INDIVIDUAL INCOME TAX RATES

Arthur Smithies, Harvard University

My assignment is to discuss the rate structure of the individual income tax. Insofar as possible I shall avoid trespassing in other fields marked out in the committee's investigation. For instance, I shall try to avoid questions such as the possibility of reducing some or all income tax rates and increasing other types of taxation to make up for the revenue loss.

On the other hand, the rate question necessarily involves matters not explicitly connected with rates. One of the committee's main concerns is the possibility of broadening the base of the income tax and thereby permitting reduction of the formal rate structure. Consequently, I cannot avoid problems such as closing loopholes, altering exemption levels and so forth-even though those matters may be dealt with more exhaustively in other papers.

One other preliminary remark: A paper such as this is necessarily speculative in character. Its conclusions rest largely on the judgment of the author, aided by deductive economic reasoning. Anything like empirical proof of the conclusions is virtually impossible. Although taxation is one of the oldest subjects of economic inquiry, it is one where decisions must still rest on opinion rather than on the definitive fruits of analysis. This is partly due to the fact that experiments cannot be performed on the taxpayer. It is impossible to hold everything else constant; and to observe his reactions to variations in the tax system.

Another reason is that much of the evidence we do have is derived from the assertion of interest groups. It is hard to find any recommendation on tax policy that runs counter to the economic, political, or ideological interests of the witness; and frequently vehemence of assertion seems to be regarded as a substitute for empirical evidence.

In fairness, I must add that the academic economist is unlikely to be free from bias. Economics has grown up with a utilitarian preference for measures that promote equality. Consequently the professional economist tends to place a heavier burden of proof on those who urge measures that increase inequality than he does on the proponents of changes in the tax laws that increase equality of the distribution of income and wealth.

The rate question can be conveniently examined by attempting to give answers to the following questions.

(1) If the income-tax revenue from a given national income is to be held constant, what changes in the rate structure are called for? Should its progressiveness be increased or decreased? Should loopholes be closed in order to permit reduction in the formal rate structure? And if so, what reductions?

(2) In a context of general tax reduction, what should be the relative priorities among various components of the income tax? Should

high bracket rates be reduced more than low bracket rates, or vice versa?

(3) In a context of a general tax increase, what should be the relative emphasis given to increases of the various income tax components?

I regard the second and third questions as of the most practical importance. They represent the circumstances under which structural changes are most likely to come about. Both questions should be asked. The answer to the one does not necessarily provide the answer to the other. For instance, a judgment that the highest brackets should not be increased does not imply that they should be top candidates for reduction.

The first question is raised because it identifies what is usually meant by tax reform. Moreover, it helps to cover some of the cases where a shift from the income tax to other forms of taxation is considered. If for instance, a reduction in the higher brackets of the income tax would justify an increase in indirect taxation, it should also justify an increase in the first bracket of the income tax. One possibility that is not covered by the questions as formulated is that tax reform may be so effective in stimulating economic activity and increasing the national income that reduction of some rates with no increase in others will result in no loss of revenue. I am deliberately excluding this as a possibility. While tax reform may have significant effects on economic growth and hence tax yields in the long run, I see little likelihood that such effects are likely to be pronounced in the short run.

In order to arrive at a judgment conceiving desirable directions of change, the tax structure needs to be examined from the points of view of the consequences of such changes on its feasibility, its fairness, effects on effort and incentives, and effects on saving and investment. Feasibility. By feasibility I mean the ability of the Government to create a tax structure that will be able to withstand the forces of erosion at the political and administrative levels. If the tax structure, once enacted, will be immediately subjected to successful attack, there are strong arguments against enacting it in the first place.

Since the problem of erosion has been exhaustively discussed in Mr. Pechman's paper, I need only refer to it in general terms. It can be taken for granted that any tax system will be continually under attack both by individuals and organized groups seeking relief from its provisions. With sales taxes, for instance, there is pressure to exempt particular commodity. (In that connection I would like to see some research done on erosion of State sales taxes.) With the income tax the pressure is to seek exemption for particular types of income or recognition for particular types of deduction from income.

Moreover, the pressure for relief increases with the severity of taxation. Everyone would agree that an income tax with 100 percent marginal rates near the top would either not survive or would be unenforceable. Consequently, progressive tax systems become decreasingly progressive at the upper end of the scale. This departure from the progressive principle that governs the middle and lower reaches of the scale is justified on feasibility grounds. The question thus raised with respect to any particular tax system is whether the feasible upper rate limit has been reached.

A further point to note is that special privileges, whether they result from taxpayer pressure or not are likely to confer their major benefits on upper-bracket taxpayers. The capital gains provisions, for instance, furnish no incentive to the bulk of taxpayers to have income treated as capital gains. Again, tax-exempt bonds provide little or no relief for the middle and lower brackets. They are so attractive to the high brackets that their price is at a level that makes them unattractive to the lower brackets, despite the exemption. Another factor is that access to special privileges may be difficult for most taxpayers. A man of modest means cannot go oil drilling on his own account, and cannot take the risk of some initial failures.

Consequently, one can expect that the gap between the formal rate structure and the effective rate structure will increase progressively as income increases. The benefits of the process of erosion will tend to be concentrated in the upper brackets. In the tax field at any rate, increasing discrepancies between fact and fiction can be regarded as undesirable; and in my view it should not be considered feasible to enact a tax system that produces them to a marked extent.

In my opinion, which I should think is widely shared, feasibility consideration alone should exclude any further increases in top bracket rates. The practical issue is whether high bracket rates should be reduced as a check to further erosion.

This raises a most difficult question of political morality. Should parties at interest be able to achieve modification of the law in their favor through their efforts to make the law ineffective. An analogous question is whether legal speed limits should be increased because they are systematically violated. The answer is not necessarily "No." If the speed limit is ridiculously low, the campaign should succeed; but if it represents a reasonable compromise among the various interests, the speedsters should not be allowed to have their way.

My judgment with respect to the income tax is that present rates should not be reduced on these grounds alone. However, I would not exclude legislative recognition of the existing situation, whereby rates were reduced and special privileges removed. The cause of political morality would be furthered by a rate structure that reflected the true relation between taxation and income.

The question of feasibility at the lower end of the scale cannot be discussed with respect to the income tax alone. Whether or not a 20 percent first-bracket rate is feasible or not depends on the total tax burden and not merely in the income tax. That is not my present subject. I merely state that I believe the present tax burden, considered as a whole, in the United States, is entirely feasible.

Fairness.-No professional economist today would consider himself entitled to make ex cathedra statements concerning the fairness of the income tax. Yet the ideas of economists in the past have profoundly affected commonly accepted notions of equity. The economist of today may be of some assistance to bodies such as the Committee on Ways and Means in formulating acceptable standards and in speculating on the standards that are likely to prevail in the future.

I am, therefore, unable to say whether the present degree of progression is equitable or inequitable; whether married couples are treated too tenderly or too harshly in relation to bachelors and spinsters; or whether the proper exemptions are allowed for dependents.

It may, however, be useful to consider how standards of equity with respect to such matters may affect the tax structure, with the continued growth of economy.

The principles of equity usually applied are the ability-to-pay principle and the benefit principle. Taxpayers should contribute in accord with their ability, but taxpayers should also pay for the services that they get from the Government.1

The ability to pay principle leads directly to progressive taxation; and the principle of progressiveness—not its degree of applicationcan be regarded as generally accepted even by the National Association of Manufacturers. Ability to pay applies most unequivocally where the Government is heavily engaged in matters such as defense, where the benefits accrue to the community as a whole rather than to identifiable members of it.

Assuming, as I do, that defense will continue to dominate the budget, ability to pay will continue to be perhaps the decisive criterion of equity. But how will it be applied as the economy grows richer? Since defense is in the common interest, it seems equitable that everyone who can afford to do so should make some contribution to it. And, as per capita income increases more people are in a position to contribute. In fact I believe that, except for particular pockets of poverty which should probably not be relieved by the tax system in any case, we have already reached a stage where every American family can contribute, and those contributions can increase.

These considerations bear directly on the question of the personal exemption. Owing to the revenue needs of the war and the postwar period the personal exemption has been held at the present level of $600 and has thus become a steadily diminishing proportion of increasing individual incomes. It has frequently been urged that this tendency be reversed.

My own view is that the trend should continue. I might even be pushed into saying the personal exemption should be reduced and eventually disappear. The theory of the exemption presumably is that a large number of families are living at a subsistence level, and cannot contribute to the cost of Government. This seems now to be an outworn notion.

Of course if the money value of the exemption is held constant, and still more if it is reduced the rate structure should be revised. As the exemption becomes a diminishing proportion of income, rates could be reduced with no diminution of revenue.

At the other end of the scale, there is no reason to believe that ability to pay will be diminished. On the other hand the process of development tends to equalize real incomes even though the distribution of many incomes remains unchanged. This is because productivity increases occur mainly in the industries producing for mass production. This fact would seem to argue against further increases in progressiveness over the whole income scale; and may argue for a reduction.

With respect to the benefit principle, the individualistic principles of our society seem to require that if people can pay for identifiable goods and services they should. This is already reflected in social

1 For an illuminating discussion of these principles see Richard Musgrave, "Theory of Public Finance," New York, 1959.