Tax Analysts®Tax Analysts®

Article Archive

February 15, 1996
Historical Document: 'Facing The Tax Problem' -- Book Two, Analysis (Part I).
Carl Shoup, Roy Blough, and Mabel Newcomber

====== SUMMARY ======

In Book Two of the report, "Facing the Tax Problem," the authors state that the U.S. tax system serves both primary and secondary aims. Revenue raising and social control are the principal goals of a tax system, providing the broad framework for discussions of tax reform. Secondary aims include tax justice, revenue stability, ease of administration, tax consciousness, and intergovernmental coordination. Using these categories, the authors evaluate the U.S. tax system of 1937.

====== FULL TEXT ======

Book Two



Section A. Revenue
Section B. Social Control

[page 57]

Section A. Revenue

Chapter 5


One of the primary aims of taxation is the raising of revenue. This problem may be considered both in general terms of taxable capacity, as is done in this chapter, and in terms of the yield that may be expected from each tax, a matter that is covered in Chapter 6.

The rapid growth of public expenditures in the United States during the past few decades and the prospect of further growth /1/ raise the question whether there is any limit to the taxable capacity of the country. This question applies to the tax system as a whole and will therefore be considered before any estimates of the revenue that may be expected from specified taxes are given.


Taxable capacity in the United States cannot be defined by a single rigid limit. It cannot be expressed in terms of a definite sum of money. Several successive limits can be set instead, which can be overstepped, one after another, under certain specific conditions. Some of these are legal, others political, and still others are economic.

Constitutional Restrictions

Constitutional restrictions on the taxing power may in a sense be regarded as limits to taxable capacity. When a state constitution forbids a locality, such as a county, to levy a property tax at more than a certain specified rate, the power of the county's tax system to

/1/ See Table 19, p. 118.

[page 58]

produce revenue is for all practical purposes strictly limited. Such
limits, however, apply only to a sub-system, such as that of a
county, within the entire system. They are therefore more properly
discussed in connection with specified taxes. Furthermore, the limit
is not a fundamental one. It can be eliminated simply by amending the
constitution. If amendment is impossible in practice, the
constitutional provisions reflect underlying political and economic
forces that must be isolated and analyzed.

Political Limits

A limit to taxable capacity may be set by political forces -- that is, by the unwillingness of the voters to pay taxes. This political limit may be reached long before the economic limits to be explained below take effect. It depends of course on innumerable factors, and it changes with them from day to day. /2/ The degree and kind of tax consciousness are major conditioning factors. /3/ A large number of taxes at low rates rather than a small number of taxes at high rates may increase the political capacity to pay. A gradual increase in taxation may yield larger tax collections in the long run than a smaller but abrupt increase. Perhaps, indeed, if the tax were imposed gradually enough, the amount that could be raised would be far greater than is commonly supposed.

Another major factor is the purpose for which the tax money is raised. The political capacity limit expands as those in control expand their demands for services. This expansion is found in an extreme form in war expenditures. In contrast, the limit will remain restricted as long as small minority groups that favor various expenditures work only for their separate interests. The limit expands when the minorities consolidate their programs.

/2/ See Note 1, Political Limits to Taxable Capacity, p. 528.

/3/ See pp. 352-64.

[page 59]


Most of the popular discussions of taxable capacity, however, seem to be concerned less with these political limits and more with what is assumed to be an economic limit. Beyond that limit, the argument runs, the people would be crushed by a tax load too heavy to bear. It is important, therefore, to determine whether an economic limit of taxation actually does exist, and, if it does, how near the existing level of taxation has come to it.

Limits Dependent on Nature of Expenditure

The position of the limit -- indeed, perhaps its existence -- depends upon the way the tax money is spent by the government.

The limit will be near if the revenues are spent in such a way that nothing is ever added to the flow of consumers' goods and services. The nation's production is then decreased and hence the income from which tax revenues flow. The process cannot be carried far without starting to dry up the sources of revenue.

Unproductive expenditure usually occurs when the government requisitions goods for destruction or pays to have productive factors held idle.

Another example is the use of tax revenues to support certain groups of people who themselves contribute nothing. The people may be government employees, and the failure to contribute is then a result of inefficient government, as when a department is overstaffed or staffed with incompetent persons. They may, on the other hand, consist of those who are idle because of involuntary unemployment, or because of old age, or injury, or sickness. The government relief program for this latter class, however, may indirectly increase the productive efficiency of the employed. They may pro-

[page 60]

duce more efficiently because they worry less about their own future
or the present insecurity of their friends or relatives. Moreover,
unemployment relief payments may result directly in an increased flow
of consumers' goods and services. Governmental spending brings into
play economic forces whose full significance cannot be given in
simple terms. Despite all qualifications, however, the support of
large numbers of idle persons is the kind of expenditure that brings
the taxable capacity limit near to the existing level of taxation.

The limit is farther away if the tax money is used to supply the community with goods and services, even though the goods and services do not include the basic necessities of life.

If the tax money is to be used to produce the necessities of life, the limit of taxable capacity becomes remote. At least, the community does not destroy its own productive power by producing too little of the things that are needed to maintain the persons who produce. The government might, for example, take so much money to spend on even such a valued service as education that the taxpayers would not be left enough money for food. The government -- the taxing-spending machine -- decides what the producing persons in the community shall do with a certain part of their effort. If they are forced to spend too much time otherwise than in keeping themselves alive and in working condition, their efforts inevitably slow down and cease. A limit of taxable capacity will then have been reached.

Limits Not Dependent on Nature of Expenditure

The economic limits to taxable capacity that have been discussed above depend on the nature of the expenditures. They are not, therefore, purely tax limits. Another kind of economic limit suggested by some observers depends, instead, simply on the volume and kind

[page 61]

of taxes. If it were passed, private business would, it is claimed,
decrease and in time practically disappear, chiefly because business
initiative would be stifled, and private savings would no longer be
accumulated. /4/

The soundness of these statements is obviously a matter of great concern. Unfortunately, the available facts with which to test them are few or are in a state of confusion. A qualified appraisal may be made by reasoning from facts that are generally known. Even this appraisal may turn out to be little more than an expression of personal opinion. Nevertheless, it is worthwhile to attempt it.

Effect of the Tax System on Business Initiative

People must be induced to risk money if the private capitalist system is to function. Of course, the investor is not the only one who assumes a risk. In practice, however, more options are open to him than to others, and the effect of tax policies on his decisions is generally both the most powerful and the most direct.

His willingness to risk money is only one part of business initiative. Another is the willingness to exert effort in promoting and managing businesses. This factor, however, seems even more uncertain and will not be covered in the present analysis.

The taxes that are most commonly accused of weakening the willingness to risk are income taxes, death taxes, and gift taxes, when they are levied at progressive rates -- that is, rates that increase as the amount to be taxed increases.

To some extent, investors weigh the chances of success and failure before placing their money. By imposing an income tax on profits, the government claims a share in the gains without offering to share in the losses. The result is something like an unfavorable shift in gambling

/4/ See Note 2, New Fabian Research Bureau Report, p. 529.

[page 62]

odds, and it may be enough to deter the investor. Whether it will or
not depends on the other courses of action open to him, and how they
appeal to him.

He may turn to some other field of business that offers less reward and also less risk. Since a tax at progressive rates, such as the federal income tax, takes a smaller proportion of the income as the income shrinks, the government will claim a smaller proportion of the less spectacular reward.

Suppose, for example, that an investor estimates roughly that he has one chance in ten of succeeding in venture A, where success will bring $500,000, and one chance in two of succeeding in venture B, where success will bring only $50,000. Suppose, further, that the two ventures are about equally attractive to him until he considers the income tax. A steeply progressive income tax that will cut down the net reward of venture A by a much larger percentage than it will the reward of venture B will induce him to choose venture B.

If personal exemption, earned income credit, etc. are disregarded and if the venture to be chosen is the investor's only source of income for the given year and is conducted as an unincorporated business, the present federal tax will cut the return from the risky venture, A, from $500,000 to $194,000. Similarly the gain from the relatively safe venture, B, will be cut from $50,000 to $40,300. A comparison of $194,000 with $40,300 is of course quite different from a comparison of $500,000 with $50,000.

Although nothing is known of the actual extent of such influences, it seems reasonable to assume that the present federal income tax is exerting an effective pressure on many investors to turn away from high-risk ventures to relatively safe fields of business.

The pressure depends on both the high rates of the tax and the progressive feature. The investor may be

[page 63]

so rich that he is already in the top parts of the income tax
schedule, where the progression is slight. If the choice between
venture A and venture B as described above is to be made by an
investor who already has a net income of $1,000,000, the problem
changes. Since that part of his income over $1,000,000 but less than
$2,000,000 is taxed at 77 per cent, in effect the $50,000 from
venture B and the $500,000 from venture A will be taxed at the same
flat rate. The rate is so high that the investor may decide to select
neither venture, but at least the degree of weighting against the
risky enterprise that was noted in the first example no longer

The investor may, of course, purchase state or municipal bonds or short-term federal securities and thus avoid entirely both the federal surtax and the federal normal tax. His investment in such securities does not necessarily deprive business of capital funds, as has been often charged. For every buyer there is a seller, and the person who sells the bonds now has the same investment problem that the buyer used to have. The only way in which the tax-exempt feature can hamper business initiative by depriving it of capital is by increasing the total amount of tax-free securities outstanding.

In other words, for state and local issues, the proper question is: how much fewer securities would the states and localities sell if the interest on their bonds were not exempt from federal income tax? This is another way of asking whether these bodies, when deciding whether to incur a debt, attach much importance to the rate of interest that they have to pay. The answer is largely a matter of opinion. The writer's present impression is that the volume of state and local debt has not been much influenced by exemption from the federal tax.

The taxes or the rates that happen to be in force at the moment are not decisive. The taxpayer looks to the

[page 64]

future. What may be is more important to him than what is. A constant
threat of a high income tax rate even if no such tax is ever levied
is a stronger deterrent to the investor and promoter than the actual
existence of such a tax over an extended period coupled with a
constantly renewed hope of repeal.

The present study cannot carry the analysis of taxation and business initiative further. No factual evidence has been gathered on the subject -- at least on conditions in the United States, -- nor has any substantial theoretical analysis been offered. Partly in order to discover whether taxation's effect on business initiative has interested economists who are not specialists in public finance, a survey has been made of some thirty or forty books by British and American authorities in the field of general economic principles, money and banking, trusts and corporation problems, etc. The period covered was 1925-36, a time when the federal income tax rates in the United States and many of the income taxes abroad were the subject of considerable popular discussion -- especially their effect on business initiative. The result of the survey was almost completely negative. /5/ The public finance specialists, on their side, have contributed practically nothing on the subject during the past decade or so in the United States. Judgment must therefore be suspended. Decisions on tax measures must for the present at least be made very largely on other grounds than the possible effect exercised by various kinds of tax on business initiative.

Effect of the Tax System on Saving

An economic limit may be reached through a discouragement of saving rather than of willingness to risk. The two are closely related, as has been seen. If the possible winnings in a risky enterprise are cut down by

/5/ These statements are based on a memorandum by C. Lowell Harriss.

[page 65]

taxation while the chances for loss remain as before, capital may not
enter that field. Perhaps the capital that would have gone into the
risky field under a different tax system will not be accumulated at
all. Thus a progressive income tax, if high enough, may tend to
restrict capital accumulation.

Taxes that do not have an especially strong effect on risk- taking may nevertheless affect capital accumulation. Even a tax system consisting wholly of taxes on commodities might therefore affect the amount of savings. Fundamentally, the answer depends on how much the taxpayers would have saved if the money had been raised by some other kind of tax-assuming the government's expenditures to be unchanged. If possible reductions in expenditures are to be considered too, the problem becomes one of comparing what the taxpayers would have done with the money and what the government does do with it. If all the taxpayers would have saved for investment all the money they now pay in taxes, and if the government invests none of it, the tax and the expenditure decrease the amount of capital that would otherwise have been accumulated.

In any case, no tax exerts a uniform influence on all taxpayers. The taxation of an estate or an inheritance, for example, may or may not lead to a decrease in the nation's savings. Some decedents might have saved correspondingly less to pass on to their heirs had it not been for the prospect of high death taxes. Likewise, some beneficiaries might have spent the money that was taken in taxes instead of saving it.

The direction of the net effect of some taxes is evident. For instance, so much of the saving done by corporations and by wealthy individuals seems to be the result of a simple failure to spend rather than a painful self-denial that an increase of the tax burden on these firms and persons would almost surely result in a net decrease

[page 66]

not in their consumption, but in their savings. Thus the community's
savings are probably less when a given amount of revenue is raised by
a progressive income tax than when it is raised by a general retail
sales tax. /6/ However, any measurement of these effects is not yet

For present purposes, therefore, the search for a taxable limit in an economic sense has chiefly negative results. Positive statements should be distrusted. The faulty logic of many current statements -- as, for example, that death taxes are necessarily capital-destroying taxes -- should be clearly recognized. Finally, decisions on tax policy will usually have to be based on tests other than the effect of taxation on business initiative and the accumulation of capital.

Tax Total Expressed as a Percentage of National Income

Popular writings during the past few years have often compared the total tax bill with the national income. Thus, if the national income is estimated at $60,000,000,000 a year, and if the total tax collections equal $12,000,000,000 a year, the writers in question point, usually with alarm, to the fact that the taxes equal 20 per cent of the national income.

From what has already been said in this chapter, it must be evident that any such percentage figure, standing by itself, has a restricted meaning. Depending on a number of factors, such as the uses to which the tax revenues are put, a percentage of 20 may or may not give cause for concern.

Two specific warnings may be given against using such percentages in discussions of the existing tax system.

The first is based on the meaning of the term "national income." The totals in common use do not include many important services that have a value but do not appear in money terms. They do not include, for example, the

/6/ A. C. Pigou, A Study in Public Finance (London, 1929), pp. 79-84.

[page 67]

value of housewives' services, or the net rental value of relatively
durable consumer goods used by their owners (homes, automobiles,
etc.). If these items were included, the percentage of the total
represented by taxes would drop sharply, especially in times of
depression, since these services are, like government services, a
relatively stable part of the total.

The other warning is based on the double counting of taxes. The national income figure is designed, in the main, to show the total amount available to individuals to spend on themselves as consumers. One method of computing the national income is to add up income that is paid out to individuals in the form of dividends, interest, salaries, wages, rents, and profits. If this figure is, for instance, $60,000,000,000, it means that $60,000,000,000 is available to ultimate consumers to meet personal expenses, such as clothing, food, and shelter. Some part of it must be used to pay for items supplied by the government. The payment is usually made in the form of taxes.

Some of the taxes, however, do not have to be met out of the $60,000,000,000. They have already been paid -- at the source, so to speak -- by business concerns, out of funds destined for investors or employees. Part of the $60,000,000,000 comes from dividend payments, for example. Suppose that $1,000,000,000 of the total taxes were paid out of corporation profits. If this $1,000,000,000 had instead been raised from taxes on the individual, the corporations would have been able to pay out $1,000,000,000 more in dividends. The national income would then have been $61,000,000,000. By a mere change in the kind of tax, the national income, as usually computed, would be increased. If total tax collections are $12,000,000,000, the proper comparison to make is between $12,000,000,000 and $61,000,000,000 -- or possibly, for certain purposes, $11,000,000,000 and $60,000,000,000.

[page 68]

This point is usually overlooked in popular discussions. The resulting percentage of tax payments is therefore too high. How much too high it is cannot be determined without further research. There is evidence, however, that, out of the $12,000,000,000 or so in taxes currently paid, at least $1,000,000,000 should be deducted if a comparison is to be made with the net figure of $60,000,000,000 for national income. This latter figure is perhaps not far from what the data for 1937 will show, but it is used here chiefly for illustrative purposes.

The business taxes in question must be taxes that cannot be shifted. In so far as the business concern shifts them to consumers in higher prices, they must be met out of the consumers' income -- in the example above, out of the $60,000,000,000. /7/

/7/ The national income paid out in 1935 has been estimated at $53,587,000,000, and the national income produced in 1935, at $52,959,000,000. Department of Commerce, National Income in the United States, 1929-35 (1936), pp. 22, 26. The statements on pp. 66- 68 above are based largely on a memorandum by C. Lowell Harriss on national income.

[page 69]

Chapter 6


Taxable capacity as discussed in the preceding chapter is of long-range importance and a matter of billions of dollars. The present chapter deals with the immediate past and the near future, and the analysis will be in terms of specified taxes and hundreds of millions of dollars.

No attempt will be made to predict outright the course of revenues over the next few years -- a task for the budget-maker, which involves, among other things, a prediction of business conditions. Another limitation to the data in this chapter concerns the effect of the tax rates on the amount of income, property, etc. reported for taxation. For example, high income tax and estate tax rates may tend to prevent the accumulation of large taxable fortunes and tend to decrease the number of large taxable incomes. At extremely high levels they may not be feasible from an administrative point of view. Uncertainty about the extent of these effects, however, makes it impracticable to take them into account in this survey.

The estimates for the income taxes and the death tax are based directly on data reported in the federal government's Statistics of Income /1/ for various years. For example, one of the estimates shows the amount of revenue that the 1936 income tax law would produce if it were applied to the incomes that were received in 1928.

The data deal chiefly with the federal government.

/1/ Treasury Department, Bureau of Internal Revenue, Statistics of Income . . . Compiled from Income Tax Returns and Including Statistics from Estate Tax Returns. Hereafter cited as Statistics of Income.

[page 70]

State and local figures are difficult, and in some cases impossible,
to obtain. Moreover, the elaborate computations involved in the
graduated-rate taxes, such as those on incomes and estates, did not
seem justified for any single state, and the computations for all the
states as a group were far beyond the resources of the survey.

Data on actual amounts yielded by various federal and state taxes are given in detail elsewhere. /2/ In the present chapter they will be used chiefly in connection with estimates based on certain assumptions. The basic summary estimates for income and estate taxes are given elsewhere, /3/ but brief summary tables appear in this chapter. Unless otherwise specified, all data on yields are given for the fiscal year ending in the calendar year stated and are evened off to the nearest tenth of a billion dollars.


Federal Personal Income Tax

In considering the estimates of the possible yield from the federal personal income tax under the rates and exemptions of the Revenue Act of 1936, the unusual height of the rates must be kept in mind. The surtax rates, which are the same as those in the Revenue Act of 1935, and which have not been changed by any of the subsequent acts up to January 1, 1937, are almost as high, on the average, as the peak rates effective in the Revenue Act of 1918. /4/ The normal tax rate, a flat 4 per cent, compares with the peak normal rates of 6 per cent and 12 per cent, in force on 1918 incomes. The average personal exemption was somewhat lower under the 1918 act than under the 1936 act. Moreover, the 1936 act, in contrast with the 1918 act, allows the personal exemptions, including

/2/ See Note 1, Federal Tax Collections, 1913-37, p. 514, and Note 1, State Tax Revenues, 1936, p. 529.

/3/ See the memorandum by Susan Burr and William Vickrey on federal income and estate tax estimates in the forthcoming volume, Studies in Current Tax Problems.

/4/ See p. 36.

[page 71]

the credit for dependents, to be deducted in computing not only the
normal tax but also the surtax.

In the first part of the discussion below, /5/ the changes made by the Revenue Act of 1936 will be disregarded. Those changes, so far as they are important for the present estimates of personal income tax revenue, were (1) the introduction of a tax on undistributed profits of corporations, which tends to force larger dividend payments, and (2) the taxation of dividends by the normal tax of 4 per cent. Both of these factors will be considered later. /6/

Changes in Rates and Exemptions: Prosperous Period

In view of the current trend of business conditions, estimates for a period of prosperity will be made first.

If the 1936 rates and exemptions were applied to incomes for 1928, the peak year of incomes reported in Statistics of Income, the personal income tax would yield $2,900,000,00 -- $400,000,000 from the normal tax and $2,500,000,000 from the surtax. This total, $2,900,000,000, is substantially more than twice as much as the tax has ever yielded. /7/ It could be raised to $3,400,000,000 if the personal exemptions were lowered from $2,500 (married) and $1,000 (single) to $1,000 and $500, and if the credit for each dependent were lowered from $400 to $200. /8/ This figure, $3,400,000,000, is nearly two-thirds of the total amount of $5,500,000,000 expected from all federal taxes for 1937, which in turn is close to the peak of federal tax revenue -- $5,700,000,000 in 1920.

Of course the amount and distribution of 1928 incomes will not be exactly duplicated in any year. In general, however, the data indicate that, if the present personal income tax is left unchanged, a year of extreme prosper-

/5/ See pp. 71-77.

/6/ See pp. 77-80.

/7/ See Note 1, Federal Tax Collections, 1913-37, p. 514.

/8/ See Note 2, Increases in Returns from Changes in Exemptions and Earned Income Credit, p. 573.

[page 72]

ity at somewhat above present price levels will produce between
$2,500,000,000 and $3,000,000,000. This estimate does not take into
account the effect of the Revenue Act of 1936 either in forcing out
added dividends to stockholders by the operation of the undistributed
profits tax, or in the taxation of dividends by the normal tax. In
1936 the personal income tax -- without these new provisions --
produced only $700,000,000. This yield reflected partly 1934 incomes
and partly 1935 incomes.

If an attempt is made to reduce the government debt, new tax revenue may be needed. Estimates have therefore been made in Table 7 of the yield from two schedules of surtax rates that are even higher than those in the 1936 act. These schedules are given in part in Table 6. They are somewhat arbitrary and are not recommendations of the writer. They are presented only because the discussion of possible increases in revenue would otherwise be too vague to be useful.

Surtax Schedule A starts imposing a surtax on $4,000-$5,000 incomes as does the 1936 act, but at a 1 per cent instead of a 4 per cent marginal rate. It increases the percentage, as incomes increase, faster than the present law does up to incomes of $90,000. On income above that amount, Surtax Schedule A and the present law are the same. For example, incomes of $9,000 to $10,000 would pay a 10 per cent marginal rate under Surtax Schedule A, compared with 6 per cent under the present law, and 30 per cent as against 19 per cent on $28,000-$30,000 incomes.

Surtax Schedule B starts as does A, but the increases are steeper than in A, and the maximum rate of the 1936 act -- 75 per cent -- is applied to all incomes above $75,000.

Surtax Schedule A is estimated to yield about $500,000,000 more than the 1936 surtax rates when applied to 1928 incomes. Thus, Surtax Schedule A, together with the normal tax of 4 per cent, would yield about $3,400,-

[page 73]


                               TABLE 6

Income Brackets      Marginal Rate              Average Rate of
                     Applicable to              Surtax on a Sur-
 (In Thousands     Surtaxable Income          taxable Income Equal
  of Dollars)      Within Bracket /b/       to Upper Limit of Bracket
                     (In Per Cent)              /c/ (In Per Cent)

                     Sched- Sched- Sched-       Sched-  Sched-  Sched-
               1936   ule    ule    ule   1936   ule     ule     ule
                Act     A      B      C    Act    A       B      C

0- 1             0     0      0     10       0     0      0     10
4- 5             4     1      1     14       1    /d/    /d/    12
9- 10            6    10     11     19       3     4      4     15
14- 15           9    15     21     29       5     7      8     18
19- 20          13    20     31     39       6    10     13     22
28- 30          19    30     41     49      10    15     21     30
48- 50          27    50     61     64      15    25     33     41
74- 75          47    52     74     75      23    34     45     51
98- 100         55    55     75     75      30    39     52     57
400- 500        68    68     75     75      57    59     70     71
750-1,000       72    72     75     75      64    65     73     73
2,000-5,000     74    74     75     75      72    72     75     75
5,000 and over  75    75     75     75      --    --     --     --

/a/ Only selected portion of the surtax rates are given. For details, see the memorandum by Susan Burr and William Vickrey on federal income and estate tax estimates in the forthcoming volume, Studies in Current Tax Problems.

/b/ These rates are the kind that appear in the law and that are commonly noted in discussions of income tax rates. For example, if a taxpayer has a surtax net income of more than $4,000, the 1936 act taxes the part of the income that lies between $4,000 and $5,000 at a surtax rate of 4 per cent.

/c/ These rates are computed from the marginal rates. They do not appear in the law and are commonly ignored in current discussions, but they are usually more illuminating than the marginal rates. For example, a taxpayer with a surtax net income of $5,000 pays a surtax of approximately 1 per cent of the $5,000, under the 1936 act.

/d/ Less than 1/2 of 1 per cent.

000,000 on 1928 incomes, and, if exemptions and the credit were
lowered to $1,000, $500, and $200, it would yield about

Surtax Schedule B, applied to 1928 incomes, is estimated to yield about $700,000,000 more than Surtax Schedule A -- or $1,200,000,000 more than the 1936 rates -- or, added to the yield of the 4 per cent normal rate, a total of $4,100,000,000. With exemptions lowered to $1,000 and $500, and the credit lowered to $200, the total becomes $4,700,000,000.

The additional revenue to come from lowering exemptions and the credit for dependents and retaining the 1936 rates is, as indicated in Table 7, about the same as

[pages 74 and 75]



[table omitted]

[page 76]

the increase from Surtax Schedule A over the 1936 rates. Under 1928
income conditions, a lowering from $2,500, $1,000, and $400 to
$1,000, $500, and $200, with no change in rates from the 1936 act,
would increase the yield of the personal income tax by $500,000,000
-- from $2,900,000,000 to $3,400,000,000. However, $400,000,000 of
the increase would come from the normal tax -- that is, in large part
from the inclusion of new taxpayers of small income -- whereas the
increases under Surtax Schedules A and B come wholly from surtax, and
almost entirely from those taxpayers with net incomes of more than
$25,000. /9/

Changes in Rates and Exemptions: Other Periods

The 1936 rates and exemptions would produce only $500,000,000 on the incomes received in 1933 -- the lowest taxable-income year of the depression -- compared with $2,900,000,000 on a 1928 base. Surtax Schedule A would yield only $100,000,000 more than the 1936 rates if both were applied to the 1933 base, and Surtax Schedule B would yield only the same increase over Surtax Schedule A. Lowering the exemptions and credits as indicated above, with 1936 rates, would increase the yield about $200,000,000 on the 1933 base. The great pressure on the taxpayer by Surtax Schedule B with the lower exemptions and credit would yield a total of only $900,000,000 on 1933 incomes.

At various stages between the 1928 and 1933 figures come the estimates based on incomes reported for the years 1924, 1931, and 1934, for which data are also given in Table 7. The 1931 figures show increases of $300,000,000 or $400,000,000 over 1933, while the 1924 figures show increases of from $800,000,000 to $1,200,000,000 over 1933 -- with 1936 exemptions and credit for dependents.

/9/ See the memorandum by Susan Burr and William Vickrey on federal income and estate tax estimates in the forthcoming volume, Studies in Current Tax Problems.

[page 77]

Estimates for Extreme Scale

Estimates were also made for a relatively high surtax schedule, called Surtax Schedule C, shown in part in Table 6, which starts with a rate of 10 per cent on the first $1,000 of surtax net income and reaches the maximum 75 per cent level at that part of the income between $70,000 and $72,000.

Obviously, such a schedule would be considered only in the event of an extreme fiscal crisis. Estimates based on it are subject to a particularly wide margin of error because of the unpredictable forces of avoidance and evasion that such rates would produce. Use of the same estimating technique that was employed for other data above gives a yield of $5,300,000,000 on 1928 incomes, of which $400,000,000 is normal tax and $4,900,000,000 surtax. If the low exemptions of $1,000, $500, and $200 are used, the total reaches $7,000,000,000, of which $800,000,000 is normal tax and $6,200,000,000 is surtax.

Estimates of yield under almost any surtax scale can be made with the help of the ready reckoning table given in another volume. /10/

New Taxes in the Revenue Act of 1936

None of the estimates above takes into account the changes made in the Revenue Act of 1936. While this act did not increase the rates of the personal income tax, it tends to increase the yield for the next few years in two ways -- by eliminating an exemption and by transferring into present years some income that normally would not appear until future years. In part, therefore, the act raises genuinely additional revenue, and in part it raises advance revenue -- that is, revenue obtained in the present at the expense of the future.

The genuinely additional revenue comes from the taxation of dividends under the normal tax of 4 per cent.

/10/ Ibid.

[page 78]

Hitherto dividends were subject only to the surtax if they came from
corporations taxable under the corporation income tax. The 1928
income statistics indicate an additional yield of $160,000,000 from
this change; the 1933 statistics, $60,000,000; and the 1934
statistics, likewise, $60,000,000. For present purposes, $100,000,000
may be used as a general estimate. These estimates assume no pressure
from an undistributed profits tax for an increase in dividends.
Details of the computation are presented elsewhere. /11/

The advance revenue -- revenue raised in the present at the expense of the future -- is the revenue that will come from the normal tax and surtax on dividends that would not have been paid out until later, if at all, had it not been for the undistributed profits tax of 1936. If this tax induces larger dividend payments, future years will show smaller dividend payments unless the corporation would have retained the earnings permanently. In the latter event, the stockholders' shares would often have shown a corresponding increase in value. Frequently this increase would eventually have been realized by sale. Thus the larger dividend payments induced by the new tax also tend to result in relatively smaller capital gains and larger capital losses for stockholders. Certain exceptions to this statement, however, might occur. The corporation might lose the money that it would retain were there no undistributed profits tax. Owing to a loophole in the law, the stockholder and his heirs might forever escape taxation if the stockholder did not sell and if no dividends were distributed as long as he or his heirs held the stock. Despite these and other qualifications, the general effect on the personal income tax is to shift the tax base from the future to the present.

The estimate problem is therefore a peculiar one. If the estimate is for the specific near-term years, 1937, 1938,

/11/ Ibid.

[page 79]

etc., the effect is a genuine revenue increase. If the estimate is an
average, long-term figure, little increase should be counted on.
However, the government will certainly gain by the amount of interest
it would have had to pay on borrowings. /12/

If all corporation earnings had been distributed as dividends in the income year 1928, the amount and distribution of the added income to stockholders would have resulted in an additional $1,400,000,000 in personal income tax under the rates of the Revenue Act of 1936. With 1933 income figures, the estimate is only $100,000,000, and with 1934 figures it is $300,000,000. These figures, shown in Table 7, are of course maxima, and are almost surely much too high, since the undistributed profits tax will scarcely force 100 per cent of earnings out in dividends.

The assumption about undistributed dividends also affects the estimate of the yield of the corporation surtax on undistributed profits. The estimate for the latter tax that is given in Table 9 below is a practical maximum (the minimum is zero). It assumes that the new tax has no effect on dividend policy.

The two maxima cannot, of course, be added. To the extent that one tax is productive, the other cannot be. The entire prospect is so uncertain that no summary of the revenue effect of the undistributed profits tax can be made, except that it will add several hundred millions of dollars -- perhaps $500,000,000 in an average year -- in the form either of corporation tax or of personal tax, largely at the expense of the yield of future years.

The rates of the undistributed profits tax are graduated, and in many cases corporations may be expected to follow the policy that is calculated to minimize the total tax. Actual revenues from the undistributed profits tax plus actual revenues from the personal tax on additional

/12/ See note g to Table 7.

[page 80]

dividends will therefore probably be less than the revenue under
either of the assumptions in the present computations. No basis
exists, however, for estimating the size of this difference.

Changes in Treatment of Capital Gains and Losses

In all the estimates given above, the definition of net income is assumed to remain unchanged as fixed by the Revenue Act of 1936. However, the inclusion of capital gains or losses in the tax base has been the subject of much debate. Capital gain or loss is, broadly speaking, gain or loss from the sale of non-business assets and of business assets aside from stock in trade and inventory generally. Should such gain or loss be excluded from the tax base? If it is included, how should it be treated?

Estimates have been made of the revenue results that might be expected from any one of four methods of treating these items. The estimates are for the income years 1928 and 1931 only, and they assume that in all other respects the tax law is that of 1936, except that the new provisions of the 1936 act affecting dividend payments are ignored. /13/ Brief summary results are given in Table 8, and the basic summary is presented elsewhere. /14/

The four bases on which the estimates are made are called: (1) the "percentage plan," (2) the "no-net-loss plan," (3) the "equal- treatment plan," and (4) the "no-tax plan." What each of these plans involves and the estimates of revenue that each would produce may be summarized as follows.

(1) The percentage plan, contained in the act of 1936, taxes a decreasing percentage of the gains according to the length of time the asset has been held. If the period is not more than one year, the full amount of the gain is taken into account in computing net income. If the

/13/ See pp. 77-78, 84-85.

/14/ See the memorandum by Susan Burr and William Vickrey on federal income and estate tax estimates in the forthcoming volume, Studies in Current Tax Problems.

[page 81]

period is more than one year but not more than two years, only 80 per
cent of the gain is taken into account; if more than two years but
not more than five, 60 per cent; if more than five but not more than
ten, 40 per cent; and if more than ten, 30 per cent.

The object of this plan is to avoid taxing a long-accumulated gain at high surtax rates merely because it happens to be realized all in one year instead of bit by bit over a number of years. Capital losses are given the same treatment, but their deduction is allowed only as an offset for capital gains of the same year, plus an extra allowance of $2,000 that is designed for the small taxpayer.

The estimates already presented show that, if the provisions of the percentage plan in the 1936 act were left unchanged, the surtax yield on 1928 incomes would be $2,500,000,000, the normal tax yield $400,000,000, and the total, $2,900,000,000.

(2) Under the no-net-loss plan, all capital gains are taxed at their full amount regardless of the length of time that the capital asset is held. Capital losses are deducted at their full amount, regardless of the length of time that the asset is held, but are allowable only to the extent of the taxpayer's capital gains of the same year. Under this plan the surtax yield on 1928 incomes would increase by $700,000,000 to $3,200,000,000, and the normal tax yield would increase by $100,000,000, giving a total increase of $800,000,000. On 1931 incomes, on the other hand, the total increase, normal tax and surtax, would be only $300,000,000.

(3) Under the equal-treatment plan all capital gains are fully taxable and all capital losses fully deductible regardless of the length of time that the asset is held. On 1928 incomes this plan would yield a total increase of $700,000,000 over the percentage plan. This increase is only slightly smaller than that of the no-net- loss plan.

[page 82]

The important difference between the no-net-loss plan and the equal-
treatment plan shows clearly in 1931 incomes. On this basis the
equal-treatment plan gives $100,000,000 less than the percentage
plan, whereas, as has been seen, the no-net-loss plan gives
$300,000,000 more than the percentage plan. This equal-treatment plan
assumes that capital losses cannot be carried over from one year to

(4) Under the no-tax plan capital gains and capital losses are both entirely ignored. If this were done, the total yield on 1928 incomes would drop sharply from the $2,900,000,000 yield under the percentage plan to $1,600,000,000 -- a decrease of $1,300,000,000. On 1931 incomes, a slight gain of $100,000,000 over the percentage plan is estimated.

In short, the government would probably be a heavy loser in revenue over a period of a decade or so if it substituted a no-tax plan for the percentage plan. It would probably lose even more if it selected a no-tax plan in place of an equal-treatment plan.

By taking the most one-sided of the plans -- the no-net-loss plan -- the government could increase its revenues even in a depression year compared with the percentage plan. In prosperous times the increase would be large. About the same increase over the percentage plan can be gained in a prosperous period under the equal- treatment plan, but in a depression year the government would have to accept a slight loss compared with the percentage plan.

Federal Corporation Income Tax

Graduation of the corporation tax rate, introduced by the Revenue Act of 1935 and continued by the Revenue Act of 1936, complicates the task of estimating the yield. However, the problem is not yet serious since the graduation is slight.

[page 83]

On 1928 incomes, as is indicated in Table 9, the corporation normal tax of 1936 would yield $1,800,000,000 -- $500,000,000 above the peak yield of 1927. The same

                               TABLE 8

                      (In Millions of Dollars)

Income    Percentage     No-Net-Loss    Equal-Treatment    No-Tax
 Year      Plan /b/       Plan /b/          Plan /b/       Plan /b/

 Surtax      2,474          3,198            3,132           1,298
 Normal Tax    432            497              489             310
             _____          _____            _____           _____
Total        2,907 /c/      3,695            3,621           1,608

 Surtax        617            845              553             711
 Normal Tax    192            235              187             213
             ______         _____            _____           _____
Total          808 /c/      1,080              740             924

/a/ The estimates are based on incomes for 1928 and 1931. Other provisions are assumed as in Revenue Act of 1936. For qualifications concerning dividends, see p. 80 and then pp. 77-78, 84-85.

/b/ For definitions of plans, see pp. 80-82.

/c/ Discrepancy of total and sum of items results from evening off to nearest million.

amount, $1,800,000,000, would be produced by the slightly different
scale of rates and provisions in the Revenue Act of 1935. On 1933
incomes the 1936 act normal tax would give only $500,000,000, and on
1934 incomes, $600,000,000. /15/

Within reasonable limits, a certain percentage increase in the average rate of the tax ought to be accompanied by a corresponding percentage increase in the yield. Thus, if the average rate were increased to 20 per cent, the yield would probably increase by about one-third. At present, the rate progresses from 8 per cent on the first $2,000 of a corporation's income to 11 per cent on that part of the income above $2,000 and not above $15,000, 13 per cent on that part above $15,000 and not above $40,000, and 15 per cent on all the income above $40,000.

/15/ Ibid.

[page 84]

About four-fifths of corporate income will, in a prosperous year such
as 1928, be income above the $40,000 level. /16/ Thus, the average
rate is close to 15 per cent.

The surtax on the undistributed net income of corporations, introduced in 1936, poses practically insoluble estimating problems. The figures given here present only a rough indication of the maximum yield that may be expected.

On the assumption that corporations would follow the same dividend practices as in the past, an income year like 1928 would produce average undistributed profits equal to 37 per cent of what the law defines as the "adjusted net income." The highest rate of the undistributed profits tax that would be applicable would therefore be 17 per cent. /17/ The amount of undistributed income would be $4,200,000,000, and the yield of the undistributed profits tax would be $500,000,000. However, if deviations from the average practice were as wide as a recent

                               TABLE 9

                      (In Millions of Dollars)

                                               Income Years
                                           1928     1933     1934 /b/

Normal tax                                1,785      499      622
Surtax on undistributed profits:
 Assuming no deviation from average /c/     546       62      254
 Assuming deviation from average /c/        804       92      375

Total corporation tax:
 Assuming no deviation from average /c/   2,331      561      876
 Assuming deviation from average /c/      2,589      591      997

/a/ Under the Revenue Act of 1936.

/b/ These estimates are subject to a wider range of error than estimates for 1928 and 1933 owing to the lack of data, at the time that they were made, on distribution of net income by income classes and on dividends received and cash dividends paid.

/c/ See pp. 84-85.

/16/ Ibid.

/17/ See p. 514.

[page 85]

sample indicates, the yield would be $800,000,000. /18/ On 1933
incomes the total corresponding to $800,000,000 drops sharply to
$100,000,000, and on 1934 incomes, to $400,000,000.

Actually, the undistributed profits tax will almost surely increase dividend distributions. Hence the direct yield of this tax will be smaller than these estimates.

In summary, the total corporation income tax as it stands in the Revenue Act of 1936 would yield, under the most favorable assumptions used here, $2,600,000,000 in an income year like 1928, and in a depression income year like 1933, $600,000,000.

Federal Estate Tax

Present Rates

In 1936 the federal estate tax produced $200,000,000. For 1937 the official budget estimate of September 2, 1936, put the yield at $300,000,000. On the basis of the returns filed in 1930, which covered chiefly the prosperous period of 1928 and 1929, the yield under the present rates would be $600,000,000, as shown by Table 11. /19/ If, instead of the filing year 1930, the filing year 1926 is used, the estimate is $400,000,000. Lowering the specific exemption from $40,000 to $10,000 would increase the yield only by $100,000,000.

All these estimates are for the tax payable by the taxpayer to the federal government -- that is, the total "tentative tax" levied by the law, minus the credit for any death taxes paid by the taxpayer to state governments up to 80 per cent of the federal tax as computed under the old 1926 provisions. This credit was in the aggregate $100,000,000 in 1926 returns and again in 1930 returns, while in 1933 returns it was only a small fraction of that amount. /20/

/18/ See the memorandum by Susan Burr and William Vickrey on federal income and estate tax estimates in the forthcoming volume, Studies in Current Tax Problems.

/19/ For basic data, see ibid.

/20/ Ibid.

[page 86]

Yields of Hypothetical Rates

Estimates of the estate tax yield are given under two sharply higher hypothetical schedules, called Schedule A and Schedule B. These schedules are shown in part in Table 10, and in detail elsewhere. /21/ As with the hypothetical income tax schedules, their use here for illustrative purposes does not imply that they are recommended. /22/ The rates shown in the schedules are for the "tentative tax," and in general they represent the combined burden of federal and state death taxes. The federal government gets only what is left after deduction of the credit for death taxes paid to state governments.

Schedule A imposes the same marginal rate as at present on estates under $5,000, but increases the rates with increases in estate more sharply than does the present law, reaching the peak of 70 per cent on estates of $5,000,000 and over instead of those of $50,000,000 and over.

Schedule B raises the present rate to 3 per cent on estates under $5,000 and increases the rates much more rapidly than in Schedule A, reaching 72 per cent on estates of $750,000 to $800,000, compared with 2.6 per cent under the present act, and going up to 75 per cent on estates of $5,000,000 and over.

The federal government would get $1,000,000,000 from Schedule A on returns filed in 1930 and $300,000,000 on those filed in 1933. The tax credit would be the same as that given above in the discussion of the 1935 act yields. /23/ Lowering the exemption from $40,000 to $10,000 would produce an added revenue in the neighborhood of $100,000,000 in either case.

The federal government would get $1,600,000,000 from Schedule B on returns filed in 1930, and $1,800,000,000 if

/21/ Ibid.

/22/ See p. 72.

/23/ See p. 85.

[page 87]

the exemption were lowered from $40,000 to $10,000. On returns filed
in 1933 it would get only $500,000,000, and $700,000,000 if the
exemption were lowered.

                              TABLE 10


Estate Brackets
after Deduction   Marginal Rate Applicable to     Average Rate of Tax
  of Specific     Estate within Bracket /b/       on an Estate Equal
  Exemption                                       to Upper Limit of
                                                     Bracket /c/

(In Thousands            (In Per Cent)               (In Per Cent)
  of Dollars)

                  1935 Schedule Schedule   1935  Schedule  Schedule
                   Act    A        B        Act     A         B

Under 5             2     2        3         2      2         3
40-50              10    12       27         6      8        16
90-100             14    26       55        10     13        29
450-500            23    42       68        18     31        57
900-1,000          29    50       72        22     40        64
4,500-5,000        53    66       74        38     56        72
9,000-10,000       65    70       75        50     63        73
50,000 and over    70    70       75        --     --        --

/a/ Only selected portions of the rates are used. The rates given in this table are the "tentative tax" rates. After the tax has been computed from them, the amount of death taxes paid to states is subtracted (within a certain limit) and the remainder only is paid to the federal government. For derails of schedules, see the memorandum by Susan Burr and William Vickrey on federal income and estate tax estimates in the forthcoming volume, Studies in Current Tax Problems.

/b/ These rates are the kind that appear in the law and that are commonly noted in discussions of estate tax rates.

/c/ These rates are computed from the marginal rates. They do not appear in the law and are commonly ignored in current discussions, but they are usually more illuminating than the marginal rates.

                              TABLE 11

                       VARIOUS ASSUMPTIONS /a/
               (Yield Figures in Millions of Dollars)

                 Returns Filed   Returns Filed       Returns Filed
                    in 1933         in 1930             in 1926
             Specific Exemption Specific Exemption Specific Exemption

Rate Schedule $40,000  $10,000   $40,000  $10,000   $40,000  $10,000
1935 Act          180      228       603      685       422      480
Schedule A        308      380       983    1,107       659      748
Schedule B        534      671     1,559    1,789     1,040    1,206

/a/ After deduction of credit for state taxes paid. Net estates are assumed as in returns filed during calendar years 1926, 1930, and 1933. For details, see the memorandum by Susan Burr and William Vickrey on federal income and estate tax estimates in the forthcoming volume, Studies in Current Tax Problems. The net estate, except where the specific exemption is given as $10,000, is as defined in the Revenue Act of 1926 as amended by later acts through 1936.

[page 88]

Other Federal Taxes

No special technique has been devised for estimating the other federal taxes, and no definite estimates are offered. Either the data at hand are too inadequate, as for the gift tax and the payroll tax, or the rate and exemption structure is so simple that no special technique is justified, as for the tobacco tax. However, a few remarks may be made on some of the chief taxes. These are based largely on recent revenue yields, as given in Table 12.

Gift Tax

The large gift tax yield in 1936 was partly the result of a rush to avoid the pending higher federal gift tax and death tax rates. The volume of gifts may drop as more states impose gift taxes. At the end of the year 1936 only three states had such taxes. On the other hand, the cumulative effects of the rate scale /24/ will result in higher rates, and thus higher taxes, for any one donor as the years pass. The net result is an uncertain outlook. A variation of $100,000,000 or so in either direction seems possible over the next few years.

                              TABLE 12

                      (In Millions of Dollars)

    Tax                      1934       1935       1936      1937 /b/

Customs                       315        345        387       402
Internal revenue:
Income tax                    817      1,099     1,413      2,303
    Personal income tax       420        527       674            /c/
    Corporation income tax    398        572       739            /c/
Estate tax                    104        140       219        268
Gift tax                        9         72       160        110
Stamp taxes etc.              294 /d/    243 /d/   252 /d/        /c/
    Telephone, telegraph, etc.
     tax                       19         20        21            /c/
    Capital stock tax          80         92        95            /c/
    Admissions tax             15         15        17            /c/

/24/ See pp. 23, 316.

[page 89]

    Tax                        1934      1935      1936       1937 /b/

Stock transfer tax               38        16        33           /c/
Check tax /e/                    41        26         2        --
  Coconut, etc., oils processed  /f/       24        28           /c/
  Other stamp taxes etc.        100        49        57           /c/
Agricultural adjustment
  taxes /g/                     371       526        72           /h/
Manufacturers' excise taxes     390       342       383      409
  Automotive taxes              299 /d/   267       299 /d/       /c/
   Auto bodies and chassis       38        44        55           /c/
   Parts and accessories          6         6         7           /c/
   Tires and tubes               28        27        32           /c/
   Gasoline                     203       162       177           /c/
   Lubricating oils              25        28        27           /c/
  Electrical energy tax          33        33        34           /c/
  Toilet preparations tax        11        13        13           /c/
  Other manufacturers' excises   46 /i/    29 /i/    37 /i/       /c/
Tobacco taxes                   425 /d/   459       501 /d/   532
   Tax on small cigarettes      350       385        425          /c/
   Tax on manufactured tobacco
      and snuff                  62        61         62          /c/
   Tax on cigars                 12        12         12          /c/
   Other tobacco taxes            2         1          1          /c/
Liquor taxes /j/                259       411        505      589
  Wine taxes                      4         6          9          /c/
  Spirits taxes                  86       189        247          /c/
  Beer etc. taxes               169       216        249          /c/
Tax on unjust enrichment /k/     --        --         --       82
 Taxes on carriers and
    employees /l/                --        --         /m/     135
 Social security taxes /n/       --        --         --      325

  Total federal tax revenue   2,987 /o/ 3,644 /o/  3,907 /o/ 5,477 /o/

/a/ See Note 1, Federal Tax Collections, 1913-37, p. 514.

/b/ As estimated in Statement by the President on the Summation of the 1937 Budget, September 2, 1936. For the estimates in the budget that was submitted January 5, 1937, which show a total of $5,636,000,000 instead of $5,477,000,000, see Note 2, Revenue Estimates in Budget of January 5, 1937, p. 538.

/c/ Data not available in segregated form.

/d/ Discrepancy between total of items and total given is due to evening off.

/e/ Lapsed December 31, 1934.

/f/ Introduced by Revenue Act of 1934.

/g/ Including cotton ginning tax and tobacco sales tax.

/h/ Declared invalid in 1936.

/i/ Obtained by subtraction.

/j/ Includes excise tax but not customs duties on imported liquor.

/k/ Introduced by Revenue Act of 1936.

/l/ Introduced in 1935.

/m/ Less than $500,000.

/n/ Introduced by Social Security Act of 1935.

/o/ The totals of the items given are $2,984,000,000 (1934), $3,637,000,000 (1935), $3,892,000,000 (1936), and $5,155,000,000 (1937). The discrepancies result from some omissions and evening off to the nearest million.

[page 90]

Stamp Taxes Etc.

The taxes lumped under this category are fairly stable in yield, except the stock transfer tax. The total seems unlikely to fluctuate much from the present $250,000,000 level.

The capital stock tax rate is $1.00 per $1,000. A doubling of the rate would add nearly $100,000,000 in revenue, provided the corporations were not given another chance to declare their value. /25/ The admissions tax yield is low because all admissions of 40 cents or less are exempt. If all above 10 cents were taxable, the yield would be increased by $80,000,000, according to Treasury estimates. /26/ A check tax of 2 cents on each bank check or draft would yield about $50,000,000. /27/

Manufacturers' Excise Taxes

The recent decrease shown in the gasoline tax yield reflects a change in rate. The gasoline tax is fairly steady in yield, and, at the existing 1-cent rate, seems likely to stay at approximately $200,000,000 a year. The yields of the other excises are not likely to fluctuate much more in the aggregate than the yield of the gasoline tax.

Probably the gasoline tax could be increased by another cent without decreasing consumption appreciably. Thus another $150,000,000 or $200,000,000 could be obtained. The tax on auto bodies and chassis is 3 per cent of the manufacturers' sales price (2 per cent for trucks). It could probably be doubled without appreciably decreasing the sales volume. The electrical energy tax applies only to domestic and commercial but not industrial consumption. The law forbids it to be passed on to consumers as a separate item. Its rate is 3 per cent, and in 1936 it yielded $34,000,000.

/25/ See pp. 271, 560.

/26/ Revenue Act, 1936: Hearings before the Committee on Ways and Means, House of Representatives, 74 Cong., 2 sess., p. 623. Hereafter cited as Revenue Act, 1936: House Hearings.

/27/ Ibid.

[page 91]

Tobacco Taxes

The tobacco taxes in the aggregate may show an appreciable and steady increase in yield, judging from their behavior in recent years.

The tax on small cigarettes is at the rate of $3.00 a thousand, which amounts to 6 cents per package of 20 cigarettes. Perhaps an increase of a cent or two would not cut consumption materially. Nearly half the states, however, have cigarette taxes, some of them as high as 3 cents a package. The other federal tobacco taxes are at a very much lower rate, and account for only 15 per cent of the tobacco tax revenue.

Liquor Taxes

The liquor tax yield is very uncertain since it depends largely on the extent to which the illegal trade in spirits is checked. Likewise, any estimate of yield under different rates is very hazardous. /28/

Social Security Taxes

The yields of the taxes on carriers and their employees and the taxes on payrolls under the Social Security Act are so uncertain that no attempt is made here to estimate them. /29/


Nothing very specific can be said on the possible yield of state and local taxes under various rates, exemptions, etc., and under various business conditions, for reasons already noted. /30/ However, a few general comments can be made.

Property Taxes

The rates of the property tax are already so high that, politically speaking, an important increase in yield due

/28/ See pp. 197-202.

/29/ See pp. 263-65.

/30/ See p. 70.

[page 92]

to higher rates can hardly be expected during the next few years. Not
only a popular feeling against such increases but also rigid
constitutional over-all rate limitations in some states are obstacles
to larger yields. /31/

Business improvement, with its accompanying increase in realty values, will tend to increase revenues. The improvement is apt to be slow, however -- more probably around 5 per cent a year than 25 per cent. The only probable source of sudden increase in revenue is the payment of huge accumulations of delinquent taxes. Many of these, however, have already been cleared up. Meanwhile, if homestead exemptions spread, /32/ and if more states abandon the property tax for state purposes, /33/ property tax collections, even in a period of improving business, will remain relatively stable around the $4,500,000,000 level.

Income Taxes

The state income taxes, personal and corporate, are yielding about $300,000,000 a year. In contrast with the property tax, this yield is capable of much expansion, since the tax is used by only about two-thirds of the states. As to the possibility or advisability of raising rates or lowering exemptions, much difference of opinion can be expected. The rates are often low. Even in conjunction with the federal income tax, the burden is scarcely severe on incomes, for example, of $10,000. A single taxpayer with no dependents, and a $10,000 income from salary, bears a combined federal and state personal income tax equal to 10 per cent of his income in Idaho, 7 per cent in Minnesota, and 9 per cent in North Carolina. These are a few examples of the heavier bur-

/31/ See the memorandum by Thomas J. Reynolds on homestead exemptions and rate limitations in the forthcoming volume, Studies in Current Tax Problems.

/32/ Ibid.

/33/ See the memorandum by Thomas J. Reynolds on state property tax rates in the forthcoming volume, Studies in Current Tax Problems.

[page 93]

dens. Further examples and details are given elsewhere. /34/
Moreover, the recent spread of state income taxes indicates a
favorable political atmosphere for the tax. /35/ If the burden of the
payroll taxes does not prove too severe, a large part of the
prospective increase in the cost of state government over the next
few years may be met by income taxation.

Other Taxes

The existing $1,000,000,00 of state and local revenue from gasoline and registration taxes scarcely promises much increase in the near future. The high level of present gasoline tax rates and the political strength of the opposition will limit increases. If motorists decide that the present highway system must be thoroughly overhauled to prevent accidents, however, they may be willing to pay part of the bill through increased highway taxes.

Less than half the states and only two large local units are using the sales tax. From the point of view of revenue the most important base is retail sales. The highest rate is 3 per cent -- used by only a few jurisdictions. Consequently, were it not for the unfavorable political reactions that the tax has lately encountered, the $400,000,000 a year now being received might be doubled or trebled.

The other state taxes, except the payroll tax for social security, and in some states, possibly, the poll tax, offer relatively little prospect of much revenue increase in the near future.

/34/ See the memorandum by Carl Shoup, Bernard L. Shimberg, and William Vickrey on combined federal and state income tax burden in the forthcoming volume, Studies in Current Tax Problems. These figures are after giving effect to the provisions concerning deduction of income taxes paid to other jurisdictions in computing net income.

/35/ See p. 45.

[page 94]

Chapter 7



Public expenditures in the United States must play an important part in determining both the kind and the rate of taxes to be imposed. Some estimate of the amount of public expenditures, for the immediate future at least, is essential before any changes in the tax system can be recommended.

The particular jurisdiction incurring these costs must also be considered. Taxes that are well adapted to central administration may be wholly unsuitable for local administration. Since the transfer of revenues from one jurisdiction to another becomes difficult when large amounts are involved, a tax, otherwise inferior, may be preferable because it is adapted to local administration.

In addition, the purposes of the public expenditures should be considered. Some governmental activities may be made largely self- supporting. Others can be best financed, perhaps, through benefit taxes. Still others must be maintained from the general tax revenue.

An attempt is made in this chapter to estimate the costs of government in the United States in the immediate future. /1/ The estimates cover the range within which it is believed that public expenditures will fall between now and 1940. No guess has been made as to the most probable expenditure within this range. Estimating government expenditures for even one year in advance is

/1/ For basic data and more detailed explanation of method, see the memorandum by Mabel Newcomer on estimated federal, state, and local expenditures in the forthcoming volume, Studies in Current Tax Problems.

[page 95]

hazardous, as recent federal budget estimates have demonstrated.
Estimates can be made of upper and lower limits, however, that will
hold, short of war and uncontrolled inflation. Such estimates are
given below. Attention must be called beforehand, however, to the
inadequacy of the data that underlie them.

The best guide to future governmental expenditures is those of the past. Though important changes can take place in a single year, probably only war or uncontrolled inflation could double, or more than double, expenditures within a one-year period. The failure of the federal government in recent years to spend as much as its budgets anticipated suggests that the machinery of government cannot be so completely revolutionized in a short space of time as to expand or contract expenditures to this extent.

Federal Expenditures

Accurate, detailed, and up-to-date records of federal expenditures are available on which estimates can be based. A single act of Congress, however, can influence the total to an important degree. The advance payment of the bonus, alone, increased the 1936 expenditures by nearly one-fourth. In the same way the passage of the Frazier-Lemke Farm Mortgage Bill would have increased the 1937 expenditures over the present official estimate by 40 per cent. It seems reasonably safe to fix the upper limit of federal government expenditures during the period under consideration at not more than double the 1936 level (barring war and uncontrolled inflation) and to fix the lower limit within a narrower range. But the margin is much too broad to make the estimate useful.

Relief expenditures will probably continue to be large for some years, but they may not continue to fall primarily on the federal government. Probably the excessive demands of the Townsend plan will not have to be

[page 96]

met. The veterans, however, having exhausted their bonus payments,
may demand -- and receive -- substantial pensions in another year or
so. Supreme Court decisions may limit governmental activities, and
therefore expenditures, or they may stimulate them. For example,
rural power programs seem to have been encouraged by the Tennessee
Valley Authority decision. The inauguration of a costly housing
program is under consideration, and the Frazier-Lemke bill for farm
mortgage relief would add $3,000,000,000 to the budget at a single
stroke. The threat of war in Europe has already led to unprecedented
army and navy appropriations. It makes future military expenditures
highly uncertain. And the cost of war itself, should it come, is
beyond prediction.

State and Local Expenditures

State and local expenditures are not subject to such wide fluctuations as those of the federal government. They are influenced, of course, by prosperity and depression, but the brunt of the costs of depression -- and also war -- will always fall on the central government. Moreover, the total of state and local expenditures cannot be materially increased by a single act of a single state or local legislative body. Federal legislation, however, sometimes effects important changes in state and local finance. The Social Security Act, for example, will probably add materially to the cost of state and local governments, as well as to that of the federal government. Expenditures of the states and localities may also increase substantially, even with prosperity, if the federal government returns to them the burden of relief.

The absence of complete and accurate data for actual state and local expenditures for any past year and the impossibility of getting any recent data at all present a greater problem than the possibility of wide fluctuations in a single year. The former publications of the Bureau

[page 97]

of the Census, Financial Statistics of States and Financial
Statistics of Cities of 30,000 Population and Over, have been
discontinued. The latest state figures are to be found in the Bureau
of the Census' Financial Statistics of State and Local Governments:
1932. These are sufficiently complete and detailed for the purpose in
hand, but they are not recent enough to reflect the full effect of
the depression. At best, four full fiscal years have passed since the
latest one recorded, and in four states the data are for the year
ending June 30, 1931. For local governments, likewise, no data more
recent than 1932 are available except for cities of 100,000
population and over. For them data can, at this writing, be obtained
for 1934, and, in part, for 1935. The deficiency is more serious for
local governments than for state governments. The inadequacy of the
most recent records available -- those for 1932 -- has made it
impossible to classify expenditures according to function in many
important cases. Without adequate knowledge of the purposes of
expenditures the difficulties of estimating future trends are

Basis of the Estimates

The estimates in this chapter for state expenditures for 1933, 1934, and 1935 are based on the actual financial statements published in such official state reports as could be obtained and on letters from state officials. Data have been obtained for forty-five states for the fiscal years ending in 1933 and 1934, and for thirty-nine states for 1935. The forty-five states were responsible for 94 per cent of all state expenditures in 1932, and the thirty-nine states accounted for 85 per cent of all state expenditures in the same year. In estimating the expenditures for all states, it has been assumed that the trend in the missing states from 1932 to 1935 was the same as that in the states for which data were found. Because the number of states for which the data are missing is small, the

[page 98]

estimate for the total should be fairly reliable. Variations in the
classification of expenditures in the different state reports may,
however, be responsible for some error.

Data are more limited for local government expenditures than for state expenditures. Consequently, the use of past expenditures as a basis of estimates for future expenditures of local units has been abandoned in favor of past revenues, for which fairly extensive data are available. /2/

Definition of Expenditure

Expenditure has been defined to conform to the Census publications' usage. For the federal government, however, the expenditures included differ in three ways. Where agencies such as the Reconstruction Finance Corporation are receiving money from repayment of former loans, the receipts have been deducted from new loans so that only the net figure has been included. Thus, the figures correspond with those reported in the annual reports of the Secretary of the Treasury. For the public service enterprises, also, the figure given is net. Third, the official federal figures include sinking fund payments.

The first two deviations from Census procedure are, for some purposes, of material importance in every year because of the large postal receipts deducted from Post Office expenditures, and in all recent years because of the large receipts from repayments of loans. They are not important, however, for the purpose in hand, since the figure for the amount to be covered by taxes is the same whichever procedure is used. The additional work involved in making the federal, state, and local figures uniform would, consequently, not be justified.

The third deviation is of course important in that it may imply a debt reduction program. Care is therefore

/2/ See Note 1, Estimates of Federal, State, and Local Expenditures, 1936-40, p. 538, for details of the estimate of local expenditures.

[page 99]

taken in the estimates to specify whether this implication is in fact


With these limitations and difficulties in mind the estimates can be approached. Since future trends must be based upon past experience, a clear picture should first be obtained of actual government expenditures -- federal, state, and local -- during recent years.


Federal expenditures have increased fairly steadily for the past twenty-five years, except for a period following the war. The post- war decline, however, spent itself by 1925, leaving the cost of the national government at $3,518,000,000 -- which even then was nearly two and one-half times as great as in 1915. Interest on the debt, payments to veterans, and higher prices were primarily responsible for the higher expenditure level. Increased population and the expansion of non-military activities also played their part. Table 13 gives the figures in detail.

Federal expenditures increased gradually from 1925 to 1930. The total increase in these five years -- from $3,518,000,000 to $3,792,000,000 -- was less than 10 per cent. Thereafter they rose rapidly with depression demands, nearly doubling in the five years from 1930 to 1935, when the total -- compiled, as noted, in accordance with Treasury technique -- was $7,272,000,000.


State expenditures, also included in Table 13, show somewhat different trends. They did not reach the high war peak of federal expenditures, and they also failed to show any post-war decline. Moreover, where federal expenditures in 1930 were only four times the 1910 level of $918,000,000, state expenditures rose in the same twenty

[page 100]

years from $323,000,000 to a level nearly seven times as high --
$2,143,000,000. /3/ Better highways and higher state

                              TABLE 13
               LOCAL REVENUES, 1920, 1925, 1930-35 /a/
                      (In Millions of Dollars)

               1920  1925  1930   1931   1932   1933   1934    1935

                         From all sources /b/

Federal /c/   6,041  3,518  3,792  3,905  4,618  4,881  7,064  7,272
State /d/       861  1,443  2,143  2,276  2,237  1,994  2,124  2,222
Local /e/     3,344  5,584  7,217  7,074  6,438  5,181  5,386  5,459
   Total     10,246 10,545 13,152 13,255 13,293 12,056 14,574 14,953

                      From taxes and loans /f/

Federal       5,453  3,318  3,548  3,651  4,556  4,731  6,947  7,135
State           702  1,209  1,830  1,973  1,920  1,677  1,807  1,905
Local         2,444  4,084  5,571  5,576  5,129  4,026  4,244  4,309
   Total      8,599  8,611 10,949 11,200 11,605 10,434 12,998 13,349


/a/ Expenditures from grants-in-aid have been credited to jurisdiction making grant. Expenditures from state-administered taxes that are distributed to local governments have been credited to local governments. State and local expenditures are governmental cost payments as defined in Statistics of State and Local Governments. Federal expenditures differ from these in including sinking-fund payments, excluding expenditures of government industries (such as the Post Office and the Panama Canal) covered by receipts from these industries, and including the net figure of government loaning units such as the Reconstruction Finance Corporation.

/b/ To avoid double counting, expenditures from money received as grants-in-aid have been omitted, and the grants-in-aid themselves have been treated as expenditures of the jurisdictions making the grants.

/c/ Obtained from annual reports of the Secretary of the Treasury. These figures differ from the total cash expenditures given in Table 2 of these reports, partly owing to the deduction of receipts from industries other than the Post Office, partly because in classifying expenditures it was necessary to substitute "checks issued" for cash in a few instances. The figures include sinking fund payments.

/d/ Figure for 1920, from M. Newcomer, "Financial Statistics of Public Education in the United States, 1910-1920," Educational Finance Inquiry, VI (1924); figure for 1925, 1930, and 1931 from Department of Commerce, Bureau of the Census, Financial Statistics of States: 1925, 1930, and 1931; for 1932 from Financial Statistics of State and Local Governments: 1932; for 1933, 1934, at 1935 estimated as explained in text. The statement in the text referring to 1915 is based on Newcomer, op. cit.

/e/ Figure for 1920 from Newcomer, op. cit.; for 1925 from National Industrial Conference Board, Cost of Government, 1923-1934 (1934); for 1930 estimated by applying the percentage increases for cities with 30,000 population and over to total expenditure reported by National Industrial Conference Board for 1929; for 1931 estimated by finding ratio of 1931 to 1932 expenditures for cities with 100,000 population and over, and applying this to 1932 expenditures; for 1932 from Financial Statistics of State and Local Governments: 1932; for 1933, 1934, and 1935 estimated as explained in text.

/f/ Obtained by subtracting non-tax revenues from net expenditures above.


standards for schools were the chief causes of this increase, but the
states were also extending their activi-

/3/ Federal expenditures figure for 1910 is from Report of the Secretary of Treasury; state expenditures figure for 1910 is from M. Newcomer, "Financial Statistics of Public Education in the United States, 1910-1920," Educational Finance Inquiry, VI (1924).

[page 101]

ties in the fields of health, welfare, and police protection. Growing
population and rising prices were other factors.

With depression, state expenditures at first increased: from $2,143,000,000 in 1930 to $2,276,000,000 in 1931. They declined between 1931 and 1933 to $1,994,000,000, and then increased again. The 1935 expenditures reached approximately the 1932 level, with a total of $2,222,000,000.

The pressure of declining tax yields was not seriously felt until 1932, but beginning in that year it was undoubtedly an important restraining influence. The rise in 1934 and 1935 is the reflection not so much of improved business conditions, as of the failure of local governments to maintain their usual share in government activities and the consequent necessity for greater state aid. The development of new taxes, especially the sales and liquor taxes, made the 1933-35 expansion possible.


Local expenditures, given in Table 13 also, show much the same trend as those of the states. As with the federal and state governments, growing population and rising prices played a part in increasing costs. The increase from the 1910 figure of $1,593,000,000 /4/ to a total of $7,217,000,000 in 1930, however, was largely the result of a growing demand for governmental services, particularly roads and schools. State expenditures for these purposes supplemented rather than replaced local expenditures. In fact, state aid was designed, for the most part, to stimulate local expenditures. Increases in local expenditures were not so great, relatively, as increases in state expenditures, but they exceeded federal increases.

With depression, local expenditures declined promptly -- from $7,217,000,000 in 1930 to $5,181,000,000 in 1933. Circumscribed by tax and debt limits, which have been

/4/ The 1910 data are from Newcomer, op. cit.

[page 102]

increased rather than relaxed in recent years, local governments have
had no choice but to cut expenditures. Some increase is apparent in
1934 and 1935 -- up to $5,459,000,000; but local expenditures in the
latter year were still far below their 1930 level of $7,217,000,000.
Very little of the increases in 1934 and 1935 was met from locally
administered taxes.

The marked decline in local expenditures that had to be met from local revenues was inevitable. Local governments have only one really important source of revenue, the property tax. /5/ But this tax had been developed before the depression to its political, if not its economic, limit by most local governments. With depression, both the economic and the political limits of this tax were greatly contracted. Even when assessed valuations were maintained at their prosperity level, it was not possible to collect the same sums from real estate owners confronted with shrinking assets and no cash income. Moreover, the normal irritation of the taxpayer was magnified by his financial difficulties. This irritation was expressed in new legislation prescribing narrower tax limits, in political pressure on local officials to reduce taxes, and in tax delinquencies when levies were not actually reduced. The result was a decline in property tax levies and an even greater decline in tax collections.

Local expenditures could be maintained under these conditions only through borrowing or aid from federal and state authorities. But borrowing likewise had its limitations. Debt limits were rigid and widespread, and many municipalities had exhausted their borrowing capacity under these limits during more prosperous years. Moreover, the limits were stated as percentages of the tax base. Such municipalities as had a substantial margin under prosperity assessments found this margin shrink-

/5/ The property tax provided 93 per cent of local tax revenues in 1932. Most of the remaining 7 per cent came from state taxes distributed to local governments.

[page 103]

ing, with no change in their indebtedness. Municipalities with no
margin, and debts falling due, had no recourse but to pay these debts
from current income. Hence the astonishing spectacle of a reduction
in municipal debts in the midst of depression, in 1933. Some
increase in borrowing took place in other years, but in no year since
the depression began was the increase very great. The total net
increase in local bonded debt in the five years from the end of 1930
until the end of 1935 was less than the average annual increase in
the preceding five years.

The only other important source that was available for localities was state and federal aid. This aid has been given liberally, and it accounts for a large part of the growth of federal and state expenditures. The increases in aid, together with the small increases from new loans, were not sufficient, however, to offset the losses from local sources until 1934, and with a rapidly growing relief burden drastic cuts in the other local expenditures were called for.

The possibility of reduction was greatest in capital outlays. Local expenditures for capital outlays are estimated to have been about $2,000,000,000 in 1930. Barring a work relief program, most of such expenditures could be deferred. Capital outlays had shrunk to less than $1,500,000,000 in 1932, and, if trends elsewhere are like those in large cities, to $800,000,000 in 1934.

Other evidence of the reduction of local expenditures is at hand. One of the heaviest local costs is for teachers' salaries. Teachers' salary scales were drastically cut, and for the most part they have not yet been restored. The largest single decrease in current expenses in cities of over 100,000 population between 1930 and 1933 was in school costs. Another factor of some importance in reducing local costs has been the centralization of administration of certain governmental functions. The outstanding, but

[page 104]

not the only, example is the transfer of schools and roads in North
Carolina from local to state hands. State expenditures for these
formerly local functions are not reflected in state aid figures since
the expenditure is direct.

Against these decreases the rising burden of relief may be set. In 1934, 82 per cent and in 1935, 86 per cent of this burden, however, fell on state and federal funds.

In the light of all these factors it is not surprising that local expenditures falling on local revenues (including state- administered taxes shared with local governments) were reduced from $7,200,000,000 in 1930 to $5,200,000,000 in 1933. They were still below $5,500,000,000 in 1935.


With this background of federal, state, and local expenditures from 1920 to 1935, the stage is set for estimating the outlays for the five years from 1936 to 1940. Two kinds of estimate have been made. The first is based on a detailed study of the kind and purpose of the expenditures of the immediate past and an adjustment of them for the future in the light of changing economic conditions, new legislation, and any other events that seem likely to occur. Since the estimates have been made for each function of government separately in most cases, they are called functional estimates.

The second is based on a projection into the future of past trends of expenditures in two periods somewhat similar to what may be expected in 1936 to 1940. These are called trend estimates.

Functional Estimates

The first estimates -- the functional estimates -- have been made on each of two assumptions: (1) that the federal government will withdraw completely from direct relief in 1939 and (2) that it will continue relief con-

[page 105]

tributions. In the first case it has been assumed that the federal
relief load diminishes very rapidly: $1,800,000,000 in 1937 (the
minimum in the budget for 1937 as of September 2, 1936),
$100,000,000 in 1938, and nothing in 1939. In the second case it is
assumed that the federal relief load is gradually diminished:
$2,300,000,000 in 1937 (the maximum set by President Roosevelt in his
revised budget of September 2), $1,800,000,000 in 1938,
$1,300,000,000 in 1939, and $800,000,000 in 1940. /6/ In either case
it is assumed that state and local relief expenditures remain
constant at $500,000,000 a year. The total relief expenditure
estimated for 1940 under the second assumption would provide public
support for approximately 4,000,000 families at the 1935 level of
relief allotments ($28 per month).

The two estimates (in Tables 14 and 16) are given because of the great uncertainty of the future relief burden. The future cost of some of the other functions is likewise very uncertain, but either the amount is too small to justify another estimate or it is too uncertain to warrant more than passing mention. The effect of an adverse Court decision in the case of social security expenditures can hardly be estimated since it cannot be determined when such a decision might be expected, nor can it be guessed when, or in what form, a substitute might be adopted. The effect of uncontrolled inflation or war is even less predictable.

In the estimates in this chapter, all totals of federal expenditures and of federal, state, and local expenditures include statutory sinking fund payments by the federal government. Social security expenditures for old age and unemployment compensation include the estimated cost of the non-contributory old-age pensions plus an amount equal to the yield of the federal and state payroll taxes

/6/ See Note 2, Federal Budget of January 5, 1937, p. 540, for later estimates of relief expenditures for 1937 and 1938.

[page 106]

for unemployment insurance, but do not include payments to the old-
age reserve fund. The summary tables 14 and 16 show three sets of
totals: expenditures from all sources, expenditures from taxes and
loans, and expenditures from taxes. For purposes of future tax
policy the second set is probably the more important. It shows the
amount that has to be raised either by tax revenue or by borrowing.
The differences between the totals in this set and in the first set
represent non-tax revenue such as fees and interest received. The
third set, expenditures from taxes, is chiefly of interest as a
historical record. It shows, for 1930, 1935, 1936, and 1937 the
amounts that were in fact covered by taxation. /7/ For 1938, 1939,
and 1940 it shows only a small difference from the second set -- a
difference that represents an assumed modest amount of borrowing by
state and local governments.

A full account of the assumptions and the technique employed is given elsewhere. /8/

Result of Assuming Federal Relief Ceases

Assuming that federal relief payments will rapidly decline and cease altogether in 1939, it is estimated that total expenditures of all three classes of government would be about $17,399,000,000 in 1936; that they would fall off only slightly in 1937 and then drop to $15,956,000,000 in 1938, recovering to $16,610,000,000 in 1939 and $17,350,000,000 in 1940. Federal expenditures were $8,880,000,000 in 1936, and are estimated to fall to $7,737,000,000 in 1937 and to show a much sharper drop to $5,307,000,000 in 1938, remaining at about that level in 1939 and 1940. State expenditures would increase each

/7/ All 1937 data are estimates. The 1937 state and local tax totals differ slightly from the data in chap. 2 because of somewhat different basis of estimate. The chap. 2 totals are minima, and in fact will probably be somewhat exceeded. See p. 9 and Note 1, State Tax Revenues, 1936, p. 529.

/8/ See pp. 121-22 and the memorandum by Mabel Newcomer on government expenditures in the forthcoming volume, Studies in Current Tax Problems.

[page 107]

year, from $2,639,000,000 in 1936 to $4,698,000,000 in 1940, with the
largest increases in 1937 and 1938. Local expenditures would also
increase every year but with less variation -- about $300,000,000 a
year after 1937. The figures are given in Table 14, which also shows
the totals of expenditures to be met from taxes and loans and from
taxes only. Table 15 gives a detailed estimate of federal
expenditures for each future year compared with actual outlays for
1930, 1935, and 1936.

Assuming Continuance of Relief

On the assumption that federal relief contributions will decline only gradually, the estimates show considerably larger figures for federal expenditures: $8,237,000,000 for 1937, $7,007,000,000 for 1938, $6,584,000,000 for 1939, and $6,152,000,000 for 1940. /9/ State and local expenditures would be the same as on the first assump-

                              TABLE 14

        TOTAL EXPENDITURES, ACTUAL (1930, 1935, 1936) /a/ AND

                      (In Millions of Dollars)

          1930      1935    1936     1937    1938    1939   1940
                                From all sources

Federal   3,792     7,272   8,880   7,737   5,307   5,284   5,352
State     2,143     2,222   2,639   3,152   4,009   4,356   4,698
Local     7,217     5,459   5,880   6,310   6,640   6,970   7,300
  Total  13,152    14,953  17,399  17,199  15,956  16,610  17,350

                                From taxes and loans

Federal   3,548     7,135   8,693   7,598   5,217   5,204   5,282
State     1,830     1,905   2,319   2,812   3,649   3,976   4,298
Local     5,571     4,309   4,580   4,910   5,140   5,370   5,600
  Total  10,949    13,349  15,592  15,320  14,006  14,550  15,180

                                   From taxes
Federal   3,548     3,469   3,929   5,501   5,217   5,204   5,282
State     1,641     1,748   2,179   2,672   3,509   3,836   4,158
Local     4,756     4,152   4,380   4,660   4,840   5,020   5,200
  Total   9,945     9,369  10,488  12,833  13,566  14,060  14,640

     /a/ For state and local governments, 1936 data are estimated,
not actual.

/9/ Total federal expenditures are estimated at $8,256,000,000 for 1937 and at $7,155,000,000 for 1938 in the federal budget of January 5, 1937. These figures include estimated additional expenditure for relief, but exclude payments to the old-age reserve account. For further details of differences between the estimates for 1938 given here and those given in this recent budget statement, see Note 2, Federal Budget of January 5, 1937, p. 540.

[page 108-109]


(In Millions of Dollars)


[page 111]

tion. Total government expenditures would, therefore, be on a higher
level each year, rising from $17,699,000,000 in 1937 to
$18,150,000,000 in 1940. Table 16 gives the figures in detail.

                              TABLE 16
        TOTAL EXPENDITURES, ACTUAL (1930, 1935, 1936) /a/ AND
                      (In Millions of Dollars)

               1930    1935    1936    1937    1938    1939    1940

                                From all sources

Federal       3,792   7,272   8,880   8,237   7,007   6,584   6,152
State         2,143   2,222   2,639   3,152   4,009   4,356   4,698
Local         7,217   5,459   5,880   6,310   6,640   6,970   7,300
  Total      13,152  14,953  17,399  17,699  17,656  17,910  18,150

                               From taxes and loans

Federal       3,548   7,135   8,693   8,098   6,917   6,504   6,082
State         1,830   1,905   2,319   2,812   3,649   3,976   4,298
Local         5,571   4,309   4,580   4,910   5,140   5,370   5,600
  Total      10,949  13,349  15,592  15,820  15,706  15,850  15,980

                                   From taxes

Federal       3,548   3,469   3,929   5,501   6,917   6,504   6,082
State         1,641   1,748   2,179   2,672   3,509   3,836   4,158
Local         4,756   4,152   4,380   4,660   4,840   5,020   5,200
  Total       9,945   9,369  10,488  12,833  15,266  15,360  15,440


    /a/ For state and local governments, 1936 data are estimated, not

                           END OF FOOTNOTE

Possible Variations

It should be repeated that the many assumptions on which the estimates shown in Tables 14 and 16 and Chart 4 are based are not beyond question. Other plausible assumptions will bring very different results. /10/

If it is assumed, for instance, that the social security program will be abandoned, and that the unemployed allowed for in the Table 14 estimate will be supported by home relief rather than work relief, the total tax requirements will be reduced from $14,600,000,000 in 1940 to $12,600,000,000. If it is assumed, further, that the federal government, as well as state and local governments, will continue to increase debts, so that at least

/10/ See the memorandum by Mabel Newcomer on government expenditures in the forthcoming volume, Studies in Current Tax Problems.

[page 112]


ESTIMATES, 1930, 1935-40

[chart omitted]

[page 113]

half the cost of the capital outlays estimated for all governments
will be met from borrowed funds, the immediate charge on taxes will
fall to $11,700,000,000.

On the other hand, if the unemployed are reduced by only one- half in the next four years and work relief at present standards is provided for all, the tax burden will increase by a little over $4,000,000,000 from the Table 14 estimates, to $18,700,000,000 by 1940. If, in addition, state and local credit should fail, or these governments should choose for some other reason to adopt a pay-as- you-go policy, so that no net increase in their debts occurred, another $600,000,000 would be added to taxes. Expenditures from all sources would then amount to about $22,000,000,000.

All the increases assumed are the result of increased governmental services. If, in addition, price levels should increase materially during the next four years, expenditures would also rise above the estimates given. This rise, however, would bring with it higher taxable incomes so that the proportion of the national income paid in taxes would not necessarily increase in consequence.

Other possibilities should also be taken into account. An expensive housing program is under consideration. The Wagner bill proposed an expenditure of $250,000,000 per year, of which only $150,000,000 would ultimately be repaid. This or a similar bill may be considered in the near future by Congress. A farm mortgage bill comparable with the defeated Frazier-Lemke bill might be passed. The veterans might demand another bonus, or pensions. And the possibilities of war and uncontrolled inflation hang over all human affairs.

One thing is certain: government expenditures and taxes will not drop to pre-depression levels. Just as expenditures after the World War failed to return to their pre-war level, so expenditures following depression will doubtless fail to return to their former level. There is the

[page 114]

same aftermath of debts. The residual relief burden is not unlike the
earlier problem of compensation for veterans. And inflation may still
come. Under these conditions the estimate of expenditures in 1940
that is approximately one-third greater than expenditures in 1930
(Table 14) is probably a conservative one. The increase in 1925 over
1915 was more than 150 per cent. /11/ The national debt has not
increased so greatly, it is true, during depression as during the
war; but the dollar has been devalued, as it was not during the war.
If the change in the gold value of the dollar is fully reflected in
the price level by 1940, the cost of the 1930 standard of government
will be greater than the costs given in the maximum estimate.

Trend Estimates

Of the second group of estimates -- those based upon past trends -- the first one (projecting the four-year 1921-25 federal period and the five-year 1925-30 state and local period) shows much lower figures than either of the two groups of functional estimates just considered. The figures of the other trend estimates (based on the four-year 1926-30 federal and five-year 1920-25 state and local periods) are very much the highest of all. The former trend estimates show total expenditures declining slightly each year, from $17,049,000,000 in 1937 to $16,908,000,000 in 1940. The latter show them rising sharply each year, from $18,760,000,000 in 1937 to $22,263,000,000 in 1940. Tables 17 and 18 give the figures for each of these estimates in detail.

First Estimate

The four-year period 1921-25 was one of declining expenditures for the federal government. There are certain similarities in conditions now and in the early '20's that might result in a similar trend of federal expenditures in

/11/ The 1915 figure is from Newcomer, op. cit.

[page 115]

the near future. At both times federal expenditures had risen very
rapidly in the preceding period to meet a national emergency, and the
government was faced with an unprecedented debt. Also, in both
periods important reductions in expenditures were promised by the
federal administration.

If the rate of decrease used for the 1936-40 estimate were the rate of the first four years of decline after the 1919 peak (to correspond to the estimated four-year decline from the 1936 peak), the estimated federal expenditures for 1940 would be much lower. In fact, expenditures would be lower than they have been at any time since the war. Such a rate of decrease cannot be expected, however, in view of (1) the debt, which is larger than at any previous time in the history of our national government, and (2) the fact that the expenditures for the 1936 peak are less than half those for the 1919 peak. Furthermore, prices in the earlier period were declining from

                              TABLE 17
                 (1937-40) ON FIRST TREND BASIS /b/

                      (In Millions of Dollars)

             1935      1936      1937      1938      1939      1940

                               From all sources

Federal     7,272     8,880     8,299     7,717     7,135     6,553
State       2,222     2,437     2,653     2,868     3,084     3,300
Local       5,459     5,778     6,097     6,416     6,735     7,055
  Total    14,953    17,095    17,049    17,001    16,954    16,908

                             From taxes and loans

Federal     7,135     8,693     8,124     7,555     6,985     6,415
State       1,905     2,090     2,275     2,460     2,645     2,829
Local       4,309     4,561     4,813     5,065     5,316     5,567
  Total    13,349    15,344    15,212    15,080    14,946    14,811


    /a/ For state and local governments, 1936 data are estimated, not

     /b/ Rate of decrease for federal expenditures 1936-40 same as
rate of decrease 1921-25 (1925 expenditure 73.8 per cent of 1921
expenditure); equal annual decreases assumed for intervening period.
Rate of increase for state and local expenditures 1935-40 same as
rate of increase 1925-30 (1930 state expenditure 148.5 per cent of
1925 expenditure, 1930 local expenditure 129.2 per cent of 1925
expenditure); equal annual increases assumed for intervening period.
Expenditures from taxes and loans have been assumed to increase or
decrease at the same rate as expenditures from all sources.

                          END OF FOOTNOTES

[page 116]

their war peak, whereas prices in the next four years are more likely
to rise than to fall.

The trend for state and local expenditures used in the first estimate -- that for 1925-30, in Table 17 -- shows an increase that was lower than for any five-year period between 1910 and 1930. A smaller increase than this can hardly be expected. The probability of rising prices and the accumulation of needs during the lean depression years tend to offset the effect of a reduced local tax base and the tendency to centralize the administration of some governmental functions. The result may well be exceptional increases in state and local expenditures.

Second Estimate

The trend for federal expenditures used in the second estimate -- that for 1926-30, in Table 18 -- shows a period of slowly rising expenditures. No such increase is expected at present. The year 1936 is regarded as a maximum for some time to come. An increase is not impossible, however, in view of (1) the growing demands on the federal government for the development of its welfare activities and (2) the devaluation of the dollar.

The trend for state and local expenditures used in the second estimate -- that for 1920-25, in Table 18 -- reflects a period of unusually rapid growth, stimulated, doubtless, by both the growing tax bases and the curb on state and local expenditure during the war period. The years from 1935-40 may well resemble those of this trend period because they follow a period when expenditures have been held in check. The fact that expenditures actually declined in the years preceding 1935 whereas they had increased in the years preceding the 1920-25 period argues for an even greater increase now than before.

Moreover, the possibility that the federal government will return the relief burden to state and local governments increases the chance of rapidly rising state and

[page 117]

local expenditures. On the other hand, tax bases -- especially those
of the local governments -- have been impaired in recent years,
whereas they had expanded just before the 1920-25 period. Rapid
recovery in the next

                              TABLE 18
                 (1937-40) ON SECOND TREND BASIS /b/

                      (In Millions of Dollars)

              1935      1936      1937      1938      1939      1940

                                From all sources

Federal      7,272     8,880     9,015     9,151     9,286     9,422
State        2,222     2,522     2,822     3,122     3,423     3,724
Local        5,459     6,191     6,923     7,655     8,386     9,117
  Total     14,953    17,593    18,760    19,928    21,095    22,263

                              From taxes and loans

Federal      7,135     8,693     8,825     8,958     9,090     9,223
State        1,905     2,162     2,420     2,677     2,935     3,193
Local        4,309     4,886     5,463     6,040     6,618     7,196
 Total      13,349    15,741    16,708    17,675    18,643    19,612


     /a/ For state and local governments, 1936 data are estimated,
not actual.

     /b/ Rate of increase for federal expenditures 1936-40 same as
rate of increase 1926-30 (1930 expenditure 106.1 per cent of 1926
expenditure); equal annual increases assumed for intervening period.
Rate of increase for state and local expenditures 1935-40 same as
rate of increase 1920-25 (1925 state expenditure 167.6 per cent of
1920 expenditure, 1925 local expenditure 167.0 per cent of 1920
expenditure); equal annual increases assumed for intervening period.
Expenditures from taxes and loans have been assumed to increase at
the same rate as expenditures from all sources.

                          END OF FOOTNOTES

[page 117]

three or four years may offer state governments the same possibility
of expanding revenues that they had in the early '20's, but the local
governments can hardly expect an equal opportunity. Consequently,
while state expenditures may rise even faster than in the years 1920-
25, it seems very doubtful whether local expenditures will rise as

Comparison of Estimates

A comparison of all four of the estimates is now in order. Table 19 gives the details both in absolute figures (for total government expenses) and in indices of increase over 1935. Chart 5 shows the relationships in graphic form.

The minimum estimate for 1940 -- the first of those

[page 118]

based on trends (Table 17) -- shows a decrease below the 1936 level
but not below the 1935 level. The probability of expenditures
returning to pre-depression levels is so slight, however, that it has
not been given serious consideration. The maximum estimate for 1940
-- the second of the trend projections (Table 18) -- is far above
even the 1936 level, and it shows no decrease in federal, state, or
local expenditures in any year. The first functional estimate (Table
14) lies between the two trend estimates in 1940, although it is
lower than either of the others in 1938 and 1939. The second
functional estimate (Table 16) lies between the two trend estimates

When the individual estimates for federal, state, and

                              TABLE 19
                   LOCAL EXPENDITURES, 1936-40 /a/

Estimate      1935      1936      1937      1938      1939      1940

                               ABSOLUTE AMOUNTS

                           (In Millions  of Dollars)

                                From all sources

   1        14,953    17,399    17,199    15,956    16,610    17,350
   2        14,953    17,399    17,699    17,656    17,910    18,150
   3        14,953    17,095    17,049    17,001    16,954    16,908
   4        14,953    17,593    18,760    19,928    21,095    22,263

                              From taxes and loans

   1        13,349    15,592    15,320    14,006    14,550    15,180
   2        13,349    15,592    13,820    15,706    15,850    15,980
   3        13,349    15,344    15,212    15,080    14,946    14,811
   4        13,349    15,741    16,708    17,675    18,643    19,612

                                  INDEX OF CHANGE

                                  From all sources

   1           100       116       115       107       111       116
   2           100       116       118       118       120       121
   3           100       114       114       114       113       113
   4           100       118       125       133       141       149

                               From taxes and loans

   1           100       117       115       105       109       114
   2           100       117       119       118       119       120
   3           100       115       114       113       112       111
   4           100       118       125       132       140       147

                          FOOTNOTE TO TABLE

     /a/ Estimate 1 is that in Table 14; estimate 2 is that in Table
16; estimate 3 is that in Table 17; estimate 4 is that in Table 18.

                           END OF FOOTNOTE

[page 119]



[chart omitted]

[page 120]

local governments for 1940 are compared, it is apparent that both of
the functional estimates are lower than either of the trend estimates
for the federal government, and higher than either of the others for
the state governments. Only the local government functional estimates
fall between the two trend estimates. Either functional estimate is
believed to be more reliable than either trend estimate.

[page 121]

Chapter 8


The estimates of revenue and expenditure given in Chapters 6 and 7 may now be summarized and compared. All years are fiscal years ending June 30 of the calendar year stated, except that some of the state and local data are for fiscal years ending in some other month but during the same calendar year.


Total expenditures -- federal, state, and local -- will fall somewhere between $17,000,000,000 and $18,000,000,000 in 1937. This total will probably fluctuate between $16,000,000,000 and $18,000,000,000 through 1940. It includes about $500,000,000 a year of statutory debt retirement by the federal government but no debt retirement by state or local governments. It includes actual payments to beneficiaries under the existing social security program. It does not include the amounts put into the old-age annuity fund by the federal government; this matter is discussed in the revenue summary below.

Non-tax revenues will reduce the total to be met by taxes, particularly for localities. If an optimistic view of the relief situation is taken, the amount of tax revenue required to cover all government expenditures without borrowing is $15,300,000,000 for 1937, $14,000,000,000 for 1938, $14,500,000,000 for 1939, and $15,200,000,000 for 1940.

Of these totals, the part to be covered by federal tax revenue, without borrowing, is estimated at $7,600,-

[page 122]

000,000 in 1937, $5,200,000,000 in 1938 and 1939, and $5,300,000,000
in 1940. The figure for state tax requirements mounts steadily from
$2,800,000,000 in 1937 to $4,300,000,000 in 1940. The local
expenditures to be met from taxes, still assuming no borrowing,
increase slowly from $4,900,000,000 in 1937 to $5,600,000,000 in
1940. Of course it is evident that for 1937, actual tax receipts will
not be sufficient, at least for the federal government, to obviate
borrowing. Moreover, it is not likely that states and localities will
abstain entirely from borrowing.

These assumptions are of course highly tentative and subject to a wide margin of error. They anticipate a continuation of business recovery, so that in 1939 and 1940 the federal government will be spending nothing on relief, and the states and localities, $500,000,000 a year. They assume no substantial rise in the price level, no increase in veterans' pensions, and no increase in the present housing or mortgage refinancing programs. On the other hand, they do allow for many kinds of increase that seem highly probable, such as that due to payments to beneficiaries of the social security program.

If the relief load should decline more gradually and the federal government should support a large share of it even in 1939 and 1940, the total annual amount required from taxation to cover expenditures would increase to nearly $16,000,000,000 in each of the years 1938- 40. All the increase is in the federal figure, which rises to $8,100,000,000 in 1937, $6,900,000,000 in 1938, $6,500,000,000 in 1939, and $6,100,000,000 in 1940. /1/


All the local units together raise almost as much tax revenue as the federal government. Until a few years ago, they raised more. The state governments raise in all

/1/ See p. 111; and the memorandum by Mabel Newcomer on government expenditures in the forthcoming volume, Studies in Current Tax Problems.

[page 123]

considerably less than either the federal government or the local
governments. Their total, however, is growing rapidly.

The federal government will obtain about $5,500,000,000 in tax revenue during the year ending June 30, 1937. This will leave a gap of $2,600,000,000 (second functional estimate). The total of state taxes during the same period will apparently be close to $2,500,000,000, and possibly as high as $2,700,000,000. Therefore, state budgets in the aggregate may be within $100,000,000 of being balanced. An estimate of total local tax collections is the most hazardous of all. Data on actual tax collections of all local units in the United States have only occasionally been assembled. A reasonable estimate for the same twelve-month period seems to be about $4,500,000,000, but the total may be as high as $4,700,000,000, which would leave the localities with about $200,000,000 to borrow. The state and local tax total of $7,000,000,000 may be compared with totals of $6,400,000,000 given by the Census Bureau for 1932 and $5,900,000,000 estimated for 1934 by the National Industrial Conference Board. The estimated grand total of federal, state, and local taxes, $12,500,000,000 at a minimum, is considerably higher than for any previous year in the country's history, including the war years. /2/

Estimates for a Year of Great Prosperity

The revenue estimates for 1938-40 differ, as has been shown, rather widely according to the different assumptions that have been made. The variations, however, have been computed only for the federal tax system. For the state totals, either the information is not available or the task of computation is too laborious. For local governments, the probable yield of the property tax, under conditions of moderate prosperity, has been esti-

/2/ See Note 1, Estimates of Tax Revenues, p. 540.

[page 124]

mated for the years 1938, 1939, and 1940 as a basis for the local
expenditure estimates.

The existing federal tax system, excluding the social security taxes, could yield $8,200,000,000 if one of the next few years were a peak prosperity year generally similar to 1928. The magnitude of this $8,200,000,000 total may be appreciated by recalling that the peak yield thus far for federal tax revenue was $5,700,000,000, in 1920, and that the lowest yield since 1929 was $1,900,000,000 in 1933. This estimate includes the effects of the extraordinary capital gains of the 1928 period and is therefore decidedly optimistic.

Federal old-age payroll taxes, under the existing law, will yield a surplus over actual payments to social security beneficiaries in each of the years 1938-40. This surplus will be about $300,000,000 in 1938, and again in 1939, and $500,000,000 in 1940. /3/ Presumably the surplus, being invested in federal securities, will make the total debt outstanding in the hands of the public less by a corresponding amount. Technically, it does not result in a "reduction" of the debt. If it is to be considered a debt reduction for purposes of calculating over-all deficits or surpluses, these figures must be added to the surplus figures or subtracted from the deficit figures given in this chapter. The federal unemployment compensation payroll tax is assumed to yield practically nothing after 1937, owing to the enactment of state taxes to absorb the credit offered by the federal tax. State payments for unemployment compensation are assumed to equal, roughly, the state credit- absorbing taxes.

Of the $8,200,000,000, the personal income tax would yield $3,100,000,000; the corporation normal tax, $1,800,000,000; the estate tax, $600,000,000; the gift tax, probably $200,000,000 or more, although it should be noted that this constitutes largely a prepayment of

/3/ Derived from Table 15, pp. 108-9.

[page 125]

estate taxes; the manufacturers' excises, perhaps $500,000,000; the
tobacco taxes, $600,000,000; the liquor taxes, $600,000,000; customs,
$500,000,000; and the stamp taxes and miscellaneous special taxes,
about $300,000,000. Included in the personal income tax estimate
above is $200,000,000 permanent revenue expected from the provision
of the Revenue Act of 1936 repealing the exemption of dividends from
the normal tax.

As has been explained, /4/ the effect of the 1936 act was to create a source of genuinely additional revenue and also a source of advance revenue -- revenue that is only temporary, since it obtains present revenue to a major extent /5/ at the cost of future revenue. If the advance revenue is included, the total rises to somewhere near $9,000,000,000. Although the tax on undistributed profits will yield genuinely additional revenue, its prospects are very uncertain. However, if an extreme figure of total possible revenue is desired, another $500,000,000 may be added, but the total, which would be somewhere near $9,500,000,000, had best not be used until some conservative adjustment is devised to take account of the economic repercussions and the administrative difficulties that such a high level of taxation involves.

Federal expenditures for the years 1938-40, exclusive of statutory debt retirement and payments to the old-age reserve fund, range, under the functional estimates, from $4,700,000,000 to $6,400,000,000, depending on the year and the assumptions made about relief expenditures. /6/ Under extremely favorable conditions, therefore, the present federal revenue system would make it possible to reduce the debt (including statutory debt retire-

/4/ See pp. 77-79.

/5/ See p. 75, Table 7, note g.

/6/ See pp. 107-11, Tables 14, 15, and 16. The maximum of $6,400,000,000 has been obtained by deducting the estimated amount of statutory debt retirement ($555,000,000) from total expenditures for 1938 under the second functional estimate (Table 16). The minimum of $4,700,000,000 has been obtained by deducting the estimated amount of statutory debt retirement ($534,000,000) from total expenditures for 1939 under the first functional estimate (Table 14).

[page 126]

ment) from about $1,800,000,000 to about $3,500,000,000 in a year
like 1928, on the basis of an $8,200,000,000 tax yield.

Estimates for a Year of a Moderate Prosperity

In a moderately prosperous year -- similar, for example, to 1924 -- the yield of the existing federal tax system (still excluding social security taxes, advance revenue, and the uncertain future revenue from the undistributed profits tax) would be about $4,900,000,000.

As shown in Table 7 in Chapter 6, the personal income tax would produce $1,300,000,000. To this may be added a rough estimate of $100,000,000 for permanent revenue obtained from repealing the exemption of dividends from the normal tax (still assuming no pressure from the undistributed profits tax for more dividends). No estimate was made for the corporation tax for such a year, but, judging from the data for 1928, 1933, and 1934, /7/ the tax would probably produce somewhat more than $1,000,000,000. As with the 1928 estimate above, nothing is included for the undistributed profits tax. Thus, the income tax total would be $2,400,000,000. The September 1936 budgetary estimate is $2,300,000,000 for the fiscal year 1937, but this includes, presumably, amounts for (1) the undistributed profits tax, and (2) advance revenue for the personal income tax as explained above. /8/

The existing estate tax, it is estimated, would produce for the federal government $400,000,000 on returns such as those filed in 1926. This may be reduced to $300,000,000 to bring it closer to 1924 conditions. Estimates for the other taxes are: gift tax, $100,000,000; manufacturers' excise taxes, $400,000,000; tobacco taxes, $500,000,000; liquor taxes, $500,000,000; customs, $400,000,-

/7/ See the memorandum by Susan Burr and William Vickrey on federal income and estate tax estimates in the forthcoming volume, Studies in Current Tax Problems.

/8/ See Note 2, Federal Budget of January 5, 1937, p. 540.

[page 127]

000; and stamp taxes and miscellaneous special taxes, $300,000,000.

Comparing these revenues with the estimated expenditures on the various assumptions, it follows that a surplus as high as $200,000,000 or a deficit as high as $1,500,000,000 might result. Actually, however, it may not be too optimistic to rely on an additional $400,000,000 or $500,000,000 of advance revenue and undistributed profits tax revenue. This would make the range of possibilities lie between a surplus of about $700,000,000 and a deficit of about $1,100,000,000.

Estimates for a Year of Depression

Finally, it is well to recall the possibilities of a year of deep depression, such as 1933. In such a year even the present federal personal income tax rates would yield only $500,000,000, including the permanent revenue from the taxation of dividends under the normal tax. The corporation normal tax would yield another $500,000,000. The estate tax would yield $200,000,000; the gift tax, perhaps $100,000,000; manufacturers' excises, $300,000,000; tobacco taxes, $500,000,000; liquor taxes, $500,000,000; customs, $300,000,000; and stamp taxes and miscellaneous special taxes, $200,000,000. The total is only $3,100,000,000. Even if advance revenue and revenue from the undistributed profits tax are added, the total would not be likely to pass $3,300,000,000. With expenditures at a $6,000,000,000 or $7,000,000,000 level -- or possibly higher, owing to emergency expenditures -- the deficit would be $3,000,000,000 or more.

Sources of Additional Revenue

If additional revenue should be needed, the income taxes and the estate taxes could yield several hundred million dollars more if the rates were raised. While the new rates would undoubtedly be heavy, they would

[page 128]

probably not be considered politically impossible in a fiscal crisis.
/9/ Lowering of exemptions promises possibilities of several hundred
million dollars more. Other parts of the system, while capable of
less spectacular yields under favorable conditions, offer chances for
substantial increases that are probably not beyond the political
limit of capacity to pay.

In general, then, if the federal government fails to balance its budget in the near future, it will probably be either (1) because expenditures rise in some manner not indicated in the functional analysis in Chapter 7, or (2) because business fails to stay above such moderate conditions as those in 1924 and as a result the relief load fails to drop markedly. In either case, the budget can be balanced if some of the more obvious revenue possibilities are employed. This statement, however, does not imply that existing or prospective expenditures are justified or that the tax load is not too heavy. No judgment at all is passed here on expenditures. The point is, rather, that the existing federal tax system has great revenue potentialities for a period of prosperity. Little reason exists for saying that new federal taxes will inevitably be necessary in the next few years if the budget is to be balanced. The answer depends upon the trend of business conditions. If, on the other hand, an effort is to be made to reduce the debt, even by only half the increase since 1930, additional taxation will be necessary unless a prolonged period of great prosperity ensues.

/9/ See pp. 73-75.

[page 129]

Section B. Social Control

Chapter 9



All activities of government react on one another, and tax policies cannot be isolated from other governmental policies. Thus, even when a tax is being levied primarily for revenue, it will be judged partly by its effect on other governmental policies. Furthermore, the primary aim in imposing taxes is frequently not revenue but regulation. In this chapter and Chapters 10 through 14 the relation between current tax policies and other important governmental policies of the United States will be examined. In the present chapter an analysis is made of the problem in general, while in the five following chapters special types of governmental policy are discussed.

Governmental Policy in Relation to Economic Activity

The prevailing attitude in the United States concerning the proper relation of government to economic life is based on a belief in the value of self-interest and free enterprise. The system of private capitalism reflects an assumption that in the main the general welfare is best promoted by giving wide play to individual freedom of choice and action. Disturbance and diversion of economic activities are thought to diminish the general welfare.

However, this faith is not unlimited, nor is it so complete as formerly. In many respects the public has

[page 130]

deliberately diverted economic activity from the courses it would
otherwise have followed. Examples are the regulation of public
utilities, banks, labor relations, foreign commerce, and so on, and
the prohibition of monopoly and other undesirable economic
activities. Thus the general principle that the government shall
exercise no influence on economic activity is qualified in practice
by the control of certain activities in which laissez faire is
thought undesirable.

Taxes disturb free enterprise, since they always have some influence on economic activity. From this point of view, one of the secondary aims in choosing among taxes would be to keep restrictive or distorting effects at a minimum. In sharp contrast, governmental control as a primary aim for taxation is designed to influence economic activity in certain specific ways, regardless of revenue.

Opposition to Social Control through Taxation

To minimize the influence of taxation on economic activity is commonly recognized as a valid and important object of tax policy. The intentional use of taxation for purposes of control, however, has been vigorously opposed.

Those who object to all kinds of governmental regulation of economic life by any method are quite consistent in opposing the use of taxation for that purpose. Others may not be so consistent. People often favor governmental use of the taxing power to promote and protect their business, while they vigorously oppose its use when it would impose any restraints on their business or affect the distribution of wealth or income in a manner unfavorable to them.

Those opposed to the use of taxation for social control advance the following arguments. Taxation deals with results rather than with causes. It is a crude method of

[page 131]

control, incapable of being used with fine discrimination and
consequently likely to result in much harm. Finally, the difficulties
of devising a sound tax system are enormously increased by adding
another primary aim to that of raising revenue.

All taxation, however, affects economic life. A refusal to consider the relation of tax policies to other governmental policies would allow some of the former to nullify some of the matter. Even if social control were not a primary aim of taxation, a valid secondary aim would be to see to it that the effects of taxation supported and did not nullify other governmental policies.

Opposition to the use of taxation for social control has not stopped such use. Society rarely refuses to use any known powerful instrument of economic and social adjustment in attaining its ends. The actual use of taxation for control is common enough to require an analysis of it as one of the primary aims of tax policy.

Types of Control through Taxation

Taxation may be used for social control in at least three ways: (1) It may be used to redistribute wealth and income in order to offset an undesirable distribution caused by the economic system. This use is discussed below in Chapter 14. (2) It may finance subsidies that are granted in order to encourage and control private activities. Subsidies to shipping companies to induce expansion of operations and subsidies to farmers to induce restriction of crops are examples. Since in these cases taxation has the purely passive role of producing revenue and the control is achieved through expenditure, this method is outside the scope of this study. (3) Finally, taxation exerts pressures upon the activities of individuals and business concerns. The protective tariff is an example. The present chapter and the four follow-

[page 132]

ing chapters are devoted to an analysis of the use of tax pressures
for control purposes.


All taxes bring about economic effects by the pressure that they exert on economic life. When taxes are imposed, activities that previously gave the maximum profit and satisfaction for the outlay may no longer do so. Thus, an individual who formerly employed labor rather than machinery might, after the imposition of a heavy tax measured by payrolls, find it profitable to buy more machinery and employ less labor. Again, a housewife who used to buy beef because it gave her the best value for her money might, after the imposition of a tax on beef, buy pork instead, which now offered her the best value.

Furthermore, some actions are made impossible by taxation. A person with a $1,500 income cannot spend so much in buying goods if he is obliged to pay a $150 tax as he could spend if no tax were imposed.

Not only do all taxes have economic effects, but -- contrary to a common claim -- the effects of the various taxes differ widely in nature and degree. In the following paragraphs some of the reasons for these differences are reviewed.

Factors Influencing Effects of Taxes on Activities

Whether the pressure of a tax will cause a change of activities depends on a number of factors, including the mobility of the tax base, the rate of the tax, the extent to which the tax is a general one, the nature of the tax base (whether the tax is on net return or gross receipts), and the situation of the person taxed.

Mobility of the tax base means the ease with which the taxpayer may change his economic activity so as to avoid the tax. A payroll tax has been mentioned that

[page 133]

might cause a chance from the employment of labor to the installation
of additional machinery. If a tax were imposed on machinery instead
of on labor, the mobility of the tax base would be less -- that is, a
shift from machinery to labor would be less profitable. The whole
working life of the machine would already have been bought and paid
for, and little money could be saved by abandoning it before it wore
out. Similarly a tax imposed in a single state on industrial income
or sales would more probably cause light manufacturing business than
public utilities or gold mines to migrate from the state. Gold can be
mined only where it exists. The public utility must be in physical
contact with customers. The manufacturing establishment, however, may
move into another state that does not impose the tax and ship its
goods to market in interstate commerce, which, under the federal
Constitution, is exempt from taxes imposed by the states.

The importance of the rate of the tax may be seen in the case of the tariff. A low tariff on the importation of some manufactured commodity -- for example, watches -- may have little, perhaps no, effect on the amount imported. On the other hand, by more than offsetting the cost advantage of foreign producers, a high tariff may completely stop the importation.

A general tax or one imposed over a wide area is less likely to affect economic activity than a special one or one imposed in a small area. A manufacturing establishment has no inducement to move if the income tax or sales tax that it wishes to avoid is levied in all states, instead of being limited to one. Similarly, a tax imposed on only one industry is likely to cause more people to shift from that industry to others than a tax imposed on all industries.

Again, taxes on net return will as a whole affect economic activity less than taxes on gross receipts. A tax

[page 134]

imposed on gross receipts may drive some concerns that are not making
profits into bankruptcy. A tax on net profits will not, since only
profitable concerns need pay it.

Finally, differences in the situations of persons influence the effects of taxes. For example, suppose that a special tax were imposed on the incomes of all practicing lawyers. The lawyers that it would induce to abandon their profession would be those who were making only about the same income that they might earn in some other occupation and who had already been debating a change. Similarly, a tax imposed on the manufacture of radios would induce some concerns to shift into the manufacture of another commodity, while others would find it more profitable to remain in the radio industry and pay the tax.

Effects of Special Taxes

Special taxes imposed on particular industries, forms of business organization, or other tax bases repress the concerns on which they are imposed, since they add to the costs of doing business. In a competitive industry that is specially taxed, either prices must be increased at once to overcome the increased cost, or firms will -- if their capital is fairly mobile -- tend to leave the industry for more attractive opportunities elsewhere. As a result those remaining in the business can increase prices. When prices are increased, some consumers will reduce purchases, while others will cease buying the product and will turn instead to an untaxed substitute. Consequently, the industry producing the untaxed substitute will be able to increase prices and will enjoy greater prosperity. Other business men in turn will be encouraged to enter it.

Similarly, when a special tax is imposed on a form of business organization or business practice, the added

[page 135]

cost of doing business in the manner subject to the tax will cause,
or at least encourage, a shift to methods not subject to the tax.

Effects of Tax Exemptions

General taxes, such as the general property tax, general retail sales tax, and general net income tax, are not in themselves well adapted for economic control. Their pressure cannot be concentrated in a sufficiently specific manner. However, special exemptions from general taxes work broadly in the same way as special taxes. If a particular plant or industry is given a special exemption from a general tax, such as the property tax, it is placed in a favored position. Either its operators can pocket the amount of the tax exemption, or they can use their lower cost position to attract customers away from competing products.

Exemption under a general tax produces its promotional effect somewhat differently from a special tax. The exemption is a special relaxation of pressure at a given point in an area of general tax pressure. The immediate effect is one of benefit to those receiving the exemption. From this benefit harm flows to competitors and to taxpayers generally to -- competitors because of the favored position in which the exempt industry or establishment is placed, and to taxpayers generally because the taxes escaped by those who are exempt must be borne by others. To summarize: in the special tax, repression at a specific point results in benefits to others; while in the exemption, release from taxation at a specific point results in injuries to others.


The use of taxation in the United States for control rather than revenue has not gone unnoticed or un-

[page 136]

limited by the courts. They have faced a dilemma in determining the
constitutionality of such legislation. If the courts restrained the
imposition of every tax that had repressive economic effects, there
would be little or no revenue for government. However, the power of
taxation to repress and destroy is great. If governments were not
limited in its exercise, they could enforce a control otherwise far
beyond their power, and they could thus warp and perhaps completely
alter the framework of economic life. To draw a line allowing free
play to legislative bodies in their revenue policies and yet not
giving them undue powers of social control through the taxing power
has been a difficult judicial task. Confusion as to the precise
location of the line is due in large part to this difficulty,
although changes in the social philosophies of the judges are also

Federal Power to Impose Taxes for Non-Fiscal Purposes /1/

The point beyond which taxation shall not be used for control has been particularly difficult to fix in federal legislation because all national regulation must be based on powers specifically delegated in the Constitution. The most important in this connection are the commerce power and the taxing power. The former, however, applies only to interstate and foreign commerce. The temptation for the federal government to use the taxing power for other than revenue purposes is therefore greater than it might be if it had more police powers.

The position of the Supreme Court of the United States on taxation for non-fiscal purposes is not entirely clear. It has frequently upheld taxes imposed in fact for such purposes. The existence of a collateral purpose in addition to that of raising revenue has been held not to

/1/ See Note 1, Cases on Federal Taxation for Non-Fiscal Purposes, p. 542, for cases illustrating the points in this section.

[page 137]

invalidate the tax. /2/ The fact that the tax is so heavy that the
business is destroyed and no revenue is produced has in some cases
been held by the Court to be immaterial. /3/ Furthermore, the Court
has at times upheld regulatory provisions that, while ostensibly
included in the law to aid in the collection of revenue from the tax,
were obviously imposed as a means of control. /4/

However, where the Court has concluded from the language of the statute that regulation was the actual purpose, it has held the charge imposed to be a penalty rather than a real tax, and thus not a proper exercise of the taxing power. This conclusion has been reached when the form of the tax measure has indicated that the charge was intended to force those on whom it was levied to conform to collateral prohibitions or restrictions imposed by the act. /5/ Another factor has sometimes been the weight of the tax, although this has in some cases been disregarded. /6/

When the Court finds the charge to be a penalty and not a true tax, the constitutionality of its imposition depends on whether Congress could under some other delegated power legally impose the restrictions that the payment was intended to enforce. /7/ In a recent case /8/ the Supreme Court has added a further restriction. It has held that a tax may not legally be employed to raise the money with which to achieve by monetary inducement a control that Congress has no power to exercise directly.

It is difficult to harmonize all the decisions of the

/2/ See Hampton and Co. v. United States, 276 U.S. 394, 412, 413 (1928); Unites States v. Doremus, 249 U.S. 86, 93, 94 (1919).

/3/ See Veazie Bank v. Fenno, 8 Wall., 533, 19 L. Ed. 482, 488 (1869); McCray v. United States, 195 U.S. 27, 60 (1904); Magnano Co., v. Hamilton, 292 U.S. 40, 45 (1934); Fox v. Standard Oil Co., 294 U.S. 87, 99 (1935).

/4/ See Felsenheld v. United States, 186 U.S. 126, 132 (1902); United States v. Doremus, 249 U.S. 86, 94 (1919).

/5/ See Bailey v. Drexel Furniture Co., 259 U.S. 20, 38 (1922); Hill v. Wallace, 259 U.S. 44, 67 (1922); Carter v. Carter Coal Co., 298 U.S. 238 (1936).

/6/ See Trusler v. Crookes, 269 U.S. 475, 482 (1926).

/7/ See Bailey v. Drexel Furniture Co., 259 U.S. 20, 38 (1922); Hill v. Wallace, 259 U.S. 44, 68 (1922).

/8/ United States v. Butler, 297 U.S. 1 (1936). In this case the Agriculture Adjustment Act was declared unconstitutional.

[page 138]

Court on the use of taxation for non-fiscal purposes. The position of
the Court has been somewhat indecisive. As a result, past decisions
often furnish precedents for deciding almost any given case whichever
way it wishes.

The federal power to use taxes for purposes of control is also limited by constitutional restrictions on the types of tax that may be employed. Indirect taxes must be uniform throughout the United States. Export duties are prohibited. Direct taxes, with the exception of the income tax, must be apportioned among the states according to population. /9/ Instrumentalities and agencies of state governments may not be taxed. These restrictions prevent the imposition of certain forms of control that might otherwise be used. If no requirement for uniformity existed, taxation could favor the development of one geographical section at the expense of another. If export duties were permitted, they could be used to reinforce the protective tariff. If apportionment of direct taxes were not required, the experiment of a modified "single tax" might be tried on a national scale. If the taxation of state governmental instrumentalities and agencies were permitted, the federal power to control state action would be greatly increased.

State Power to Impose Taxes for Non-Fiscal Purposes /10/

To the states are reserved all powers not delegated to the federal government and not denied to the states. The Constitution specifies some restrictions on state taxation, and others have been developed by the Supreme Court. States may not tax federal instrumentalities or impose burdens on interstate commerce, nor may they go beyond the limits set by the Fourteenth Amendment.

The Court has developed restrictions on state taxation, especially in its interpretation of the clauses of

/9/ See Note 2, Uniformity Requirements for Federal Taxes, p. 512.

/10/ See Note 2, Cases on State Taxation for Non-Fiscal Purposes, p. 542, for cases illustrating the points in this section.

[page 139]

the Fourteenth Amendment concerning due process of law and the equal
protection of the laws. A state may not arbitrarily discriminate
among persons and tax objects, but it may classify them for taxation
according to some basis that the Court deems reasonable. Thus, in the
chain store tax cases the Court decided that the distinction between
chain stores and independents was a reasonable basis of
classification that justified the imposition of special taxes on
chains. The distinction between a large volume of sales and a small
volume, however, was held not to be a reasonable basis to justify the
imposition of a progressive sales tax. /11/ By finding that the basis
for classification is not a reasonable one, the Court may restrain
the use of taxes that otherwise could be used for non-fiscal
purposes. In effect, the Court is in a position to uphold those forms
of control that it considers reasonable and to eliminate those that
it considers unreasonable.

The federal Supreme Court apparently applies to the state governments the same rules for the weight of taxation and the imposition of penalties that it applies to the federal government. However, because of their reserved powers, the states have police powers under which regulation may be imposed. If the Court decides that a statute imposes a penalty and not a tax, the statute may still be upheld if it is deemed a proper exercise of the police power.

State constitutions, as interpreted by state courts, in general place no limit on the rate at which a tax may be imposed, regardless of non-fiscal effects. However, in a few states local license taxes must not be prohibitory or unreasonably repressive. Furthermore, if the state court finds that the "tax" charge is imposed for regulation instead of as a true tax, the rate of the charge must be related to the power of the state to impose the

/11/ See p. 183.

[page 140]

regulation. The rate of a "tax" is likely to be considered by the
court as one of the factors determining whether the charge is a tax
or is imposed for regulation.


Economic and administrative limitations are added to those on the legal power of the federal and state governments to use taxation for control purposes. The tax may be economically unsuited to the regulatory task to which it is put. For example, a retail sales tax on all commodities would probably reduce the consumption of liquor, but it would have too many other important consequences to warrant its adoption for the specific purpose of reducing liquor consumption. Again, a tax may place too heavy an administrative load on government. Heavy taxes to restrict the consumption of liquor, which might be very effective if the taxes could be perfectly applied, might fail because the administrative officials were not able to cope effectively with smuggling and bootlegging. /12/

/12/ See Note 1, Liquor Tax Evasion, p. 566

[page 141]

Chapter 10


The promotion of industries that are deemed by government to be of special public interest has been perhaps the chief non-fiscal use made of taxation in the United States. By far the most important example is the protective tariff, but special internal taxes and tax exemptions are also employed. The subject will be discussed under those three divisions.


From the beginning, the tariff was understood to be imposed for promotional and control purposes as well as for revenue. The preamble of the first revenue act mentioned a necessity for "the encouragement and protection of manufactures." Similar statements appeared commonly in the titles or preambles of subsequent tariff acts. In a case before the Supreme Court in 1928, a tariff measure was attacked in part on the ground that it set out "to encourage the industries of the United States." According to the argument, it was therefore unconstitutional, since revenue was the only object for which Congress could levy customs duties. The Court held that this argument was nullified by the whole history of tariffs, and that it was contrary to the settled point of law that, "so long as the motive of Congress and the effect of its legislative action are to secure revenue for the benefit of the general government, the existence of other motives in the selection of subjects cannot invali-

/1/ See Note 1, Bibliographical Note for the Protective Tariff, p. 543.

[page 142]

date Congressional action." /2/ In 1933 the Court even went so far as
to state that protective duties are laid under the commerce power
rather than under the taxing power, although they may be laid under
either. Apparently the tariff is now legally in the position that it
has actually held since the Civil War: it is an instrument for
protection and promotion, with incidental revenue aspects.

Injury to Domestic Producers

The protective tariff may seem to promote some industries without injury to any other domestic industries. This appearance is due to the mechanism of international trade, which hides the effects of the tariff.

All sales to foreign countries must sooner or later result in an equal volume of purchases from foreign countries. Unless one seller is willing to give his goods away in exchange for promises that are never fulfilled, there must be an exchange of goods and services, at the same time or later, direct or roundabout. The world's supply of gold is much too small to finance unbalanced trade for long. In the absence of tariffs, there develop certain industries, naturally favored by the kinds of resources, skills, etc. of the country, that sell their products abroad. An equal amount of low-cost products of other countries is imported in payment. When a protective tariff is introduced, imports decrease, and consequently the foreign countries must reduce by approximately the same amount their purchases from the exporting industries. Thus the restriction of imports diminishes or destroys the foreign market for certain domestic products. The domestic producer of the protected commodity is benefited both at the expense of the domestic consumer, whom he can charge a higher price for his product, and at the expense of the domestic producer for the foreign market.

/2/ Hampton and Co. v. United States, 276 U.S. 394, 412 (1928).

[page 143]

If this promotion of one domestic industry at the expense of others were more obvious, the application of a protective policy would probably be more strongly contested than it has been. However, the effect on the domestic producer for the foreign market is diffuse and indirect. The obvious and direct sufferers are the foreign producer, the importer, and the consumer. Foreign producers have no domestic political influence, importers are neither numerically nor politically important, and the injury to each consumer is piecemeal and hidden in prices. The result is a greater willingness to encourage and promote specific industries through external taxes, such as the tariff, than through internal taxes.

Tariff Beneficiaries in the United States

In the United States manufacturing has in general been the chief beneficiary under the protective tariff. Agriculture has been mainly an exporting industry. Therefore, it has been injured rather than benefited. During the earlier decades of the federal tariff the economic life of the country was largely agricultural, and the result of the tariff was to diversify industries and to establish manufacturing by promoting infant industries.

However, after the Civil War and especially toward the end of the century, American manufacturing industries were well past their infancy. During the last half century many of them have endeavored to organize into huge industrial combinations for monopolistic control. The tariff apparently has facilitated the growth of combinations and the imposition of high prices, by excluding the foreign competition that might have kept prices at a competitive level. Thus, while the Federal government has for more than forty years attempted to destroy and prevent industrial combinations and monopoly prices, it has retained, in the tariff, an instrument that encourages them.

[page 144]


The repressive effects of internal taxes imposed to promote special industries are more obvious than the effects of tariffs because an internal tax is laid directly on a domestic industry. Taxed industries object strenuously to discriminatory taxation and can easily show that one group of citizens is being deliberately favored at the expense of others. This is a powerful complaint politically, and unless the political importance of the complainants is small it is likely to secure relief. For this reason internal taxes have been used only rarely for promotional purposes. Another reason is that few commodities have exact substitutes. A tax imposed on a commodity would thus not divert demand to another commodity with any precision. The diversion might be small and might also be widely scattered among many commodities in such a situation. Discrimination between producers of the same commodity would be difficult, in view of constitutional requirements of uniformity.

Bank Note Tax

There are, however, some examples of the use of internal taxation for promoting one business at the expense of another. Perhaps the earliest was the tax of 10 per cent imposed on state bank notes beginning in 1866, which is still in effect. /3/ The national banks had been set up during the Civil War partly to serve as a market for government bonds and partly to give the country, through their note issues, a uniform currency of standard value. Congress soon observed that the full benefits of the national bank notes could not be attained without legislation to drive the competing state bank issues out of circulation. This aim was achieved by the tax.

/3/ See Note 2, Bibliographical Note for Bank Note Tax, p. 544.

[page 145]


The Supreme Court upheld the tax, although it was clearly so heavy that it would not produce revenue and would destroy the note- issuing business of the state banks. The Court affirmed that the responsibility of the legislature for excessive taxation was to the people and not to the courts, and, further, that Congress had the power to regulate money and might use the taxing power for that purpose.

Oleomargarine Taxes

Another example is the taxation of oleomargarine. /4/ In 1886 Congress imposed a tax of 2 cents a pound on oleomargarine and heavy license taxes on oleomargarine manufacturers and dealers. Imported oleomargarine was subjected to a special internal revenue tax of 15 cents a pound in addition to the tariff. Congress apparently had no hostility to the domestic oleomargarine industry as such, since there was no tax on the product manufactured for export. The purpose was, of course, to protect and encourage the dairy industry, which saw in oleomargarine a threat to its markets.

This act was amended in 1902, when the general rate on oleomargarine was fixed at 10 cents a pound. For oleomargarine free from any artificial coloring that caused it to look like butter, however, the rate was lowered to 1/4 cent a pound, and license taxes on dealers were greatly reduced. In 1931 the low rate was restricted to oleomargarine free from yellow color, whether artificial or not. The obvious immediate objectives of the rate on the colored product were not only to prevent the misrepresentation of oleomargarine as butter but also to discourage the use of oleomargarine. As its manufacturers complained in attacking the constitutionality of the act, the white product did not appeal to public taste. The Court upheld the tax on the grounds that it was

/4/ See Note 3, Bibliographical Note for Oleomargarine Taxation, p. 544.

[page 146]

imposed for revenue and, whatever the effects, the judicial branch of
government had no reason to interfere.

While the federal tax on colored oleomargarine was sufficiently high to drive it from the market, the uncolored product continued to be sold for private consumption, a quantity of coloring matter being sold with the oleomargarine for home coloring.

Some states, however, have endeavored to eliminate all use of oleomargarine within their borders. For example, the state of Washington imposed a tax of 15 cents a pound on butter substitutes. The tax was attacked in the courts but was unanimously upheld by the Supreme Court of the United States. The principles followed were the same as those stated in the opinion upholding the federal tax. A number of other states have adopted taxes similar to the Washington tax.

Agricultural Adjustment Taxes

In the examples given above, an industry was favored by taxing the product of a competing industry. The Agricultural Adjustment Act was intended to promote the prosperity of certain agricultural industries by reducing the supply of their products and also by paying direct subsidies. /5/ The taxes were essentially of two kinds -- the processing taxes and the cotton and tobacco control taxes.

The processing taxes were excise taxes imposed on concerns processing the agricultural commodities. They differed from the many other excise taxes imposed by the federal government in two essentials. First, their rates were originally determined by a formula administered by the Secretary of Agriculture, although later the amounts were fixed by statute. Second, the proceeds were paid as bonuses to the producers of specified agricultural products who limited their production in accord-

/5/ See Note 4, Bibliographical Note for Agricultural Adjustment Laws, p. 544.

[page 147]

ance with the desires of the Department of Agriculture. The taxes
were, however, like the other excises in being merely a source of
revenue, in the direct sense, not a measure of control.

The taxes in the acts regulating the cotton and tobacco industries, however, were used as a direct means of control. They were imposed on production in excess of the quotas allotted to growers. The rates were fixed at 50 per cent of the value of the excess cotton and from 25 per cent to 33-1/3 per cent of the value of the excess tobacco.

In the Supreme Court decision on the Agricultural Adjustment Act the processing taxes were held unconstitutional. The Court held that a tax was invalid if it was levied to finance a control that the federal government had no power to exercise. Such a tax was deemed analogous to taxes held invalid in other cases because they were penalties. Since the Agricultural Adjustment Act decision held that the federal government lacked the power to regulate agriculture, the taxes on cotton and tobacco, although they were not involved in the case, were almost certainly also unconstitutional. They would probably have been considered penalties.


Under the heading of tax exemptions will be included various methods of lightening the tax burdens of favored concerns or industries. The difference between special taxes and tax exemptions is essentially a matter of degree. The imposition of special taxes on one or a few concerns or industries is the equivalent of exempting the many, while the exemption of one or a few is the equivalent of taxing the many.

Tax exemptions may be very narrow or very wide. The narrowest perhaps is the exemption of a single plant or concern from a general tax, such as the property tax.

[page 148]

The widest is the special lightening of the total tax burden on
certain broad groups of concerns by failing to impose certain taxes
that would rest heavily upon them. The latter is not an exemption in
the legal sense, but it is included in this discussion because of the
difficulty of drawing a line. Economically it is similar to
exemption. In both cases, the method is to create a point or area of
lower tax pressure in a field of relatively higher pressure.

Exemptions and unequal taxes may have two effects. One is to encourage the location of a given economic activity in certain taxing jurisdictions. The other is to encourage the economic activities that are granted the exemption and to discourage other economic activities.

Exemptions Affecting Location

Where people and goods are free to pass from one taxing jurisdiction to another, as in the United States, differences in taxes and exemptions have the effect of encouraging the location of a given economic activity in the jurisdiction where the tax burden is lightest.

Some industries have difficulty in changing their locations to escape heavy taxes. When a farmer packs his goods and moves to a distant state with his moving vans placarded "driven out by high taxes," or when the owner of a stone quarry threatens to move his business out of the state if a certain tax is imposed, the protest is not regarded seriously. /6/ An essential element of the businesses cannot be moved -- in the one case land, in the other raw material. Similarly, the location of railroads, public utilities, mines, and retail stores is affected relatively little by the weight of taxation.

Other industries, however, are relatively mobile. They can often be completely moved out of the high tax area without sacrificing any essential business element. Manufacturing establishments, stock exchanges, motion

/6/ See Note 5, Tax Pressure of Mobile Industries, p. 545.

[page 149]

picture studios, and grain elevators, for example, are able to
convince the public that they can make good their threats to migrate.

State legislatures may have no desire to favor an industry, such as warehousing, or a whole field of industry, such as manufacturing, but the mobility of such industries gives them a peculiarly strong bargaining power in the determination of tax policies. It is immaterial that the same threats to migrate are used on all state legislatures and that the threats are not often made good. So long as the possibility appears to exist, these industries will have an advantage in state and local taxation.

Interstate tax competition of this kind is a reason that some states have, for example, failed to adopt corporation income taxes, or have allowed the specially low assessment of manufacturing machinery. Florida in 1924 passed a constitutional amendment forbidding the imposition of income or inheritance taxes /7/ and advertised the fact widely. Many states have provided special exemptions from the general property tax for particular industries. /8/ A survey in 1929 revealed constitutional or statutory provisions that granted some measure of tax exemption to manufacturing property in fourteen states. The exemptions were for limited periods, usually not over five or ten years. The exemption was rarely in the constitution; sometimes it was provided by statute; frequently it was granted only by vote of the municipal legislature or by the electors.

Such exemptions are a method of intercity, as well as of interstate, competition. In addition to legal exemptions, municipalities sometimes vote exemptions beyond their legal power, and local assessors frequently assess

/7/ Later repealed in respect to the inheritance tax. See Note 5, Tax Pressure of Mobile Industries, p. 545.

/8/ See Note 6, Bibliographical Note on Exemption of Industrial Property, p. 545.

[page 150]

industrial property at low rates with the tacit consent of the

If one state or locality grants an exemption, the natural tendency is for other states and localities also to grant exemptions. For the most part these competitive exemptions and underassessments probably offset each other. The net result is a general exemption to the industries in question. Consequently they are favored at the expense of less mobile industries.

Exemptions Not Primarily Concerned with Location

Not all exemptions of industries are primarily the result of competition with other jurisdictions. Some can be traced to the political strength of the affected group or the belief that the existing economic system does not give sufficient incentive for development. Thus agricultural property of several kinds is granted special exemptions in many states. /9/

Exemption of Dwellings

A few states exempt homesteads occupied by the owner from property taxation, the exempted value being limited in amount. /10/ One reason for the spread of homestead exemption is the desire to encourage home ownership and to discourage tenancy.

Occasionally, in times of housing shortage, exemptions have been offered to encourage the building of new houses. /11/ A New York statute passed in 1920 permitted municipalities to provide for such exemptions. By a New York City ordinance passed in February 1921, exemption from assessment for real estate taxes until 1932 was granted to all housing on which construction was begun before April 1, 1922, and finished within two years. Exemptions were up to $1,000 per room for a

/9/ See Note 7, Bibliographical Note on Agricultural Tax Exemption, p. 546.

/10/ See the memorandum by Thomas J. Reynolds on homestead exemptions and rate limitations in the forthcoming volume, Studies in Current Tax Problems.

/11/ See Note 8, Bibliographical Note on Exemption of Improvements, p. 546.

[page 151]

maximum of 5 rooms in single-family houses and 5 rooms per apartment
in multi-family houses. Later the period of construction was
extended. The degree to which this exemption stimulated housing
construction is subject to debate. A housing boom certainly followed
the passage of the ordinance, but a similar boom began about the same
time in other cities. In any event, measurement of the effect of the
exemption is scarcely possible.

Exemption of All Improvements

The proposal for the permanent exemption of improvements -- as contrasted with temporary exemption of the type cited above -- has its roots in the single-tax philosophy developed and popularized by Henry George and his followers. The single-tax advocates propose that practically the full economic rent of land and natural resources shall be taken by government in taxes, and that no taxes shall be imposed on improvements, personal property, incomes, or other tax objects.

Two main arguments are presented for the single tax. First, the tax is just, because land and natural resources were originally a gift of nature, while their increase in income and value has been due to social development and the increase in population. Second, the tax is economically desirable, because land and natural resources are available for use in production without the expenditure of labor, saving, or management. Taxes placed on the rent of land, in contrast with other taxes, will not, it is said, discourage economic activity and enterprise but will encourage them.

The opposition to the single tax has been largely based on the grounds of justice /12/ and inadequacy of revenue. It has been so effective that the single tax in its pure sense is not an issue anywhere in the United States. The

/12/ See pp. 290-91.

[page 152]

rent of land, even if taken in full, would probably be insufficient
today to supply the demands of government for revenue. The evidence
on this point, however, is perhaps not conclusive. Furthermore, the
public has shown no inclination deliberately to impose taxes equal to
the full economic rent. Instead, the issue has arisen in the
differentiation between land and improvements through the partial or
total exemption of improvements from ad valorem taxation.

Such an exemption has been used in the Canadian Northwest. In 1913 Pittsburgh enacted a graded tax law that provided for the gradual reduction, from 1914 to 1925, of the percentage at which improvements were assessable for city taxes. The final assessment ratio was fixed at 50 per cent of value. Land continued to be assessable at 100 per cent of value.

The extent to which exemption of improvements has actually resulted in the growth of improvements is difficult to measure. It was generally believed that the effect in Canada was substantial. The effect of partial reduction in Pittsburgh has been debated. On theoretical grounds, exemption of improvements almost certainly stimulates building construction, although no general agreement exists on the desirability of such stimulation.

The exemption of improvements might be expected to play an important part in the movement for tax limitations on real estate. If real estate tax reduction is to serve as a method of social control, it may take the form of a reduction of taxes on improvements in order to encourage construction and widespread ownership of buildings. Distinctions have been made in some cases between urban and rural land, but no tax limitation law has distinguished between land and improvements, chiefly, perhaps, because such laws have been sponsored largely by farm organizations and by organizations of real estate dealers and large landowners.

[page 153]

Chapter 11


This chapter deals with the control of business organization and practices through taxation. In contrast with the preceding chapter, no account is taken of the kind of product or service.


Under a strict laissez faire policy a government avoids encouraging large units as against small ones, or vice versa. No sentiment has developed in the United States in favor of abandoning this policy because it fosters unduly small business units. A fear has grown, however, that it permits business units to become too large. It is widely believed that through their economic power large concerns can drive smaller ones out of business even when the latter are really more efficient. The recently passed anti-price- discrimination law amending the federal anti-trust laws /1/ appears to reflect this belief. Also it is thought that an undesirable and even dangerous degree of economic, social, and political control is concentrated in large business units.

Taxes Favoring Small Units

Taxes relating to chain stores and holding companies, and certain other taxes discussed below, have an incidental bearing on bigness in industry. Other tax measures, however, have been directly aimed at throwing

/1/ The so-called Robinson-Patman Act, 15 U.S.C.A. secs. 13 et seq. (Supp. 1936). See also "The Legality of Price Discrimination under the Robinson-Patman Act," Columbia Law Review, XXXVI (December 1936), 1285-1321.

[page 154]

the balance of competitive advantage in favor of small units or
preventing its being thrown in favor of large units.

For example, the taxes imposed on net incomes of corporations by the federal government and by six states are at progressive rates. /2/ In three of the states the progression stops at the $6,000 income level. Only in South Dakota does it extend into brackets above $15,000. The maximum rate of 8 per cent in that state starts at $318,000. The federal rates under the revenue acts of 1935 and 1936 are progressive up to $40,000, the lowest rate under the 1936 act being 8 per cent and the highest rate 15 per cent. Prior to the Revenue Act of 1932 a credit or exemption of $3,000 was allowed to corporations with incomes of $25,000 or less.

The federal tax levied on undistributed corporate earnings /3/ allows a minimum of $5,000 of retained earnings to be taxed at the lowest rate, which may give as much as $675 advantage to corporations with small earnings.

Gross receipts taxes, although usually imposed at flat rates, occasionally are imposed at progressive rates. For example, the gross receipts taxes imposed on telephone companies in Wisconsin are graduated from 2-1/2% per cent to 8 per cent, according to the volume of receipts. /4/ In Louisiana a retailers' license tax is imposed in such a manner as to constitute a progressive rate on sales. All progressive taxes on sales or gross receipts appear to be seriously in danger of being declared unconstitutional since the United States Supreme Court held the Kentucky progressive gross sales tax unconstitutional. /5/

/2/ Based on Tax Systems of the World, pp. 128-29.

/3/ See pp. 14, 16.

/4/ Wisconsin Statutes, 1929, sec. 76.38.

/5/ Stewart Dry Goods Co. v. Lewis, 294 U.S. 550 (1935), discussed on pp. 183, 186-87.

[page 155]

Taxes Favoring Large Units

Some tax measures favor the large over the small unit. It has long been recognized that the administration of the general property tax commonly discriminates against small owners, since, according to most studies, the percentage of assessed to full value is higher for small properties than for large properties. /6/

Taxes on corporations at the time of their organization or their entrance into the state frequently favor large corporations. The taxes are commonly based on invested capital, and the discrimination results from four features often present: namely, (1) the fixing of a minimum tax, (2) the setting of brackets within which a flat tax is imposed, (3) regressive tax rates, and (4) in a few states, an upper limit to the amount of the tax. In a very few cases a flat amount is charged, which gives the maximum advantage to the large corporation.

Annual taxes on capital stock are usually imposed at flat rates, but some have the regressive features that are found in the corporation organization and entrance taxes.

Local license taxes are often levied at flat amounts, regardless of the size of the business, or at rates that are regressive when compared with any measure of business size.


Prior to the nineteenth century the association of persons in a corporation was a privilege granted by a government only under very special circumstances. During the past century, however, owing to the growth of business needs and the competition of states for revenue, incorporation with almost unlimited powers has been

/6/ See Silverherz, op. cit., pp. 211-12.

[page 156]

made practically a common right, although at law it is still a
privilege granted by government.

This development has resulted in two opposing attitudes toward the corporation. /7/ One is that since the corporation has been accepted as a common right the state should not encourage either incorporated business or unincorporated business at the expense of the other. They should not, for example, be subject to different tax burdens. The other attitude is that since corporate organization is a special privilege it should be taxed as such. The purpose of this chapter is not to decide what, if any, type of business organization should bear a heavier tax burden. Rather, it is to find out whether tax discrimination exists that tends to encourage or discourage the use of some particular form of business organization.

Tax discrimination exists if the investor bears, either directly or through a tax on business, a higher tax on income derived through one form of organization than through another. Some types of taxation do not affect the form of business organization. Taxes on real estate, tangible personalty, and sales, for example, are imposed regardless of whether the business is incorporated. Taxes on income, profits, and capital stock, on the contrary, have generally been designed to rest more heavily on incorporated than on unincorporated businesses.

Special Taxes on Corporations

Certain taxes are imposed on corporations but not on unincorporated businesses. The chief examples at present are (1) state taxes imposed at the time of organization or entrance into business in the state; (2) annual state taxes on the capital stock or the value of corporate shares in excess of the assessment of corporate property; (3) the combination federal capital stock and excess

/7/ See pp. 259-60.

[page 157]

profits tax; /8/ (4) federal and state income taxes; and (5) the
federal tax on undistributed profits. /9/ The organization, capital
stock, excess profits, and undistributed profits taxes are clearly
special burdens resting on corporate investors. The income tax
presents a more complex problem.

The federal government and many of the states impose on corporations income taxes that are not imposed on unincorporated businesses. Only in New York State has a special net income tax been imposed on unincorporated business in addition to the personal income tax. The New York tax is an emergency tax, applying only to the years 1935 and 1936. A $5,000 exemption is allowed, and the rate is 4 per cent, as compared with a corporation rate during the same years of 6 per cent. /10/ Federal and state individual income taxes require a partner or an individual proprietor to include in his personal return his entire share of the profits of the business. This requirement stands, regardless of whether any or all of the profits are kept in the business instead of being taken out for the taxpayer's personal use.

Co-ordination of Corporation and Personal Income
Taxes: Earnings Paid Out

The fact that income taxes are imposed on corporations and not on unincorporated businesses does not necessarily mean that the income tax system as a whole -- corporate tax plus individual tax -- discriminates against corporations. For example, dividends paid by a corporation may be exempted from taxation in the hands of the individual stockholders. Moreover, discrimination depends on the rates of the tax on individual incomes and the amount of the investor's total income. Another im-

/8/ For a description of this tax, see Note 9, Present Excess Profits Tax, p. 560.

/9/ For a description of this tax, see pp. 14, 16, and Note 4, Details of Undistributed Profits Tax, p. 513.

/10/ New York Tax Laws, arts. 9-A, 16-A.

[page 158]

portant factor is the proportion of profits that is paid out
currently in dividends instead of being retained in the business. For
the moment, however, it will be assumed that none of the profits are

A number of state income tax laws provide that dividends received from corporations subject to the state's income tax are exempt from individual income tax. This provision lessens tax discrimination between incorporated and unincorporated businesses but cannot eliminate it entirely, except in chance cases. Some stockholders receive so little income as to be exempt from the individual tax. The tax paid by the corporation, however, lessens their dividends or decreases the book value of their shares. Thus they bear a tax that they would not have had to bear if their funds had been invested in an unincorporated business. Furthermore, the losses of the incorporated business cannot be offset against the non- business income of the individual or his income from some other business enterprise. The losses from unincorporated businesses can, on the contrary, be offset in this manner.

A few states give a tax credit for the corporation tax paid, instead of exempting the dividends. In these states somewhat the same considerations apply. Discrimination against the corporate form results also from the fact that a state does not, generally, make any allowance for corporate income taxes paid to other states or for dividends received from corporations that are taxed by other states but not by the state in question. /11/

The existence of these sources of discrimination must be borne in mind throughout the following discussion.

Co-ordination When Rates Are Both Equal and Flat

When dividends are exempt from the individual income tax, the effect on tax discrimination depends largely

/11/ See the memorandum by Carl Shoup, Bernard L. Shimberg, and William Vickrey on the aggregate burden of federal and state income taxation, in the forthcoming volume, Studies in Current Tax Problems.

[page 159]

on the rates of the corporation and the individual taxes. If the
individual rate and the corporation rate are the same and are flat,
not progressive, the tax on the corporation amounts merely to a tax
imposed and collected at the source of the income. Then relatively
little discrimination between incorporated and unincorporated
business results.

A somewhat different situation was presented under the 1913 and 1916 federal income tax acts. They exempted corporation dividends from the individual normal tax but not from the individual surtax. The normal rates on individuals and corporations were the same. Exemption of dividends from the normal individual tax almost equalized the burdens on corporate and non-corporate business if all earnings were paid out when received. The burden was not quite equalized, however. For corporation earnings, the individual surtax was applied to the dividend after payment of the corporation tax, while for unincorporated business earnings, the surtax was applied to the income before the deduction of the individual tax. The advantage was thus slightly in favor of incorporated business -- except where, as previously explained, the personal exemption of the individual from the normal tax or the offsetting of business losses against other income was sufficiently important to wipe out this advantage or replace it by an advantage in favor of unincorporated business.

Co-ordination When Rates Are Unequal or Progressive

If the individual rates are lower than the corporation rates, the burden is heavier on incorporated business. For example, since 1919 the federal corporation rates have been higher than the individual normal rates. The greatest discrepancy was in the year 1935, when the highest normal rate was 4 per cent and the corporation rate was graduated from 12-1/2 per cent to 15 per cent. When all earnings were promptly distributed as dividends, the

[page 160]

excess of the corporation tax rate over the individual normal rate
was disadvantageous to incorporated business, since stockholders
receiving corporate dividends were subject to a higher total of taxes
than were investors in unincorporated businesses. The Revenue Act of
1936 repealed the exemption of dividends from the normal tax and to
that extent increased the discrepancy. The increase was, however,
slightly offset by the reduction of the corporation rates under
$40,000. The undistributed profits tax passed in the same act is not
a factor so long as all earnings are distributed.

Rates on individual incomes might be progressive, with some rates lower and others higher than those on corporation incomes. If dividends were exempt in the hands of individuals and earnings were all distributed, the relative advantages of incorporated and unincorporated businesses would vary for different persons. Persons with large incomes would be taxed more heavily on incomes from unincorporated business. Persons with small incomes would be taxed more heavily on incomes from incorporated business.

Co-ordination of Corporation and Personal Income
Taxes: Earnings Retained

In the preceding paragraphs it has been assumed that all earnings are promptly paid out as dividends. If the earnings are retained by the corporation, however, and no tax is levied on undistributed profits, the situation is quite different.

Two cases must be distinguished. In the first case, the total income of the investor is so small that the personal tax on his earnings from unincorporated business is less than the corporation income tax on an equal amount of business income. Here, the income tax system clearly favors the unincorporated form. Even without a later taxation of income derived from the corporate

[page 161]

shares, the tax on corporation income is larger than the tax on
income from unincorporated business.

In the second case, the total income of the investor is so large that his earnings from unincorporated business are taxed at a higher rate than the corporation income tax on an equal amount of business income. Here, the tax advantage may shift from the unincorporated form to the corporate, depending on how the retention of the earnings affects the taxes of later years.

The possibility of saving taxes through the retention of earnings by the corporation exists because a corporation is a legal entity distinct from its stockholders. By the Supreme Court's interpretation of the Sixteenth Amendment, the earnings of the corporation do not become taxable income of the stockholders until distributed to them or until "realized" in the form of a capital gain by the sale of the stock. /12/ Thus, if corporate profits are allowed to accumulate as surplus, the stockholder avoids individual income taxation. Where his total income is sufficient to make him subject to high taxation, he is placed under great pressure to urge the retention of earnings.

Postponement versus Permanent Avoidance

It may be objected that the withholding of dividends merely postpones the payment of the tax and does not really avoid it. This objection may or may not be true. Four cases must be distinguished.

In the first case, earnings are withheld and reinvested by the corporation for one or more years, and paid later as dividends. In this case, even if the same tax is paid, it is paid at a later date. The individual stockholder, through the corporation, has the use of the amount of the tax as an earning asset for the period during which

/12/ Lynch v. Hornby, 247 U.S. 339, 343 (1918); Lynch v. Turrish, 247 U.S. 221, 229 (1918); Eisner v. Macomber, 252 U.S. 189, 219 (1920).

[page 162]

the tax is delayed. He gains the amount of the interest or earnings
on this asset during the period.

At the present-day levels of individual surtaxes, this advantage is of genuine importance. For example, suppose a person who has income subject to the federal tax in the highest bracket -- 79 per cent -- is faced with the alternative of (1) receiving $100 as a dividend, paying the tax of $79, and investing the remaining $21 so that it will earn 5 per cent after all taxes except the individual tax, or (2) leaving the $100 in the corporation for one year and then receiving it plus a 5 per cent return after corporation tax. At the end of the year, after paying his individual tax, he has in the first case $21.22, or net earnings of $0.22. In the second case, he has $22.05, or net earnings of $1.05. As a result of postponing dividends his earnings are nearly five times as much. If the highest bracket of his income tax were 50 per cent, his net earnings would be twice as great when dividends were postponed as when they were paid immediately and invested.

Aside from the interest factor, the stockholder also gains if the dividends are distributed in a year when his total income is smaller and so subject to lower surtax rates. He loses, however, to the extent that the earnings, distributed in a lump sum, force the tax base into higher brackets than an equal sum distributed in annual installment payments. He gains, again, if the dividends fall in a year when his tax return would otherwise show a loss. Any change in statutory tax rates, of course, affects him also. These factors, however, are largely matters of chance unless the stockholder is in control of the corporation.

In the second case, the earnings are permanently withheld and reinvested by the corporation, but the investor sells his stock. In this case, the reinvestment of earnings normally results in a somewhat proportional increase in

[page 163]

the value of the stock. When the investor sells the stock, he is
subject to taxation on his capital gain. Aside from the interest and
surtax factors, the amount of the tax is affected by the provision
for taxing only part of the gains, from 80 per cent to 30 per cent,
depending on how long the securities have been held. /13/

In the third case, the earnings are withheld and reinvested by the corporation and then lost through business reverses. No tax is collected from the individual since he never received the income. If he had received a dividend, he would have had to pay a tax even if the dividend was subsequently reinvested and lost, but in this case the tax would have been somewhat offset by the reduction in whatever capital gain (or increase in whatever capital loss) he might have realized from sale of the stock.

In the fourth case, the earnings are withheld permanently, and the stockholder does not sell the stock before his death. The tax is avoided entirely since neither he nor his heirs pay an income tax on the increase in the value of the stock prior to the time of his death. Much of what would have gone to pay income taxes goes to pay estate and inheritance taxes, but a substantial amount of avoidance is possible.

In view of the possibilities of tax avoidance in these cases, persons with large incomes would have reason to prefer investment in corporations to investment in unincorporated concerns, provided the dividend policies were subject to their control or were controlled in their interest. This inducement to incorporate depends not so much on the type of business and its size -- considerations that seem most important from the point of view of efficient operation -- as on the total income of the persons who happen to be engaged in the undertaking.

The 1936 tax on undistributed corporation earnings

/13/ See pp. 80-81 for details of this tax.

[page 164]

greatly reduces the advantages of incorporation to the wealthy
stockholder if no taxable dividends are paid. The tax amounts at a
maximum to 20.5 per cent of the taxable income of the corporation
less the corporation normal tax. This, of course, reduces either the
amount of dividends finally paid or the market value of the stock. As
a result, the income of the stockholder with a small total income is
subject to much higher taxation than if it had come from an
unincorporated business. The advantage of the corporation to
stockholders with large incomes becomes less, and the line between
loss and gain is drawn at a much higher income level.

Methods of Reducing Inequalities

Taxation of Distributable Shares

As long as the progressive individual income tax is retained, only one way of eliminating the tax discrimination between incorporated and unincorporated business for every investor appears feasible. The stockholder's share of the undistributed corporation profits must be included in his taxable income while his share of corporation losses must be allowed as a deduction from his taxable income, and, in addition, either corporate business as such must be exempted from income taxation or unincorporated business must be taxed at the same rates as incorporated business. By this method the income from both corporate and non-corporate business is taxed to the individual whether it is distributed or not.

Constitutional difficulties, however, prevent the general application of this method to federal taxation in view of the Supreme Court's interpretation of the Sixteenth Amendment, whereby the earnings of the corporation do not become taxable income of the stockholders until distributed to them as dividends or "realized" as capital gain by the sale of stock.

[page 165]

If the stockholder were required to state in his return his share of the corporation's earnings for the year, the administrative difficulties would be great. The accounting complexities of tracing undistributed income and loss through intricate holding company arrangements to final individual stockholders would be extremely costly. The existence of non-cumulative preferred stock would make the determination of distributive shares difficult because the ability of the boards of directors to determine the time of paying dividends would affect the allocation of profits between preferred and common stockholders, although the tax law might prescribe rules for determining the ownership of the income. If the corporation's statement of income were not accepted by the tax authorities, final settlement might necessitate the reopening of a multitude of individual returns.

The undistributed profits might be reached instead by taxing all capital gain accrued, even if unrealized, during the year. The individual's net income would then reflect, not only retained corporate earnings, but the many other factors that affect the amount of capital gain accrued during a year. The share in the year's profits would not appear as a distinct item in the personal return. /14/

In either case, stockholders would have to pay the tax on income not received in cash, with a resulting financial strain that would play into the hands of majority stockholders who wished to force the sale of securities held by a minority.

Forced Distribution of Corporate Earnings

Another method would force the payment of all earnings to stockholders in the form of taxable dividends. It would have to be coupled, of course, with a repeal of the corporation income tax or with equal taxation of both kinds of business.

/14/ For a loophole that would have to he closed, see p. 302; see also pp. 477-83.

[page 166]

This method would come much nearer to equalizing corporate and non-corporate taxation for each investor than the existing federal income tax. However, equalization would not always be obtained, even under this plan. Losses of a corporation could not be passed on to the stockholder in the form of "negative dividends" deductible from his other income in computing his personal income tax. Incidentally, outside the field of tax policy, the government would be discriminating between corporations and partnerships in the control that it exercised over the distribution of business earnings in general.

Equivalent Tax on Undistributed Earnings

Still another method would make the relative attractiveness of corporate and non-corporate business approximately equal for a composite of investors as distinguished from individual investors. It would involve repeal of the corporation income tax and the imposition of a tax on the undistributed profits of the corporation -- not, however, at a rate heavy enough to force all earnings out as dividends. Theoretically, the rate would simply be set at the point where the total taxes on undistributed profits would equal the tax that would be levied on them in the hands of individual stockholders were they distributed.

The difficulty with this method is that, while the attractiveness of the two types of business might thus be equalized for investors in general, it would not be equalized for individual investors. Some investors would find the taxes imposed on income from unincorporated businesses much higher, while others would find taxes on income from corporations much higher. Incidentally, corporations with different dividend policies would no doubt attract stockholders of different income levels.

[page 167]

Allowances to Unincorporated Business

A fourth method that has been suggested for equalizing corporate and non-corporate tax burdens is to place the earnings of unincorporated businesses on the same basis as the earnings of incorporated businesses. The individual would be taxed only on that part of the proprietorship or partnership earnings that was actually taken out of the business by him. /15/

This provision would have to be accompanied by an unincorporated business tax equal to the tax on corporations. Many difficulties, however, would ensue. To be strictly comparable, the losses of partnerships and of proprietorships would not properly be deductible from income outside the specific enterprise. Furthermore, the facility with which the distribution of proprietorship and partnership profits could be controlled by the owners would not only raise very serious administrative problems but would in itself constitute a substantial tax advantage to unincorporated business. Serious discrimination between business income and other income would result from exempting the reinvestment of business savings.

Furthermore, the desirability of encouraging the reinvestment of income within the business unit is dubious. This problem has been acute in the case of corporations; /16/ the suggested policy would extend it to all business.

Finally, the postponment or avoidance of the tax has certain questionable features from the point of view of a just distribution of the tax burden, even assuming that all forms of business profits are given the same chance of postponement or avoidance. The points at issue -- irregularity of income, increases in value of property held until death, and the failure of a business tax to take into ac-

/15/ Harley L. Lutz, "Coordination of the Personal Income Tax and the Business Tax under the Model Plan for State and Local Taxation," Bulletin of the National Tax Association, xx, No. 3 (December 1934), 86-88.

/16/ See pp. 168-74.

[page 168]

count the personal situations of the business owners -- are discussed
elsewhere. /17/


The dividend policies of corporations are matters of public interest. They affect the amount and direction of tax preference between incorporated and unincorporated businesses. /18/ They also influence the effectiveness of the progressive rates of the individual income tax, since the withholding of earnings nullifies the progressive rates or greatly alters their effect. Also, the amount of revenue derived from the income tax is affected.

Economic Consequences of Dividend Policies

Furthermore, dividend policies have important economic effects outside the immediate tax field. The recent passage of the federal tax on undistributed corporation profits has provoked a great deal of controversy concerning those effects. Most of the arguments lack factual support and involve highly theoretical points of a fundamental nature. Only a brief and incomplete statement indicating the character of the controversy can be given here.

On the one hand it is urged that if all earnings were promptly distributed as cash dividends corporations would be prevented from expanding naturally. This point would apply particularly to small corporations since they have less ready access to organized capital markets. Therefore, it is claimed, not only would industrial growth slacken but also competition would be weakened. Monopoly would be encouraged, since small corporations would have greater difficulty in growing large enough to compete effectively. Furthermore, it is urged that without the accumulation of reserves during

/17/ See pp. 303, 302, and 306-7, respectively.

/18/ See pp. 157-68.

[page 169]

years of good earnings corporations would find it impossible to pay
interest and dividends during periods of depression. Thus
bankruptcies would be increased and depressions intensified.

In answer to the last argument some have maintained that corporate weakness in depression is socially desirable. It would result in quicker price-cutting and liquidation, which are said to be prerequisites of recovery.

Those who favor distribution of dividends claim that withholding them results in two kinds of misapplication of funds. In the first place the corporation does the investing without consulting the individual stockholders, who would in at least some cases spend the dividend instead of investing it. In the second place the corporate directors are likely to reinvest the funds in the firm's business, while, if the stockholders made the investment on their own initiative, they would at least in part invest in other establishments and industries that would promise larger returns. These two types of misapplication, it is urged, result in socially uneconomical investment, with a tendency to overdevelop certain lines of production. Industrial maladjustments and depression result.

Furthermore, it is asserted that corporations do not immediately reinvest all the withheld earnings in plant during depression and recovery years but hold part of it in liquid form and invest it during boom years. This concentration of investment in one period of the business cycle, it is maintained, adds to the intensity of booms and to the severity of the following depressions.

The payment of dividends in shares of stock instead of cash would, so far as these arguments are concerned, be the equivalent of retaining the earnings.

It will be observed that many of the arguments on both sides involve conflicting theories of the causes and

[page 170]

remedies of business depressions. This is not the place to
investigate either the fundamental issues involved in such theories
or the results or desirability of dividend policies. Rather, the
purpose of this section is to examine the effects of tax laws on the
choice between paying dividends and keeping the earnings in the

Influence of Individual Income Tax on
Dividend Payments

A previous section discussed the savings in individual income taxes that might result from the retention by a corporation of its earnings. These possible savings, of course, exert a pressure on directors to withhold earnings.

The strength of the pressure to withhold earnings has undoubtedly differed in different corporations. At one extreme is the fairly closely held corporation, where a few large stockholders can adjust dividend policies to achieve their maximum personal benefit. At the other extreme is the large corporation with such a wide stock distribution that no wealthy stockholders are in a position to dominate dividend policy. Even in the latter case, however, stockholders who faced the liability of paying high individual surtaxes would press less actively for dividends than they otherwise would. In the absence of pressure to pay dividends, pressure to retain earnings, which is continuously present, would have more effect.

The individual income tax is thus a powerful influence for withholding corporation earnings from the stockholders. Moreover, it increases in strength as the individual tax rates increase.

Measures to Encourage Dividend Payments

Because of the resulting reduction in revenues -- and in some instances for other reasons as well --, attempts have been made in the income tax laws to offset the pressure to withhold earnings. Indeed, some of these attempts

[page 171]

go farther and aim at forcing more dividends than would otherwise be

Taxation to Individuals of Undistributed Corporate Earnings

The influence of taxation on dividend payments could be avoided by including in individual taxable income the entire "distributive" share of corporate earnings -- that is, not only the amount distributed but also the amount undistributed. This method was used to some extent until it was prevented by adverse Court decisions. /19/ The federal income tax imposed during the Civil War (act of 1864) provided for such an inclusion. The federal acts of 1913, 1916, and 1917 provided that, in case of the fraudulent use of a corporation for the purpose of evading the tax on individuals, the distributive share of the corporation's income should be taxed to the stockholders. /20/

Beginning with the 1918 act, it has been provided that stockholders might choose such taxation in lieu of taxes imposed as penalties on corporations deemed to be accumulating profits to prevent imposition of the surtax, and more recently in lieu of special taxes on personal holding companies. /21/ In the British income tax law, the administrators of the tax may require the inclusion of undistributed income in individual returns subject to surtax. /22/

Taxation of Undistributed Profits in Special Cases

Attempts have been made to stimulate dividend payments by imposing on the corporation a tax that is measured by the amount of undistributed profits, and that is sufficiently heavy to make distribution preferable to payment of the tax.

/19/ See p. 161.

/20/ William KixMiller and Arnold R. R. Baar, Consolidated United States Income Tax Laws (Chicago, 1923), p. 419.

/21/ Ibid.; Revenue Act of 1936, secs. 101, 351.

/22/ Andre Bernard, Income Tax in Great Britain, Including a Description of Other Inland Revenue Taxes, 70 Cong., 1 sess., House document no. 332.

[page 172]

The application of such a tax in certain cases began with the Revenue Act of 1918 and has continued ever since. Heavy taxes are provided for any corporation "formed or availed of for the purpose of preventing the imposition of the surtax upon its shareholders or the shareholders of any other corporation, through the medium of permitting earnings or profits to accumulate instead of being divided or distributed." The fact that a "corporation is a mere holding or investment company, or that the earnings or profits are permitted to accumulate beyond the reasonable needs of the business" is prima facie evidence of such use. /23/ Still heavier taxes are imposed on "personal holding companies" of not more than five persons. /24/ In both cases, as already noted, the special taxes do not apply if all stockholders include their entire distributive shares of corporation earnings in their individual returns.

The tax on personal holding companies is capable of fairly precise application. However, the tax imposed on corporations used to prevent imposition of the surtax is extremely difficult to administer. Intent is hard to prove, and the point at which the accumulation of earnings and profits is beyond the reasonable needs of the business is difficult to determine. The management of a corporation can almost always point to some need, present or future, certain or contingent, that makes the accumulation of earnings desirable. As a result, little success has attended efforts to enforce the special tax on corporations used to avoid individual surtaxes. Recent decisions of the Board of Tax Appeals upholding Treasury Department assessments, however, may point to greater success in future enforcement.

General Taxation of Undistributed Profits

The extension of the undistributed profits tax to all corporations, except a few specifically exempt, was

/23/ Revenue Act of 1936, sec. 102.

/24/ Ibid., sec. 351.

[page 173]

adopted in the Revenue Act of 1936. As with the taxes of narrower
scope discussed above, the purpose is to force distribution of all,
or nearly all, the earnings.

The imposition of a general tax on undistributed earnings has been under consideration at least since 1917. The revenue act of that year imposed a tax of 10 per cent upon that portion of the undistributed profits that was not actually invested and employed in the business, retained for employment in the reasonable requirements of the business, or invested in obligations of the United States issued after September 1, 1917. /25/ The tax thus applied to uninvested and unneeded profits rather than to all undistributed profits. It was not included in the Revenue Act of 1918 or in those of the following years, although, as indicated above, a special, punitive tax was imposed on profits withheld to evade the surtax.

An effort was made to substitute a tax on undistributed corporate income for the corporation income tax in 1921, but without success. /26/ In 1924 the Senate amended the revenue bill to include taxes on undistributed corporate profits, the rates ranging up to 40 per cent. The proposal was, however, lost in conference. /27/

A committee of the National Tax Association on the simplification of the federal income tax, reporting in 1927, proposed a flat-rate tax on undistributed corporation income, to be "considered a premium tax paid by the corporation for its stockholders, in exchange for retaining the earnings in the business and thereby postponing the normal and surtax until a future day." /28/ The rate suggested was 10 per cent. The Joint Committee on Internal Revenue, reporting in 1927, did not propose a tax on undistributed corporation earnings but pro-

/25/ Revenue Act of 1917, sec. 1206-(2), quoted in KixMiller and Baar, op. cit., pp. 1029-30.

/26/ Congressional Record, 67 Cong., 1 sess., October 17, 1921, pp. 6833, 6865.

/27/ Congressional Record, 68 Cong., 1 sess., May 7, 1924, pp. 8031-33.

/28/ "Simplification of the Income Tax: Report of Committee of the National Tax Association," Proceedings of the . . . National Tax Association, 1927, p. 134.

[page 174]

posed to allow the corporation, in computing its net income, a
deduction of a fraction of the dividends paid, with certain
limitations. /29/ Apparently the tax was not again considered by
Congress until recommended by the President in his message of March
3, 1936.

The Revenue Act of 1936

The Revenue Act of 1936 contained two provisions bearing on the dividend policies of corporations. One eliminated the exemption of dividends from the normal tax rate of 4 per cent. The tendency of this action is to discourage the payment of dividends.

The other provision taxed undistributed corporate earnings. This tax starts with a 7 per cent rate for an amount of undistributed net income not in excess of 10 per cent of the adjusted net income and has a maximum rate of 27 per cent for the amount of undistributed net income in excess of 60 per cent of the adjusted net income. /30/

This measure tends to encourage the payment of earnings to stockholders in the form of taxable dividends. However, it does not counteract in a precise manner the effect exercised by the individual income tax in encouraging the withholding of dividends. If the corporation is controlled by persons who have small total incomes, the tax on undistributed profits may be much heavier than necessary merely to counteract the pressure of the individual income tax. It may therefore induce the distribution of more dividends than would be distributed if no income tax at all existed. If the corporation is controlled by persons whose total incomes are large and who pay surtaxes in the higher brackets, the tax on undistributed corporation profits may not entirely counteract the pressure of the individual income tax for the retention of profits.

/20/ Report of the Joint Committee on Internal Revenue Taxation (1928), 1, 55-56.

/30/ See pp. 14, 16, for further details.

[page 175]

Possible Dividend Policies

With the tax on undistributed corporation earnings in effect, a corporation may adopt any of several possible policies in regard to its earnings.

It may pay the tax and not distribute the dividends. This would burden all common stockholders proportionately. The contrasting effects on stockholders with small total incomes and those with large total incomes have already been indicated. The effect on the corporation would be to reduce the amount available for the payment of debt, expansion, or other purposes by the amount of the tax.

A second course of action would be to pay the earnings in the form of cash dividends. This would minimize the tax for the stockholder of small total income, but it might make the effective tax higher for the stockholder of large total income. The corporation would be prevented from reinvesting the earnings without definite action by the stockholders themselves. Large public corporations, with widely distributed stock, in good financial position, and with no particular need for expansion, might therefore be expected to adopt this second course of action.

If the corporation needed to retain its cash and yet desired to avoid the undistributed profits tax, it might, as a third possibility, offer the right to buy more stock at less than market price when it paid cash dividends, or, fourthly, it might offer dividends in cash or stock at the option of the stockholder. A fifth possibility arises from the language of the act and the interpretations of the Treasury Department that allow any distribution of stock dividends that is taxable to individual stockholders to be used as a method of dividend payment to avoid the tax on undistributed profits. /31/

/31/ Treasury Decision No. 4674, Approved August 6, 1936.

[page 176]

In general, the rule of the Supreme Court has been that, if a
dividend to stockholders is to be taxable, it must change the
relative interests of the stockholders. /32/ Such a change
constitutes "realization" of income, which must occur before a tax
can be collected. Apparently, therefore, the payment of preferred
stock to common stockholders, where preferred stock of the same rank
is already outstanding, or of common stock to preferred stockholders,
if such payment is permitted, fulfills the conditions necessary for
taxability. Thus, it is possible to distribute all earnings in the
form of taxable dividends without taking any assets out of the

Certain difficulties, however, stand in the way of using the preferred stock method to pay dividends. The financial structure of the corporation might eventually become top-heavy with preferred stock. The sale of preferred stock to provide funds to pay the individual's income tax might seriously affect the prices of such shares. The value of the preferred shares might be difficult for the tax administrators to determine. Stockholders might be averse to receiving dividends in the form of shares in view of the necessity of selling them to secure funds. Since there is no allowance for the losses of bad years, the distribution of preferred stock for all earnings in good years would gradually result in a loss of equity by the common stockholders. The fact that few corporations made use of the dividend in preferred stock during the closing months of 1936 indicates that such difficulties as these may have a powerful effect in practice, at least during the first few months.


The holding company as a method of business organization is a relatively recent development, dating from 1889. Holding companies are incorporated by the states.

/32/ Koshland v. Helvering, Sup. Ct., May 15, 1936.

[page 177]

Interstate competition in granting corporate charters has prevented the states from exercising any substantial control over holding company organization or operation. The federal government can exercise control only as the affairs of the holding companies relate directly to interstate commerce or to other matters under federal jurisdiction.

Although the value of the holding company as an instrument of corporation finance is unquestioned, doubts have arisen, especially in recent years, about the social desirability of some of the uses to which it has been put. These misgivings have apparently influenced recent federal tax legislation. /33/

Effect of Consolidated Returns

One recent change in federal tax laws relates to the filing of consolidated returns for the corporation income tax. Under the federal income tax laws from 1918 to 1932, closely affiliated corporations were permitted to file consolidated returns without penalty. The advantage of using the consolidated return was that, if one corporate subsidiary was incurring losses while others were making profits, the losses would be deducted from the profits and only the balance of net profit would be subject to tax. In order to come within the class permitted to file consolidated returns, it was necessary that 95 per cent of the stock of each of the corporate units involved be owned directly by one or more of the others and that the common parent corporation own directly at least 95 per cent of the stock of at least one of the other corporations. /34/

Since 1932 the policy in regard to consolidated returns has been reversed. The revenue act of that year imposed an additional tax for the years 1932 and 1933 of 3/4

/33/ See the tax message of President Franklin D. Roosevelt, New York Times, June 20, 1935.

/34/ Revenue Act of 1932, sec. 141(d).

[page 178]

of 1 per cent where consolidated returns were used. /35/ The
additional rate was raised to 1 per cent in 1933, /36/ and to 2 per
cent in 1934. /37/ Even more important, the 1934 act abolished the
privilege of filing consolidated returns for all except railroad
corporations. The 1936 act extended the privilege to street and
interurban railway corporations and wiped out the additional tax.

Taxation of Dividends

The treatment under the federal corporation income tax of dividends received by one corporation from another has also been changed. The income tax law of 1913, which imposed a tax of 1 per cent on corporation income, did not exempt dividends received by a corporation. Holding company systems thus paid a tax for each company in the chain. Beginning with the war income tax of 1917 /39/ and the Revenue Act of 1918, /40/ however, dividends received by corporations were fully deductible. The Revenue Act of 1935 reduced the deduction from 100 per cent to 90 per cent, /41/ and the Revenue Act of 1936 further reduced the deduction to 85 per cent. /42/

Certain circumstances preceding the changes in dividend deductions indicate that they were directed against holding companies. A prominent Senator had, in 1934, proposed eliminating the dividend deduction. /43/ The proposal was made again at a White House conference on January 21, 1935, where methods of attacking holding companies were discussed. /44/ President Roosevelt in his message of June 19, 1935, called for the ultimate elimi-

/35/ Ibid., sec. 141(c).

/36/ National Industrial Recovery Act of 1933, sec. 217(e).

/37/ Revenue Act of 1934, sec. 141.

/38/ Revenue Act of 1936, sec. 141.

/39/ Sec. 4.

/40/ Sec. 234(a)(6).

/41/ Sec. 102(h).

/42/ Sec. 26(b).

/43/ Senator William E. Borah, New York Times, March 2, 1934.

/44/ New York Times, January 23, 1935.

[page 179]

nation of unnecessary holding companies by taxation.

Furthermore, the Revenue Act of 1935 facilitated the simplification of holding company systems. It exempted from recognition under gain or loss provisions any dividends of property received by a corporation in complete liquidation of a corporation controlled by it. /45/

The 1936 change in dividend deductions was made ostensibly to prevent avoidance of the progressive rates of the corporation income tax laws of 1935 and 1936 by discouraging division of the income of a business among several subsidiary corporate units. However, substantial avoidance would be unpractical for a large corporation since its business would have to be divided among numerous small subsidiaries. Furthermore, such avoidance could be prevented by taxing subsidiaries at the highest rate of taxation applicable to the whole affiliated group.

A number of holding companies, some of them parts of very large corporation systems, have been eliminated since the passage of the 1935 act. In some cases the decrease in dividend deductions was specifically stated to be a cause of the action. /46/

Federal Capital Stock Tax

The federal capital stock tax /47/ also tends to discourage holding companies. The extent to which this tax is increased by the existence of holding companies depends on the value of the stock declared by the corporation. Whether the declared value of capital stock will be near its true value will depend in part on how much of an excess profits tax is expected to be paid on dividends received. If holding company executives do not expect the fraction of dividends subject to tax to increase beyond the present 15 per cent, they are not likely to declare the value of their stock at nearly so much as its

/45/ Sec. 110. The Revenue Act of 1936, sec. 112, tightened the provisions.

/46/ See Note 1, Recent Eliminations of Holding Companies, p. 547.

/47/ See Note 9, Present Excess Profits Tax, p. 560.

[page 180]

true value. Thus, the effect of the capital stock tax in increasing
the taxes of holding company groups will be reduced. If stocks are
declared at full value and are earning at the rate of 6 per cent on
that value, the capital stock tax amounts to 1-2/3 per cent of net

State Taxes

State taxes on intangibles, capital stock, and incomes also have possibilities for discouraging the use of holding companies. However, the opportunity of incorporating in practically any state and the resultant interstate competition are likely to prevent highly restrictive state taxes from being passed, and from being effective if they are passed.

Tax on Undistributed Corporation Profits

The tax on undistributed corporation profits, imposed in 1936, may conceivably encourage the use of holding companies. For example, if an operating corporation with $1,000,000 earnings after corporation normal tax wishes to distribute $250,000 to its stockholders, it may (1) pay $250,000 directly to the stockholders, or (2) pay part of its earnings to a holding company that in turn pays $250,000 to the stockholders. If the $250,000 is paid directly by the operating corporation to the stockholders, the tax on undistributed profits is $137,500. If $500,000 is first paid to the holding company, which in turn pays $250,000 to the stockholders, the undistributed profits tax on the operating corporation is $75,000, the corporation normal tax on the holding company is $10,090, and the undistributed profits tax on the holding company is $35,633.35. The total tax by this method is $120,723.35, or less by $16,776.65 than when profits are distributed directly to stockholders. However, the holding company would have difficulty proving that it was not being used to avoid taxation, and that it should

[page 181]

not therefore be subject to the especially heavy taxation provided in
the income tax law for such corporations.

Adequacy and Flexibility of Taxation to Discourage
Holding Companies

The previously discussed federal and state taxes apply, at least in their main aspects, to businesses or corporations in general and not especially to holding companies. Taxes might be applied specifically to holding companies as such by any one or more of a variety of methods -- gross receipts, income, capital stock, and so on. It is not clear, however, that such legislation would be constitutional. In so far as it was, the federal government would have the power to destroy the holding company completely by the use of these taxes. A further decrease in the deduction of dividends received by corporations in computing their income taxes would probably be a sufficiently heavy penalty to eliminate most holding companies.

Taxation to discourage holding companies, however, is not sufficiently flexible. It cannot be adapted to discourage only the holding companies that are deemed to be socially undesirable and to spare those that are socially desirable. It would scarcely be constitutional, even if it were administratively feasible, to classify holding companies for taxation according to the control that they exercise over industries and the manner in which that control is exercised.

Indeed, taxation is much more likely to discourage the harmless holding companies than the harmful ones. If, for example, a holding company form of organization is being employed to control a large body of capital by means of a relatively small investment, only the threat of bankruptcy is likely to induce those in control to change their form of organization. The simplification of the corporate structure would make it impossible for

[page 182]

them to continue their control. Consequently, simplification is not
likely to be forced by taxing provisions, such as those of the
federal income tax, that impose a tax of only about $2.25 on each
$100 of dividends received by a holding company. The relatively
unnecessary and socially insignificant corporate units that have been
continued for their minor convenience are the ones first affected.

The relation of the holding company to the final operating company presents an opportunity for discrimination. The pyramiding of two or more holding companies is rarely necessary for operating purposes, and it facilitates manipulation. Accordingly, the corporation income tax might distinguish between the first intercorporate payment of dividends, and the second and subsequent intercorporate payments of dividends. That is, dividends paid to a holding company by the operating company would be exempt or taxed at a lower rate than dividends paid by one holding company to another. Such a tax would strongly discourage the pyramiding of holding companies and would discourage single holding companies only slightly -- if at all.


As chain merchandising organizations have grown in importance in the United States, great political pressure has been brought upon legislatures to increase the ability of independent merchants to compete with the chains. Usually this end has been sought through special state taxation of chains.

Adoption of State Chain Store Taxes

Taxes somewhat like present-day chain store taxes date from 1917. However, the special taxation of chain stores dates from 1929, when three states passed such laws. The most influential was that of Indiana. Under this law a single store under one ownership or manage-

[page 183]

ment was taxed $3.00 annually. If there were from two to five stores
under one ownership or control within the state, the tax was $10.00
on each of these additional stores. The tax per store increased with
additional stores, reaching a maximum rate of $25.00 on the twenty-
first store. /48/ Neither the kind of goods sold, the size of the
stores, the volume of sales, nor net income affected the amount of
the tax.

Another important act was passed by Kentucky in 1930. Strictly speaking, the tax that it imposed was a progressive one on gross sales rather than on chain stores. Sales of one organization were considered as a unit whether or not the organization was a chain. The rates began at 1/20 of 1 per cent of gross sales not exceeding $400,000 and rose to 1 per cent of the excess of sales over $1,000,000.

These two laws served as models for those passed by other states. The Supreme Court of the United States upheld the Indiana law in 1931, /49/ but in 1935 it declared the Kentucky law unconstitutional. /50/

Since the approval of the Indiana law, a rapid spread of chain store taxes, usually of the Indiana type, has occurred, as well as a very considerable increase in the rates at which they are imposed. On January 1, 1937, chain store taxes were in force in twenty states. /51/ The highest maximum rate imposed in any state was $750[.]

Constitutionality of Chain Taxes

In upholding the constitutionality of the Indiana law, the justices of the Supreme Court divided five to four. The majority held that the differences in organization, management, advertising, and other business procedures between chains and other stores were sufficiently pro-

/48/ Tax Research Foundation, Federal and State Tax Systems (3rd ed.; Chicago, 1932), p. 114.

/49/ State Board of Tax Commissioners v. Jackson, 283 U.S. 527 (1931).

/50/ Stewart Dry Goods Co. v. Lewis, 294 U.S. 550 (1935).

/51/ American Retail Federation, Chain Store Taxes (Washington, January 1, 1937).

[page 184]

nounced to justify the special classification and taxation of chains.
The minority maintained that mere difference is not a proper basis of
classification for a revenue tax; that the ground of the difference
must have a reasonable and just relation to the object of the
legislation, and that no such relation was present in this case
because all the differences upon which the majority rested their
decision frequently existed among independents and among chains as
well as between the two types.

The Court divided in the same way in upholding a West Virginia act. /52/ This decision contained the significant statement that it was not a matter of concern to the Court that the tax was so high as to exceed the net profits of some of the stores taxed. So long as the state had the power to tax, the rate of tax was immaterial.

Adequacy of Chain Taxes for Control

Destruction of Chains

The decisions of the Supreme Court open the door to the destruction of chain organizations by state taxation, if state legislatures are so inclined. The tax can be made so high that the chain cannot afford to operate, and accordingly must reorganize as a new system of retailing or retire from the field entirely. In the petroleum industry the taxes in at least one state appear to have had substantially this effect. /53/ It is possible, of course, that the Court may, in the future, reverse itself.

Equalization of Taxation

The chain taxes now in effect are apparently not flexible enough to achieve their less extreme objectives. They do not, for example, equalize the burden of taxation on chains and independents. It has sometimes been claimed,

/52/ Fox v. Standard Oil Co., 294 U.S. 87-99 (1935).

/53/ "Iowa Driving out Chains: Indiana Standard Begins Leasing All Iowa Service Stations," National Petroleum News, XXVII, No. 22 (May 29, 1935), 24-H; "Anti-Chain Tax Helps Victim," Business Week, No. 347 (April 25, 1936), pp. 9-10.

[page 185]

notably in Wisconsin, that the personal property tax on merchants
discriminates against independents in favor of the chains. The
contention is based on three grounds. First, the local assessor
cannot place an accurate value on the property of the chains because
their books are located in the main office of the company, usually in
some distant city. Second, when goods are assessed at cost -- the
usual custom with merchants' stocks --, the inventories submitted by
chains carry the goods at their original cost without including the
costs of transportation or of carrying on the wholesale function,
while independent stores are assessed at the prices at which their
goods were purchased from local wholesalers. Third, the chain has a
more rapid turnover than the competing independent, and accordingly
the tax is less for the same volume of business.

An investigation of specific instances /54/ indicated that these claims are well founded. However, in interpreting tax comparisons, it should be noted that a "typical" chain does not exist. The popular conception of the chain store applies only to a handful of nation- wide, multi-unit organizations, which are the principal object of attack. The many regional, local, and small national chains are often overlooked. In total, however, they have more stores than the large national chains. Furthermore, rates of turnover vary widely among chains and among independents. /55/ Thus, if there is tax inequality between chains and independents, it probably varies widely in amount.

If the chains have a lighter tax burden under the personal property tax, it is owing either to the insufficiency of the personal property tax as a method of business taxation or to inadequacies of administration.

/54/ See F.H. Hardy, "Wisconsin's New Chain Store Tax and Its Relation to Personal Property Taxation," Bulletin of the National Tax Association, XIX, No. 3 (December 1933), 66-72.

/55/ See Reavis Cox, "Inadequacies of Chain Store Taxation," Bulletin of the National Tax Association, XIX, No. 5 (February 1934), 132-41.

[page 186]

It is scarcely conceivable that inadequacies of administration --
highly variable at best -- can be offset with any degree of accuracy
by a special tax imposed without regard to the amount of the personal
property tax or the degree of inefficiency in administration. Such
taxation is simply throwing on an extra tax for good measure to make
sure that the taxes will not be too small.

If equality of tax burden is the objective, a more logical method of achieving it would be to decide what form of business taxation would be fair to both types of merchandising organization and to impose such a tax in lieu of the personal property tax.

Equalization of Competition

Another objective of chain store taxation is the equalization of competition between chains and independent stores. This aim presents equally difficult problems. It is based on the assumption that the peculiar advantages of the chain form of organization can be measured and offset by taxation. They have not been, however, by the taxes imposed so far.

With the exception of the Louisiana statute -- which has not yet been passed on by the United States Supreme Court -- chain store taxes do not take into consideration the total number of stores operated by the chain organization but only the total number within the individual state. A local chain, having few of the advantages accruing to large, national chain organizations, might thus be required to pay a higher rate of taxation on its stores than a national chain that enjoyed all the chain advantages but had relatively few stores within the state.

Furthermore, the rates of progression in chain taxes show no sign of having been based on any careful analyses of the competitive situation. No consideration is given to the nature of the business. Since the decision

[page 187]

on the Kentucky law, /56/ the taxes have rarely differed according to
the size of stores. A variety of different elements determines such
competitive advantages as chain stores may have over independent
stores. To set up the common ownership of multiple retail outlets as
the sole element in determining the tax is to make the equalization
of competition an impossibility. /57/

/56/ Stewart Dry Goods Co. v. Lewis, discussed on p. 183.

/57/ Cox, op. cit., p. 140.

[page 188]

Chapter 12


Although there are a variety of other ways in which taxation has been used for social control, only three will be discussed here -- namely, the control of speculation, the enforcement of industrial codes, and the control of labor relations.


Control of speculation has been an occasional governmental policy. The term speculation is used here in the [illegible] sense of the purchase or sale of anything with the expectation of a profit through later sale or purchase. The profit is thus a capital gain rather than a periodic income while the asset is held. /1/

Since the speculator ordinarily buys for a fairly rapid turnover, even a light tax on transactions will reduce his profits and increase his losses so much that his activity may be discouraged.

Although several tax measures influence speculation, only one of those discussed below seems to have been designed for that purpose.

Taxes on Stock and Commodity Transfers

The federal government imposes a tax on security transfers, amounting to 4 cents on each certificate of $100 par value or less for shares selling at less than $20, and 5 cents where the shares sell at $20 or more. This tax is to be reduced to a flat rate of 2 cents on July 1, 1937.

/1/ For a list of definitions of speculation see Alfred L. Bernheim and Margaret Grant Schneider, eds., The Security Markets (New York, 1935), p. 12.

[page 189]

Five states impose taxes on stock transfers. The most important from the point of view of revenue is the New York State tax, which is levied at rates of 3 cents a share on stock selling at less than $20 a share, and 4 cents a share on stock selling at $20 or more a share. /2/ The maximum tax on a transaction in New York is thus 9 cents a share, or 1/10 of 1 per cent of a share selling at $90. This tax is perhaps too small materially to affect most types of speculation, but substantial increases in the tax would rapidly make certain types of speculation unprofitable.

The federal government also imposes a tax on the sales of produce for future delivery. The tax is 3 cents for each $100 of sale, to be reduced to 1 cent on July 1, 1937. The tax is burdensome upon a special class of speculators, the floor traders, who make "complex turns" in the market within a few hours in the hope of profit of one or two "points." It is probably not particularly burdensome on speculation generally at the [illegible] and prospective rates. It could, however, be [illegible] effective instrument to destroy transactions in commodity futures. It might also be adjusted to fall more heavily on certain types of transactions, leaving others, which are considered more desirable, with a lower burden or none.

Such an adjustment has been made for some types of cotton futures. Under the Cotton Futures Act of 1916 a special tax of 2 cents a pound is imposed upon all sales of cotton futures except those made under three carefully defined types of contract. No revenue has been received under the act, but the result has been to force commodity exchanges to confine themselves to the specified contract types. A similar tax on sales of grain futures was

/2/ McKinney, Consolidated Laws of New York, Annotated: Supplement, 1936, Book 59, Tax Law (Brooklyn), secs. 270, 270-a. On July 1, 1937, the rates become 1-1/2 cents and 2 cents, respectively.

[page 190]

imposed under the Future Trading Act of 1927, /3/ but this was
declared unconstitutional as being part of an unconstitutional
regulatory measure. /4/

Income Taxation of Capital Gains

The capital gains provisions of the federal income tax law also have an effect on speculation. When an asset is sold within one year of purchase, the income or loss is included in the computation of taxable income at 100 per cent; if held between one and two years, 80 per cent; if held between two and five years, 60 per cent; if held between five and ten years, 40 per cent; and if held over ten years, 30 per cent. Anyone who is considering the sale of a security or other commodity will therefore have a differential tax advantage if he holds the security up to ten years.

Since the amount of capital loss deductible in computing net income is limited to capital gains plus $2,000, the tax does not act as an inducement to sell securities in order to create a loss with which to reduce taxable income from other sources. Before this provision was enacted, the law apparently did act as such an inducement. At present the tendency is, rather, to encourage a spread of capital losses over years in which there are also capital gains.

The argument is sometimes made that any tax on capital gains encourages a longer holding of securities. The validity of this argument depends on circumstances. If the security owner has come to the conclusion that the price of a security will decrease, the fact that the must share with the government his gain from selling it should not, if he acts rationally, deter him from selling. If he does not sell at the time and the price goes down, his income, as well as his tax, is lessened when he finally does sell. The psychological dislike of paying

/3/ Act of August 24, 1921, chap. lxxxvi, 42 Stat. 187.

/4/ Hill v. Wallace, 259 U.S. 44 (1922); Trusler v. Crookes, 269 U.S. 475 (1926).

[page 191]

taxes may of course overbalance his rational judgment in such a

If the future price of securities is uncertain, however, he may be deterred from selling by another consideration. If he sells the security, takes his [illegible], and pays the tax, he will have a smaller capital than if he does not sell. In effect, he has an opportunity to earn a return on the tax as long as payment is withheld. The importance of this factor is dependent on the length of time involved and the degree of uncertainty about future prices. /5/

The same considerations apply in the case of real estate so far as the income tax is concerned.

Taxes on Real Estate Transfers

In most sales tax laws, the sale of real estate has been exempted. The federal government imposes a small tax on real estate transfers. The gross income tax law of Indiana imposes a 1 per cent tax on the gross receipts from real estate sales. /6/


In August 1935, Congress passed the Bituminous Coal Conservation Act, /7/ which set up a code of fair practices for the regulation of the industry. To enforce the code, an excise tax of 15 per cent of the sale price of coal at the mine was imposed, with provision for a drawback of 90 per cent of the tax to coal producers who accepted, and acted in compliance with, the code. The Supreme Court declared this act unconstitutional, /8/ holding the tax to be a penalty to enforce compliance with regulations that Congress did not have power to impose.

/5/ For a more thorough analysis of the effects of taxation of capital gains the reader is referred to a forthcoming study on this subject by Robert M. Haig. With respect to the time element, see p. 162.

/6/ Act of 1933, chap. 1.

/7/ Act of August 30, 1935, chap. dcccxxiv, 49 Stat. 991.

/8/ Carter v. Carter Coal Co., 298 U.S. 238 (1936).

[page 192]

Presumably, if the Court had found that Congress had the power to
regulate in this field, a tax useful in enforcing the regulation
would have been upheld.


Under the Constitution the federal government is denied the power to control labor relations or working conditions -- except as they may relate to interstate commerce. In the absence of direct power, however, taxation has been used on a few occasions to achieve this end. The infrequency of the use is probably due to a number of discouraging Court decisions.

States have made little if any use of taxation to control labor relations, probably because they have other powers that can be used more effectively.

Taxes to Control Employment Conditions

Persons handling white phosphorus, as in match manufacture, are often afflicted with an occupational disease in which the jaw of the victim is eaten away. Certain substitutes for white phosphorus have been discovered that do not involve the health hazard but that are somewhat more expensive to use. A voluntary effort to eliminate the use of white phosphorus by American match manufacturers failed because of the refusal of 5 per cent of the industry to agree. /9/ On recommendation of President Taft, /10/ Congress imposed a prohibitive tax of 2 cents per 100 matches on white phosphorus matches in 1912 and forbade their exportation. /11/ The act was not litigated.

An attempt to eliminate child labor by means of taxation was blocked by the Supreme Court. Congress

/9/ J.R. Commons and J.B. Andrews, Principles of Labor Legislation (rev. ed.; New York and London, 1927), p. 357.

/10/ Message of December 6, 1910, quoted in D.D. Lescohier and E. Brandeis, History of Labour in the United States, 1896-1932 (New York, 1918-1935), III, 361.

/11/ Act of April 9, 1912, 37 Stat. 8.

[page 193]

passed a law in 1916 forbidding the shipment of products of child
labor in interstate commerce. /12/ The Court declared this act
unconstitutional on the ground that it was not a valid exercise of
the power of Congress to regulate interstate commerce, and was
instead a veiled method of regulating conditions of manufacture. /13/
Congress passed an act in 1919 imposing a tax of 10 per cent on the
net profits of persons or corporations employing children below
certain specified ages. /14/ The tax was not on the labor or on the
products of the labor but was imposed on all the production of a
manufacturer knowingly employing any child labor. The act contained
rather detailed provisions for administration, which pointed to
regulation rather than revenue as its purpose.

The Supreme Court, through Chief Justice Taft, held the child labor tax act unconstitutional on the grounds that the tax was not a genuine tax but a penalty to enforce a regulation that Congress had no power to impose. /15/ In this case the Court developed the distinction, discussed above, /16/ between a tax and a penalty.

The Bituminous Coal Conservation Act, previously discussed, /17/ also represented a further attempt of the federal government to employ taxation to control labor relations. The code that was contained in the act included provisions that were considered important by the miners' union. The passage of the act averted a nation-wide coal strike. Invalidation of the act by the Supreme Court was in harmony with the decision in the child labor tax case. It further indicated that the Court would not approve taxation to enforce regulations otherwise beyond the power of Congress to impose.

/12/ U.S.C. 432, 64 Cong., 1 Sess.

/13/ Hammer v. Dagenhart, 247 U.S. 251 (1918).

/14/ Act of February 24, 1919, 40 Stat. at L. 1057.

/15/ Bailey v. Drexel Furniture Co., 259 U.S. 20 (1922).

/16/ See p. 137.

/17/ See p. 191.

[page 194]

Social Security Act Tax Credit

The Social Security Act of 1935 included still another tax that was not imposed primarily for revenue purposes. It levied on certain employers an excise tax, measured by wages, to begin at 1 per cent and rise to a maximum of 3 per cent in 1938. Against this tax the employer was permitted to credit contributions paid into approved state unemployment funds up to 90 per cent of the tax. /18/

The Court has not yet passed on the constitutionality of the Social Security Act. Therefore, it is not known whether the provisions are to be viewed as a penalty to force state action in a field outside federal jurisdiction or as a legitimate exercise of the taxing power. The crediting provisions are modeled on those of the Revenue Act of 1924 and later years that allow a credit of state death taxes against the federal estate tax. This device has been upheld by the Supreme Court. /19/ There are, however, significant differences between the two acts. The death tax credit relates to purely revenue measures and makes no attempt to induce states to adopt policies falling under their reserved police powers. The tax imposed by the Social Security Act was obviously imposed to induce states to adopt unemployment compensation laws. Even the type of state law must be approved by the federal government before the tax credit is granted. While the precedent of the inheritance tax case may be followed, the line drawn in the child labor tax case /20/ is at least approached.

/18/ Act of August 14, 1935, chap. dxxxi, pars. 901-3, 49 Stat. 620.

/19/ Florida v. Mellon, 273 U.S. 12 (1927).

/20/ See pp. 192-93.

[page 195]

Chapter 13


Almost every government regulates the consumption of commodities that are deemed socially harmful. Taxation has been used frequently to promote this policy.

Taxation imposed to reduce harmful consumption is sumptuary taxation, and it should be distinguished from taxation imposed to protect or regulate business. The latter may accomplish its aim through reduction in the purchase and consumption of commodities, but the purpose of the reduction is not to prevent harm that might result from consumption.

Sumptuary taxation should be distinguished also from modern luxury taxation. Although luxuries were often thought harmful in the Middle Ages and legislation was passed to discourage their consumption, modern luxury taxes are imposed not for this reason but because the purchase of luxuries is an indication of ability to pay. The primary purpose of luxury taxes is revenue. Any consequent reduction in consumption is incidental rather than intentional.

Few truly sumptuary taxes exist in modern tax systems. The taxes imposed on alcoholic beverages constitute much the best example of sumptuary taxation. Taxes on habit-forming narcotic drugs, alcoholic beverages, and tobacco may all be considered as having sumptuary aspects. The pressure of the appetite for habit-forming drugs, however, is so strong that taxation cannot be considered a very significant aspect of a

[page 196]

system for regulating their consumption. Social disapproval of
cigarettes may possibly have been a factor in the imposition and
retention of especially high rates of cigarette taxation, but it is
now practically extinct.

Taxes are used to regulate consumption in a number of ways. One is to increase the price of the commodity and thus cut down consumption because of the expense. Another is to reduce the number of places where the commodity can be purchased and thus reduce the number of outlets and the resultant temptation to buy. A third is to make domestic manufacture unprofitable and thus to bring the whole industry under federal control through the federal power over imports. A fourth is to use the administrative regulations set up for the collection of the tax as an instrument of control.


Under the commerce clause of the Constitution the federal government has the power to regulate the importation and exportation of narcotic drugs and their sale in interstate commerce. It lacks the power directly to regulate or prohibit their manufacture and intrastate sale. However, a system of federal narcotic control has been set up on a combined basis of the commerce power and the taxing power.

None of the taxes imposed on narcotics is expected to raise the price to consumers so high that they will cease using the drugs on account of expense. Even the internal revenue tax of $300 per pound, imposed on the manufacturers of opium for smoking purposes, /1/ was not expected to raise the price, since most of the opium used was being smuggled into the country. The tax was definitely intended, however, to prohibit the manufacture of smoking opium, and not to raise revenue. It was considered by its backers to be the only way in which

/1/ Act of January 17, 1914, chap. x, sec. 1, 38 Stat. 277.

[page 197]

Congress could constitutionally accomplish its purpose. /2/ This tax
is an example of the third method mentioned above of using taxation
for control.

An internal revenue stamp of 1 cent an ounce on narcotic drugs and occupational taxes on dealers are the basis for a detailed system of regulations. These illustrate the fourth method noted above. Although there can be no question that the primary purpose of the narcotic tax act /3/ was regulation rather than revenue, the Supreme Court upheld it as constitutional. /4/ It insisted, however, on strict construction of the act. /5/


Beginning with the tariff levied on distilled spirits in 1790, liquor taxation in the United States has been imposed both for revenue and for control of consumption. The relative emphasis on control has differed between federal and state legislation and from time to time.

Prior to the prohibition period revenue was the more important objective in federal liquor taxation. The plan of administration was designed solely to raise revenue, and the excise tax ($1.10 per proof gallon) was by no means prohibitory.

Although the liquor taxes of the states and localities were imposed primarily for revenue, the burden of controlling the liquor traffic fell on these jurisdictions, and their licensing systems generally appear to have had control as the main objective. This tax objective seems to have increased in importance with the general movement against the liquor traffic.

Since the repeal of prohibition, attention has again

/2/ Congressional Record, 63 Cong., 1 sess., June 24, 1913, p. 2168.

/3/ Acts of December 17, 1914, chap. i, sec. 1, 38 stat. 785; February 24, 1919, chap. xviii, sec. 1006, 40 Stat. 1130; February 26, 1926, chap. xxvii, sec. 703, 44 Stat. 96, 97; Code of U.S., Title 26, sec. 1040, 1383-90.

/4/ United States v. Doremus, 249 U.S. 85 (1919); Nigro v. United States, 276 U.S. 332 (1928).

/5/ Linder v. United States, 268 U.S. 5 (1925); United States v. Daugherty, 269 U.S. 360 (1926).

[page 198]

been devoted to liquor control. Its most important aspect at present
is the elimination of the illicit liquor traffic.

In this field control is likely to conflict with revenue as an aim. If restriction of consumption is desired, much higher taxes are needed than those needed to achieve a maximum of revenue, assuming no serious problem of illicit traffic. Recently, conflict has arisen between revenue and the control of bootlegging. It has been claimed that the tax on spirits must be lowered in order to drive bootleggers out of business, before it can be raised to the maximum revenue level or to a level that will greatly restrict consumption.

It may be objected that the reduction of consumption through high prices is a class measure. Only the poor are deprived of intoxicants, while the wealthy can consume as freely as before. If the truth of this objection is granted, two considerations may keep it from being fatal to a high price policy. First, the social problems of liquor have always been more serious among the low income groups of society. Second, in the absence of a strong social approval of enforcement, any other method of keeping down consumption is also in effect a class measure. As the experience of the prohibition period indicates, the elimination of bootlegging is impossible where there is vigorous demand from financially able buyers.

Another aspect of control is the shifting of consumption from beverages with a high percentage of alcohol to lighter beverages. This was urged upon Congress when liquor taxation was under consideration after repeal, on the ground that the personal and social problems arising from the use of alcoholic beverages were associated almost entirely with the stronger liquors. /6/ It was pro-

/6/ Tax on Intoxicating Liquor: Joint Hearings before the Committee of Ways and Means, House of Representatives, and the Committee on Finance, United States Senate, 73 Cong., interim, 1, and 2 sess., p. 359. Hereafter cited as Tax on Intoxicating Liquor: Joint Hearings.

[page 199]

posed that a progressive tax on alcoholic content be imposed instead
of the substantially proportional tax actually adopted.

Foreign Success with High Taxation

The experience of certain other countries has indicated that taxation is an effective instrument of liquor control. American observers have drawn two conclusions from this experience: (1) that total liquor consumption can be reduced by taxation, and (2) that consumption can be shifted from ardent spirits to the lighter wines and beer.

The experience of Great Britain is most often cited to support these conclusions. Since 1913 the taxes on spirits have increased in Great Britain more than 400 per cent, with an even greater relative increase in the tax on beer. /7/ The tax on spirits is nearly $13 per gallon, /8/ as contrasted with a federal tax in the United States of $2.00 per gallon, and an average total federal and state excise of $3.00. /9/ Official British statistics show a decrease since 1913 of 60 per cent in the consumption of spirits, 41 per cent in the consumption of beer, and, furthermore, a decrease of 74 per cent in convictions for drunkenness. /10/

The whole change, however, cannot be attributed to the high prices caused by taxation. The economic depression, dole regulations, a reduction in the hours of sale, the spread of education, and an increase in facilities for amusement have also taken place during this period and have been partially responsible for the change. However, the costliness of becoming intoxicated is held by

/7/ D. W. McConnell, "Liquor Traffic," Encyclopedia of Social Sciences, IX, 505.

/8/ The British tax on beer is equivalent to about $14.35 per United States barrel of 31 gallons, Roswell Magill, L. H. Parker, Eldon P. King, A Summary of the British Tax System with Special Reference to Its Administration (Washington, 1934), pp. 30-31.

/9/ Paul Studenski, "Liquor Taxes and the Bootlegger," National Municipal Review, XXIV, No. 1, Supplement (January 1935), 71.

/10/ Raymond B. Fosdick and Albert L. Scott, Toward Liquor Control (New York, 1933), p. 37.

[page 200]

observers to be an important cause of the improvement. /11/ Similar
results have been observed in Denmark. /12/

The adoption of a similar policy in the United States would not necessarily be equally successful. The size of the country, the free movement of people within it, and the length of its seacoast and borders make the control of the liquor traffic through taxation more difficult than it is in a country like Great Britain. /13/

Suitability of Different Taxes for Control

To discourage the consumption of alcoholic beverages, the tax must be capable of special application to the liquor industry. It must either raise the price of liquor, reduce the number of places of sale, or reduce the effort to sell. Many taxes, such as the inheritance tax and the real estate tax, are too general in their effects to be useful in liquor control.

Special taxes might be imposed on the incomes or profits of those engaged in the liquor business. /14/ A tax levied on excess profits due to monopoly or speculation might be justified on the ground that governmental policy was responsible for these gains or that the liquor industry was suspect and should be subject to special restrictions. /15/ The tax would probably not be shifted to consumers. Therefore, it would probably discourage the growth of greater monopoly and would prevent the industry from developing rapidly in financial strength, but it would scarcely reduce consumption.

A heavy uniform tax on all net income from the liquor industry might, and probably would, be shifted to consumers in higher prices, at least in part. It is not likely, however, as some have contended, that such a

/11/ J. J. Mallon, "Drunkenness vs. Drink," Survey, LXVII (February 1, 1932), 489.

/12/ McConnell, op. cit., p. 505.

/13/ See p. 339.

/14/ See Note 1, Special Liquor Profits Tax, p. 547.

/15/ See pp. 271-72.

[page 201]

tax would substantially weaken the profit motive and thus destroy the
driving power behind most of the abuses of the liquor traffic. Some
producers might be eliminated, but those who remained would still
want to make as much profit as possible. Only through government
operation can the profit motive be eliminated. /16/

The income tax would exert less pressure to build up profits through a large volume of sales than would high, flat-rate occupational licenses, since the tax under such licenses is fixed and the profit on an increased volume is clear gain. However, the income tax would have less tendency to reduce the number of sales outlets than would flat-rate license taxes.

The tendency of flat-rate license taxes to encourage large-scale units would limit competition and create a semi-monopoly. Prices could then be increased to an uncertain extent. The pre-prohibition policy was distinctly a high license policy. On the whole, much lower licenses have been favored since repeal, probably owing to the desire to discourage bootleggers. A sharp differential in favor of sales outlets handling only beverages of lower alcoholic content appears distinctly useful in encouraging a shift to such liquors.

Taxes levied uniformly on manufacture or sale have the greatest chance, according to economic analysis, of being entirely shifted. Therefore, they offer the greatest possibilities of decreasing consumption by increasing prices. Geographical uniformity and ease of administration are furthered by levying the taxes on the manufacturer. Taxes on the manufacturer have been opposed, however, on the ground that in the process of shifting they are pyramided, and thus consumers are overburdened.

The use of a federal tax to assist states in enforcing

/16/ Government operation is urged by Gulick, Tax on Intoxicating Liquor: Joint Hearings, p. 163, and by Fosdick and Scott, op. cit., chaps. v and vi.

[page 202]

their liquor laws was exemplified by a tax of $1,000 imposed on all
persons selling liquor in violation of any state law. This was
enacted during the prohibition period, when the federal government
had the power to regulate the liquor traffic, and was retained after
repeal. The Supreme Court of the United States has declared the law
unconstitutional on the grounds that the $1,000 payment was not
really a tax but a penalty imposed in a field in which the federal
government had no power to legislate. /17/

/17/ United States v. Constantine, 296 U.S. 287 (1936).

[page 203]

Chapter 14



Public expenditures and the taxes imposed to finance them always change the distribution of wealth and income. This is because the benefits that a person derives from governmental expenditures seldom exactly equal his tax burden.

However, it is possible to change the distribution of wealth and income without effecting a redistribution. This distinction between redistribution and change in distribution must be emphasized -- especially so because in recent public discussions the possibility of such a distinction has been ignored.

Redistribution means taking from some and giving to others. A change in distribution includes redistribution. It also includes, however, the process by which some persons get something that is newly created and not taken from someone else. For example, sometimes the value of a governmental service that is rendered to a particular group in the community is much larger than its cost. Suppose the tax to support this expenditure is levied solely on the benefiting group. The net result will be a change in the relative distribution but not an actual decrease in anyone's total income or wealth. The benefits conferred by the fundamental functions of government are probably worth more to every taxpayer than their cost to him in taxes. A change in distribution -- but not redistribution -- occurs also when the tax money col-

[page 204]

lected from someone is wasted without increasing anyone's income.

Frequently, however, the total income or wealth of some people is decreased by taxation while the total income or wealth of other people is increased through the less vital functions of government. In this case redistribution takes place. A similar process occurs in the payment of bonuses, pensions, doles, subsidies, and interest and principal on debt.

The difficulty of measuring expenditure benefits and tax burdens makes it impossible to ascertain precisely the amount, and sometimes even the direction, of either redistribution or a change in distribution.



A change in distribution of income can be achieved by altering the pattern of the tax system or the expenditure system while keeping the total volume of taxes and expenditures unchanged. It can also be achieved by changing the total volume of taxes and expenditures. To secure a more equal distribution of income (1) the taxes may be so changed as to lighten the tax burden on the mass of people, or (2) the total volume of expenditures may be kept unchanged but the purposes for which they are made may be adjusted to produce greater benefit for the great mass of people, or (3) the volume of expenditures benefiting the great mass of people may be increased.

One way of using method (1) above is to replace taxes that are regressive by taxes that are progressive, or to increase the progressiveness of the taxes. /1/

/1/ If a tax, after being shifted to those who finally bear its burden, takes a greater percentage of the income of large income receivers than of smaller income receivers, the tax is progressive in incidence. If it takes a greater percentage of the income of small income receivers than of larger income receivers, the tax is regressive in incidence. If in its incidence the tax takes the same percentage of all sizes of income, the tax is proportional in incidence.

[page 205]


Redistribution of the flow of current income can be achieved more quickly through taxation and governmental expenditure than can redistribution of accumulated wealth. People usually pay their taxes out of new increments of wealth -- that is, out of income -- and keep intact their accumulated principal. Governmental expenditures for services ordinarily add to the current incomes of people rather than to their accumulated wealth.

In the course of time, however, these shifts in income distribution are likely to affect the distribution of wealth. The taxed individual may save less than he otherwise would because of the reduction in his income, while the individual receiving governmental services or subsidies may be able to save more. This process is a redistribution not of existing wealth, but of wealth accumulated in the future.

Inheritance and estate taxes are sometimes said to achieve redistribution of wealth in a direct way, as other taxes do not, because they are often so heavy that investments must be sold in order to provide funds for their payment. However, such forced sales do not result in redistribution of wealth in the sense in which the term is ordinarily understood. Ownership of the investment passes from the taxpayers to someone else, who purchases it out of either current income or uninvested savings. In one sense, this transfer of ownership may be redistribution, but it is not direct redistribution to recipients of government services or bounties. Also, it is not necessarily redistribution from a person in one income group to a person in another income group.

The sale of an investment to pay inheritance and estate taxes may affect adversely the creation of new wealth. The ownership of existing wealth has been

[page 206]

transferred, but no new saving has taken place. To the buyer,
however, the purchase may be in lieu of investment in newly created

Redistribution of existing wealth could be accomplished by a capital levy in kind followed by a distribution of the wealth thus secured. Some of the difficulties of such a distribution are discussed below. /2/


The effects of a given tax or expenditure measure on distribution may be incidental to some other purpose or they may instead constitute the main purpose of the measure.

The essential functions of government are required by all people. The effects that the financing of such functions exercise on distribution are purely incidental.

The government may conclude that many citizens greatly need certain services or goods -- not comprised in the essential functions of government -- that they cannot afford. If it is thought that the general public interest in supplying them outweighs the public interest in a light tax burden, the government will probably supply them at public expense. Free public education, public recreational facilities, health clinics, and work relief projects are examples. The result is a change in the distribution of income. A redistribution of income also probably occurs. Though accompanying the achievement of the public purpose, however, it is largely incidental. There is no desire to reduce anyone's income either absolutely or relatively but only to increase the amount of the income of some. The performance of such public activities is everywhere recognized as a proper function of government. A great difference of opinion exists, however, about the extent to which government should engage in them.

/2/ See pp. 213-14.

[page 207]

On the other hand, a government may believe that some persons' incomes are too large, at least compared with the incomes of others. Taxation and expenditure can be used to reduce the large incomes. Here, redistribution becomes a primary aim of taxation. Recently proposed "share-the-wealth" programs fall in this class, as do efforts to prevent the perpetuation of "economic dynasties" by taxation.


Whether a primary aim of the tax system should be greater concentration of wealth and income, or greater equality, or a minimum change, depends in part on the effects of redistribution and in part on the attitude of people toward inequality.

Effects of Redistribution

Opinions on the nature of the effects of redistribution are divergent. The various points of view can only be outlined briefly here.

Effect on Human Satisfactions

The idea of diminishing utility is commonly accepted. Most people assume that the more money a person has, the less utility or satisfaction he will get from an increase in wealth. Although such satisfactions are subjective feelings and cannot be compared, a dollar is assumed to give less satisfaction to a wealthy person than to a poor one. If this is true, the total satisfactions and consequently the economic welfare of the community as a whole increase with an increase in the equality of wealth and income. It is assumed, of course, that no overbalancing adverse effect occurs.

Effects on Production: Traditional View

Opinions differ as to whether such an adverse effect does occur. Economic welfare depends not only on the

[page 208]

manner of distribution but also on the volume of production. The
volume of production depends in turn largely on the amount of
machinery and other capital goods that increase the efficiency of
labor. Capital goods, however, can be produced, usually, only if the
community is willing to do without a certain amount of goods for
current consumption -- that is, if the community is willing to save.
In general, a larger fraction of their total income is saved by
people receiving large incomes than by those receiving small incomes.
To this extent, production and thereby economic welfare tend to be
promoted by the concentration of income.

However, if economic progress is to result from private saving based on inequality, the increased volume of consumers' goods that finally results from the use of the new capital must be sold. The sale will be made possible, according to the usual economic analysis, through the reduction of prices. Thus, all persons will share in the benefits of increased production. Rates of interest will also decrease as the supply of capital increases, and this decrease tends to keep the rate of capital growth from becoming excessive.

If this reasoning is sound, the use of taxation and public expenditures to redistribute wealth may result in a reduction in the rate of growth of production and, if carried far, in a decline in production.

It is also claimed that high taxes on the wealthy for the purpose of redistribution would discourage business initiative and risk-taking, and thus decrease production. It seems reasonable to believe that beyond some point, impossible to define, this would be the result. /3/

Effects on Production: Socialist View

Socialist writers do not accept this analysis of the economic system. Their argument has recently gained

/3/ See pp. 61-64 for a discussion of the effects of taxation on business initiative.

[page 209]

rather wide acceptance. They claim that, when income and wealth are
concentrated, the amount of saving and investment is so great that
consumers' goods are increased more rapidly than the power of
consumers to buy them. As a result, it is held, depressions occur,
and perhaps eventual collapse of the capitalist system may follow.
According to a variant of this argument, saved funds are not invested
because there is no prospect of profits through sale of consumers'
goods; the idle saved funds result in unemployment and inability to
purchase consumers' goods; and depressions follow.

Those who hold this view urge that redistribution of income and wealth does not diminish but actually increases production. Goods that would otherwise glut the market are then purchased. In turn, production and employment are encouraged. Furthermore, according to this view, the limiting factor on investment is not saving but profitable opportunities for investment, which grow out of the sale of consumers' goods. Redistribution, by encouraging the purchase of consumers' goods, thus not only stimulates immediate production, but also makes investment profitable and so increases future production beyond what it would otherwise be.

Effects on Production: Conclusions

These two views of the effects of inequality of distribution of wealth and income on the increase of production are fundamentally different interpretations of the operation of modern capitalism. They go far outside the field of taxation proper, and no attempt can be made here to analyze or evaluate them at length. Perhaps the truth is that a certain degree of inequality less than now exists would be the optimum beyond which equalizing distribution would be socially harmful.

In the opinion of the writer, if redistribution, especially of accumulated wealth, were carried to its logical

[page 210]

extreme -- to a point of approximate equality of incomes -- profound
changes in the economic system would occur. It would almost certainly
result in a decrease in the volume of privately saved capital and
would probably mean the end of private capitalism. The persons
receiving the redistributed wealth would probably have small incomes.
To such persons receipt of more income at once would probably be more
important than continued ownership of wealth. Much of the wealth
would be liquidated as rapidly as possible. The amount of private
wealth and consequently of private production would decline. To
offset the decline, government saving and participation in productive
industry would be necessary. The end result would be socialism.

Effects on Liberty

Some contend that the use of taxation to equalize the distribution of wealth and income is an aid to liberty, while others hold it to be a menace to liberty. On the one hand, it is claimed that liberty can exist only when there is approximate economic equality. When the economic opportunities of the many are controlled by the few, economic liberty declines and political liberty follows. On the other hand, it is asserted that taxation itself is a restriction of liberty since it takes away the freedom of the individual to determine how his income shall be expended. These two arguments are obviously based on two different views of what kind of liberty is important.

The Desire for Equality

The problem of the distribution of income and wealth is further complicated by differing personal attitudes. Relative incomes as well as absolute incomes are important. Undoubtedly some people -- and if concentration were carried to extremes perhaps nearly everyone

[page 211]

would be in this group -- would prefer a somewhat lower income in a
society where there was approximate equality of incomes to a somewhat
higher income in a society where other people had very much higher
incomes. Furthermore, people probably differ widely in the importance
that they place on equality. Some indeed may still get satisfaction
and security from having people of great wealth to whom they can look
as superiors.

Because of the complicated and conflicting factors involved, different people will have widely differing opinions on whether taxation should be directed toward reducing inequality, toward increasing it, or toward maintaining the existing condition.



Under some other economic system the redistribution of income could be used to effect an almost complete equality of incomes. If private capitalism is to continue, however, the amount of redistribution that can actually be accomplished is limited.

In the first place income that does not exist obviously cannot be redistributed. The amount subject to redistribution is thus limited to the amount of the income. Persons advocating redistribution often grossly overestimate the income of the nation. If the entire income of the nation in the most prosperous year (1929) had been distributed equally among the entire population, the income of each person would have been only about $673, or $2,692 for a family of four. /4/

The redistribution of income is also limited by the existing forms of income. Income is not money, although it is ordinarily received in that form. Real income is the goods and services that money will buy. A change in the distribution of income would mean a certain amount of

/4/ Maurice Leven, Harold G. Moulton, and Clark Warburton, America's Capacity to Consume (Washington, 1934), Table 1, p. 148.

[page 212]

change in the nature of goods and services produced, and this process
of change would involve not a few months or a year but rather many
years. Changes in the forms of capital and the organization of
business for production would have to be made. Rapid increases in
taxes for the purpose of government expenditure would result in a
shift of workers and in a shift in demand for the products of
industry. Some capital would thus be wasted, and some businesses

Limits to the redistribution of income also result from the volume and the types of governmental expenditures. If the amount of governmental expenditures is not increased, the amount of redistribution can be no greater than the volume of governmental expenditures. The workable maximum is much less, however. Many governmental expenditures are made for purposes that benefit all groups in society, not the poorest groups only, and the taxes cannot practicably be loaded only on the very wealthiest, to the extent of 100 per cent of the top parts of their incomes.

If the volume of governmental expenditures is not fixed, the amount of redistribution might be increased greatly, both through services and through subsidies. However, governmental services ordinarily give to the recipients types of income that do not replace the things that they have been buying -- especially those in the low income levels. Education, parks, and modern highways, for example, are forms of free income, but they do not reduce the need for food, clothing, and shelter. Thus, unless government supplies the common necessities of life -- as it does in public business or the direct relief of poverty -- its services are relative luxuries, and the labor power of the country may be diverted from the production of necessities. Cash payments, on the other hand, while they have undesirable features, at least place in the hands of individuals the wherewithal to buy neces-

[page 213]

sities, and any diversion of productive effort that results will
probably be to the production of necessities.


There are also limits to the redistribution of wealth by taxation. First, of course, as with income, no more wealth can be redistributed than exists. In 1922 -- when the volume of wealth was about what it is today -- the total wealth of the United States amounted to about $3,200 for each person, or $12,800 for a family of four. /5/ One-eighth of this was owned by government and so could not be redistributed. /6/ Second, most wealth is not money but land, buildings, and other physical things. Over half (55 per cent) of the 1922 wealth was in the form of real estate alone. /7/ Accordingly, a redistribution of wealth cannot in the very nature of the case be a redistribution in money. Any attempt directly or indirectly to redistribute existing wealth in money means that it cannot be transformed into cash except by selling it -- an impossibility on any large scale because of the limited money supply --, or by using it as the basis of loans or money issues. The latter procedure, however, will inevitably result in rising prices, falling real incomes, and all the other effects of inflation.

If existing wealth is to be redistributed, it must be in kind. Money or credit might be used as a lubricant for the process of redistribution, but at the conclusion of the process the existing assets, not money, must be in the hands of the new owners. The most direct way would be for the government to collect a capital levy in the form of the assets themselves and divide them among the people to whom redistribution is to be made.

/5/ Wealth figure of $353,000,000,000 from National Wealth and Income, a Report by the Federal Trade Commission, 69 Cong., 1 sess., Sen. Document No. 126, p. 1 (hereafter cited as National Wealth and Income). Population figure of 110,688,000 from W. I. King, The National Income and Its Purchasing Power (New York, 1930), p. 47.

/6/ National Wealth and Income, p. 2.

/7/ Ibid., p. 29.

[page 214]

Much of the property is in non-corporate form, however, and in
ownership units so large that they would have to be broken up. Thus,
a taxpayer might give an apartment house to the government in payment
of a capital levy imposed to redistribute wealth. To break this
property into distributable pieces would necessitate incorporating
it. Practically all large wealth ownership would have to be
incorporated in order to be divided.

Taxpayers would attempt to pay their taxes in the types of asset that they had least desire for in their altered financial circumstances. The government would probably have an excellent collection of high priced estates, jewelry, and yachts, none of which would be acceptable to persons to whom distribution was being made. The government would have some difficulty in distributing even investments directly to their new owners, because persons in the financial circumstances of the latter would have different standards of preference for investments from those prevailing on the market prior to the capital levy. For example, they would probably not wish speculative investments. If they were forced to take investments that they did not want, the political storm following any losses can easily be imagined.

To meet these problems, the government might assume the ownership of the assets and issue participating certificates to those receiving the distribution. If the amount involved were large, this would make the government a large industrial owner with a substantial stake in the operation of industry. To that extent, private capitalism would almost surely be replaced by socialism.

[page 215 blank]

[page 216]