation. It is, in fact, only the presumption against the principle as such which protects us against its abuse. This is particularly true where, as in the case of progressive taxation, every argument which can be advanced in favor of some progression is equally valid in favor of more progression. The idea merely points in a direction in which it is thought desirable to deviate from a standard. What at first limits it is no more than the unfamiliarity of it. But it always justifies a little more of the same than before. The principle itself indicates no halting point and the “good judgment”13 of the people to which its defenders are usually driven to resort as the ultimate safeguard are merely the opinions shaped by past policy. In its cumulative effects the successive decisions on what is just in the light of the principle will always go far beyond what its initial sponsors thought desirable.
It is sometimes contended that taxation proportional to income is as arbitrary a principle as progressive taxation and that it has merely a greater apparent mathematical neatness but little else to commend itself. There are, however, fairly strong arguments in its favor. Not only is there still much in the old argument that, since almost all economic activity benefits from the basic services of government, these services form a more or less constant ingredient of all we consume and enjoy, and that, therefore, the more a person can command of the resources of society, the greater will also be his gain from what government has contributed to make these services possible. But more important is the fact that proportional taxation leaves the relation between the net remunerations of different kinds of work unchanged. This is not quite the same as the famous old maxim that “no tax is a good tax unless it leaves individuals in the same relative position as it finds them”14 because it stresses the effect not on the relation between individuals but on the relation between the net remuneration for particular services performed, which is the economically relevant factor. It also does not, as might at first seem, beg the issue by simply postulating that the proportional size of the different incomes should be left unchanged.
While there might be a difference of opinion on the question whether the relation between two incomes is left unchanged if they are both reduced by the same proportion or if they are reduced by the same amount, there can be no difference on the question whether the net remuneration received for two services, of which the one was, before taxation, larger, equal or smaller than the second, stands after taxation still in the same relation to the second. This, however, is the crucial issue with regard to which the effects of progressive taxation are fundamentally different from that of proportional taxation. It is, of course, the reward received for the use of particular resources which determines their allocation, and what is important is that taxation should leave these relative rewards unchanged. Progressive taxation, however, alters them very considerably by making the net reward received by the owner dependent on what else he has earned during some arbitrary period, such as a year. If, before taxation, a surgeon gets as much for an operation as an architect for planning a house, or a salesman selling ten refrigerators as much as a photographer for making forty portraits, this will still be true if equal proportional deductions are made from these payments. But with progressive taxation of incomes this relation may be violently changed. Not only will services which before taxation receive the same reward leave very different net rewards to those who rendered them; a much larger payment for one service may indeed leave less to him who rendered it than a smaller payment to another person.
This means in the first instance that progressive taxation inevitably offends against what seems to me the most basic principle of economic justice, that of “equal pay for equal work.” If what two lawyers are allowed to retain from their fees for doing exactly the same work, or two surgeons from their fees for performing the same operations, depends on their other earning during the year, they will in fact, probably, derive very different profits from their efforts. The man who has worked very hard or who has for other reasons been particularly successful during the year will receive a much smaller remuneration for further effort than the one who has been idle or unlucky. And, indeed, the more the consumers value a man’s services, the less worth while it is made for him to exert himself further.
The fact that the taxation of a given sum earned will vary with the time rate at which such earnings accrue to the recipient is the source of most of the injustices and the cause of the misdirection of resources which present taxation produces. There is no need to dwell here on the familiar and insoluble difficulties which, as a result of progressive taxation, arise in all instances where effort (or outlay) and reward are not approximately synchronized but where the former are expended in the expectation of a distant and uncertain result—in short, in all instances where human efforts take the form of long and risky investment. No practical scheme of averaging incomes can really do justice to the problems of the author or inventor, the artist or actor, who reap the reward of perhaps decades of effort in a few years. Nor will it be necessary to stress once again the discouraging effects progressive taxation must have on the willingness for the more risky type of capital investments. That such taxation discriminates against the risky ventures which are worth while only because in the case of success they will bring a return big enough to compensate for the great probability of total loss is so obvious that it should not need emphasis. But it may perhaps be said that what little truth there is in the alleged exhaustion of investment opportunities probably is very largely the result of a fiscal policy which in this manner directly eliminates a wide range of ventures from the field which can be undertaken by private enterprise.
That this sort of taxation is so generally approved is closely connected with the fact that our society has come to think of an appropriate income as the only legitimate and socially desirable form of reward, and further, to think of this income not as related to the value of the particular services rendered but as conferring what is regarded as an appropriate status in society. This comes out very clearly in such arguments, frequently used in support of progressive taxation, as that no individual can be worth more to society than, say, $20,000 a year.15 That this contention lacks any foundation and appeals solely to unreflecting emotion and prejudice would at once become clear if it were stated in the form of saying that no act any individual can perform in a year, or for that matter in an hour, can be worth more to society than $20,000. Of course it can and sometimes will have many times that value. There is no necessary relation between the time an action takes and the benefit society may derive from it.
The whole attitude which regards large gains as unnecessary and socially undesirable springs from the psychology of people who are used to sell their time for a fixed salary or fixed wages and have come to think as a remuneration of so much per unit of time as the normal thing. But, while this method of remuneration has become the only practicable one in an increasing number of fields, it is reasonable only where people sell their time to use it at another man’s direction. But it is senseless with respect to men whose task it is to administer resources and whose main aim is to increase the resources under their control from their earning. For them to control resources is a condition for practicing their vocation just as much as the acquisition of certain skills and knowledge is such a condition in the professions. Profits and losses are a way of redistributing capital among them more than merely a means of providing their current sustenance. The conception that current net receipts are normally intended for current consumption, though natural to the salaried man, is alien to the thinking of one whose aim is to build up a business. Even the conception of income is in his case largely an abstraction forced on him by the income tax. It is no more than an estimate of what, in view of his expectations and plans, he can afford to spend, rather than an objective fact. I doubt whether a society consisting mainly of “self-employed” individuals would ever have come to take the income concept for granted as we do, or would ever have thought of taxing differently a given amount earned according to the time rate at which such earnings accrue.
It must appear somewhat doubtful, however, whether in a society which will recognize no other rewards than what to its majority appears a very ample income and do not admit the acquisition of a fortune in a comparatively short time as a legitimate form of remuneration for certain kinds of services, it is possible in the long run to preserve a system of private enterprise. Though there may be no difficulty in widely dispersing the ownership in well established enterprises among a large number of small capitalists, the building up of a new enterprise still is and probably always will be bound up with the control of large resources by a few individuals. New developments will, as a rule, still have to be backed by a few persons intimately acquainted with the field, and it is certainly not to be wished that all further evolution should be dependent on the existing financial and industrial corporations.
I do not wish here to enter into the much discussed question of the effect of progressive taxation on the amount of new capital formation—not because this seems to me unimportant but because another influence on capital formation seems to be equally important and less generally appreciated. It is the effect on the locus of capital formation. It is one of the advantages of a competitive system that successful new ventures are likely for a short time to bring very large profits with the result that new capital is being formed in the hands of the very people who have the best opportunity of employing it. The large gains of the successful innovator meant in the past that the man who had shown the capacity of profitably employing capital in new ventures would soon be able to back his judgment with his own means. Much of the individual formation of new capital, since it is offset by similar capital losses of others, is in this connection more usefully regarded as part of a continuous process of redistribution of the capital of society than as a profit which constitutes part of the net income of society. The taxation of such profits at more or less confiscatory rates amounts therefore in effect to a heavy tax on this turnover of capital which is part of the driving force of a progressive society.
One of the consequences of the discouragement of individual capital formation at the points where there are temporary opportunities for very large profits is a serious restriction of competition. As the whole system tends to favor corporate as against individual saving it strengthens the position of the established corporation against newcomers and tends to create quasi-monopolistic positions. By making the rise of new entrepreneurs more difficult it unquestionably assists, presumably against the intention of its advocates, the concentration of industry.
An even more paradoxical and socially grave effect of progressive taxation in this field is that this instrument, intended to decrease inequality, in effect helps to perpetuate existing inequalities and eliminates one of the most important compensations for the kind of inequality which is inevitable in a private enterprise society. It does this by greatly reducing vertical mobility because it diminishes the chances of rising from one class to another.16 That the rich were not a closed group but that the successful man might in a comparatively short time become the owner of large capital resources used to be the redeeming features which did most to mitigate the psychological effects of inequality. The chances of rising into the class of the wealthy are today, however, in some countries such as Great Britain, probably already smaller than they have been at any time since the rise of .modern industrialism. One significant effect of this is that the administration of more and more of the world’s capital is coming into the hands of men who, though they enjoy very large incomes and all the facilities they can wish for, have never on their own account and at their personal risk controlled substantial property. Whether this is altogether an advantage to society remains to be seen.
At the rates to which progressive taxation ascends in some countries it means in effect that greater equality is brought about by setting a ceiling to the net income anybody may have available for spending. (In Great Britain, during the war and immediate post-war years, the largest net income anyone could earn was approximately £500, or $14,000—though this was partly mitigated by the fact that capital gains were not treated as incomes.) We have seen that in view of the insignificant contribution which progression in the higher brackets makes to revenue, this can be justified only on the ground that it is regarded as in some sense socially undesirable that anyone should command such a high income. But what is a large income in this sense depends, of course, on the views of the particular community and thus in the last resort on its average wealth. The consequence of this is that, on the whole, the poorer a country is, the lower it tends to set the limits on permissible incomes or the more difficult it will make it for any of its inhabitants to reach the levels which in wealthier countries are still only moderate incomes. Or, in other words, the poorer a country is, the more difficult it will make it for all its citizens to get rich. This fact stands out very clearly in any international comparison of income tax rates on different incomes expressed in a common unit, say, the dollar—though Great Britain with her exceptionally severe progression somewhat upsets the rule. A rough comparison of this sort shows, for instance, that an average income tax rate of 25 per cent and 50 per cent respectively was reached by a family with three children in the countries named at the following incomes...

One need merely to conceive of the same principle being applied to the different regions of any one country to appreciate its implications. It certainly throws a curious light both on the moral basis of the belief that the view of the majority of a community should be entitled to set a limit on what are to be regarded as “excessive” incomes, and on the wisdom of those who believe that in this manner they will assist the increase of well-being of the masses. Can there be much doubt that poor countries, by preventing individuals from getting rich, will also slow down the general increase of wealth? And does what applies to poor countries apply any less to the rich ones?
Any discussion of the relation between rich and poor in our own environment is so strongly charged with emotional attitudes that it is generally useful to examine the principles involved with reference to differences between national groups. If we do this, can there be serious doubt that today the prospect of the relatively backward people of raising their standard of life is very much better because there exist more advanced people who have developed the techniques they can apply; and that their prospects would be very much poorer if progress of wealth in other parts of the world had, by some kind of international taxation, been kept to a level not too much ahead of their own? This is not the place to go into any systematic examination of the connections between inequality and progress. But the point which must be briefly mentioned is that a substantial part of the larger income of the more advanced people is spent on financing the cost of experimentation and that the results then become available to the others without all the losses due to the recurrent investment in blind alleys, etc. Is it not clear that not only the advanced but also the more backward people would be still at a much lower level if from the beginning the more successful had not been allowed to pull ahead but if any incomes far above the rest had at once been taxed away for redistribution among the poor? And is the role of the rich within any given nation in this respect really very different from that of the few wealthy nations in the world as a whole?
In the last resort the whole problem of progressive taxation is, of course, an ethical problem and the real question in a democracy is whether the support it now receives would continue if people fully understood how it acts. That in many respects it is based on principles which these people would not approve if they were put in the abstract is probably true. Neither that a majority should be free to impose a discriminatory tax burden on a minority, nor that as a result identical services are very differently remunerated, nor that for a whole class, merely because its incomes are out of line with those of the rest of the community, the normal incentives are practically removed, are principles which can be defended from the point of view of justice. If, in addition, one considers the waste of energy and effort to which progressive taxation induces in so many ways and only on a few of which we have here touched, it should not seem impossible to convince reasonable people of its un-desirability. Yet experience shows how rapidly habit blunts in these fields the sense of justice and how the mere fact that a principle has once been applied for some time makes it easy to carry it to extremes.
It is indeed one of the strongest arguments against progressive taxation that it is so difficult again to abandon once it has been introduced. There would probably be no danger and no justified objection if a majority decided to grant an economically weak minority some relief in the form of proportionally lower taxation and the main principle at which one should probably aim is that the majority which determines the burden of taxation should also bear it at its maximum rate; because once it is admitted that it is right that a majority impose a heavier proportional burden on a minority, there seem to be no limits to the length to which this will be carried.
The problem of erecting a barrier which will stop this process of drift is greatly complicated by the fact that, as we have seen, so far as personal taxation only is concerned, some progression is probably both legitimate and desirable. But is there any principle which we can hope will be adopted and which will prevent that the opportunity thus opened will be abused? It is hardly to be expected that an attempt to limit the scale of progression to some particular maximum figure would, in this respect, be effective. Such a percentage figure would be as arbitrary as the principle of progression itself and would be as readily changed when there was need for additional revenue.
What would be needed is a principle which, while limiting the maximum rate of direct taxation in relation to the total burden of taxation, will keep the possible progression of total taxation within narrow limits. The most reasonable limit of this kind would seem to be that the maximum admissible (marginal) rate of direct taxation be fixed at the proportion of the national income which the government takes in taxation, so that, if the government took 30 per cent of the national income, 30 per cent would also be the maximum rate of direct taxation. If a national emergency raised this proportion, the maximum tax rate would also be raised and similarly be automatically reduced when the over-all tax burden was reduced. The application of this principle would still leave some progression of total taxation in existence, since those paying the maximum rate would also pay some indirect taxes which would bring their total taxation above the average for the community. The application of this principle would have the not inconsiderable advantage that every budget would have to be prefaced, as it were, by an estimate of the percentage of the national income which the government proposed to take in taxation. This percentage would provide a sort of standard rate which for the lower income classes would be reduced in proportion as they were taxed directly. The net result would probably be a slight over-all progression, in which the marginal taxation of the largest incomes could, however, never exceed the rate at which incomes were taxed on the average by more than the amount of indirect taxation and in fact (since the limit would apply to the marginal rate of direct taxation) by considerably less.