― 188 ―

III. Shape of Rate Scale

14.15. From submissions to the Committee it is evident that the shape of the rate scale, so far as it determines how much more tax a high income bears than a low, is a subject on which the community holds strong but differing opinions. The appropriate degree of progression in the rate scale of personal income tax is a subordinate aspect of the question of the degree of progressivity proper to the tax structure as a whole. As the analysis of Chapters 3–5 suggested, preoccupation with the progressivity of income tax obscures the realities of the problem. The discussion of progressivity requires an appreciation, virtually impossible on the basis of present statistics, of the impact of all taxes, and not merely of one of them. This becomes a consideration of increasing importance with the imposition of capital gains tax, the restructuring of company tax and estate and gift duties, and the introduction of a broad-based consumption tax—in other words, all the measures later proposed in this report. Underlying the implications for progressivity of the measures taken—or not taken—in these areas are the trends in revenue requirements, the evolution of expenditure policies, and the basic developments in the economy over the years.

14.16 Some consideration of the present rate scale is nonetheless called for. Table 14.B may conveniently serve as a starting-point to identify three ranges of net income:

  • (a) The low range, up to an income of, say, $3,000. In 1971–72, 39.8 per cent of taxpayers fell within this range, providing 9.4 per cent of total revenue from personal income tax. Because of continued inflation and recent changes in tax rates, the corresponding percentages are likely to be noticeably lower in 1974–75.
  • (b) The high range, above an income of, say, $20,000. Only 0.4 per cent of taxpayers fell within this range in 1971–72, but they accounted for 7.9 per cent of total revenue. Somewhat higher percentages may be anticipated in 1974–75.
  • (c) The intermediate range between the two extremities, encompassing the majority of taxpayers (59.8 per cent) in 1971–72 and supplying the bulk of tax revenue (82.7 per cent). A higher percentage of taxpayers is likely to come within this range in 1974–75, but the percentage contribution to revenue may not be greatly different from what it was in 1971–72.

Low Range

14.17. One of the first questions of detail to settle in any analysis of the income tax scale is the level of the initial step. At present this step, at 1 per cent, is effectively abolished, except in a few cases, by the exemption of the first $1,040 of taxable income, so that up to this level there is a marginal (and average) rate of zero. After the shading-in arrangements already described, a marginal rate of 7 per cent resumes at $1,061 where the average rate is 1.3 per cent. As noted, in the United Kingdom system the initial marginal rate now stands at 33 per cent. The contrast is striking.

14.18. The most important advantage of an initial step of some magnitude is not that it raises much tax from the taxpayers whose total income is confined within this step; it is rather that a sizeable minimum average rate is thereby struck for all higher incomes. Without a significant rate on the lower steps, much higher marginal rates than otherwise must be imposed further up the income scale if substantial revenue is to be raised from the later income brackets with their greater numbers.

14.19. On the other hand, it may seem quite wrong and vertically inequitable to impose any significant income tax on very low incomes, and this points to a near-zero rate on such incomes, whatever the unwelcome implications for marginal rates higher up. In discussion of tax matters it is often suggested that this dilemma can be escaped

  ― 189 ―
by one or other of two devices. First, a universal personal allowance can be provided, as a deduction from taxable income, tax only being paid on what remains: as indicated in a footnote to Table 14.A, the United Kingdom, the United States and Canada provide for a universal personal allowance. Or, secondly, a non-reimbursable tax rebate can be given cancelling out the tax that would otherwise be paid at these levels: New Zealand has recently introduced a rebate along these lines. But these are in fact exactly equivalent to making the first step a zero-rate one; and they mostly serve to conceal the abrupt rise in the effective marginal rate when, at the point of exhaustion of the allowance or the rebate, the marginal rate of the tax scale begins to be effective. Universal tax allowances or non-reimbursable tax rebates are merely techniques for altering the actual progressivity of the tax scale. It seems altogether simpler and less confusing to determine the progressivity of the income tax on its merits in the one obvious place: in the scale. The issue of the amount of tax to be levied on low incomes should not be obscured by artificialities.

14.20. It seems essential to the Committee that decisions about the level of the first step, and the general shape of the rate scale for low income earners, should be made in the light of some clear understanding of the kinds of people it is who have these very low taxable incomes. That information is currently not available. But it is reasonable to suppose that they are persons who are realising capital or receiving exempt incomes from social services benefits or scholarships, or have been working part-time or for only part of the year. And many others may be a statistical illusion arising from a person with more than one job preferring to appear as two persons to the Commissioner. At a time when the minimum wage is over $4,000 a year, it is clear that few could in fact survive on these lower incomes if the latter were a true measure of available resources for essential consumption.

14.21. The Committee would not recommend any increase in the exemption limit above the present figure of $1,040. Moreover, it sees considerable merit in an initial marginal step rather higher than at present and extending further up the income scale. The initial step might be set at 20 per cent. In view of the fact that a marginal rate of 20 per cent now applies on increments of taxable incomes between $3,000 and $4,000, this would mean a first step extending up to a taxable income of $4,000. (The band of taxable income above $1,040 over which there must be shading-in would, of course, have to be rather wider than at present.)

14.22. An increase in the height and width of the initial marginal step in the manner suggested would in itself involve a substantial increase in tax revenue, since the higher marginal rates on initial income would apply to all taxpayers. It would be important to ensure that additional revenue was used, as far as possible, to protect the living standards of specific categories of taxpayers in the low range and at the bottom end of the intermediate range, notably those with families or dependent on social service benefits. This will be especially necessary if, in line with what was recommended in the previous chapter, social service benefits, including child endowment, are made taxable.

14.23. In paragraph 12.6 the Committee has recommended that the dependent spouse allowance be converted to a rebate, which might be set at $300. However, if an initial marginal rate of 20 per cent is introduced, as is proposed here, the spouse rebate will need to be of the order of $600 to ensure that a married person on a taxable income of $4,000 pays no more tax than at present.

14.24. The proposals in paragraphs 14.21–14.23 are addressed to the more immediate future. The Committee has indicated, however, that as a long-run objective it favours placing less reliance on personal income tax. It might therefore be possible to

  ― 190 ―
visualise an initial marginal rate on taxable incomes up to $4,000, fixed in the first instance at 20 per cent, being in time gradually reduced. Indeed, there would be considerable merit, if a simpler tax system is to be achieved, in at some stage zero-rating the initial marginal step, so that persons on low incomes pay no income tax at all and persons on high incomes are exempt from tax on the initial increments of income. It would then of course be necessary, as has been indicated earlier, to rely exclusively on taxable grants to assist low-income families.

High Range

14.25. It is sometimes argued that there is considerable scope for redistributing the income tax burden from the lower to the higher income groups. That is to say, tax reductions on lower incomes could be financed by tax increases on higher incomes. However, from the data of the distribution of the tax burden by grade of net income summarised in Table 14.B, it is readily apparent just how little scope there is for such redistribution. The 0.4 per cent of taxpayers in 1971–72 with net incomes in excess of $20,000 received only 3.0 per cent of taxable income, contributing 7.9 per cent of total tax collected. Given that the average rate of tax on a taxable income of $20,000 was over 40 per cent, further significant increases in revenue could not have been obtained from taxpayers in that range. An increase to 75 per cent in the average tax liability of a taxpayer on $20,000 would have left him with after-tax income of $5,000 and involved a near 100 per cent marginal tax rate. Such a change for all taxpayers with taxable incomes in excess of $20,000 would have added only about 3.5 per cent to total revenue, and then only if the taxpayers concerned had continued to earn the same income as before—a dubious assumption given the confiscatory marginal tax involved.

14.26. Hence in establishing what the maximum income tax rate should be, factors other than simply contribution to revenue must be the chief concern. As already indicated in paragraph 14.15, the question of the income tax scale inevitably merges with the wider issue of the progressivity of the tax system as a whole and the character of other taxes in the system. Within a narrower compass, considerations of some relevance in fixing the top rates of personal income tax include the relationship between the systems of personal and company income taxes, the effect of taxation on incentives to work and save, and the incentives provided by high rates of taxation for tax avoidance and evasion.

14.27. As is pointed out in Chapter 16, there may be some advantage in linking the maximum rates of personal and company income tax. This was, for example, a key element of the proposal of the Carter Commission for full integration of personal and company income tax, and West Germany is considering a scheme to link the maximum marginal rates of both taxes. Such a link affords a means of limiting the purely tax advantages of retaining undistributed profits in the hands of companies: the tax burden on undistributed profits would equal the top marginal rate of income tax and for most taxpayers this rate of tax would exceed the rate of personal income tax. Such a relationship, however, involves setting the maximum rate of tax applied to all personal income in the light of considerations relevant only to dividend income, which might easily lead to inappropriate results.

14.28. In fixing the maximum marginal rate of tax, as well as in framing the rate scale lower down, some account must be taken of possible effects on the incentive to work. A taxpayer's incentive to work is affected by both the ‘income’ and ‘substitution’ effects of income tax. Because income tax reduces a taxpayer's net of tax income, he will to that extent have an incentive to increase his work effort to recoup at

  ― 191 ―
least some of the income taken in tax. The extent of this income effect will be determined by the taxpayer's average rate of tax: the higher the average rate, the greater will be the influence of the income effect. However, the substitution effect works in the other direction. The effect of income tax is also to reduce the return from additional work effort, and the higher the marginal rate of tax the greater will be the incentive to reduce work effort and substitute leisure in view of the increased relative attractiveness of leisure.

14.29. Economic theory is thus inconclusive as to whether high rates of income tax adversely affect incentives to work. Commonsense reasoning suggests that the higher are marginal rates of tax in relationship to average rates the more is the substitution effect likely to outweigh the income effect; but psychological and sociological forces obviously play a major role in determining the work ethics of particular individuals, while the ability of taxpayers to vary the number of hours they work may be limited by such factors as the length of the miniumum working week or the state of demand in the labour market. It is little wonder that empirical evidence on the effects of income tax on incentives to work has tended to be inconclusive and of hardly any guidance to policy-makers.

14.30. The implications of income tax for personal saving are clearer, for here the substitution and income effects tend to reinforce each other. High marginal rates of income tax encourage the substitution of immediate consumption for saving; high average rates, especially at the upper end of the income scale where saving features more prominently, mean lower disposable incomes and hence less opportunity to save. But it must not be overlooked that a large part of saving today is undertaken in the business sector and by government; and the Committee has already expressed the view in Chapter 4 that incentives to save, like incentives to work, are important but not crucial in deciding how progressive the tax system ought to be.

14.31. What it is possible to be rather more certain about is the encouragement high marginal rates of income tax give to avoidance and evasion. The large number of professional people who own farms is not necessarily a reflection of a love of the land: tax considerations often loom large. High marginal rates of income tax substantially increase the reward from an activity of this kind, and attempts to avoid the workings of the law pose serious problems for the equitable administration of the tax system. As far as loopholes in the legislation or in the administration of the law are concerned, a frequent criticism of the tax system raised in many countries with high rates of taxation is that the tax system is steeply progressive only in form and that in substance many wealthy taxpayers are able to avoid the full effects of the progressive rate schedule. The Committee has been concerned in various of its recommendations in this report to limit the scope for tax avoidance.

14.32. Evasion of the law raises separate though to a large extent interrelated issues. Taxpayer compliance is an essential feature of the administration of a comprehensive income tax levied on a large number of taxpayers; and while evasion will always be a problem for tax administration, the problems can be expected to be more severe the lower the willingness of the public to accept that the tax system is a fair and equitable one. Except in special situations such as war-time, a high burden of income tax is unlikely to be favourably received by the population at large. This is a factor of considerable practical relevance to the design of a progressive rate schedule and explains in some measure why the Carter Commission in Canada and tax authorities elsewhere have been attracted by the prospect of a maximum marginal rate of 50 per cent.

  ― 192 ―

14.33. The Committee, too, sees advantages in top marginal rates of tax lower than at present. As a first step it might be possible, in the more immediate future, to lift the points on the income scale, in the top range, where the highest marginal rates start applying: for example, the 64 per cent rate could cut in at a taxable income of $40,000 instead of the present $20,000, and the 67 per cent rate at $80,000 instead of $40,000. But in line with the Committee's view that more reliance should be placed on the taxation of goods and services and less on personal income tax, a reduction in the maximum marginal rate to the region of 50 per cent would be thought an appropriate long-term target.

Intermediate range

14.34. A significant feature of the intermediate range, which encompasses the majority of taxpayers and provides the bulk of revenue, is the high marginal rates of tax applying over the greater part of the range. Thus by the time taxable income reaches $8,000, the marginal rate is already 48 per cent, and by the time it reaches $12,000 the rate is as high as 55 per cent.

14.35. It follows that any attempt to raise substantial additional revenue from this range, if unaccompanied by increases in marginal rates in the low range, would necessarily involve lifting marginal rates over a wide section of the intermediate range to excessively high levels. For example, 10 per cent more income tax revenue, secured exclusively from net incomes in excess of $10,000, would, on the basis of 1971–72 figures, have meant an increase in average tax rates of the order of 50 per cent and a rather steeper increase in marginal tax rates. Instead of paying at a marginal rate of 48 per cent, an individual on a taxable income of $10,000 would have been subjected to tax at a rate of over 80 per cent on each extra dollar he earned.

14.36. Recent restructuring of the rate scale to relieve the tax burden on the low range, and on the bottom end of the intermediate range, has brought the issue into sharp focus. As Table 14.C reveals, taxable incomes between about $5,500 and $11,000 now face significantly higher marginal tax rates than in 1973–74: indeed, on a taxable income of $7,000 the increase is more than 16 per cent.

TABLE 14.C: AVERAGE AND MARGINAL TAX RATES, 1973–74 AND 1974–75: Selected Incomes

1973–74   1974–75  
Taxable income   Average tax rate   Marginal tax rate   Average tax rate   Marginal tax rate   Change in marginal tax rate between 1973–74 and 1974–75  
per cent  per cent  per cent  per cent  points  per cent 
1,500  6.4  12.7  3.0  7.0  -5.7  -44.9 
2,000  8.4  17.2  4.0  14.0  -3.2  -18.6 
3,000  12.0  22.0  7.3  20.0  -2.0  -9.1 
4,000  15.2  30.3  10.5  26.0  -4.3  -14.2 
5,000  18.3  33.3  13.6  32.0  -1.3  -3.9 
6,000  21.0  35.7  16.7  38.0  +2.3  +6.4 
7,000  23.3  37.9  19.7  44.0  +6.1  +16.1 
8,000  25.3  41.8  22.8  48.0  +6.2  +14.8 
10,000  28.9  48.2  27.8  52.0  +3.8  +10.8 
12,000  32.1  54.6  31.8  55.0  +0.4  +0.7 
16,000  37.7  60.3  37.6  60.0  -0.3  -0.5 
20,000  42.2  64.0  42.1  64.0  0.0  0.0 
40,000  53.1  66.7  53.1  67.0  +0.3  +0.5 

  ― 193 ―

14.37. The upward adjustment of marginal tax rates in this latest restructuring has a number of disturbing implications. It may well inhibit work effort in view of the fact that, for many taxpayers, higher marginal rates have been accompanied by lower average rates: for such taxpayers both the income and substitution effects referred to in paragraph 14.28 will exert a depressing effect on work effort. Moreover, to the extent that marginal tax rates are built into claims for higher incomes in the manner discussed in Chapter 6, increases in marginal rates over much of the intermediate range will tend to foster more rapid inflation. Finally, raising marginal rates to higher levels may encourage avoidance and evasion among a larger circle of taxpayers.

14.38. It was in some degree because of its concern to keep marginal rates in the intermediate range as low as possible that the Committee in earlier paragraphs proposed a higher and wider initial marginal rate bracket and some modification of marginal rates in the high range. These measures will permit wider tax brackets and less steeply rising marginal rates over the intermediate range. Any long-run shift towards less reliance on income tax will facilitate further moves in this direction.

14.39. As indicated in Chapter 6, it is taxpayers in the intermediate range, especially in the top half of the range, who have suffered most from the unplanned restructuring of tax liability produced by inflation. In the next section the question of adjusting the rate schedule for the effects of inflation is examined more closely.