II. Tax and Non-Tax Policies in Redistribution

4.17. Even if the current distribution of income after tax were known more accurately than it is and were judged to be satisfactory, it would still of course be necessary to discuss whether the existing tax structure is contributing to the result in as simple and efficient a way it could. But if it may be assumed that society desires eventually to achieve a more equal distribution, it becomes necessary explicitly to discuss the relative effectiveness of the policy instruments available. Of these the taxation system is by no means the only one actually or potentially employable in Australia, and it is probably less all-powerful than appears at first sight. There is thus a very real problem in finding the right mixture, the most efficient balance, between the various kinds of instruments that are or can be brought into use. Hence the Committee cannot reach recommendations about the tax system without referring briefly to the other policies available, though any detailed examination of them would involve an unwarrantably wide interpretation of its terms of reference.

4.18. It is evident that were the government to possess effective and wide-ranging controls over all earned incomes and all payments for rights in land and industrial property and financial assets (rents, dividends, interest, etc.), it would thereby have a dominant control of the distribution of income. The equity objective of taxation would largely become superfluous except as regards the distribution of property income between persons as determined by the initial and evolving distribution of ownership of the property itself. The tax system could be operated instead primarily with the objectives of simplicity and efficiency in mind. Such control over incomes may effectively be the situation in totalitarian economies and they can have very simple tax systems. While it is here assumed that control to this extent will never exist in this country, it is worth recording that in so far as, within Australia's political framework, government or other public agencies have some direct influence upon the

  ― 30 ―
evolution of relative gross incomes, the redistributional tasks of the tax system can be to that extent lightened.

4.19. Policies towards monopoly and restrictive practices of all kinds are plainly relevant. The distortions of the market tackled by such policies usually, though not always, tend to increase inequality.

4.20. Policies towards land-tenure and urban development, even though primarily directed towards the achievement of decent urban conditions in the right places, may have important implications in one area of great difficulty for tax policy, that of capital gains from transactions in land.

4.21. Much social service expenditure, for example on health and counselling of one kind or another, even when not involving direct cash transfers is evidently important in preventing extremes of poverty.

4.22. Finally, and perhaps of the greatest general significance, there is the State's part in determining the amount and influencing the composition of education, general and technical and professional. At the lower end of the scale some persistent poverty can be attributed to failures at the level of elementary education. Higher up the scale much of the difference in earned incomes certainly derives from the scarcity of skills that are costly in time and money to acquire. This is a large issue, and not without many intricacies, but the point must be made that in the long term education policy is a powerful weapon in the pursuit of greater equality in incomes—perhaps indeed the most powerful available in a democratic community.

4.23. To return now to the use of taxation as a redistributive measure, one remark should be made at the outset. When contemplating any table showing the distribution of income gross, the taxes paid and grants received at each level, and the resulting distribution net, it is tempting to suppose that simply by altering taxes and grants an exactly equal alteration can be made in after-tax incomes, i.e. that the gross distribution can be taken as something fixed and upon which one may operate at pleasure. This, of course, is a mistake. A change in the tax pattern will in general have an effect upon both the size and the distribution of gross incomes, and the final result to net incomes will be affected by this interdependence, or rather by the combined outcome of a whole set of interdependencies. Some judgments about those tangled and still controversial issues must be offered.

4.24. It is important to distinguish between the initial effect of a change in the progressivity of the rate structure and the permanent effects. Obviously a large change coming suddenly is liable to have sharp impact effects which may or may not last and which may even be reversed. In this part of the analysis it is necessary to consider only the more permanent steady differences that might be made to gross and net incomes by more or less progressive tax structures, after any initial reactions to alterations in the present structure had worked themselves out. The case of an increase in progressivity via personal income tax can be taken by way of illustration. Three types of permanent reaction have to be considered: upon work effort, upon savings, and upon gross income claims.

4.25. Few issues have been more elaborately argued at the theoretical level or more inconclusively tested empirically (or are more difficult to test) than the effects of varying tax levels upon the individual's choice of work against leisure. It is generally held that a high marginal rate of tax will limit the desire to work, but the real question is whether the effect will be large or negligible. It need not be doubted that a marginal rate of 80 per cent or more, imposed on the marginal earned income of a man who

  ― 31 ―
can live comfortably on a few hours' work, will inhibit him from working longer unless he has an exceptional delight in the kind of work he is doing. But within the range, more relevant to practical discussion, of say 40 to 70 per cent, the matter is problematical. It is evident on all sides that most well-to-do people now earning incomes that are taxed towards the top end of the scale do work hard and long and it may be judged that pride and interest in their work are dominant reasons for this. The effects of tax rates may well be small. At the lower income levels the consequences of medium and high marginal tax rates are perhaps much more mixed, and very much more dependent on personal and job characteristics. Hence in general the Committee is doubtful whether, within the kinds of limit that need to be considered seriously, changes in the progressivity of the tax scale would have important effects upon work incentives.

4.26. Any progressive tax system must to some extent be inimical to savings. This follows from the familiar fact that within the working-age groups the proportion of income saved, generally speaking, increases with income. But the precise sensitivity of savings to tax rates on any permanent basis is extremely hard to judge, and the difficulty is compounded in inflationary conditions when the volume and pattern of savings are likely to be dominated by varying expectations about potential returns, real and monetary, and about the stability of income. It is also to be remembered that in the industrial sector (the needs of which are often in mind when the importance of saving to growth is being emphasised) much investment is internally financed and its productivity may be determined as much by its quality as its quantity. The Committee regards incentives to save, as it regards incentives to work, as important but not crucial factors in deciding whether the tax system should be made rather more or rather less progressive.

4.27. A still more enigmatic problem is that of the connection between the progressivity of tax rates and the occupational and hierarchical spread of gross incomes. In conditions such as those of Australia at present and in the foreseeable future, where a high proportion of incomes from effort is in one way or another collectively determined with considerations of ‘fair relativities’ very much to the fore in the debate, it is at least theoretically conceivable that any attempt to compress (or expand) established differentials in real income may be defeated by countervailing adjustments in gross incomes. This would be so if it were relativities net of tax that were involved in ideas about fair shares, and market forces were no obstacle.

4.28. It is a reasonable inference from the very frank specification of ‘fringe benefits’ which takes place in the recruitment of senior staff in many areas of the economy, that in that range levels of real income net of tax are a prime factor in settling remuneration. At other levels the matter is not so clear. Taxation is not overtly considered by the Arbitration Commission. But facts that will be relevant to the intensity of emotion behind a dispute do not need to be explicitly stated to enter the minds of negotiators. The notion of taxes as a burden may be misleading but it has a deep hold. When an increased wage, arrived at by either a Court award or direct negotiation, proves disappointing when received in the pay packet with tax deducted, consequences to the next claim may be expected. Explicit bargaining in ‘real’ terms, net of tax, is spreading overseas and the higher the general level of taxation becomes the more probable it is that it will come to Australia, more especially as, with continuing inflation, everyone is becoming used to thinking in ‘real’ terms. The matter is further considered in Chapter 6.

  ― 32 ―

4.29. There are limitations on the possible extent of such tax-offsetting income adjustments. As the relative prices of the services of those who secure an upward shift of gross income increase, it will be to the interest of their purchasers to economise and to change techniques so as to use more of other, now relatively cheaper, factors of production. But this may not always be possible, even in the long run, and the power of interest groups may be such as to hamper it when it is.

4.30. The kinship is obvious between the argument here and the questions of relative shares (and inconsistent group claims about them) that are central to the debate on the causes of the inflation now endemic in western economies. There seems unfortunately to be as little hope of agreed conclusions in this debate as in the wider one.

4.31. It is about the relationship between taxation and earned incomes that these doubts principally arise. Taxes on property income can be expected to prove more or less progressive in accordance with the rate structure. So can estate and gift duties, if adequately tightened up, and so too would a wealth tax if technically feasible. Hence the Committee does not come quite to the conclusion that the distributions of income net of tax and of property are not significantly amenable to tax policy when all repercussions are taken into account. It does however believe that the distributions are by no means so readily malleable as may appear at first sight.