Tax Analysts®Tax Analysts®

Article Archive

February 15, 1996
Historical Document: 'Facing The Tax Problem' -- Book Two, Analysis (Part II).
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. Tax Justice
Section B. Fluctuation in Revenue
Section C. Ease of Administration
Section D. Tax Consciousness
Section E. Intergovernmental Co-ordination

[page 217]

Section A. Tax Justice

Chapter 15


The secondary aims of a tax system, to be discussed in Chapters 15 to 24, supply the grounds for choice among alternative tax measures, once the primary aims have been determined.


More lip service has been paid to justice than to any other secondary aim of the tax system. With rare exceptions, everyone protests his eagerness to pay his "fair share" of the taxes. This conception of justice, however, is too vague to be of much use. It must be broken down into more specific elements.

Justice cannot always be sharply distinguished from other secondary aims. For instance, economic efficiency may be another secondary aim of a tax system. In so far as a system fails in the long run to promote economic efficiency, it also probably fails to promote certain aims of justice. On the other hand, some widely held aims of justice may run directly counter to economic efficiency. In the present discussion, the study of justice is confined to the more immediate distribution of the tax burden, without particular regard to the long-term economic consequences.

A discussion of tax justice first assumes that a given total amount of revenue is to be spent in a given manner. The question then is: what is the just way to distribute the tax burden that is required to raise this revenue? The answer must take account of the pattern of ex-

[page 218]

penditures. For example, the amount of money to be raised from
automobile taxes may depend on the volume of highway expenditures.
This survey uses the expenditure pattern that will probably exist in
the United States during the next few years. /1/


At least two major difficulties beset anyone who tries to think in terms of tax justice. In the first place, opinions differ on the objectives to be reached if a tax system is to be just. However, certain objectives -- although they are to some extent mutually contradictory -- have gained widespread acceptance. An examination of the characteristics of the system that relate to these objectives will help anyone to form his own ideas on how just the system is.

In the present section the existing system is examined for the following four characteristics:

(1) The sort of rigidity that (a) makes it impossible to rectify, by certain tax decreases, former injustice, as when a reduction of the property tax cannot help the one who was originally injured, since he no longer owns the property; or (b) results from the creation of innocent vested interests that would be harmed by certain tax increases, as when an investor buys into a company -- for example, a chain store -- and then soon sees the value of his investment drop because the government unexpectedly levies a heavy special tax on his company.

(2) Charges for government services that bear some relation to the cost that each user causes or the direct benefit that he receives, as when highway users are required to pay a gasoline tax or a motor vehicle license tax.

(3) Special taxation of certain sources of gain because

/1/ See pp. 94-120.

[page 219]

of their moral implications, as when excess profits from monopoly are
specially taxed.

(4) Adaptability to differences in the personal status of taxpayers, as when progressive rates and allowances for marital status, age, etc., are used.


The second difficulty that besets the student of tax justice is the necessity of considering the effect of the tax system as a whole -- a subject of which he can know very little.

A proposal regarding a certain tax -- for example, a lowering of the personal exemptions of the income tax -- might be thought unjust if there were no other taxes in the system, but it might be thought just if the tax were only one of many. Furthermore, of two tax systems composed of many taxes, the proposal might increase the justice of one and decrease the justice of the other.

Knowledge of the system as a whole must be obtained, because justice involves the relations of individuals, and tax burdens on individuals are made up not of one tax but of the whole system of taxes. Thus, it may be necessary to ascertain, for example, how much of the total tax burden, federal, state, and local, is borne by John Smith, a middle-aged lawyer with a yearly income of $9,000, who has a wife and four children to support, who drives his car 10,000 miles a year and lives in a city where he must pay a private agency to have his garbage removed.

Unfortunately, no one has yet succeeded in describing exactly how the burden of the entire tax system is actually distributed among individuals, or among geographic groups, income groups, or any other elements of the population. Any estimate of Smith's total tax burden must include more than personal income taxes and other taxes that he pays directly. Some of the taxes

[page 220]

paid by others -- sales taxes, for instance -- may be shifted to him
in the form of higher prices for the goods that he buys. He in turn
may be shifting to others some of the taxes that he pays directly.

Not enough is known about the shifting of taxes and the production habits and consumption habits of the populace to make possible any accurate statement of the distribution of the tax burden at the present time. Reasonable assumptions only can be made. The problem is presented in detail in the next chapter, with some illustrations of the range within which a few of the answers may be expected to fall.

The difficulties of ascertaining the distribution of the tax burden create an unfortunate temptation to conclude that principles of tax justice are not vitally important from the practical point of view. However, the difficulties must not be exaggerated. The skilled observer can make some progress toward an intelligent judgment. Strong probabilities can be found even when no provable facts exist. Having formed objectives of justice, the observer will frequently be able, with a fair degree of confidence, to brand a tax proposal as unjust even though he cannot be sure of what a complete and final marshalling of the facts would show. In the present state of tax knowledge, he may not always be able to decide where the balance of justice lies among entire tax systems -- at least, among systems as closely similar as many of those in effect today. However, he may frequently be able to gauge the probable effect of some one tax measure on his conception of the justice of the system as a whole.

[page 221]

Chapter 16


The justice of a tax system depends in large part on the relative real tax burden that each person bears. This burden varies even among individuals who have the same income. The income tax itself will differ for two individuals of the same income if the sources of income and the size of family differ. Other taxes will vary with place of residence, amount and kind of property owned, and habits of expenditure.

Consequently, it is impossible to say that the taxes paid by every man with a $1,000 income, for example, will amount to a specific percentage of his income. The most that can be said is that, if several men with $1,000 incomes get their incomes from the same specified sources, spend them in the same manner, live in the same locality, and own property of the same kinds and amounts and assessed at the same rates, the taxes that they will pay will amount to practically the same sum.

Moreover, even a precise statement of taxes paid is inadequate. The tax burden of an individual can be measured only after knowing what amount of the taxes that he pays is shifted to others, and what amount of the taxes of others is shifted to him. In other words, the test is taxes borne rather than taxes paid. But this involves many debatable assumptions about the shifting of taxes.

In view of these difficulties, very little confidence can be placed in the tax burden figures given in the recent flood of literature on "unseen" or "hidden" taxes. /1/

/1/ See Note 1, Tax Burden Studies, p. 548.

[page 222]

An attempt is made in this chapter, however, to measure the tax
burdens of hypothetical persons. Owing to the same difficulties,
these estimates are not strictly representative of the income groups
from which the cases have been selected. They are offered, rather, to
indicate some of the problems involved in estimating tax burdens and
to show the variations in estimates that will result from changing
the assumptions. They are given in this chapter in summary form and
are explained in detail in another volume. /2/

The taxes assumed are those in effect in the calendar year 1936, and allowance has been made for those in effect for only a part of the year. The tax burden is taken to be the amount of taxes paid that the taxpayer has been unable to shift to others or to discount in advance, plus the amount of taxes shifted by others to him. No account is taken of benefits that the taxpayer may have received in return for these taxes. On the other hand, the tax burden figures do not include special assessments, fees, or prices charged by government-owned enterprises.


In selecting the hypothetical cases for which the tax burden has been estimated, a wide range of incomes, from $500 to $1,000,000, has been chosen. In other respects families have been assumed to be as much alike as Is reasonable in view of the differences in their incomes. All the families have been assumed to be of the same size -- a man, his wife, and two minor children. All are home owners; all have cars; all use tobacco; and all the urban families use liquor. All within the given state live in the same rural or urban community, as the case may be, and are, in consequence, subject to the same property tax rates.

/2/ See the memorandum by Mabel Newcomer on the distribution of the tax burden in the forthcoming volume, Studies in Current Tax Problems.

[page 223]

Two states have been selected to indicate the differences in tax burden that may result from differences in state and local tax systems. New York has been chosen as a state that has developed income taxation to a marked degree, with personal, corporation, and unincorporated business income taxes. Illinois has been chosen as a state that is leaning heavily on the sales tax. New York has no general sales tax, except for the local tax in New York City, and Illinois has no income tax. Another important difference is that Illinois still taxes personalty whereas the New York property tax is limited to real estate. Neither state uses the property tax for state purposes, but state property taxes have been so generally reduced during the depression that these states can hardly be regarded as exceptional in this respect. /3/

The amount of expenditure and the amount of savings assumed for each family, and also the broad classifications of expenditure, follow the proportions indicated for the different income classes in the Brookings Institution study, America's Capacity to Consume. /4/ The details of expenditure have been suggested by various budget studies, especially those of the United States Bureau of Labor. No allowance has been made for differences in the cost of living in New York and Illinois, the same expenditures being assumed for the families in both states. The slight difference in taxes that would result from such an adjustment does not seem to justify this additional refinement. These hypothetical budgets are not presented as typical. They are considered to be reasonable for each income class, but the patterns of expenditure will vary widely even among families of the same income group, particularly in the higher income classes.

/3/ See the memorandum by Thomas J. Reynolds on the state property tax rates in the forthcoming volume, Studies in Current Tax Problems.

/4/ Leven, Moulton, and Warburton, op. cit.

[page 224]

The kinds and value of property owned by each family have been determined as described elsewhere. /5/ As with expenditures, the figures are believed to be reasonable for each income class but not necessarily typical.

Substantial variations in the tax burden could be computed by using different, but still reasonable, assumptions concerning the amount and nature of each family's property and expenditures. Since this study must come to an end, however, and since it is primarily concerned with problems more closely related to the tax system, only one set of expenditure and property assumptions has been used. Such variations in burden as are given below result from changes in two kinds of assumptions: those concerning the shifting of taxes and those concerning changes in the ownership of property and income within the family, so far as such changes might be made deliberately in order to reduce the family's taxes. It is, of course, true that the nature of the tax system may also affect the nature of expenditures and the amount and nature of taxable property held. However, the effect is less than that on the manner in which the ownership of property is distributed within a family, since the possibility of reducing the tax burden is less and habits of consumption must be changed in order to reduce it.


Taxes restricted to narrow fields and taxes of very minor revenue importance have not been included in these estimates.

In order to keep the problem reasonably simple, it has been assumed that all the families that hold stock are investors in a single, domestic, manufacturing corporation. Thus, special taxes on other types of corporation

/5/ See the memorandum by Mabel Newcomer on the distribution of the tax burden in the forthcoming volume, Studies in Current Tax Problems.

[page 225]

that will fall ultimately on the stockholder, such as the New York
tax on banks, have been omitted. The omission of such taxes need not
reduce the estimated tax burden below the amount borne by an average
investor, however, since the manufacturing corporation pays taxes
that are roughly equivalent.

Certain taxes, mostly specific consumption taxes, have been omitted as too small to justify estimates. Although they enter into the ultimate tax burden, they amount to less than 6 per cent of federal tax collections, and less than 1 per cent of state and local tax collections.


Five series of assumptions have been made about tax shifting and intra-family changes in property holdings. The tax burden for each family has been computed under each series. These series are described in Table 20. In general, it has been assumed that income, death, and gift taxes are not shifted; that business taxes, other than those on net income and capital stock, may be shifted to the consumer in part or entirely; and that land taxes may be capitalized in part or entirely. For income, death, and gift taxes some variation has been assumed, not because of shifting, but because of changes in the ownership of property and income within the family. Only business net income, capital stock, mortgage, and stock transfer taxes remain unchanged with respect to both shifting and intra-family distribution of ownership through the five series of assumptions.


Table 21 is a tax consciousness table rather than a tax burden table. It gives the taxes paid by each family directly to the government, including those taxes whose burden does not really rest on the family. However, the taxpayer will probably regard the total of these taxes

[page 226]


[table omitted]

[page 227]

[table omitted]

[page 228]

as his particular burden since he pays all of them directly to the
government as taxes.

Tables 22-25 give the various estimates of burden in detail. Table 22 tells part of the story of the true tax burden in terms of Series I assumptions. It shows only the taxes that are not shifted. However, it includes, besides the taxes paid directly to a government official, those that are paid to a non-governmental agency but that are more or less clearly labelled taxes, such as the gasoline tax. The latter taxes are probably also regarded by the taxpayer as a part of his burden, but he will rarely make the necessary calculation to determine the exact amount, and he will therefore be somewhat less conscious of it. Table 23, likewise restricted to the Series I assumptions, contains only the concealed taxes, such as customs duties, that are shifted to the taxpayer. They probably play no part in his tax consciousness.

                              TABLE  21
                            (In Dollars)

Occupation:                      Farmers             Earners
Income /a/          500      1,000     2,000     1,000     2,000

                                New York Families

Federal               0          0         0         0         0
State                 9         13        13         9        15
Local               162        234       384       114       223
  Total             171        247       397       123       238

                                Illinois Families

Federal               0          0         0         0         0
State                 8         10        10         8        12
Local               103        140       200        90       188
  Total             111        150       210        98       200

                          [table continued]

                 Salaried            Salaried    Corporation
Occupation:       Worker   Merchant   Worker      Officials
Income /a/        5,000      5,000    20,000   100,000 1,000,000

                                 New York Families

Federal              21         19       805    13,515   506,438
State                49        261       765     6,358    62,937
Local               546      1,365     1,068     7,495    37,000
  Total             616      1,645     2,638    27,368   606,375

                                 Illinois Families

Federal              25         31       881    15,809   562,736
State                25         38        53        78        90
Local               473      1,730       981     5,138    24,750
  Total             523      1,799     1,915    21,025   587,576

     /a/ This income figure is income actually received. It is not
suitable as a base for computing the percentage of income paid in
taxes. The proper base for a percentage computation is "potential
income" -- that is, the income that would have been received if the
taxes had not existed --, since some of the taxes are in fact paid
out of this potential income, not out of income actually received.

[page 229]

The total burden under the assumptions in Series I consists of the figures in Table 22 plus those in Table 23, and this total is shown in Table 24.

The total burden under each of the five series of assumptions is shown in Table 25.

                                   TABLE 22
                                   TO OTHERS
                       BASED ON ASSUMPTIONS IN SERIES I
                                 (In Dollars)

Occupation:                      Farmers             Earners
Income /a/          500      1,000     2,000     1,000     2,000

                                New York Families

Federal              /b/         2         7         5         8
State                 5         18        26        23        33
Local                32         45        73        86       167
  Total              37         65       106       114       208

                               Illinois Families

Federal             /b/          2         7         5         8
State                 3         13        18        20        27
Local                19         27        40        72       154
  Total              22         42        65        97       189

                          [table continued]

                 Salaried            Salaried    Corporation
Occupation:       Worker   Merchant   Worker      Officials
Income /a/        5,000      5,000    20,000   100,000 1,000,000

                                 New York Families

Federal              45        181       920    16,248   615,148
State                74        286       876     6,987    99,668
Local               410        410       801     4,202    18,120
  Total             529        877     2,597    27,437   732,936

                               Illinois Families

Federal              49        193       997    18,543   671,444
State                43         67        96       399    13,590
Local               394        394       742     3,525    17,250
  Total             486        654     1,835    22,467   702,284

     /a/ This income figure is income actually received. It is not
suitable as a base for computing the percentage of income paid in
taxes. The proper base for a percentage computation is "potential
income" -- that is, the income that would have been received if the
taxes had not existed --, since some of the taxes are in fact paid
out of this potential income, not out of income actually received.

     /b/ Fifty cents.


For comparative purposes the tax burden has been estimated as a percentage of income in Table 26 and Chart 6. The income figure used is not the actual income shown in Tables 21-25, but the income that the person would have received if the land tax had not been capitalized /6/ to the extent assumed, and if certain other taxes, which, it has been assumed, have been shifted to him, had not been imposed or had been shifted elsewhere.

/6/ See pp. 238-40.

[page 230]

                              TABLE 23
                            (In Dollars)

Occupation:                 Farmers                 Earners
Income /a/          500      1,000     2,000     1,000     2,000

                                   New York Families

Federal               7         12        18        25        42
State..               4          8        18        12        24
Local..              12         24        53        30        60
  Total              23         44        89        67       126

                                  Illinois Families

Federal               9         16        25        30        51
State..              13         22        40        33        39
Local..              12         24        53        30        60
  Total              34         62       118        93       170

                          [table continued]

                 Salaried            Salaried     Corporation
Occupation:       Worker    Merchant  Worker       Officials
Income /a/        5,000      5,000    20,000   100,000 1,000,000

                                   New York Families

Federal             254         73     3,079    15,060   257,167
State..             127         64     1,183     5,934    94,494
Local..             150        150       600     3,000    30,000
  Total             531        287     4,862    23,994   381,661

                                  Illinois Families

Federal             293         95     3,441    16,844    285,282
State..             124        115       415     1,745     16,702
Local..             150        150       600     3,000     30,000
  Total             567        360     4,456    21,589    331,984


/a/ This income figure is income actually received. It is not
suitable as a base for computing the percentage of income paid in
taxes. The proper base for a percentage computation is "potential
income" -- that is, the income that would have been received if the
taxes had not existed --, since some of the taxes are in fact paid
out of this potential income, not out of income actually received.


                              TABLE 24
                          TOTAL TAX BURDEN
                            (In Dollars)

Occupation:                 Farmers                 Earners
Income /a/          500      1,000     2,000     1,000     2,000

                                   New York Families

Federal               7         14        25        30        50
State..               9         26        44        35        57
Local..              44         69       126       116       227
  Total              60        109       195       181       334

                                   Illinois Families

Federal               9         18        32        35        59
State..              16         35        58        53        86
Local..              31         51        93       102       214
  Total              56        104       183       190       359

                          [table continued]

                 Salaried            Salaried     Corporation
Occupation:       Worker    Merchant  Worker       Officials
Income /a/        5,000      5,000    20,000   100,000  1,000,000

                                   New York Families

Federal             299        254     3,999    31,308    872,315
State..             201        350     2,059    12,921    194,162
Local..             560        560     1,401     7,202     48,120
  Total           1,060      1,164     7,459    51,431  1,114,597

                                   Illinois Families

Federal             342        288     4,438    35,387    956,727
State..             167        182       511     2,144     30,292
Local..             544        544     1,342     6,525     47,250
  Total           1,053      1,014     6,291    44,056  1,034,269

                          FOOTNOTE TO TABLE

/a/ This income figure is income actually received. It is not suitable as a base for computing the percentage of income paid in taxes. The proper base for a percentage computation is "potential income" -- that is, the income that would have been received if the taxes had not existed --, since some of the taxes are in fact paid out of this potential income, not out of income actually received.


[page 231]

This "potential income" varies with the varying assumptions in the
different series.

                              TABLE 25
                            (In Dollars)

Occupation:                 Farmers                  Earners
Income /a/          500      1,000     2,000     1,000     2,000

                                   New York Families

    I..              60        109       195       181       334
   II..              79        118       198       150       276
  III..              79        118       198       150       276
   IV..             105        163       277       174       324
    V..              60        109       195       181       334

                                 Illinois Families

   I..               56        104       183       190       359
  II..               67        103       163       147       279
 III..               78        122       197       147       279
  IV..               84        130       207       166       316
   V..               56        104       183       190       359

                          [table continued]

                Salaried             Salaried     Corporation
Occupation:      Worker     Merchant  Worker       Officials
Income /a/        5,000      5,000    20,000   100,000  1,000,000

                                   New York Families

    I..           1,060      1,164     7,459    51,431  1,114,597
   II..             967      1,246     7,679    52,662  1,155,622
  III..             967      1,246     7,679    52,662  1,155,622
   IV..           1,089      1,467     8,075    55,619  1,180,487
    V..           1,060      1,243     8,814    69,972  1,439,551

                                 Illinois Families

   I..            1,053      1,014     6,291    44,056  1,034,269
  II..              915      4,587     6,288    44,341  1,062,966
 III..              990      2,875     7,488    50,341  1,167,966
  IV..            1,006      4,736     6,599    46,334  1,079,997
   V..            1,053      1,104     7,587    63,851  1,336,201


/a/ This income figure is income actually received. It is not suitable as a base for computing the percentage of income paid in taxes. The proper base for a percentage computation is "potential income" -- that is, the income that would have been received if the taxes had not existed --, since some of the taxes are in fact paid out of this potential income, not out of income actually received.



If the foregoing estimates are taken at their face value, it is apparent that everyone is bearing a substantial burden of taxation. As shown in Table 26, the smallest burden under any conditions is estimated to be 8 per cent of potential income, for one of the Illinois farm families, and the largest exceeds 100 per cent, in the extreme case of a corporation official where the family has an actual income of $1,000,000 but no steps are taken to distribute the income or property among members of the family in order to lessen income or death taxes. Even in this extreme case, the tax burden is in excess of 100 per cent of potential income only when it is assumed

[page 232]

that the family attempts to keep the family property intact, on the
death of the head of the family, through insurance sufficient to
cover death taxes.

                              TABLE 26

Occupation:                 Farmers                 Earners
(In Dollars) /a/:   500      1,000     2,000     1,000     2,000

                                 New York Families

    I..            12.0       10.9       9.8      18.6      17.2
   II..            15.0       11.4       9.7      15.4      14.2
  III..            15.0       11.4       9.7      15.4      14.2
   IV..            19.1       15.3      13.2      17.6      16.4
    V..            12.0       10.9       9.8      18.6      17.2

                              Illinois Families

    I..            11.2       10.4       9.2      19.3      18.2
   II..            12.7       10.0       8.0      15.0      14.2
  III..            14.6       11.6       9.5      15.0      14.2
   IV..            15.6       12.4      10.1      16.9      15.9
    V..            11.2       10.4       9.2      19.3      18.2

                          [table continued]

Income           Salaried            Salaried     Corporation
(In Dollars) /a/: Worker   Merchant   Worker       Officials
                  5,000      5,000    20,000   100,000  1,000,000

                                 New York Families
    I..            20.8       23.9      31.6      44.3       84.5
   II..            18.8       24.5      30.6      44.0       83.6
  III..            18.8       24.5      30.6      44.0       83.6
   IV..            20.9       27.8      31.9      45.7       84.5
    V..            20.8       25.6      37.4      60.3      109.1

                              Illinois Families

    I..            20.5       20.6      27.2      38.4       81.0
   II..            17.7       52.9      26.5      38.6       79.9
  III..            18.9       40.7      30.1      41.6       81.4
   IV..            19.3       54.0      27.7      39.9       80.8
    V..            20.5       22.4      32.8      55.7      104.6


/a/ This income figure is income actually received. It is not the
base used for computing the percentage of income paid in taxes. The
base used for the percentage computations is "potential income" --
that is, the income that would have been received if the taxes had
not existed --, since some of the taxes are in fact paid out of this
potential income, not out of income actually received.


Other assumptions might reduce burdens somewhat, but as long as tax revenues amount to 20 per cent or 25 per cent of the national income, /7/ and the larger part of these revenues comes from other taxes than personal income and death taxes, few citizens can hope to escape with nominal contributions.

Comparison by Income Classes and Occupations

All the estimates indicate that the tax burden is regressive for those income classes not subject to income and death taxes. This regressivity is largely due to the

/7/ See pp. 3, 67.

[page 233]



[chart omitted]

[page 234]

assumptions that the ratio of expenditures to income is greater for
small incomes than for large incomes and that the rate of assessment
is higher for small properties than for large properties. For the
higher income groups, whose tax burden is dominated by personal
income, corporation income, and death taxes, the burden is definitely

For families of the same income, the wage earners have a heavier tax burden than the farmers, owing primarily to the higher property taxes in urban areas. If none of the tax on land is capitalized, however, the farmer has a somewhat heavier burden since he has a larger investment in land. For the $5,000 incomes, the merchant's burden is consistently heavier than the salaried worker's, even when it is assumed that business taxes are entirely shifted, owing to the merchant's larger property holdings and, in consequence, his heavier death taxes. When it is assumed that business taxes are only partly shifted, the difference in the burdens of these two $5,000 income families becomes very great.

New York Compared with Illinois

In a comparison of the tax burdens in New York and Illinois, the New York burden is shown to be substantially heavier. For the wealthier group, however, the lower federal taxes in New York partly compensate for this difference. The burden of state and local taxes, under the assumptions in Series I, is $165,000 less for the Illinois taxpayer with a $1,000,000 income than for the New York taxpayer with an equal income. But the Illinois taxpayer loses $84,000 of these gains to the federal government. The deduction of state taxes from taxable income in calculating the federal tax means that more than half these state taxes are "burdenless" for those individuals whose income reaches into the upper brackets. That is, if the taxpayer did not have any state

[page 235]

taxes to pay, his federal taxes would increase by more than half the
state tax that is here assumed for him. And the poorer individuals
who do not gain from this credit bear relatively little more in state
and local taxes in New York than in Illinois, the New York system
being more highly progressive than the Illinois system.

Results of Changing Assumptions

Comparing the difference in estimated burdens under the different series of assumptions, it is perhaps surprising to find that radical differences in assumptions result in comparatively small changes in burdens in most instances, as shown in Tables 25 and 26. A tax may or may not be shifted from producer to customer or employee; but as long as the same individual serves in all three capacities -- as stockholder, consumer, and wage earner -- some share of the tax will come his way, and, while his interests will not be equal in these three capacities, some will at least partly counterbalance others.

The assumption that only half, instead of all, the consumption and business property taxes and one-third, instead of all, the payroll tax are shifted to the consumer changes the estimated burden only 1 per cent or less of the potential income in nine of the twenty cases. Only the Illinois merchant who bears half the retailers' occupation tax will find a radical change in his burden (Series II as compared with Series I). The assumption that only one-quarter of the retailers' occupation tax is borne by the merchant, one-quarter being shifted back to the producer, increases the burdens of producers appreciably and brings substantial relief to the merchant (Series III as compared with Series II). The assumption that only half the land tax has been capitalized instead of the entire sum increases everyone's burden, but especially the farmer's (Series IV as compared with Series II).

[page 236]

The New York farmer (except the $2,000-income farmer) and merchant have their lowest percentage burdens when it is assumed that all the taxes on business property and products are shifted and all the land tax is capitalized (Series I and Series V). The other New York families have their lowest burdens when it is assumed that only half of such taxes are shifted (Series II). In Illinois the farmer with the $500 income, the merchant, and the corporation official with the $100,000 income, have their lowest burdens when it is assumed that all taxes on business property and products are shifted and all the land tax is capitalized (Series I). The other Illinois families have their lowest burdens when it is assumed that only half the business property and consumption taxes are shifted (Series II).

The tax burden of the higher income groups rises materially when it is assumed that all the property and income are in the possession of the head of the family and a joint income tax return is made for husband and wife (Series V). Under these conditions the families with $1,000,000 incomes would be unable to provide adequate insurance to keep their estates intact at death. These families can maintain their estates intact, however, and enjoy a free income of over $200,000, after allowing for all concealed taxes, simply by dividing the property and income between husband and wife (part of the assumption in Series I). Also, they may reduce their taxes even further by sharing property and income with their children or by transferring part of their property to a personal holding corporation. It is possible for a family to be so wealthy that they cannot under any circumstances, by legitimate means, maintain their estates intact at death under the present tax system, but the Statistics of Income record few such cases.


Estimates based on innumerable other assumptions or different combinations of the foregoing assumptions

[page 237]

might be interesting and profitable. Those presented above, however,
should be sufficient to indicate the general character of the 1936
tax burden in the two states in question, and, within wide limits,
the weight of this burden. The estimates under the different
assumptions do not fix upper and lower limits for individuals, or
even for the average of the income and occupation groups in question.
Keeping the sources of income and the amounts and kinds of
expenditure and property constant throughout has set limits on the
variations in burden that do not in fact exist. Moreover, when it has
been assumed that half of a tax is shifted, it has been assumed that
this fraction is shifted in every case. Actually, of course, if
exactly half of the entire sales tax were shifted, it might still be
true that one merchant was successful in shifting the entire tax
while another shifted none at all. No such variation is taken into

It seems safe to conclude: first, that the tax system as a whole is regressive for the lower income groups, and distinctly progressive for the upper income groups; second, that the burden of the New York system exceeds that of the Illinois system, but that the weight of the New York system is moderated to an important degree by the resulting reduction in federal taxes for New York citizens; third, that urban tax burdens exceed rural tax burdens; and, fourth, that the large number of taxes that can be shifted have so diffused the burden that it would be practically impossible for any individual to escape a substantial tax burden, even though it would be quite possible to pay no tax directly. The findings of this study are too uncertain to warrant any conclusions beyond these limited generalizations.

[page 238]

Chapter 17


The mere passage of time makes certain objectives of tax justice harder to achieve. A large part of the tax system of the United States has become affected in this way. Two difficulties are involved. First, it is often impossible to rectify past injustice by reducing present taxes. Second, the imposition of new taxes or higher rates may bring an element of injustice merely because it represents a change in the system.

The first difficulty can be illustrated by the real estate tax. If an injustice was done in the past by some increase in the real estate tax, it may be impossible to rectify it by lowering the tax if the property has meanwhile changed hands.

The second may be illustrated by the excess profits tax. However desirable this tax may appear from many points of view of justice, it may on occasion be harmful to legitimate vested interests.

These points need not be the controlling ones in tax justice. Whoever considers them important, however, will discover that they make the tax system of the United States as it now exists more difficult to improve from the point of view of justice alone.


The Nature of Capitalization

The example just given of the real estate tax involves tax capitalization, a process that depresses real estate

[page 239]

values. Under tax capitalization, taxes that are not payable until
future years burden the taxpayer in the present year. The burden is
felt in a decrease in the capital value of some piece of property
owned by the taxpayer. The decrease is equal to the total of all the
expected future taxes on the property -- the total, that is, reduced
to a present value. The prospective tax payments of all future years
are thus said to be "capitalized" into a present amount, which
correspondingly depresses the value of the property.

Usually only a part of the tax is capitalized. Indeed, capitalization is a reflection, generally, of a difference between two taxes. It is a part of the higher tax -- the part represented by the excess over the lower tax -- that is capitalized. Moreover, the amount capitalized always depends on what buyers and sellers think that the future taxes will be, not on what the taxes actually turn out to be. Finally, capitalization may occur before the tax law is passed. The "present" year, the year of capitalization, is the year when the buyers and sellers first become aware of the likelihood of the future taxes. Consequently, capitalization may occur even if the tax is in fact never imposed, if only the likelihood is strong enough.

So long as he does not try to sell his property, the owner may not realize what has happened to its value, for capitalization can of course occur whether he realizes it or not.

If the burden of the prospective taxes is capitalized and thus rests in the present year on the present owner, no further burden develops for him or anyone else in the future years. The same tax cannot be a true burden twice -- once in the present year and again in the future years when it becomes due. Whoever owns the taxed property in the future experiences no true burden from the tax at that time, since the burden has all been concentrated on

[page 240]

the one who owned the property when the tax was proposed or enacted.

Capitalization of the Real Estate Tax

Capitalization of prospective taxes is of particular importance in the United States because of the great dependence on the real estate tax. This tax, or at least part of it, is particularly susceptible of capitalization.

The greater part of the United States seems to have passed the "present year" stage, when real estate tax rates are raised and the prospective added burden through future years is capitalized. Instead, the dominant practical issue in property taxation at the moment is rate reduction. The "future years" in the example above have arrived. In order to form an opinion on the justice of a rate reduction, it must be ascertained whether the taxes currently paid by the current owners are a true burden on these owners or whether the current taxes simply represent, in whole or in part, a burden that was borne some time ago by past owners. Thus, although capitalization is essentially an anticipatory adjustment, it has a logical place in a chapter devoted to the influence of the past.

A Hypothetical Example of Capitalization

An extreme hypothetical example may serve to clarify the issue. Suppose that, in addition to the existing real estate tax, a special tax of 2 per cent per year on capital value were levied by the city of New York on all land bordering on Forty-second Street, one of the main cross-town business arteries. Suppose further that the new tax would apparently remain in force indefinitely. Land-owners in the area affected would like to raise their rents in order to meet the new tax, but they would probably be unable to do so. If they had already been getting all the land rent that they could, they would not be able to

[page 241]

raise the rent now. Building owners using the land could not pay a
higher rent, since they could not demand higher rents from their
tenants, because, in turn, the tenants could not raise their sales
prices to their customers. The customers could and would buy their
goods and services elsewhere, from firms not too far distant from
Forty-second Street.

The ultimate result would be more complex. In general, however, the owners of the land would be faced with a perpetually lower net income from their properties, after paying all taxes including the new discriminatory tax. This outlook would correspondingly depress the selling value of the land. The decrease would, of course, be much more than the amount of one year's tax. The whole series of future tax payments would be taken into account by any prospective buyer: the impact of all these future taxes would be compressed into the present moment.

If a piece of land yields $100,000 a year to its owner, after paying all expenses including taxes, and if the going rate of interest for this kind of investment is 5 per cent, the land will sell for $2,000,000 ($100,000 a year equals 5 per cent on this sum); but, if a new non-shiftable, discriminatory tax of $10,000 a year is levied on the land, the net yield drops to $90,000 a year, and the selling value -- interest still being at 5 per cent -- drops to $1,800,000. The selling value has dropped not by $10,000, a single year's tax, but by $200,000, that is, by the whole future series of tax payments discounted at 5 per cent.

Justice under the Hypothetical Example

If the land is, indeed, sold for $1,800,000 after the new tax has been levied, the new owner cannot logically complain that the special tax is a discriminatory burden on him. The tax has not affected him financially. Had it not been levied, he would have had to pay $2,000,000 for the

[page 242]

land. Because it was levied, he was able to obtain the land for
$1,800,000. The saving of $200,000 exactly offsets the burden
represented by the new tax -- $10,000 a year.

The new owner would, of course, gain $200,000, or, in other terms, $10,000 a year for an indefinite period, if the tax were now removed. Such a gain, however, would be "undeserved." The new owner would be getting something for nothing just as truly as if the city had picked him at random from a street crowd and given him a perpetual annuity of $10,000 a year.

On the other hand, attention must be given to the unfortunate landholder who was the owner of the property at the time the tax was imposed. In effect, he was burdened by a tax of $200,000. No longer the owner of the land, he cannot now benefit by elimination of the tax. To rectify the injustice, if there is any, the city must keep the new tax in force, find the former owner of the land, and turn the yield over to him -- an unlikely procedure.

Alternative Tax-Free Fields of Investment

The amount of tax capitalization that has occurred under the real estate tax in the United States is highly uncertain. If capitalization is to occur, the tax must discriminate against one particular form of investment. If all property is taxed alike, certain repercussions on the rate of interest, the amount of saving, and the general direction of investment may result, /1/ but no capitalization, as that term is used in the present chapter, will occur.

Moreover, if the money from a special tax on special property is spent to benefit the property, little or no net change may occur in its value. For example, if a property tax on a parcel of real estate is $100, of which $40

/1/ See pp. 64-66.

[page 243]

is spent in such a way as to benefit the property enough to offset
the burden of $40 of the tax, only the remaining $60 can be
capitalized; and, if all other investment opportunities of equal
amount are taxed $25, only $35 of the $60 can be capitalized. /2/

In most parts of the United States, real estate is taxed more heavily than the other chief forms of investment -- merchants' inventories, farm animals, shares of stock, bonds, etc. Even allowing for the benefits received by real estate from governmental services, /3/ some part of the increasing property tax of the past few decades has probably been capitalized and has correspondingly depressed values below the levels they would otherwise have reached. To this extent, the tax now rests as a true burden not on every present owner of real estate, but on two groups of people: (1) such present owners as have continued to hold their properties during the period of rising taxes and (2) former owners who sold after the tax increases had exercised their depressive effect on values.

More capitalization is possible for a tax on land than for a tax on buildings. Land usually wears out slowly or not at all, while buildings must eventually be replaced. Presumably no one will erect a new building to replace an old one unless the consuming market is so strong that rentals will show an indication of covering all costs, including the increased real estate tax. The tax on the new building, then, will tend to be shifted to tenants in the form of higher rentals. To the extent that shifting exists, there is no occasion for capitalization.

Proportion of Old Owners, and Past Stability of Tax Rate

Reduction of the real estate tax may on occasion give a bonus to new owners, but this reduction may be considered necessary in order to treat fairly the old owners

/2/ See Note 1, Relatively Tax-Free Fields of Investment, p. 549.

/3/ See Note 2, Capitalization, Benefit, and Shifting, p. 550.

[page 244]

who have held throughout the rising tax period. Thus, unless a way is
found to distinguish between old owners and new owners, a lowering of
the tax rate may represent justice for one group and something more
than justice for another. The longer the tax has been in force at an
unchanged rate, the greater the proportion of new owners will be, and
the less the pressure, from the point of view of the resting place of
the burden, to reduce or abolish the tax.

Data on the degree of stability of property tax rates throughout the past ten years are consequently of interest in a discussion of capitalization. Details presented elsewhere show that there has been no appreciable rise in state property tax rates in many states since 1925, or in the local rates in several large cities over a somewhat shorter period. /4/ Increases in the ratio of assessed valuation to true valuation may have caused a real increase in the burden in some of these cases. On the whole, however, the real estate tax load seems to have been constant or decreasing long enough so that a large proportion of present owners consists of those who have capitalized, in purchasing, much of the present property tax.

Capitalization of Other Taxes

The effects of tax capitalization in the United States have not been thoroughly studied. Scientific discussion of the subject has been confined to the property tax, and even here it has been scanty. Further study might reveal the extent of capitalization in several other taxes, such as the "severance taxes" in some states on the extraction of natural resources, chain store taxes with their depressive effect on the price of chain store securities, some part of the corporation income tax with its depressive effect on corporation shares compared with bonds and

/4/ See the memorandum by Thomas J. Reynolds on state property tax rates in the forthcoming volume, Studies in Current Tax Problems, and Note 4, Fluctuation in Urban Property Tax Rates and Assessed Valuations during the Depression, p. 525.

[page 245]

with partnership interests, and such high-rate commodity taxes as
those on liquor and tobacco, so far as they prevent the manufacturers
from expanding to a more profitable size. Moreover, special commodity
taxes or service taxes of any kind may result in tax capitalization,
with effects on ownership interests. These taxes present a somewhat
different problem from a real estate tax, particularly a land tax.
Their effects in reducing the volume of sales will be discussed in
the following paragraphs.


If a special tax is imposed on a particular commodity, its producers or distributors may have to raise the price so high, in order to get the money to pay the tax, that they meet with stiff resistance from consumers. They may find that unit sales, and perhaps even dollar sales are seriously curtailed. Some firms may have to go out of business, and all firms will probably be affected to some extent. If the tax is imposed in a period of business improvement, sales may fail to decline, but unit sales will still be less than they would have been without the tax. The restrictive effect is felt not only by the producers and distributors of the taxed commodity but also by others, such as suppliers of raw materials.

The injury resulting from the restrictive effect can be avoided in part if the tax is imposed so gradually that the decrease in consumption is matched by part of the depreciation of capital that would occur anyway. By the time the depreciation has occurred, the investors are on notice not to replace all of the capital in that particular business.

The kind of tax in question here cannot be paid out of existing profits, either because its scope includes firms without profits or because it is too large. It must, there-

[page 246]

fore, be passed on, in whole or in part, to someone else -- usually
to consumers, through increased prices.

Effect of the Passage of Time

If the tax has been in force for a considerable period of time, its repeal or a reduction in its rate may do little to help the stockholders, creditors, wage earners, etc. who have been injured by its repressive effect. Therefore, no such repeal or reduction may be called for on the grounds of justice.

Two types of firm must be distinguished -- those that were driven out of the business completely, and those that have stayed in on a reduced scale of operation. A third class is conceivable -- firms that stayed in with sales unaffected by the tax. In practice, however, probably few of this type could be found.

The firms driven out of business are within a short time -- perhaps a year or two -- beyond help by a repeal or lowering of the tax. The owners and creditors will probably have scrapped, or sold at a heavy loss, the specialized assets, such as certain kinds of store fixtures and machinery. The unspecialized assets, such as warehouses and factory buildings, will have been turned to another use or allowed to deteriorate through abandonment. The special skill of the personnel -- financial heads, factory managers, skilled employees, etc. -- will not have vanished, but many of these people will have found occupation elsewhere.

As to the firms that have remained in the business, at least two questions are pertinent.

First, has that part of the firm's capital equipment and labor force that was rendered idle by the tax been scrapped or otherwise made unavailable to handle the business that might return under rate reduction or repeal? If the answer is yes, little opportunity exists to rectify past injustices by tax relief. Of course the owners and creditors

[page 247]

would not object to tax relief. Expanded sales might give an
additional profit even though new capital had to be raised.
Nevertheless, most of the damage would be beyond repair.

Second, has the business changed hands since the tax was imposed? If so, was the price so low that the buyer was clearly assuming that the tax would continue indefinitely? If the answer is yes, the process of tax capitalization has made it impossible to eliminate past injustices simply by reducing the tax.

Effect of Certain Existing Taxes

Practically all the commodity and service taxes levied by the federal government have been in effect uninterruptedly only since 1932. Perhaps tax reduction might yet aid many of the owners, creditors, etc. who were injured by whatever depressive effect on sales the taxes exercised. On the other hand, most of the taxes are at rates that represent less than 5 per cent of the retail price. With rates so low, a marked effect on sales is unlikely.

The three commodity taxes that carry heavy rates are those on cigarettes, spirits, and gasoline. The motor vehicle license tax, also, is substantial.

Taxes on Cigarettes

The federal tax on "small," or "Class A," cigarettes (the usual size) is equivalent to 6 cents on a package of twenty. However, the rate has been unchanged since 1919. Ownership of the largest companies has subsequently changed hands, to a considerable extent, through the sale of shares on the stock exchange. Presumably, the purchasers of cigarette company shares have been buying on the assumption that the tax would not be reduced or repealed. All these factors indicate that in its relation to the present cigarette manufacturing firms, the depressive

[page 248]

effect on sales, although substantial, may perhaps not be considered
a just reason for lowering the cigarette tax.

The rate of the cigarette tax is indeed high compared with the rates on some other commodities -- cigars, for example. The cigar tax is equivalent to about 4 per cent of the retail price, including tax, while on the same basis the cigarette tax is equivalent to nearly 50 per cent of the retail price. A mere comparison between commodities, however, is misleading from the point of view of justice to the producers and distributors. The persons affected, not the commodities taxed, must form the basis of comparison. Lowering the tax rate might not rectify the injustice in certain parts of the tobacco industry; in other parts it might.

Twenty states now levy taxes on the wholesale or retail sale of cigarettes within their borders, at rates ranging from 1.3 cents to 5 cents a package. Merchants in the taxing state are often affected when the cigarette manufacturers are not, since many smokers curtail their local patronage but purchase in tax-free states or through tax- free channels of interstate commerce. One of the taxes was first imposed in 1936, three in 1935, none in 1934, and one in 1933. Most of them, therefore, have been in force for a period of time that is considerable, from the point of view of the present discuss1on.

Taxes on Spirits

The federal tax rate on distilled spirits is high compared with most federal commodity tax rates: $2.00 a proof gallon -- that is, $2.00 on a gallon containing 50 per cent alcohol. It has been in effect since the repeal of prohibition. None of the business that legal producers are now losing because of the tax could have originated before the tax was imposed. The same consideration applies to the spirits taxes levied by so many of the states,

[page 249]

at rates that commonly range from 80 cents to $1.00 a gallon.

Taxes on Gasoline and Motor Vehicles

To the federal gasoline tax of 1 cent a gallon must be added state taxes ranging from 2 cents (Connecticut, District of Columbia, Missouri, and Rhode Island) to 7 cents (Florida and Tennessee), and, in a few states, county and city taxes. In a sense, however, the high rates may have increased the sale of gasoline instead of tending to decrease it, since most of the gasoline tax revenues have been used to build and maintain highways.

In a similar sense, automobile license taxes may have increased the sale of automobiles, by helping to build highways. The rates are fairly substantial compared with ordinary commodity tax rates. For example, the New York State tax amounts to $15 a year on a private passenger car weighing 3,000 pounds unladen. Such a car sells for about $1,000. If it is given a life of seven years, the seven taxes of $15 each may be discounted at 5 per cent, to give a total present- value tax burden on the new car of $87, which is 8.7 per cent of the retail price.

Other special taxes are either at rates too low, or of too restricted an application, to justify comment here.


The discussion up to this point has related only to proposals to maintain, reduce, or eliminate a tax. Consideration will now be given to proposals to increase a tax or to levy a new tax. The point at issue is the "innocent vested interest" that would be harmed or destroyed.

In general terms, the problem is one of consistency of action on the part of the government. Suppose that those who engage in a given activity have for many years been left untouched by any special tax, and that no indication has been given that any such tax will ever be imposed.

[page 250]

If the government now suddenly reverses its policy and, moreover,
claims to base its action primarily on grounds of justice in the
distribution of the tax burden, the taxpayer may claim that the
government is illogical, and is defeating its own end. If, however,
the government's proposal is supported by arguments that do not bear
on justice -- for example, an assertion that the change will increase
the productivity of the economic system --, the presence of an
innocent vested interest cannot prove the government's position to be
illogical. The matter simply becomes one of comparing two or more
basically different aims.

Real Estate Tax

In many jurisdictions, innocent vested interests have been built up that would be injured by an increase in the property tax rate. During the past decade most real estate taxpayers have probably had little occasion to expect markedly higher tax rates. A general tendency toward stability in rates has been manifest, with a growing movement toward rate reduction in some states. Consequently, those who have purchased real estate, or, what comes to much the same thing, have refused to sell their holdings, can hardly be accused of improvidence if they have done so on the assumption that tax rates would not rise.

Any increase in the rate would be all the more serious, from the point of view of justice, because of the tendency for at least part of the increase to be capitalized. Capitalization is not, however, so important in the distribution of the burden when the tax rate is unexpectedly increased as it is when the rate unexpectedly decreases. /5/ An unexpected rate decrease discriminates between two groups of property owners: (1) the old owners who have held their property through one or more unexpected rate increases,

/5/ See pp. 241-42.

[page 251]

and (2) the new owners who have bought only after the rate increases
and hence have been able to allow for part or all of them in the
purchase price. An unexpected rate increase, on the other hand,
treats these two groups of owners alike. In so far as property
ownership is widespread in the community, the injustice of an
unexpected rate increase is considerably mitigated.

Income Tax

If the rate of the personal income tax is raised unexpectedly, no vested interest is disturbed for most taxpayers. However, an investor who has just put some money in a risky venture may claim that a sudden rise in income tax rates to very high levels is unfair. The government does not compensate him if he loses, and it now proposes to take far more of his gains if he wins than he had thought it would when he made his decision. The odds, so to speak, have shifted unexpectedly.

Similarly, an unexpected rise in rates may be relatively hard on the person whose income is irregular. Perhaps he has entered into investments or activities where the gain or fee, though practically certain, is realized in a lump sum at the end of a number of years instead of piecemeal year by year. He is thus forced far up into high surtax rates, except for capital gains under the present federal law. Changes in the definition of taxable income may disturb vested interests -- for example, changes in capital gains and losses provisions, and introduction of a tax on undistributed profits. In general, however, the income tax usually has less relation to the vested interest problem than the real estate tax.

Excess Profits Tax and Customs Duties

Aside from the property tax and the income tax, the vested interest problem is of immediate importance in the United States chiefly in connection with the excess

[page 252]

profits tax and the customs duties. Both taxes are liable to sudden
and marked change. Both of them raise complex problems of justice in
relation to consistency of action.

An excess profits tax is necessarily impersonal, since an "excess" profit is "excess" no matter who gets it. /6/ The resulting conflict with vested interests creates a perplexing problem in justice.

Suppose that investor A puts $100 into a new concern, that twenty years later his share of the earnings amounts to $1,000 a year, and that A then sells his share in the business to B for $20,000. Suppose that an excess profits tax is now unexpectedly imposed on the concern, decreasing the net profits, after taxes, to such an extent that B gets only 2 per cent on his investment instead of the expected 5 per cent. The person who really benefited by the excessiveness of the profit was A, and he escaped the tax entirely because he happened to sell out before anyone suspected that the tax would be introduced. Investor B, on the other hand, is deprived of part of what he regards as his normal and justifiable return of 5 per cent.

If the excess profits tax is being imposed because it will bring certain economic benefits to the community at large or will relieve certain classes from an undue burden that has nothing to do with the existing tax burden, B's complaints may be logically over-ruled. If, on the contrary, the excess profits tax is proposed simply as a means of distributing the existing tax burden more equitably, B has a forceful argument. How forceful it is will depend on the circumstances -- for example, the number of present investors of the A type and the number of the B type, and the number of the B type that are really not investors in a long-term sense, but speculators who may be reasonably asked to bear the consequences of shifts in governmental policies as they expect to bear the conse-

/6/ See p. 273.

[page 253]

quences of other unforeseen events. Perhaps a personal element could
be injected into the excess profits tax by distinguishing among
various owners in the same business, and exempting the profits going
to those who purchased their shares within a certain period of time.
Such an attempt would of course be out of place if the purpose of the
tax were something more than an equitable redistribution of the
existing total tax burden.

The present federal excess profits tax can hardly be discussed from any of these viewpoints since it is primarily designed as a check on evasion under the capital stock tax. /7/ The combination of the two taxes is also useful in levying a special tax on unforeseen gains -- a matter that does not involve vested interests. If a genuine attempt is made to tax excess profits from monopoly, however, the vested interest problem will have to be faced.

The protective tariff has almost certainly lowered the average standard of living in the country. Nevertheless, few economists wish to see it completely and immediately removed. Partly, their caution arises from fear of the effects of any sudden transition on the productive power of the community. Partly, too, it arises from a desire not to deal too harshly with the large amount of investment that has been built up behind the tariff walls on the assumption that protection would be continued.

Importance of Vested Interests

No one would argue that vested interests, however legitimate, must forever bar changes in the tax system that are desirable either from other points of view of justice or for economic or political reasons. Indeed, the term itself has acquired a connotation of undue assertion, by the few, of highly technical legal rights, at the expense of justice to the many. Almost any tax change will necessarily disturb some vested interest. /8/ The chief prob-

/7/ See Note 9, Present Excess Profits Tax, p. 560.

/8/ See Note 3, Remote Vested Interests, p. 551.

[page 254]

lem is usually one of choosing among the different interests, and
deciding upon the speed of change. From the point of view of tax
justice, destroying a vested interest gradually over two generations
is a very different matter from destroying it suddenly.

[page 255]

Chapter 18


Persons who directly use and benefit from certain governmental services should pay a benefit charge, according to a generally accepted standard of tax justice. A benefit charge directly burdens only the beneficiary: for instance, the fare that a passenger pays on a municipal street-car, or a gasoline tax if the revenue is devoted to highway construction and maintenance.

Benefit charges differ greatly. Some require the direct beneficiaries as a group to cover the entire cost of the service; others require them to cover only part. The cost can be divided among the individual beneficiaries in any one of several ways. In some cases great pressure is put on the direct beneficiary to use the service; in others, he is free to reject the service and the charge that goes with it.

Several classes of benefit charge are common in the United State. They are (1) prices, as when the localities operate street-car systems as a private enterprise would; (2) fees, as when the localities charge for marriage licenses; (3) special assessments, as when the localities charge the real estate owners for laying sidewalks or installing sewers in certain neighborhoods; and (4) benefit taxes, as when the state levies a gasoline tax on all users of the highways. All these are obviously user charges -- the cost is paid, at least in part, by the user.

Closely akin to the user charges is something that might be called a "responsibility charge." Thus, the pro-

[page 256]

duction and consumption of liquor can be shown to be responsible for
a certain amount of social expense -- the expense of law enforcement,
for example. A special tax on liquor is therefore sometimes supported
on the grounds that those who are responsible for the expense should
defray it -- the business should "pay its way." Aside from the liquor
tax and certain minor state and local license charges, however, this
reasoning has been given little application. Moreover, in practice,
no effort has been made in the United States to determine the height
of the liquor tax rate by the magnitude of such social expenses.


Benefit charges express an attitude that existing differences in wealth, income, and opportunity shall be taken for granted. In other words, the distribution of wealth and income is not likely to be greatly changed if the government finances new activities on a benefit basis. According to benefit financing, the user of the service must pay the government for the cost that he causes it by demanding the service. In its extreme form this doctrine assumes that if the would-be user is too poor to meet the cost, he cannot have the service.

If a peddler, for example, is so poor that he is unwilling to pay a license fee charged to cover the cost of regulation, the philosophy of the benefit charge says that he has no place in the business. If a would-be motorist is so poor that he is unwilling to help build and maintain a substantial part of the highway system by contributing revenue through the gasoline tax, he is deprived of the use of the highways. The same can be said about the other chief benefit charges in use in the United States: fares for street railways; charges for municipally supplied water, electricity, and gas; charges for the use of the mails; special assessments; wage taxes to build up reserves for old-age pensions; and perhaps even part of

[page 257]

the real estate tax, in so far as it covers the expense of such items
as garbage and ash collection and fire protection. /1/ At least a
trace of the buying-selling relation that characterizes private trade
is always found in a benefit charge: to get the thing, the user must
pay for it, or at least for part of it.

In so far as a governmental activity is not financed by benefit charges, the result is a tendency to change the distribution of wealth and income. The distribution is changed, that is, from what it would be if the government did not undertake the activity. For example, the greater part of elementary education in the United States is not financed by direct benefit charges. The distribution of wealth, income, and opportunity is therefore different from what it would be if elementary education were left to private business and so were financed by fees based chiefly on the cost that each pupil causes. /2/

If the benefit charge is not used, the economic gaps between groups in the community may widen as well as narrow. In other words the distribution may be from the poor to the rich, through taxation of the poor to support services rendered to the rich. For some services, a change in distribution is likely to occur because of the technical impossibility of allocating benefit or cost. For example, the cost of most of the police force cannot, for technical reasons, be allocated among the beneficiaries of this service on the basis of cost caused.


A relatively small part of the present expenditures of the federal government is financed by benefit charges. The postal revenues, totalling between $600,000,000 and $700,000,000 a year, /3/ are the most important example.

Some advocates of the recently imposed federal gaso-

/1/ See Note 1, Minor Examples of Benefit Taxation and Non- Benefit Earmarking, p. 552.

/2/ See p. 206.

/3/ Annual Report of the Postmaster General, Fiscal Year Ended June 30, 1935, p. xi.

[page 258]

line tax support it on the ground that the federal government spends
an equal or greater amount on highways. Thus, the persons who bear
the burden of the gasoline tax are given a special benefit in return.
A similar argument has been offered for the federal manufacturers'
excises on the production of automobiles, tires and tubes,
accessories, and lubricating oil. The federal gasoline tax, at 1 cent
a gallon, produces much more revenue than the other automobile
excises combined. In the fiscal year 1936, the former produced
$177,000,000, compared with the latters' $122,000,000.

In the future, if the existing social security law remains in effect, the taxes on employees and on payrolls to finance the contributory old-age pension plan will be substantial benefit charges.

The state governments are getting $1,000,000,000, or more than one-third of their annual revenues, from taxes that are levied as benefit charges -- the gasoline tax and the motor vehicle license taxes. About 18 per cent of the gasoline tax revenues in 1935 and 12 per cent of the motor vehicle tax revenues in the same year, however, were used for other than street or highway purposes. /4/ The only other benefit charge of importance as a state revenue is the unemployment compensation taxes in force at the end of 1936 in about two-thirds of the states.

The local governments, in contrast with the federal and state governments, obtain an appreciable amount of money from regularly recurring revenue other than taxes.

The ninety-four cities having a population of over 100,000 reported for the fiscal year 1934 a total revenue of $2,700,000,000. Ten per cent of this amount came from the earnings of public service enterprises, and another 6 per cent from interest, rents, highway privileges, and

/4/ These figures are derived from two statements issued by the Department of Agriculture, Bureau of Public Roads: Disposition of State Motor-Fuel Tax Receipts, 1935, issued November 9, 1936; and Disposition of State Motor-Vehicle Receipts, 1935, issued November 9, 1936.

[page 259]

earnings of general departments. /5/ This latter group of revenues is
also important to state governments and to other local units. /6/

Special assessments are levied on real estate to defray certain types of expenditure that tend to enhance the value of real estate in a given area. At times they have been of major importance, especially in city and special district financing. Lately, however, this method of financing has played a distinctly minor role, at least in large cities. In 1934 the same ninety-four cities noted above collected only 2 per cent of their revenues from special assessments.

The most important benefit charge, from the point of view of total revenue, is probably that part of the local real estate tax that can be justified on a benefit basis, but no data are available to indicate its size for the country as a whole. /7/

Some of the proponents of the corporation income tax look upon it as a sort of benefit tax, though not so specifically linked to benefits as the charges just noted. The government grants the favor of incorporation, and it maintains by its authority the framework of an orderly business life within which the corporation works. The corporation income tax, the argument runs, merely gives the government its due compensation for these services.

This point of view is exemplified by the New York State tax law. Corporation profits are taxed at 6 per cent and, with unimportant exceptions, dividends received from taxed corporations are not exempt from the personal income tax. The federal Revenue Act of 1936 made dividends subject to the normal tax for the first time. In

/5/ Department of Commerce, Bureau of the Census, Financial Statistics of City Governments: 1934, Processed Press Release of November 27, 1935, p. 1.

/6/ Department of Commerce, Bureau of the Census, Financial Statistics of State and Local Governments: 1932, Table 1, p. 7.

/7/ see Edwin H. Spengler, "Is the Real Estate Tax a Benefit Tax," Memorandum Number Five, Report of the New York State Commission for the Revision of the Tax Laws: Submitted February 15, 1932, Legislative Document (1932) No. 77.

[page 260]

view of the confusion of issues with the undistributed profits tax,
it is not clear that Congress has definitely decided to regard the
normal corporation income tax as an impersonal tax rather than a
means of complementing the personal income tax. /8/ Because of the
extremely general nature of the benefit that is involved, the
corporation tax as a benefit tax will not be further analyzed in this

Extent to Which Cost Is Covered by the Charge

Benefit charges in the United States today fall between two extremes. At one extreme is the purchase-and-sale relation of private business. At the other is the lack of direct connection between payment and service that characterizes the major part of governmental activities.

To find out how near to either extreme any particular charge lies, it is pertinent to ask: does the revenue collected equal the amount that the government spends on the service? If income equals outgo, the activity as a whole is, superficially, close to being on a business basis. If income greatly exceeds outgo, the so-called benefit charge is really in part a general tax. If income falls far short of outgo, the deficit part of the expenditure must be deemed a matter of such general public interest that it can be charged against the general tax fund.

Judged by this test, most of the major types of benefit charge in the United States either are not fully on a business basis -- for example, the automotive taxes in the states -- or are on a basis not clearly defined as yet -- for example, the taxes in the Social Security Act.

Highway Taxes

The federal government's revenue from the gasoline tax equalled 75 per cent of the total federal-aid highway expenditures (regular aid, plus emergency and public

/8/ See pp. 306-7.

[page 261]

works aid) in the fiscal year 1933, increased to 90 per cent in 1934,
and dropped to 58 per cent in 1935. If the excises on lubricating
oil, tires and tubes, automobiles, and parts and accessories are
added, the percentages become, respectively, 105, 133, and 96. /9/
Over the past decade as a whole, federal automotive taxes of all
kinds have fallen far short of federal highway expenditures. The
discrepancy is justified by some observers on grounds of national
defense, postal use, and work relief.

The element of general public interest in federal highway construction cannot be isolated from the individual benefit element. It is not as if so many miles of highway, costing so much, were clearly to be paid for on a business basis, while another stretch, with its own definite cost, was clearly chargeable in full to national defense. Not even the work relief element can be isolated. Every mile of highway that the government has built in the past few years probably obviated some unemployment that might have called for work relief.

In 1929 in several States, state revenues from the gasoline tax and the motor vehicle license tax were covering all highway expenditures by the state road department. The distinction between state highways and local highways is, however, somewhat artificial. Some states were in addition giving money to local units to spend on local roads. From a broad point of view, therefore, the state and local roads may be considered together, and state and local benefit charges may be considered together. In 1929 current state and local benefit charges for highways outside cities were covering only about half the current construction and maintenance costs of such highways, but this proportion would be materially increased by regarding part of the rural real estate tax as a benefit road tax. Adding federal aid for highways to the receipts does not change the percentage greatly. The proportion

/9/ Federal Aid for Highways, 74 Cong., 2 sess., Senate Report No. 1976, p. 14.

[page 262]

varied from state to state, but in no state did motor tax receipts
equal or exceed the state and local expenditures on rural highways.

The situation has of course changed somewhat since 1929. State and local highway expenditures were cut during the depression. Moreover, as noted above, /11/ in some states part of the revenues from the state gasoline tax or the motor vehicle tax have been formally allocated -- "diverted" -- to some non-highway fund, such as the general fund or the education fund. Even in these states, however, no real diversion exists if total state and local highway expenditures, minus federal aid for highways, are greater than total highway benefit taxes, including those diverted revenues.

Some special business taxes and special license taxes on motor vehicles not used for pleasure have been developed in recent years in several states. These are sometimes levied at high rates, but the aggregate revenue is apparently still small. /12/

Postal Charges

The Post Office Department as a whole is run either at an appreciable loss or without loss, depending on whether certain controversial items are included or excluded from the total expenses properly chargeable against the sale of stamps. At least some of the expenses, such as those incurred in the free carriage of mail matter for Congress and all other governmental agencies, reading matter for the blind, and newspapers for delivery within the county of publication, appear logically chargeable against taxpayers in general, not against purchasers of stamps. Even if the two groups of people are

/10/ "Report of the Committee of the National Tax Association on Taxation of Motor Vehicle Transportation," Proceedings of the . . . National Tax Association, 1930, pp. 138-43.

/11/ See p. 158.

/12/ See Note 2, Special Minor Highway Taxes, p. 553.

[page 263]

almost identical, the burden is of course distributed differently.

Old-Age Benefit Taxes

The federal old-age benefit payments are being financed by two "taxes with respect to employment." Whether the old-age benefits are in fact being fully financed on a benefit-charge basis depends upon at least two important factors. First, does the burden of both taxes really fall upon the employees who eventually get the benefit? Second, will the revenue from the taxes actually be adequate?

One of the taxes is legally levied on the employee ("income tax on employees"), and the other is legally levied on the employer ("excise tax on employers"). In both taxes, however, the employer is responsible for the payment to the government. The only difference is that in the employees' tax the employer can deduct the tax from the wage, while in the other tax he cannot. He may, of course, set the wage at a lower level than he would have done if no tax had been levied. In many cases he will not be able to pay his "excise tax" out of his existing profits and will have to get the money from consumers by charging higher prices, or from suppliers by paying lower prices, or from employees by lowering wages. In the absence of special study on the allocation of the burden of this tax, no definite statement can be made.

The taxes apply to wages paid after December 31, 1936, and mount gradually from a combined rate of 2 per cent in 1937, 1938, and 1939 to a combined rate of 6 per cent in 1949 and thereafter. In each year the rate is equally divided between the employees' tax and the employers' tax. Uncertainty of course exists on the adequacy of these rates. They are expected to build up a reserve fund so large that the interest from it, together with the annual yield of the two taxes, can meet the benefit payments specified in the law. If the amount thus obtained is in-

[page 264]

sufficient, the present law requires Congress to make up the balance
from some other source. The amount of the old-age benefit payments is
not made dependent on the amount of money received from the two
taxes. Indeed, the proceeds of the taxes are not specifically pledged
to the old-age fund at all, but no doubt exists, of course, that they
were enacted with the fund in mind.

The "old-age benefits" must be sharply distinguished from the "old-age assistance" payments. The latter are made by the states with the help of federal money, and they presumably go only to needy persons. Like relief payments in general, the old-age assistance payments cannot, by their nature, be financed on a benefit-charge basis. If the beneficiaries were able to defray the cost, they would not need the payments. The federal government has made no special provision at all for financing its share of the assistance payments. It may be able to meet a considerable part of the charge by borrowing from the reserve fund built up for the old-age benefit plan.

Unemployment Compensation Taxes

The unemployment compensation plan provided by the Social Security Act is even more uncertain than the old-age benefit plan in its benefit-charge aspects. The federal "tax on employers of eight or more" levied by the act is based on payrolls. Before allowing for the credit to be noted below, it rises from 1 per cent in 1936 to 3 per cent in 1938. None of the money, however, will be used to make unemployment compensation payments.

The chief purpose of the tax is to induce states to levy similar taxes. To this end, the law allows such state taxes as a credit against the federal tax otherwise due, with certain limitations. For business depressions as severe as the recent one the funds that can be raised by the state taxes will almost certainly prove inadequate for

[page 265]

even the unemployment relief load, to say nothing of the total
unemployment load, including those unemployed who are not in need of

Special Assessments

A local special assessment may sometimes be large enough to cover the cost of special benefit projects such as the laying of a sidewalk or the lighting of streets. In some cities, however, as in New York City, a certain improvement may be financed only in part by a special assessment on the property in the specially defined benefit area. The rest of the cost will be met by an addition to the real estate tax rate applicable to all the property in the borough or the city. The assumption is that the special benefit element is covered by the special assessment, and the element of general public benefit by the general rate increase.

In recent years many local improvements of a kind that often stimulate local land values seem to have been financed entirely without the aid of special assessments -- mainly by federal grants and federal loans.

Real Estate Tax and Miscellaneous Fees

The local real estate tax bears an interesting relation to such services as garbage and ash collection and other local services that give some sort of special measurable benefit. The revenue from the local real estate tax alone is probably more than large enough in almost all the communities in the United States to cover the total cost of garbage and ash collection, sewer service, and fire protection. It is doubtful whether the revenue in some communities is large enough to cover also the cost of police protection and other services that some observers think are of measurable special benefit. Opinion differs whether these services can be logically financed by a particular kind of tax, or whether, on the contrary, they are more

[page 266]

like the services of a battleship, which cannot be allocated. /13/

Local prices and fees, from the scattered evidence available, are not far from covering the cost of the services in question, but practice varies widely from one part of the country to another.

In summary, the revenue system of the United States -- federal, state, and local -- reveals relatively few instances where a certain kind of governmental service is paid for wholly by the particular group of persons whom the service is supposed primarily to benefit.

The Distribution of Benefit Charges among Members of
the Benefited Group

Another question may be asked in comparing benefit charges with the purchase-and-sale relation of private business. How is the benefit charge distributed among members of the benefited group? If a benefited group has been found, and if part of or all the cost of the service is to be covered by a special charge on the group, some principles must be devised to determine how much one member of the group shall pay compared with another.

Postal Charges

In a way, the question is simply a refinement of that discussed above. /14/ For example, the postal service as a whole is practically self-supporting. Whether it is really run on a benefit basis, however, depends on whether each of the sub-activities -- the first- class mail, the second-class mail, and so on -- is supported on this basis. Official data indicate that they are not. First-class mail paid $109,000,000 more than enough to cover its estimated share of the cost in the fiscal year 1935. The estimate is, however, very uncertain because of the difficulties of al-

/13/ See Note 3, Development of the Benefit Theory in Property Taxation, p. 553.

/14/ See p. 160.

[page 267]

locating certain costs. /15/ The deficits shown were as follows:
second-class mail, $86,000,000; third-class mail and fourth-class
mail; each, $21,000,000; foreign mail, $33,000,000; special services,
$13,000,000; and penalty and franked mail, also $13,000,000.

To what extent the users of the various classes constitute different groups is not known. In any event, some important economic problems are raised by the encouragement of certain forms of activity (for instance, newspaper publishing) and the discouragement of others (for instance, certain kinds of direct-advertising campaigns).

Automotive Benefit Charges

The other benefit-charge items show no such clear departure from the benefit-charge basis. For some of them no comprehensive attempt to split up the total cost among the various sub-services has been made. Although highway engineers, for example, have carefully studied the relative part of total highway costs that can be attributed to various types of truck, bus, and pleasure vehicle, no statement for the country as a whole shows how the costs caused by a particular type of vehicle compare with the revenues received from the vehicles of that type.

The gasoline tax is the major highway charge. It is only a partially accurate instrument for distributing burden on a cost- causing basis. Fuel consumption, within certain ranges, increases at a slower rate than speed, and wear and tear on the road may increase faster than speed. Therefore, fuel consumption and gasoline tax paid may increase at a slower rate than the wear and tear on the road. Fuel consumption also may not keep pace with the damage caused by an increase in weight, and it fails to take adequate account of differences in tire equipment.

/16/ Annual Report of the Postmaster General, Fiscal Year Ended June 30, 1935, pp. 102-4.

[page 268]

In addition, much of the highway construction is to accommodate the
occasional driver, who is responsible for the peak loads on week-ends
and holidays, and who pays little gasoline tax in the course of a

The gasoline tax alone, therefore, cannot distribute highway costs in proportion to the responsibility for them. Most of the states supplement it by a substantial annual motor vehicle registration tax. Under this tax the amount payable depends not on mileage but usually on gross weight (weight of vehicle plus weight of estimated or actual load). Some states use net weight or horse-power.

The federal excise taxes on producers of automobiles, tires and tubes, etc. are not very helpful in distributing highway costs on a cost-caused basis. /16/

Other Benefit Charges

The amount of cost caused by each beneficiary is of course easy to discover when the cost consists of money payments made directly to the beneficiaries. From this point of view, the old-age benefit plans present no such problem as the gasoline tax and the motor vehicle tax. The last two taxes must be studied carefully to determine how well they measure each aspect of wear and tear on the highway.

The results of the federal unemployment compensation legislation depend upon the response of the states to the pressure put upon them by the federal general-fund tax of 3 per cent on payrolls.

Special assessment practice varies considerably in its details from one locality to another, but, in principle, the allocation is to be made on the basis of relative benefit as measured by increase in property values.

/16/ See Note 4, Federal Automative Excises as Distributors of Benefit Taxation on a Cost-Caused Basis, p. 554.

[page 269]

The Degree of Compulsion on the Beneficial

The degree of compulsion that is exercised on the so-called beneficiary to accept the service and pay the share of the cost allocated to him is another important question. If the amount demanded by the government is no more than the cost of the service, the government is at first sight not levying a general tax. However, if the beneficiary does not want the service -- indeed, might pay something to avoid receiving it --, the tax must seem to him as much a general charge as his income tax. Superficially, the transaction may be on a business basis. From the beneficiary's point of view, it is not.

Various degrees of compulsion are found in benefit charges. Frequently the compulsion is like that exercised in collecting general taxes. A property owner who finds himself in a minority cannot refuse to take the service and thus avoid a special assessment voted by the majority. He must pay his share or forfeit his property. The same lack of option to the individual is to be found under the property tax. Any one property owner must accept and pay for the garbage collection service, for instance, whether he wants it or not. The old-age benefit plan, too, gives as little option as the special assessment. A large majority may not wish to save for their old age, but their alternative is to give up their jobs and seek employment in some activity that is not covered by the Social Security Act.

The highway taxes offer somewhat more freedom of choice since no one need pay them if he stays off the highways. In a broader sense, however, it is doubtful whether anyone escapes all of either of these taxes. Everyone must buy clothes, food, and shelter that have been transported in whole or in part in trucks that have paid highway taxes. The taxes must usually be shifted

[page 270]

-- to consumers or others -- if the business concerns are to continue
to function.

The Post Office charges are about on the same level of compulsion as the highway taxes. Both of them represent compulsion similar to that exercised by a private monopolist of a widely used commodity or service. So do many of the local fees (for example, the marriage license fee, court fees) and prices (for example, charges for government-supplied electricity and water).

[page 271]

Chapter 19


Special taxes are commonly advocated on gains derived by methods that violate prevailing standards of economic morality. /1/ Outstanding examples are unexpected gains (windfalls) and monopoly profits.

The present tax system in the United States places relatively little emphasis on this source of gain.

The federal Revenue Act of 1936 imposes a tax on "unjust enrichment." This is aimed chiefly at business firms that profited from the decision of the Supreme Court that held the Agricultural Adjustment Act invalid. The "unjust enrichment" taxed by the act is that of firms that got more gross receipts than they otherwise would have, by including a processing tax in their selling price, and then avoided the expense of the tax either by not paying it, or by getting a refund or credit from the government, or by being reimbursed for the tax (when it was removed) by vendors who had passed it on to these firms.

The existing federal "excess profits tax" was designed primarily to prevent evasion of the capital stock tax. /2/

The Vinson Act of 1934 requires contractors or subcontractors to relinquish to the federal government any profit in excess of 10 per cent of the total price in any contract over $10,000 for constructing naval vessels or aircraft, or parts for them.

Under the Gold Reserve Act of 1934, the federal government was able to obtain devaluation profits by taking gold from private hands and paying for it in non-redeem-

/1/ See Note 1, Economic Implications of Taxes on Excess Profits, etc., p 555.

/2/ See Note 9, Present Excess Profits Tax, p. 560.

[page 272]

able money. This procedure, although equivalent in many respects to a
100 per cent tax on the profits from gold devaluation that would
otherwise have accrued to private interests, is not usually thought
of as a tax.

The Silver Purchase Act of 1934 imposed a "transfer tax" on all transfers of silver bullion, the amount of the tax being 50 per cent of the profit on the transaction. The purpose was to prevent speculators from benefiting greatly from the change in the government's silver buying policy.

Like the federal government, the states have refrained from basing taxation to any large degree on the moral implications of income. Local governments have also rejected proposals for such special taxes as the land increment tax and the single tax on land rent.

Nevertheless, a detailed examination of the problem is advisable -- especially because of the effectiveness of the federal excess profits and war profits taxes during the World War, at least in raising revenue, /3/ and because of the growing discussion of monopoly profits.

Some gains that have been generally considered undesirable from society's point of view have not been the object of agitation for special taxes. Little sentiment exists for special taxation of gains from illicit operations, such as illegal manufacture and sale of liquor, burglary, kidnaping, and prostitution. Receipts from these sources are taxable under the ordinary income tax. Perhaps a special tax on such activities would be considered too clear a confession of failure to suppress them more directly. Moreover, the illicit income is subject to fines and other penalties. Aside from occasional taxes on pari-mutuels, no special tax is levied on gains from pure chance, such as gambling gains and the discovery of property lost by others. Whatever the explanation, the proposals for, and enactments of, special taxes on socially

/3/ See p. 29.

[page 273]

undesirable gain have been confined almost entirely to excess profits
on business capital and to landed property and income.


All the taxes of this kind have at least two characteristics that distinguish them from the usual type of income tax.


The most striking characteristic is impersonality. In contrast with the personal income tax, the death taxes, and the gift tax, the amount of the tax on socially undesirable gains is determined without regard to the personal situation of the taxpayer. He may be rich or poor, possessed of a large family or none at all, inclined to charity or addicted to dissipation -- the tax is the same. Apparent exceptions, such as the failure of Great Britain and the United States to tax the excess profits of farmers during the war, must be attributed to administrative and political influences. /4/

The basic reason for impersonality is that the income or property in question should not be allowed to get into private hands, no matter whose hands. Another reason, sometimes, is that the gain came with little effort or sacrifice. If it was thus "unearned" in a true sense, the argument runs, it can be surrendered in taxes with little sacrifice, no matter who the taxpayer may be. However, as the discussion below will indicate, this reason may call for a somewhat less impersonal tax than the other.

/4/ For excess profits taxation in war-time, see T.S. Adams, "Should the Excess Profits Tax Be Repealed?" Quarterly Journal of Economics, XXXV (May 1921), 363-93; Robert Murray Haig, The Taxation of Excess Profits in Great Britain, A Study of the British Excess Profits Duty in Relation to the Problem of Excess Profits Taxation in the United States: The American Economic Review, X, No. 4, Supplement (December 1920) (hereafter cited as Haig, Taxation of Excess Profits in Great Britain); Sir Josiah Stamp, Taxation During the War, pp. 146- 212.

[page 274]

Restriction to Gains from Capital

The other distinguishing characteristic of these taxes is that they apply almost wholly to income from capital, not to salaries, fees, and other income from personal services.

The reason for restricting the tax to income from capital is obvious in the case of special taxes on land -- the single tax and the land increment tax. It is not so obvious in the case of the excess profits tax. The explanation may be that its practical application on a wide scale originated during the World War. At that time the governments had to heed the popular protest against firms that were making profits to which they were thought to have no moral right. The special taxation of "excess" was thus directed toward income from business capital. Great Britain exempted professional men from its excess profits tax. Crude attempts in the United States to apply the profits tax to incomes from fees and salaries revealed how unfitted the prevailing conception of excess was for this purpose.


Matters of administrative technique, revenue yield, etc., important though they are, only tend to confuse the issues until the term "excess" is defined. Some of the more vexing problems that are labeled "administrative" are often really basic problems of tax justice. The concept, excess, is a curious one because it is so appealing at first and then so stubbornly resists analysis. Almost everyone thinks he can say off-hand what he means by an excess profit, but if he is asked to be specific he is apt to become badly confused. The idea has shown so much vitality in theory and practice that it must have something substantial behind it, but careful searching is needed if this is to be isolated and judged.

[page 275]

Just Price

The excess profits tax theory concedes that, up to a certain point, profits from any business transaction are not excessive. In this respect the special taxation of excess can be distinguished from the special taxation of land increment or, under the single tax, land rent. In the last two cases the test is based on kind rather than amount. To some extent the point at which profits become excessive is dependent upon the prices charged by the taxable concern for its products. The excess profits tax is therefore directly related to the problem of the "just price," which has a long and involved history.

Excess profits taxation is in some ways an alternative to price fixing. Whether it is more desirable will depend upon the particular kind of transaction in question. The argument for price fixing may be the stronger if the transaction is one in which the government is the purchaser, since it can readily fix the price in this instance. Otherwise, effective price fixing may be impossible. If it is possible, a necessity may arise for rationing the existing supply among consumers. /5/ The factors that should guide the choice between the fixing of maximum prices and an excess profits tax deserve more consideration, however, than can be given to them here.

Sometimes an excess profits tax may be used to complement the government's policy of price fixing. For example, if rate regulation of public service corporations proves too inflexible to prevent untrue profits, an excess profits tax may be useful. The recent suggestion by the Secretary of Agriculture that a special processing tax be levied on sugar refiners, who were said to be benefiting unduly from the sugar quota system, is another case in point. /6/ Possibly an excess profits tax would also serve to complement patent and copyright grants.

/5/ See Note 2, Excess Profits Taxation as an Alternative to Rationing, p. 556.

/6/ New York Times, January 14, 1937.

[page 276]


In practically all discussions of excess profits it is said or implied that monopoly is the basic cause. The term "monopoly" is not usually restricted to the extreme case of absolute control by one concern of the supply of an article or service. It refers rather to the exercise of enough control over supply to set higher prices and reap higher profits than are needed to induce production. "Imperfect competition" is in some ways a better term to use, but "monopoly" still carries to the general reader more of the connotations needed in the present discussion.

Sources of Monopoly Powers

Monopoly powers or imperfect competition of course has many different sources. Some are the result of a government grant, with no regulation or imperfect regulation of prices charged. Others arise from large-scale production and the division of labor (as in the production of automobiles), the cost of transportation (as in the cement industry), the ownership of natural resources (for example, the aluminum industry), and differentiation of product through trade-mark advertising (for example, the cigarette industry). Sometimes monopoly powers arise from temporary shortages of certain commodities, such as those caused by the huge demands unleashed by war, when such gains as they may produce are called "war profits." /7/

The lack of competition is sometimes due to the deliberate efforts of a monopolist; sometimes, on the other hand, to factors over which the monopolist had no control. A rapid rise in prices will find some dealers, by chance, more fortunately situated than others. The difference between anticipated monopoly and unanticipated

/7/ See Note 3, War Profits Taxation, p. 556.

[page 277]

monopoly may have some bearing on the kind of excess profits tax that
should be levied. /8/

Monopoly Powers and Monopoly Profits

Monopoly powers, however, do not always result in excess profits. Sometimes all they can do is to prevent heavy losses from competition. These losses may occur in industries where much of the expense (for example, depreciation) does not represent a current drain on cash, and, therefore, a temptation exists to cut prices below the long-term cost level.

Co-operative control of output through trade associations, price leadership, sharing of the market, stabilization of prices, price discrimination -- these and other practices, obviously widespread, involve monopoly powers. These activities may produce few or no monopoly profits recognizable as such by the casual observer. The benefits to the firms may take the form, not of excess profit, but of normal profit on excess capacity. Another possible result is a higher average profit over the term of a business cycle without an apparent excessive profit during any one period. /9/

The obvious existence of monopoly power, therefore, does not inevitably mean that monopoly profits exist on a scale wide enough to make an attempt at excess profits taxation advisable. Direct data on the extent of monopoly profits would be helpful, but, unfortunately, few are available. /10/

The relations between excess profits and "just price" and monopoly may help with the problems of policy that are faced in drafting an excess profits tax law. For example, the absence of a dividing line between excess

/8/ See p. 280.

/9/ The statements on pp. 276-77 above are based largely on a memorandum by C. Lowell Harriss on monopoly profits and taxation.

/10/ Ibid.

[page 278]

profits and non-excess profits will not be a deterrent. The lack of a
definition of free competition has not deterred attacks on monopoly.
The monopoly aspects of excess profits will not be the sole guide-
post in the drafting and application of the law, but they will
exercise great influence. For another thing, not all that is commonly
regarded as monopoly will be subject to an excess profits tax. The
prospector or the patentee, for example, may have a compelling
argument for exempting part or all of his monopolistic profits.


Scope of Business Unit Taxable

Combination of Different Kinds of Business

Suppose that a certain individual owns and operates two businesses, a grocery store and a printing establishment, and that the grocery store is losing money while the printing establishment yields monopoly profits. If an excess profits tax is applied collectively to the businesses of this particular individual, the losses of one business tend to offset the high profits of the other, and no tax may be due. If the tax is applied to each business separately, however, the printing establishment pays a tax.

The usual solution of this situation, at least in principle, is to tax each business separately. If monopoly profits are to be specially taxed because they represent socially undesirable income, the fact that the beneficiary of the monopoly is losing money in some other venture seems beside the point. The same problem exists and the same solution applies if the taxpayer is a large corporation that conducts complex and interlocking enterprises of which some are monopolies and some are not.

Excess profits tax laws, however -- at least in the United States and Great Britain --, have paid little atten-

[page 279]

tion to the separation of monopoly elements from non-monopoly
elements within the same ownership. Whether drafters of future excess
profits tax laws find it one of their major tasks will depend upon
how generally the kinds of business -- or kinds of transaction, to be
more strict -- that yield monopoly profits are found in close
association with other kinds of business. Only further research can
ascertain the importance of this particular problem. /11/

Combination of Business Assets and Non-Business Assets

When the assets of an enterprise include a greater amount of cash or securities than the business requires, a similar set of problems arises. The income from cash and securities is almost always well below the normal profit percentage allowed by an excess profits tax law. The rate of profit for the enterprise as a whole is thus lowered, and it pays a smaller excess profits tax, or none at all. The effect is accentuated if the rates of the tax are graduated in accordance with the rate of profit.

Many enterprises can utilize this method of avoiding the tax. Profits can be invested in low-yielding securities in the name of the business instead of being distributed to owners who would normally invest them in a like manner. The enterprise may even borrow in order to buy such assets if the excess profits tax counts borrowed money as part of invested capital. /12/

To prevent this kind of tax avoidance the excess profits tax must refuse to recognize as part of the business unit any assets not needed in conducting the business. The administrative difficulties caused by attempting to draw such a distinction are considerable. They are especially troublesome when the business is unincorporated and the sole proprietor or the partners are accustomed to

/11/ See Note 4, Separate Businesses under Single Ownership, p. 557.

/12/ See pp. 280-81.

[page 280]

merging their personal and business bank accounts, brokerage
accounts, etc. After attempting for a while to include
proprietorships and partnerships in the scope of the war-time excess
profits tax, the United States restricted it to corporations.

Unanticipated Gains

If the excess profits tax is aimed primarily at "windfall" gains, particularly those from inflation, different considerations may influence the lawmakers. Inflation may cause an individual to lose money on balance, even though it benefits some of his enterprises. The legislator must decide whether he is more concerned over the net status of the individual, or over the fact that above- normal profits are being made in certain branches of business, necessarily at someone's expense.

Capital Entitled to Only a Fixed Return

When a business unit is financed partly by creditors or preferred stockholders, analogous perplexities arise. These contributors of capital are entitled to a fixed return only. Shall the return on the proprietors' capital, or, in a corporation, the return on the common stockholders' shares, be combined with these fixed returns of creditors and preferred stockholders when the rate of profit of the business unit as a whole is being determined?

If the proprietors are making an excessive profit on their own capital contribution, considered separately, a merging of their proceeds with those of the creditors and preferred stockholders will pull down the average rate of profit subject to tax, and the excess profits tax will be reduced or wholly avoided.

Here again the degree of impersonality desired in an excess profits tax must be the guide. Complete impersonality requires that no account shall be taken of the differences between creditor and owner. It therefore requires

[page 281]

that the excess profits tax shall apply to profits before interest is
deducted, in contrast with the income tax, and also before preferred
dividends are deducted. Correspondingly, the invested capital against
which the business income is measured will include contributions by
creditors and preferred stockholders. /13/

If the chief purpose of the tax is to obtain a portion of the monopoly profits for the government, the entire amount of invested capital, no matter who the contributors are, is the significant factor. If, however, inflation gains are the chief point of attack, impersonality in applying the tax may not be the governing consideration. /14/ The case for counting only the rate of return on the proprietors', or common stockholders', capital is strengthened. Common stockholders who make handsome percentage gains by holding a thin equity during inflation may then be taxed accordingly, without regard to the moderate average percentage return of all capital contributors to the firm.

In any event the method followed during the World War in Great Britain and in the United States was illogical. The rate of return for tax purposes was influenced by the return on preferred stockholders' capital but not by the return on creditors' capital. This equivocation sometimes had disturbing consequences. /15/

Scope of Unit versus Capital Entitled to
Only a Fixed Return

If the excessiveness of the profit is ascertained by considering only the return on the proprietorship, or common stock, interest, puzzling problems may arise over the scope of the business.

A concern with $1,000,000 in ordinary business assets

/13/ See Note 5, Excluding from Invested Capital All Capital with a Fixed Return, p. 558.

/14/ See p. 280.

/15/ See Note 6, Disturbing Consequences of Distinguishing between Various Types of Fixed-Income Capital p. 558.

[page 282]

and $500,000 in securities that are obviously unnecessary for the
business will be taxed, under the principles noted above, /16/ in
accordance with the rate of return on $1,000,000, not on $1,500,000.
The $500,000 in securities and the income from it will be ignored.
If, now, the liability side of the balance sheet shows $600,000
common stock and $900,000 bonds, a conflict becomes evident.
According to the rule that the tax shall be based only on the
profitableness of the common stock capital, not more than $600,000
and the return from it should be considered. According to the rule
that only the business assets shall be considered, the return on
$1,000,000 of certain specified assets must be the basis of the tax.
The conflict must be resolved in some way. Shall the amount be
$600,000, or $1,000,000, or $400,000 (the stockholders' pro rata
share of the business assets), or some compromise among the three? If
it is less than $1,000,000, which part of the $1,000,000, with its
earnings, shall be excluded?

Unless a definite connection can be found between the borrowed money and certain specified assets, which in turn possess definitely allocable earnings, no one answer appears sounder than another. /17/ The dilemma simply reveals an irreconcilable conflict caused by considering the excess profits tax simultaneously as an impersonal tax based on the nature of the assets and a personal tax based on the status of the investors.

Length of Accounting Period

A highly seasonal business will in certain months show profits that might be considered excessive if the losses of the slack months were forgotten. Similarly, there are concerns that have highly profitable years and years of loss. The year-to-year fluctuations are caused by

/16/ See pp. 279-80.

/17/ See Note 7, Conflict between Inadmissible Assets and Inadmissible Liabilities, p. 559.

[page 283]

general fluctuations in business conditions and by alternate periods
of experimentation and realization, as in businesses based on
invention or on prospecting for natural resources.

If a month's gain can properly be offset by another month's loss, a similar offset might seem the only consistent way to treat year-to-year fluctuations. There is an important difference, however. The monthly offsets are confined within the seasonal period -- a year. No logical period of definite length for allowing offsets is obvious in the second case, except in a few businesses -- such as cattle-raising -- that have a production cycle longer than a year. Shall the excessive gains of one year be reduced, for tax purposes, by averaging them with the losses, or subnormal gains, occurring five years later -- ten years later -- fifteen years later? The government cannot postpone demand for payment until the concern's life is ended. Several devices have been adopted to meet this problem. The chief ones are (1) basing each year's tax on the average yearly profits for the past few years, usually two or three, and (2) calculating each year's tax separately, but giving the taxpayer a refund, as the British did, in the years when his profits drop below the excess profits line. /18/

The present federal income tax law does not allow the losses of one year to affect the gains of another in defining taxable income. For this reason, among others, the net income figure obtained under the present income tax law is unsuited for use in an excess profits tax.

In the abstract, the end of the concern's life may seem a satisfactory stopping point. A final reckoning may then be made of the excessiveness of the concern's profits over its whole existence. Additional tax payments or tax refunds could be made before the accounts were finally

/18/ See Note 8, Allowing One Year's Loss as a Subtraction from Another Year's Gain, p. 559.

[page 284]

closed. Actually, the end may be impossible to ascertain with
sufficient accuracy. If a radio manufacturer gradually evolves into
an electric refrigerator manufacturer, the date when the radio
business ended may be difficult to fix. The radio profits and losses
cannot well be merged with those of the refrigerator operations
unless the principle of separate taxation of separate businesses is
abandoned. /19/ The principle is equally relevant whether the two
kinds of business occur simultaneously or in sequence.

Moreover, to treat the concern's entire life as a unit may conflict with certain ideas about the taxation of monopoly profits. A firm may maintain a monopoly position through several years and then lose it. Should the profits of the monopoly period be reduced, for tax purposes, by averaging them with subnormal profits or with losses that the firm eXperiences after its grip is broken? The question is at least debatable.

Obvious difficulties arise whether the law adheres to the strictly annual standard or extends the accounting period to cover the life of the company. The decision will probably be a compromise.

Determination of Invested Capital

Whether a profit is excessive is usually determined by
comparing it with the amount of invested capital from which it flows.
Therefore, a taxpayer finds it to his advantage to make his invested
capital figure as large as is legally possible.

Sometimes the law is so drawn that the value of the monopoly privilege or whatever it is that enables the taxpayer to make excess profits may be reckoned as part of the invested capital. Thus the invested capital grows with the profits, and the rate of return never appears ex-

/19/ See pp. 278-79.

[page 285]

cessive. To develop a conception of excess that will prevent
avoidance of the tax through this writing-up of invested capital is
one of the chief tasks of the proponents of excess profits taxation.
An analogous problem has been faced in setting capital values for
public utility rate regulation.

Using Records of the Past

The direct method of setting the capital value is to find out the amount that was paid to obtain the assets of the concern, in other words, the capital put in either by creditors or owners, directly or by the reinvestment of earnings. The task of ascertaining this amount is usually difficult and sometimes impossible. The chief reasons for this unfortunate situation are (1) the length of time that has to be covered by the investigation and the corresponding size of the task, (2) the lack of records, (3) the difficulty of placing a dollar value on certain assets that have been contributed in kind instead of in money by owners or creditors, and (4) the difficulty of ascertaining how much of the capital has been lost through depreciation, obsolescence, etc.

Moreover, a figure based on the actual amount put in may permit the writing-up of capital value. The original monopolist may have sold out to the present owners at a high price. In such a case, the invested capital figure must somehow exclude the dollars that were paid for the very excess profits privilege that is to be taxed. Otherwise, the new owners are in effect treated as having a vested interest that must go tax free. /20/

Finally, if the price level has been fluctuating, the invested capital figure will vary from one concern to the next depending on the time at which each happened to acquire its assets.

/20/ See pp. 251-53.

[page 286]

Using Current Records

None of these objections is serious in determining changes in invested capital after the tax has been initiated. Each year's changes can be taken care of as they come. Records can be examined promptly before they are lost or destroyed. The existing income tax demands an estimate of the value of new assets in kind and the amounts lost by depreciation and obsolescence so that an excess profits tax adds no administrative burden for these items. Sales at a price reflecting monopoly privilege or a rising price level will subject the vendor to a tax -- probably inadequate -- on his capital gain, and on a falling price level the disadvantage of a lower capital figure will to some extent be offset by allowance of a capital loss. Nevertheless, the problem is not solved so long as a possibility exists that the valuation for the first tax period includes something that it should not -- namely, an amount representing the capitalized value of the excessive profits. The depreciation and obsolescence charges usually will not eliminate this element, since it will ordinarily not appear on the books at all and so will furnish no basis for depreciation or obsolescence.

If most of the excess profits are based on a recent and unforeseen upheaval, such as a war, the problem is somewhat less difficult. Because the profits were largely unforeseen, there is less likelihood that the recent book values will include a capital figure representing them. Furthermore, the excessiveness of the profits may for a time be determined chiefly by comparing them with the annual profits that the concern made before the war instead of comparing them with the invested capital. As the concern expands or contracts, and as new concerns develop, this method of determining excess profit becomes less valid. Over a period of only three or four years, however, it may be used extensively, as in Great Britain during the World War.

[page 287]

Immediate Inventory versus Delay

The surest way to avoid including in the first valuation something that should not be there is to make an inventory of all the assets of all the business firms as of the day when the tax is initiated. Cost of reproduction new, less depreciation and obsolescence, would presumably be the basis of the valuation, with such provisions for eliminating the influence of changes in price levels as might be desired. The magnitude and difficulty of the task are obvious.

The only other way seems to be to wait for a number of years before imposing a tax, so that tax authorities can gather data on the amount of new capital going into each concern and the amount of capital coming out. When the process of wearing out and replacement has gone far enough, an invested capital figure can be set, based on the records of capital inflow and outflow. This capital figure could then be adjusted for assets, such as land, that the concern could show had remained intact in the business throughout the period. No adjustment, however, would be allowed for any capital element that represented the monopoly privilege or any other profit source that it was desired to tax.

Something approaching the waiting plan has been in use in the United States since 1933 in the form of the so-called excess profits tax, which is merely an adjunct to a corporation capital stock tax. The rates are moderate, and the fictitious capital value used in the first year is adjusted from year to year in accordance with actual capital inflow and outflow. However, no provision has yet been made for ultimate and accurate valuation of extremely long-lived or permanent assets, such as land, or for ultimate exclusion of capital values based on monopoly privileges. /21/

/21/ See Note 9, Present Excess Profits Tax, p. 560.

[page 288]

Fair-Earnings Level and Tax Rate

If the framers of the concept of "excess" reach a decision on (1) the scope of the business unit taxable, (2) the treatment of capital entitled to a fixed return only, (3) the length of the accounting period to be used, and (4) the method of ascertaining invested capital, they must then decide (5) what percentage of capital is a fair return, and (6) what rate or rates shall be applied to the excess.

The capital percentages and tax rates that were used in the war taxes in Great Britain and the United States are not good guides. In both countries the levels set reflected an admitted failure to adjust other elements in the excess profits tax structure. The British authorities granted a high percentage of return on capital invested in mines before applying excess profits taxes, partly because the tax law allowed no deduction for depletion. They granted higher percentage rates of return to unincorporated concerns than to incorporated concerns because the law allowed no deduction for hypothetical salaries of working proprietors and partners. The United States law failed to allow refunds for bad years, in contrast with the British law, but the severity of this treatment may have been counteracted to some extent by exempting a higher percentage of return on capital, or perhaps by a lower tax rate, than otherwise.

Fair-Return Percentage

If the provisions discussed thus far have been framed as satisfactorily as possible, the capital return percentages do not need to be called upon to redress errors. They can instead reflect simply what is to be considered a normal return on capital in view of the risk involved and the time period covered.

A strong case can be made for varying the capital re-

[page 289]

turn percentage from industry to industry according to risk involved.
Indeed, variations from firm to firm within an industry are
similarly defended. A large concern may run less risk than a small
one, owing to its wider range of activity. Of a large number of small
wildcatters in an oil field, all but a few may show heavy losses, and
the few, large profits. If the wildcatters merge into one company,
the consolidated account may show no heavy loss and no great gain,
but simply a moderate profit. To demand an excess profits tax from
the few that got large profits before the consolidation and to decide
that no excess profits exist after the consolidation seems a
questionable procedure. At least it is inconsistent with the view
that emphasizes the impersonality of the tax. /22/ If consistency is
sought by an exemption of the few fortunate wildcatters in the pre-
consolidation stage, the wild-catters in general must be permitted a
high percentage of return on capital, or they must be allowed to mark
their capital values up by using, not the actual cost of their
properties, but their market values.

The excess profits tax has not usually been considered a business tax imposed because of the benefits that government gives to business. Since government activities may give particularly large benefits to some concerns, however, an excess profits tax, with low rates starting at a low percentage level, might be used as a supplementary business tax.

Tax Rate

The tax rate to be applied is not really a part of the problem of defining excess profits. Its height, however, may well be made to depend on the degree of success reached in formulating such a definition. A 100 per cent tax rate on excess profits is not feasible when there is danger that business enterprises will be ruined through

/22/ See p. 273.

[page 290]

errors in administration or failure of the law to provide for unusual
cases. This danger will not disappear unless the definition is easy
to understand and nearly perfect from the point of view of justice,
or unless it errs decidedly in favor of the taxpayer. /23/ Moreover,
a 100 per cent rate may result in undesired effects on business
initiative and other economic factors.

A graduated rate scale may be justified because excess profits can never be separated from fair profits with complete accuracy. Therefore, because of the chance of error, the rate that is levied close to the dividing line might be made moderate. Otherwise, graduation in the excess profits tax must depend upon a distinction between two excess profit dollars, both clearly excess but one representing a higher percentage return on capital than the other.


Special taxes on land have been urged on the ground that private ownership of land or its proceeds is partly or wholly undesirable from a social point of view. To some extent the charge of undesirability is based on grounds of economic influences. /24/ Another argument, however, holds that private ownership of land has no moral justification, since land is the result of no man's work or sacrifice and its value is derived from community growth and action.

The Single Tax and the Graded Tax

The extreme advocates of the single tax propose to finance all government expenditures by a tax on land rent, at a rate as high as 100 per cent, if necessary. Considerable debate over this issue has arisen from time to time. At the moment the tax gives no indication of

/23/ See Note 3, War Profits Taxation, p. 556.

/24/ See pp. 150-52.

[page 291]

being an important political issue in the United States except
possibly in a few states where it is linked with other measures.

Variants of the single-tax doctrine call for the taxation of improvements on and in the land, but at a rate lower than that on land. For some years Pittsburgh has levied its real estate tax in this manner. /25/ On the whole, even this graded tax has been given little consideration. The recent demand for a reduction of real estate rates has not differentiated between land and improvements.

The chief difficulty with proposals for the single tax and the graded tax, from the point of view of justice, is the damage that would occur to innocently acquired vested interests. /26/ One's judgment on the justice or injustice of those proposals will depend chiefly on the importance that one attaches both to these vested interests and to the distinction between destroying them quickly and destroying them over one or more generations.

The Land Increment Tax

Instead of taxing existing land rent at an especially heavy rate, and disturbing innocently acquired vested interests, /27/ some students favor a special tax on increases in land values. Recently the tax has been proposed as a substitute for special assessments. /28/

It is not always clear whether the tax would apply only to future increases. Suppose that at the time the tax is enacted a landowner holds a parcel of land worth $15,000, which he had purchased many years before at $10,000. If, after the tax has been enacted, he sells the land for $18,000, is he to be taxed on $8,000 or $3,000? If he is to be taxed only on $3 ,000, and other landowners

/25/ See p. 152.

/26/ See pp. 249-54.

/27/ See pp. 250-251.

/28/ Spengler, "The Increment Tax versus Special Assessments," Bulletin of the National Tax Association, XX, No. 9 (June 1935), 258- 61; XXI, No. 1 (October 1935), 14-17; XXI, No. 6 (March 1936), 163- 67; and XXI, No. 8 (May 1936), 240-44.

[page 292]

are to be treated similarly, it is impossible to avoid the difficult
administrative task of compiling a special inventory of all land as
of the date of enactment.

A variant of this plan requires valuation of all land as of some prior year when values in general were higher than in the year of enactment. This plan is advanced when the year of enactment coincides with a depression in real estate values. The burden on the taxpayer would be lightened, but the difficulty of a land inventory is increased if the valuation must be as of some date in the past.

Many land increments are part of a development that causes land decrements in other areas. Thus, when a subway opens up new territory, it relieves pressure on other parts of the city and tends to cause a fall of rental values there. When increments and decrements occur at approximately the same time but affect different parcels of land -- as in the subway example --, the only way to allow the one to offset the other is to give to the unfortunate landowner the proceeds of the tax levied on the fortunate one. Otherwise, the question arises whether it will be consistent to allow decrements to offset increments when both occur in the same parcel of land.

If land that was worth $15,000 when the tax was enacted drops to $10,000 a few years later and then rises to $18,000 a few years later still, how much of the increment is to be taxed? If the land is sold at the $10,000 level and sold again at the $18,000 level, a tax on $8,000 -- not on $3,000 -- may seem reasonable. If the land is not sold at the $10,000 level, but is sold at the $18,000 level, application of the special tax on $3,000 -- not on $8,000 -- may seem reasonable. Can any reason be given, however, for making allowance, in effect, for the decrement in the latter case and not in the former? To make the difference in treatment depend on a change in ownership is to make the tax a somewhat personal one -- that is,

[page 293]

the amount of the tax becomes dependent on personal circumstances
instead of solely on the fact that a land value increment has

Consistent treatment of land decrements is more important now than it was in the early days of the country when the increments were destined to be so much larger than the decrements.

The problems just noticed do not necessarily lead to a condemnation of land increment taxation, but they suggest that proposals for it should be made so definite that the particular philosophy of tax justice that it represents can be identified.

A comprehensive and effective excess profits tax might be made to reach increments in land -- at least, increments that were much above a normal interest rate. The burden might not be so heavy as most advocates of increment taxation would desire, but, if the excess profits tax succeeded in taking as much as 60 per cent or 70 per cent of all annual increments above 5 per cent or 6 per cent, little need would be felt for a special increment tax. Another substitute for increment taxation that has been proposed is a super-tax to be applied to the entire value of the land when an increase has occurred. /29/

/29/ See Seligman, Studies in Public Finance, pp. 272-77.

[page 294]

Chapter 20


The standards of tax justice discussed under capitalization, vested interests, special benefits, and socially undesirable income do not depend on the entire economic and social situation of the taxpayer. The conclusions are usually the same whether the taxpayer, when all his income and all his wealth and all his obligations are considered, is found to be wealthy or poor.

A common standard of tax justice, however, demands that some account should be taken of economic and social differences among individual taxpayers -- insists, as it is usually phrased, that the individual's tax burden should be adapted to his "ability to pay." Some taxes are ready instruments for achieving this purpose. Others are singularly perverse, and even tend to have a directly contrary effect. This chapter analyzes the ability of the major taxes to take account of the taxpayer's personal status.


The property tax has been narrowed from its original status as a general tax on almost all a person's possessions until in most jurisdictions it now rests in practice on little but real estate. It cannot, therefore, take account of so large a part of the economic world of most taxpayers as the income tax can. Even if all forms of property were taxed both in principle and in practice, the tax would suffer from a major handicap. A person's real estate can, under established law, be taxed only by the jurisdiction within which it is located, regardless of

[page 295]

where the owner lives. On the other hand, personal property, with
some exceptions, is taxed at the residence of the owner. A small
taxing jurisdiction, therefore, may find it impossible to tax many of
its residents on the basis of their total property ownership.
Moreover, the residence test for personal property is apt to result
in the formation of "tax colonies." Wealthy taxpayers living in these
"colony" areas pay very low rates, since the expenses of the
community can easily be met by a low rate on the large personal
property tax base. The property tax, either state or local, has
severe limitations as a general ability tax, under the existing rules
of jurisdiction.


Unlike the income tax, the property tax in the United States cannot rely on a graduation of rates to gain adaptability to personal factors. /1/ Instead it must depend on exemptions of various kinds.

A few states, most of them in the South, have exempted from the state tax small parcels of real estate used for residential purposes ("homesteads"), but only Florida has exempted the homesteads from most of the local tax also. /2/ This device may be effective in reducing the tax burden on home owners with small incomes, but it does nothing for tenants with equally small incomes. Moreover, unless particular care is taken, the taxes levied to replace the lost revenue may weigh as heavily on the small home owners as did the property tax.

Some states segregate special classes of real estate, such as timber lands and mines, for taxation at special rates or on special bases. The reasons commonly given are the desire for conservation of natural resources, and justice in the distribution of the tax burden.

/1/ See Note 1, Graduated Property Tax Rates, p. 561.

/2/ See the memorandum by Thomas J. Reynolds on homestead exemptions and rate limitations in the forthcoming volume, Studies in Current Tax Problems.

[page 296]

Personal property is sometimes exempted or taxed at special low rates. However, this exemption often reflects administrative defeat at the hands of recalcitrant taxpayers, rather than an attempt to further tax justice. New York State exempts personal property altogether, and about one-fourth of the states -- scattered throughout the South, East, and Far West -- exempt all intangible personalty, such as stocks and bonds. A few states tax almost all kinds of personalty, at least in theory, but they sometimes tax different kinds at different effective rates. /3/

The exemptions commonly granted to charitable, religious, and educational organizations, and those occasionally granted to business (usually manufacturing) are chiefly to encourage certain types of activity. The business exemptions often result from interstate competition. /4/

The use of property tax exemptions for redistributing the burden is hindered by the difficulty of knowing where the burden lies and where it will lie after the change is made. Although this difficulty is common to all taxes, it is especially important for the property tax because of the size and universality of the burden. /5/

The universality of the property tax burden is often ignored in current tax discussion. Many members of the community, it is sometimes said, are "exempt" from taxes, since they are so poor that they pay no income taxes, death duties, or gift taxes, and own no taxable property. They are, however, exempt only in the sense that they have no direct contact with the tax collector. They must live in dwellings of some sort, and most of them are probably paying more in rent than they would be paying if no real estate tax had been levied for the past decade or so. The increase in rent due to the tax may be negligible for rural tenants in small scattered country

/3/ See Note 2, Special Treatment of Personalty, p. 562.

/4/ See pp. 147-50.

/5/ See pp. 353-54.

[page 297]

dwellings. Little or none of the tax on land -- in contrast with the
tax on buildings -- is shifted to tenants, and in a declining
community even part of the tax on buildings may have no effect on
rental price. In general, however, a large part of the real estate
tax enters into the rental price of dwellings. In addition, it
influences the price of all other commodities and services through
its burden on office buildings, factories, transportation lines,
warehouses, retail stores, etc. No one, however destitute he may be,
escapes the direct or indirect burden of the real estate tax for long
unless he breaks all contact with the economic world and lives a
self-sustaining life as a hermit.

Variations in Rates

Variations that the property tax rate has shown from time to time and place to place indicate that the tax has some adaptability to the differing economic and social positions of taxpayers. Even this elementary characteristic, however, has been greatly impaired in Michigan, New Mexico, Ohio, Oklahoma, and West Virginia, where the state constitutions have been recently amended to set a stringent maximum limit on the rate. The limit, which can usually be exceeded only by special vote of the electorate, applies to the combined property tax rate levied by all the layers of government. It is so low -- for instance, 1 per cent in Ohio -- that the combined rate reaches the maximum almost everywhere in the state. A similar loss in flexibility is found in other states where the rates for various political subdivisions and various public purposes are limited by statute or constitution. /6/

Discriminatory Valuations

In many parts of the country the property tax takes on a sort of perverse impersonality. One example is the

/6/ See the memorandum by Thomas J. Reynolds on homestead exemptions and rate limitations in the forthcoming volume, Studies in Current Tax Problems.

[page 298]

arbitrary discriminations that occur when taxable values are set. The
assessment of values for the property tax is often so faulty that
some properties are valued at two or three times as much as other
properties that are almost identical. Valuation requires a high
degree of individual judgment. To know what a parcel of land, for
example, is worth, the assessor must in theory know what future
income it will yield -- or would yield, if rented -- and what
interest rate should be chosen to capitalize the future yield into a
present value. Actually, most assessors probably do not even attempt
to follow this method. Instead they use a mixed basis of past
valuations, current sales of similar properties, and, in some cities,
formulae based on more or less rigid front-foot value and depth and
position factors. The better the assessor, the less the injustice,
but whether even the best possible assessor could really cure the
evils of the situation is doubtful. In this respect, the property tax
contrasts with such levies as the liquor tax, which require very
little individual judgment of the assessors.

Because the property tax is so heavy, and because the amount levied on any one piece of property depends so much on the exercise of the official's judgment, the tax carries the constant threat of a more crushing unfairness to a greater number of people than any other tax in the country. The meticulous detail with which the income tax laws, rulings, and court decisions provide for calculation of the precise tax due, to the cent, contrasts strongly with the amount of discretion left to the assessor under the real estate tax. The latter tax involves a far larger sum in the aggregate, and it is relatively devoid of standards whereby taxpayer and tax collector can check each other's computations.

Proposals have been made to change the base of the property tax from capital value to income, partly in the

[page 299]

hope that the assessor can do a fairer job if he has merely to
estimate the income for the year just past. The income basis is
employed in Great Britain for the local "rates," or property tax, but
specialists in real estate taxation do not agree on the extent to
which it would reduce discrimination in this country. A gross income
base, or gross earnings base, has been substituted by a few states
for the capital value tax on certain public service corporations.
Also, New Hampshire, Ohio, and Tennessee use income rather than
capital value for certain intangibles. Otherwise, the income base,
gross or net, for the property tax has made no headway in the United

Representative Property and Multiple Taxation

There is multiple taxation of a kind where both intangibles and the tangibles upon which they rest are taxed -- for example, when both mortgage bonds and the mortgaged properties are liable. As a result the state and local tax burdens on two pieces of real estate may differ simply because one piece is represented by an intangible, such as a bond, while the other is not. The total tax burden on a mortgaged farm equals the tax on the farm plus the tax on the mortgage, while an unmortgaged farm feels the weight only of the tax on farms.

Multiple taxation of this kind is sometimes another example of a perverse sort of impersonality. Sometimes, of course, it can be defended as a consistent part of a consciously developed system. It cannot be, however, when it results from the conflicting actions of two or more jurisdictions -- usually two or more states. /7/ In at least twenty-six states, however, so much intangible property is exempt or is taxed at such low rates that the

/7/ See pp. 367-68.

[page 300]

amount of multiple taxation in those jurisdictions can scarcely be
considered serious. /8/


In contrast with such taxes as those on admissions, imports, and real estate, the income tax covers the major part of the taxpayer's economic life. Moreover, tax students agree that it is not, in general, shifted by the taxpayer to someone else. Consequently it can be refined to take account of many factors that determine how much one taxpayer should contribute compared with another.

Source and Use of Income

Should a man's tax burden be lightened because he is supporting a large family? Personal exemptions and credits for dependents can be allowed under the income tax -- but scarcely under the admissions tax. Should the government favor those who devote part of their income to charity? The income tax may allow the taxpayer to deduct his charitable contributions in computing his net income, but the customs duties that are borne by each taxpayer can scarcely be modified thus. Should the man who lives on interest and dividends be required to pay more than the man who receives the same amount from his own labor? The income tax can discriminate in its treatment of the two kinds of income, whereas the basis for the distinction cannot exist under a real estate tax.

Refinements such as these are widely used under the income tax. The federal government and all but three of the income-taxing states allow deductions or credits according to the taxpayer's personal status (whether he is married or single, head of family, etc.) and the number

/8/ See Tax Systems of the World, pp. 94-95.

[page 301]

of his dependents. The three excepted states restrict their tax to
income from securities. Similarly, the federal government and almost
all the states permit a limited deduction for gifts to charitable and
other similar organizations. The distinction between "earned" and
"unearned" income has again been incorporated in the federal personal
income tax law after a lapse of two years.

Unified Tax versus Series of Taxes

The income tax may be a unified tax, such as the federal tax in the United States. It may, in contrast, consist of a series of taxes, each with its own rates and exemptions. Such taxes are found in France and Italy, along with certain special income taxes that apply to all the taxpayer's income.

This contrast further illustrates the adaptability of the income tax to the personal status of the taxpayer. The series of taxes in France and Italy reflect the view that justice is best served by distinguishing sharply between different types of income. On the other hand, the unitary taxes of Great Britain and, in the United States, the federal government and of most of the income-taxing states imply that the individual is not to be split up into distinct segments. One taxpayer is not to be subdivided into a salary-earning taxpayer, a dividends-receiving taxpayer, a landlord-taxpayer, and so on.

The only important agitation for a series type of income tax in the United States in recent years has come from those who favor separate taxation of capital gains, such a profits from sale of securities. Among the states there are some examples of series income taxes, but these seem to have developed chiefly as a result of the decay of the property tax on intangibles. /9/

/9/ See p. 296.

[page 302]

Capital Gains and Losses

The provisions governing the taxation of capital gains and losses under the income tax give evidence of a struggle over the extent to which the personal status of the taxpayer should be taken into account.

Under the present federal law, the taxation of gains is dependent on the general economic situation of the taxpayer -- that is, the gains are added in with income from other sources, and to the net result the provisions concerning personal exemptions and rates are applied. In other words, the advantage that any given taxpayer would get from the exemption of capital gains depends on his total income, his family status, etc.

The provisions concerning losses, on the other hand, tend to ignore the personal status of the taxpayer. The losses can be set off only against capital gains of the same year, plus an extra $2,000. For example, a person with a salary of $15,000 and a net capital loss of $20,000 is taxed on $13,000 ($15,000-$2,000).

The taxation of capital gains departs in still another way from a consistent recognition of personal status. The gain accruing to a capital asset up to the time of a taxpayer's death is exempt, largely, it appears, because of constitutional limitations on the definition of income. Still other points indicating the artificiality of present capital gains taxation can be found. /10/

Accounting Period

The degree to which one accounting period shall be isolated from another is important in adapting the income tax to the personal status of the taxpayer. One taxpayer, for example, may show a large loss one year and a large income the next year, but, for the two-year period as a whole, the same total income as another

/10/ See note 5, p. 191, and pp. 434-91.

[page 303]

taxpayer who returns a moderate net income each year. Under the
federal income tax law, the first taxpayer cannot use the loss of the
first year to offset any part of the income of the next, so he pays
more tax than the other. In effect, the taxpayer is split
artificially into several chronological personalities, and to this
extent his true personal status is ignored.

Even if no yearly loss is ever involved, the problem remains. Some taxpayers get their income regularly and some irregularly. Owing to the progressive rate scale, an irregular income is at a tax disadvantage. A gain that has been accumulating over a number of years and then is realized and taxed all in one year is subject to heavy surtax rates compared with a gain of the same total size that is realized bit by bit, year by year. For capital gains and losses, the federal law attempts to meet this problem by decreasing the amount of the gain that is taxed (or the loss that is deducted) as the period of accrual lengthens. /11/

Tax-Exempt Securities and Salaries /12/

The exemption of income from certain securities and salaries makes it difficult to adjust the income tax to the relative personal situations of individuals. It is particularly damaging to the justice of a progressive surtax.

Interest on state and local bonds cannot be taxed by the federal income tax. The total amount of such bonds outstanding as of June 30, 1935, was more than $18,000,000,000. Apparently, however, only one- fourth or one-fifth is in the hands of individuals subject to surtax. The rest is held chiefly by banks, insurance companies, and public funds.

Because these bonds are tax exempt, the states and lo-

/11/ See p. 190.

/12/ The basis for statements made in this section is a memorandum by J. Wilner Sundelson on tax exempt securities.

[page 304]

calities do not have to offer so high a rate of interest as they
otherwise would. The federal government thus loses in revenue, and
the states and localities gain by a saving in interest payments. The
two amounts are not equal, however. A wealthy taxpayer gains much
more by his federal tax avoidance than he loses in accepting lower
interest. This is true principally for the reason that there are so
many state and local bonds outstanding that they have to bear
interest rates high enough to attract buyers from the low-income
groups, corporations, and others that are not subject to the high
personal income surtaxes. Under the 1936 federal income tax rates,
for an investor with a $50,000 income a 2 per cent yield on a tax-
exempt state or local bond corresponds to a 3 per cent yield on a
bond that is not tax exempt. For an investor at the $250,000 level,
the corresponding percentages are 1 and 3. Actually, there is no
such spread between interest rates of taxable bonds and exempt bonds.
The states and localities have to pay more than 1 per cent -- even
more than 2 per cent -- to compete with the taxable 3 per cent bond
of the same maturity and soundness.

The wealthy investor therefore does, on balance, escape some of the burden that he would bear if no tax-exempt securities were outstanding. In other words, the effectiveness of the higher surtax rates is impaired. The surtax rates might be made even higher so that the net burden would be the same as it would be with no tax exempts and a lower surtax rate. At best, however, this is a crude method of adjustment. It assumes, of course, that the wealthy investor would not put all his capital in tax-exempt bonds. This assumption seems justified, since data from estate tax returns indicate that taxpayers with incomes above $60,000 have been receiving only from 10 per cent to 20 per cent of their income in tax-exempt form, despite the high tax rates.

[page 305]

Salaries of state and local officials are exempt from the federal income tax, with some qualifications, depending on the status of the official and the unit for which he works. Obviously, no adjustment of the tax rates can affect this privilege. Perhaps, however, the state and local taxpayers, not the state and local employees, are the beneficiaries -- that is, perhaps the exemption simply results in a lower salary scale.

Apparently income from future state and local bond issues or from state and local salaries can be taxed only if the federal Constitution is amended. Such an amendment is unlikely, unless it is proposed as part of a general adjustment of federal, state, and local fiscal relations. The states have little to gain directly from such a move. Any plan to tax the income from securities already issued is open to the charge of bad faith.

The matter of tax justice is probably the most important element in the tax-exempt problem, but other elements are present. The federal government is getting from $100,000,000 to $200,000,000 less in revenue than it would if present rates were maintained and the state and local exemptions on outstanding securities were abolished. Furthermore, states and localities may be induced to issue a slightly larger amount of securities than they would if the tax-exempt privilege were removed. /13/

Income from federal bonds is exempt from state personal income taxes and from many state corporation income taxes. This raises a similar but much less important problem. The state income tax rates do not go high enough to be seriously distorted in their effects by the existence of exempt investments. The exemption of federal salaries from state taxation is more important in that here, again, no change of rates can touch the tax exemption privilege.

/13/ See p. 63.

[page 306]

Federal exemption of federal bonds is probably not a serious problem at the moment. About half the federal debt is exempt from federal surtax, but almost all of this is short-term paper held largely by institutions, which in any event bears very low rates of interest. Similarly, state and local exemption of state and local securities is relatively unimportant.

Imputed Income

The income taxes in the United States are not adjusted to the taxpayer's personal economic power in so far as it is represented by something other than current transfers of money or barter. For example, a person who owns some securities but not his home must pay an income tax on the security income, but a person who owns his home and no securities does not pay any income tax on the annual value of his home. If the security owner sells his securities and buys a home, he lessens his income tax. The income taxes could include an assumed income from the home owner's dwelling, as in Great Britain. /14/ Administrative problems in the United States would make this adjustment difficult, for the time being at least.

Gifts and Inheritances

The income tax does not, under existing federal and state laws, take account of the taxpaying ability represented by a gift or an inheritance. Whether it can be designed to do so is debatable. /15/


Corporation Income Tax

The corporation income tax is much less adaptable than the personal income tax to the economic or social status of the ultimate taxpayer. It is essentially a cruder

/14/ See Magill, Parker, and King, op. cit., p. 20.

/15/ See pp. 313, 315. See also Note 3, Liberality in Deductions from Gross Income, p. 562.

[page 307]

taxing instrument, since it strikes a group rather than an
individual. The group that it is intended to strike may be the owners
of the common stock of the corporation, the consumers of the
corporation's products, the corporation's employees, or the people
who sell to the corporation. In practice, the corporation income tax
usually rests on the owners of the common stock. /16/

The corporation income tax is sometimes used chiefly as a supplement to the personal income tax. The profits of the corporation are taxed before they flow out of the corporation as dividends, and then the dividends are not taxed in the hands of the stockholders so severely as they would otherwise be.

Holders of stock in a large corporation are of various kinds -- some are rich, some are barely well-to-do; some have large families, some have small families. The ordinary corporation tax cannot take these facts into account. It is levied on the corporation as a unit, and it reduces every stockholder's dividend by the same percentage. Whether the corporation tax is at a flat rate or at graduated rates, whether it is a tax on all net profits or simply undistributed profits, common stockholder A's share in the company's earnings is cut by precisely the same percentage as the share of common stockholder B.

In other words, provided that the burden rests on common stockholders, and not on employees, customers, or others, a corporation's income tax is equivalent to a tax at a uniform flat rate on every one of its common stockholders. A tax at a uniform flat rate may be advisable under certain circumstances, but it can scarcely be considered highly adaptable to the personal status of the individual.

However, the difficulties involved in any plan that relies wholly on personal income taxation, without a supplementary corporation income tax, are serious. They

/16/ See p. 310.

[page 308]

have been discussed above in connection with governmental policies as
they affect corporations as against unincorporated concerns. /17/

Progressive Rates

Despite the fact that progressive rates on a corporation's net income are equivalent to a uniform flat rate on all its holders of common stock, six of the thirty states that impose a tax on the net income of corporations employ a progressive rate scale, and the federal government in the revenue acts of 1935 and 1936 has adopted the progressive-rate principle for corporate incomes in a moderate way. Perhaps these scales reflect a belief that corporations with large profits have common stockholders with large incomes, and corporations with small profits have common stockholders with small incomes. In so far as this is true, a progressive corporation tax will impose a heavy uniform burden only on rich stockholders and a low uniform burden only on poor stockholders. Common stock ownership is probably not distributed in so simple a pattern, however. /18/

Non-Resident Tax

If a corporation is owned by non-residents, the state cannot reach the owners directly with a personal income tax. In this case, the problem of correlating a corporate tax with a personal income tax does not arise. A corporation operating in New York may be owned by individuals residing in New Jersey. New York may claim that it has a right to tax all income arising within its borders, even the part that goes to non-residents. It can enforce its claim only by requiring the corporation to pay a tax, since it has no jurisdiction over the persons of the New Jersey stockholders. Consequently no question can arise of coordinating the New York corporation tax

/17/ See pp. 164-66.

/18/ See Note 4, Graduated Corporation Income Tax Rates, p. 562.

[page 309]

with the New York personal income tax, for these corporations.

Taxation of Unincorporated Business Net Income

New York State is the only taxing jurisdiction in the United States that employs a special business tax on unincorporated concerns, based on net income. Extension of the business net income tax in this way has been supported on the grounds of the general benefit to business that results from government. /19/ It is also a way of reaching non-resident owners of all the businesses, incorporated and unincorporated. Of course the latter reason alone is not adequate to support either of the New York taxes as they now stand, since they both apply to business profits going to residents, and the residents are not given any relief on that account under the personal income tax.

Any system that uses a general tax on business income taxes certain kinds of ownership income more heavily than other kinds, and more heavily than creditor income. A dollar of profit earned by an unincorporated concern or a dollar of profit earned for a corporation common stockholder is taxed twice -- once under the business income tax and again under the personal income tax. A dollar of interest income is taxed only once -- in the hands of the creditor. A dollar of preferred stock income is also taxed only once -- in the hands of the owner of preferred stock. There is this difference, however, between the interest income and the preferred stock income: the corporation pays a corporation income tax on the profit earned for the preferred stock. The amount available for the common stockholders is of course reduced correspondingly, except to the extent that the common stockholders can float the bond and preferred stock issues at lower interest and dividend rates because

/19/ See pp. 259-60.

[page 310]

of the relative tax exemption. In a restricted sense, then, the
common stockholders are triply taxed.

This comparison of the burden on various kinds of investors will seem irrelevant to those who support the business tax on highly impersonal grounds.

Incidence of Business Net Income Taxes

All the arguments cited above for the taxation of business income assume that the tax is not passed on to consumers, suppliers, or anyone other than the owners of the concern. This assumption is generally accepted among professional students of taxation, and such statistical data as are available seem to support it. The reasoning runs as follows: granted, first, that a fair degree of competition exists; second, that there are always some firms that are barely getting along and are making little profit or none at all; third, that the successful firms, no matter how successful they may be, do not dare to put their selling prices much above the level at which the no-profit firms sell, for fear of losing trade -- then a net income tax cannot be added to prices and thus shifted to consumers, since a tax of this kind applies only to the successful firms. The concerns that are barely getting along pay little or no income tax, and hence they receive no impulse from the income tax to raise their selling prices. Since they do not raise their prices, the taxed firms cannot raise theirs.

Some business men insist that the corporation income tax is considered as a business cost, just as are wages, rent, etc. and is consequently taken into account in setting selling prices. This argument overlooks the fact that this particular business expense differs from wages, rent, etc., in that some of the firm's competitors do not bear it. Competition makes it difficult for a business firm to raise its selling prices just to take care of an expense that many of its competitors do not have.

[page 311]


Both the inheritance tax and the estate tax -- the two chief forms of death tax -- appear capable of many refinements, and thus highly adaptable to the personal status of the taxpayer.

Under the inheritance tax each heir or other beneficiary pays a tax on his share. The government therefore has an obvious chance to impose a heavier tax on one type of heir -- say, a nephew -- than another -- say, a son. The estate tax is levied as one tax on the entire estate. It, too, can be refined to discriminate among beneficiaries, but not so readily for the more extreme kinds of differentiation.

A majority of the states impose inheritance taxes, despite the example set by the federal estate tax. Some states impose both taxes, using the estate tax as a "deficiency levy" to take up whatever part of the credit against the federal tax is left unabsorbed by the inheritance tax. /20/

Influence of Death Taxes on Testator's Decisions
Concerning Beneficiaries

The adaptability of either tax is more apparent than real. Provisions that distinguish between sons and nephews, levy progressive rate scales resembling those of the personal income tax, and grant exemptions varying in amount from person to person are based on an assumption that the legislator knows who bears the real burden of the tax.

To pick an example at random, when Pennsylvania's legislators put a 2 per cent rate on inheritances received by direct heirs and a 10 per cent rate on inheritances received by collateral heirs, they probably did so in the belief that the direct heirs and the collateral heirs would

/20/ See pp. 26-27.

[page 312]

really bear the burden of the tax that was imposed upon them.
Suppose, however, that a resident of Pennsylvania wishes his son and
his nephew to share equally in what is left of his fortune after all
death taxes are paid. Since the state taxes nephews at a higher rate
than sons, he will have to leave the nephew more, and the son less,
than if the state treated both alike. The only effect of the
discriminatory rate, in such an instance, is to change the terms of
the will. The heavy tax on the nephew does not really harm him, and
the purpose of the rate differentiation is at least partly defeated.

On the other hand, the discrimination may work in the same direction that it was intended to, but more severely. The decedent- to-be may be determined not to let the state get much of his property. In view of the heavy tax on his nephew, he may leave his nephew less, and his son more, than he would have if the rates had been equal.

The discriminatory tax may stimulate the decedent-to-be to work harder or spend less in order to pass on to a nephew a certain predetermined amount net after taxes. In this case the heavier tax rests as a burden on the decedent-to-be, not on the nephew. Still further, a residuary legatee may be the one to suffer, to satisfy the testator's wish not to have the nephew's position -- net after tax -- really diminished by the discriminatory tax.

Rates Progressing by Size of Inheritance

Progression of the rates according to the size of the inheritance may also fail to reach its objective, which seems to be based roughly on the proposition that the richer a person is, the higher the tax rate should be, since the loss of a dollar means less to him. /21/ Therefore, the argument continues, a person receiving a large inheritance should pay at a higher rate than a person receiving

/21/ See pp. 320-21.

[page 313]

a small inheritance -- for instance, one receiving $100,000 should
pay more than twice as much as one receiving $50,000. A poor person,
however, may receive a large inheritance, and a wealthy person a
small inheritance. In so far as the feelings of individuals can be
compared, a dollar of inheritance tax may mean more to the one who
gets the large inheritance than to the one who gets the small
inheritance. However, the one who gets the large inheritance will be
taxed at a higher rate under the kind of inheritance tax progression
that is common in the states and that was suggested for the federal
government in 1935.

In practice, the problem may not be a serious one, since there is a readily observable tendency for large inheritances to go to those who are already well-to-do.

Abroad, the progressive rate scale of the inheritance tax has at times been made dependent on the amount of other wealth or income of the heir. This procedure is a step toward a "general ability tax," which allows all the major elements of ability to interact in determining the amount of the taxpayer's personal tax.

Even if the beneficiaries have identical amounts of other wealth and income, the progressive rate scale may not achieve its aim. The testator may fix the terms of his will or regulate his activity and his spending in such a fashion as to negate the influence of the progressive rates.

Justification on Impersonal Grounds

From one point of view, the death taxes are impersonal. Like excess profits, the thing taxed is considered fit for special taxation, regardless of the particular persons that may be involved. The immediate family may depend on the legacy for support, but they can be cared for by proper exemptions. The tax on all the rest of the property passing by death is, from this point of view, considered as a tax either on windfalls or on savings that

[page 314]

were made with little sacrifice. If the beneficiaries are the ones
that really bear the burden of the tax, it is justified on a windfall
basis. If the testator is the one who bore the real burden, by
earning more or spending less than he would have done had no death
taxes existed, the tax is supported on the grounds that the saving
must have been very easy or it would not have been accomplished, in
view of the remoteness of the end.

Few inheritances, however, are entirely unexpected, and some are of the nature of earned income. On the other hand, the strength of the argument need not depend on pure windfalls -- rather, upon the windfall element in a partially expected inheritance. Perhaps, too, the distinction noted in excess profits is valid here. Some of the non-windfall inheritances may be regarded as gained by methods frowned upon by society, as when someone receives an inheritance simply because he chanced to be a remote relative or a casual acquaintance of the testator. /22/

Other Devices That Give Adaptability to Death Taxes

When two or more deaths occur in the same family within a short time, so that the estate becomes exposed to quickly repeated taxation, the law commonly grants partial exemption or makes some other allowance. The federal government exempts from its estate tax or gift tax any estate that has paid either of those levies within five years. About one-fourth of the forty-seven states that have death taxes make a similar allowance. /23/ Another plan, accomplishing the same purpose in advance, makes the amount of the inheritance tax depend in part on the age of the beneficiary -- the older the beneficiary, the less the tax. This method is not used in the United States.

/22/ See pp. 271, 556.

/23/ This statement is based on a memorandum by Carl Shoup and Bernard L. Shimberg on provisions in death tax laws affecting the liquidation of estates.

[page 315]

The equitable distribution of an inheritance tax between a life tenant and a remainderman, or between one with a vested interest and one with a contingent interest, offers difficulties that are far from satisfactory solution. The chief problem is that of assigning values to the various interests.

As an extension of the "unearned wealth" idea, an unusually heavy rate might be applied to property that the decedent had received as a gift or inheritance. This method is not employed in the United States, but some authorities believe that it is both practicable and justifiable. /24/

Attempts to integrate the death tax with the entire personal situation of the beneficiary may some day lead to inclusion of inheritances as a part of the income subject to income tax. However, any such measure should probably be adjusted to lessen the burden that results from crowding so much income into one year and consequently bringing such high surtax rates into play. Such an adjustment is made, in principle, for capital gains. /25/

These and other matters of personal status under the death taxes seem destined to be much more widely discussed in the next few years than they have been in the past.


The gift tax in the United States is a child of the death duties. Wealthy persons were avoiding death taxes by donating property during their lifetime to children and other heirs. The gift tax, which applies only to gifts between living persons, could be chiefly a "policeman tax." Under this plan, its rates would be set high enough to discourage gifts. The transfers of property would instead yield their revenues through the death

/24/ See Shultz, The Taxation of Inheritance (Boston, 1926), pp. 311-13 (Rignano plan).

/25/ See p. 190.

[page 316]

tax. Congress has not viewed the gift tax in this light. Rate
graduation under the federal gift tax copies that under the federal
estate tax, but the rates are set at three-fourths the estate tax
rates. Much money has poured over this low gift tax dam, particularly
when a prospective change of rates in either the estate tax or gift
tax has made existing gift tax rates seem especially attractive.

Only three states now levy a gift tax, but more will probably do so unless the federal government raises its gift tax rates in relation to those of the estate tax.

The donor's loss of interest on the gift tax thus paid in advance of death is sometimes given as the reason for the relatively low gift tax rates. However, the matter is more complicated than this argument implies. The gift tax rate scale is cumulative -- that is, a donor goes up the scale into higher and higher rates year by year as he continues to make gifts. Let it be supposed that he gives $45,000 a year to his son for ten years. The first year's gift costs him nothing at all in gift tax, because of an initial $40,000 exemption, and an annual exemption of $5,000 for each person to whom he makes a gift. The second year's gift will cost him little -- no matter how much he intends to give away in the aggregate, eventually --, since he is starting at the bottom of the progressive rate scale. The third year's gift costs more, and so on. To the wealthy man intent on minimizing the total of gift tax plus death tax, the early gifts much more than pay for themselves in death tax avoided. In effect, the gift chops off what would otherwise have been the top part of the estate, taxable under the death tax at a high rate -- perhaps, 40 per cent or 50 per cent -- and transforms it into a gift taxable under the gift tax at a very low rate -- 4 per cent or 5 per cent. In view of this comparison, the interest loss incurred by the early payment of the tax is insignificant.

[page 317]

The gift tax presents some interesting problems of its own in relation to the personal status of the taxpayer. For example, if some exemption is to be granted to allow for the ordinary maintenance that a father may be deemed to owe his children, the problem arises of preventing large-scale avoidance of the tax by small but numerous gifts judiciously distributed over the years.


The taxes imposed on the production, sale, or use of commodities or services -- for example, the tobacco tax, the liquor tax, the customs duties, and the sales tax -- can take relatively little account of the taxpayer's personal status. The persons whom they are intended to burden are usually once or twice removed from the one who pays the tax to the government; that is, the tax must be shifted, ordinarily by increasing the selling price of the commodity or service. The necessity for shifting destroys most of the possibilities for refinement.

Exemptions and Refunds

The merchant cannot be expected to discriminate among his customers to the extent that the government can discriminate among its income-taxpayers. A state imposing a general retail sales tax might wish to exempt all persons on the relief rolls from its burden. However, the exemption of retail sales to all such persons would put a heavy task on retailers, who would have to distinguish between their customers on relief and other customers. The tax officials, too, would have a sizeable task in checking the accuracy of the retailers' returns. An alternative would be to allow relief recipients to submit to the state their claims for refunds of the tax, with receipts, etc.

Both exemption and refund operate on the assumption that the tax is really shifted to the customer. It is always

[page 318]

difficult to determine whether the assumption is fulfilled in any
given instance. Separate charging of the tax, so common under retail
sales taxes, is no guarantee that the tax is being shifted. A
retailer might sell an article at $1.00, for instance, if there were
no sales tax, while, under a 2 per cent sales tax, he sells it at "98
cents plus 2 cents tax."

Despite these difficulties, both exemptions and refunds are used in a few of the taxes on commodities and services. As of 1934, twenty-eight states allowed refund of gasoline tax paid on gasoline devoted to certain non-highway uses, and two other states permitted the sales of such gasoline to be made tax free. /26/ Both procedures, but particularly the latter, have led to widespread complaints of fraud. With certain restrictions not pertinent here, the federal processing taxes of 1933-36 and the California retail sales tax act of 1933 allowed refunds to certain organizations or persons who had purchased taxed articles and distributed them for relief or other charitable purposes.

Discrimination by Commodity or Service

Sometimes the commodity or service taxed is sold in a variety of grades, or at a variety of prices. It is often feasible to distinguish among the different grades or prices for tax purposes. Thus, the federal tax on admissions does not apply when the admission price is 40 cents or less. This provision benefits the moving-picture public that goes to the cheaper theatres. In this manner, the admissions tax is slightly adaptable to differences in economic or social status -- the exemption can be set at various levels depending on the degree of discrimination desired. The tax levied on manufacturers of cigarettes allows no such exemption, although the tax is much heavier than the admissions tax. The variety in grade

/26/ Crawford, op. cit., p. 20.

[page 319]

and price of cigarettes is not nearly so great as that found among
moving-picture theatre admissions, and the cigarette tax has
correspondingly less adaptability for distributing the tax burden.

Similar in principle to restriction of the admissions tax to high-priced admissions is the imposition of a tax on a commodity or service that is purchased only by a small group of taxpayers. Such a tax has a highly selective effect. The recent, but short-lived, federal tax on yachts, for instance, offers a better chance to impose a special burden on the very wealthy than does a tax on bread. A system of commodity taxes made up of a number of these selective taxes, each carefully chosen because of the particular group that it burdens, might seem to offer considerable possibility of adjustment to the personal status of taxpayers. However, the number of such highly selective taxes that are administratively feasible is small, and the experience of the federal government during and shortly after the war, and again during the recovery years after 1932, has shown that the revenue from these sources is almost negligible. The fact that the commodity or service has a restricted use indicates that the tax rate must be high if it is to yield much revenue, but the rate must not be so high as to render evasion inevitable. At practical rates the yield is small. The 10 per cent federal tax on certain club dues and initiation fees yielded only $6,000,000 in 1936; in contrast, the 3 per cent federal tax on electricity sold to domestic and commercial (not manufacturing) consumers yielded $34,000,000 in the same year.

The major commodity taxes, in contrast with the obvious luxury taxes, are useful primarily for placing on the lower economic groups a burden that cannot for administrative or political reasons be distributed satisfactorily by a more refined instrument, such as the income tax. A sales tax, for example, is a relatively heavy bur-

[page 320]

den on poor persons, since they spend a much greater proportion of
their income on sales-taxed articles than the wealthy do.

The assumption implicit in these statements, that the tax is shifted, is a reasonable one, after allowing for a transition period when the tax first comes into effect. In the process of shifting, some business men may lose trade. The price of the taxed commodity is probably not raised at once by the precise amount of the tax above what it would be if some other tax were levied instead, or if some corresponding expenditure were eliminated. However, it may in general be assumed that, if the cigarette tax, or liquor tax, for example, were cut sharply, the retail price of cigarettes or liquor would drop by a roughly corresponding amount, and that a tax increase would be followed by a price increase. These are adequate assumptions for a discussion of this general nature.


There is no agreement on the general aim to be kept in mind in adapting taxes to economic and social differences among taxpayers. It is agreed, however, that any adjustment to these differences should be designed on the basis of ability to pay. Specific interpretations of this phrase, however, differ greatly. Moreover, it is not yet possible to measure the degree to which the system departs from any of them -- either because they are too vague or because the tools of measurement are lacking.

Equality of sacrifice is one of the ability-to-pay doctrines. It requires that any individual, rich or poor, feel the same sacrifice from his tax payments as any other. The present technique for measuring the actual distribution of the tax burden is so imperfect /27/ that no attempt can be made here to indicate how far the present system fails to meet that standard. Even if the actual distribu-

/27/ See pp. 221-22, 236-37.

[page 321]

tion were known, there would still be the problem of determining how
much tax for each taxpayer would represent an equal sacrifice for

Similar considerations apply to another standard adopted by some students -- that is, minimum sacrifice, or least aggregate sacrifice. Another standard is supplied by the faculty theory, advanced by Seligman, who takes into account both the sacrifice involved in paying the tax to the state, and the ease, through the existence of privilege, with which the taxpayer obtained the resources for paying the tax. Seligman justifies the higher taxation of unearned income largely on the basis of "the principle of privilege." /28/

The importance of these concepts should not be underestimated. As they become more clearly defined, and as the instruments for measuring the distribution of burden are improved, they may be of much greater use in all serious attempts to analyze a given tax system.

Whatever standards are chosen, they do not, strictly speaking, apply to any one tax, though they are often loosely spoken of as if they did. They can be applied only to a tax system as a whole. Little meaning is conveyed by saying merely that a personal income tax with a certain scale of rates, exemptions, etc. conforms to the speaker's interpretation of the principle of ability to pay. The real issue is: how does some tax system as a whole appear when this particular tax forms a part of it? The conclusion must obviously depend upon the composition of the entire system.

/28/ Seligman, Essays in Taxation, pp. 338-42. For recent discussions of the minimum-sacrifice and equal-sacrifice principles, see Pigou, op. cit., pp. 60-63, 75-117, and J. P. Jensen, Government Finance, pp. 211-15.

[page 322]

Chapter 21


Some taxes fluctuate in yield much more than others when business activity or prices in general rise and fall. The extent of the fluctuation cannot always be measured by the figures of actual yield. These often reflect something more than changes due to business activity and prices. They may, for example, reflect changes in the tax law and changes that affect the particular thing taxed but that do not affect business activity in general or prices in general -- for example, the spread of the cigarette habit as it affects the yield of the cigarette tax. Often the figures cannot be recast to eliminate these extraneous factors, and a study of fluctuations due to general business and price changes is therefore handicapped by a lack of refined data.

Recent fluctuations, chiefly in federal taxes, that have been associated with cycles in business activity and prices will first be examined. Sufficient data are not readily available to analyze collections under the most important non-federal tax -- the property tax -- in the same way. /1/ Attention will then be given, in view of the current interest in the subject, to the effects of a prolonged upward price movement that is not necessarily followed by a return to the starting point. Such a movement is usually held to be a characteristic of inflation.


Sensitive Taxes

The income taxes, the death taxes, and some of the transfer taxes, such as the stock transfer tax, have been

/1/ See Note 1, Fluctuations in Property Tax Collections, p. 563.

[page 323]

particularly sensitive to cyclical changes. By disregarding for the
moment the influence of other factors -- longterm trends, and minor
changes in rates and other changes in the tax law -- a general idea
of the degree of sensitivity can be obtained from the data on actual
yield shown in Table 27. Where no data are shown, it is because
changes in the tax were so great as to destroy comparability.

The absolute magnitudes of the decreases in the corporation income tax and the personal income tax are significant. In 1933 the combined yield of the two taxes was almost $1,700,000,000 less than in 1930 -- a decrease of 69 per cent -- despite the increase in rates and certain other provisions of the Revenue Act of 1932, which affected part of the 1933 receipts. The decrease in the income taxes alone accounted for almost the entire decline -- nearly 50 per cent -- that was shown by federal tax yields from 1930 to 1933.

Much of the decrease was caused by a provision no longer in the law: the unrestricted deduction of capital losses against income from any source in the same year. Important remaining causes are the instability of net income, even aside from capital gains and losses, and the progressivity of the rate scale for the personal surtax. /2/ The corporation normal rate has recently been made progressive, and the new undistributed profits tax is also progressive. The net result of these two innovations has been to increase the instability of the income tax system. /3/

The estate tax was a minor source of revenue during this period. Moreover, the decline in its yield was not so severe -- 54 per cent from 1930 to 1933.

The tax on club dues and initiation fees illustrates the sharp decline that may occur in the yield of some of the luxury taxes. In this case it was 46 per cent from 1930 to

/2/ See pp. 332-33.

/3/ See Note 2, Fluctuations in Progressive Rate Yields in Corporation Taxes, p. 563.

[page 324]

1933. The continued decline of the yield after 1933 probably reflects
certain methods of avoidance that have been recently developed -- for
example, lowering dues and increasing service charges.

The yield of the stock transfer tax dropped by almost the same percentage as the yield of the personal income tax, from 1930 to 1932, and it has moved erratically since 1933.

                              TABLE 27


                      (In Millions of Dollars)

                                             Tax on      Tax on
                      Personal             Club Dues     Capital
Fiscal  Corporation    Income     Estate   and Initi-     Stock
 Year    Income Tax      Tax       Tax     ation Fees   Transfers

1925         916         --         --          --        12.8
1926       1,095         --         --          --        17.1
1927       1,308        912         --          --        16.7
1928       1,292        883         60          --        24.2
1929       1,236      1,096         62        11.2        37.6
1930       1,263      1,147         65        12.5        46.7
1931       1,026        834         48        11.5        25.5
1932         630        427         47         9.2        17.7 /d/
1933         394        353 /b/     30 /c/     6.7        33.2
1934         398         --         --         6.0        38.1
1935         572         --         --         5.8        15.7
1936         739         --         --         6.1        33.1

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

/b/ Despite a sharp increase in rates and certain other changes designed to increase the yield, in the Revenue Act of 1932.

/c/ Despite a sharp increase in rates in the Revenue Act of 1932. This increase, however, probably had little effect on 1933 collections, owing to lag in payments.

/d/ Effective June 21, 1932, the rate was doubled for shares selling at less than $20, and increased by 150 per cent for other shares.

Non-Sensitive Taxes

Most of the non-sensitive taxes now levied by the federal government are minor excises and special taxes that were not in force during 1928-32. An exception is

[page 325]

the customs duties, shown in Table 28. The yield of the customs
responded little to prosperity in the period 1926-30. The decline
from 1931 to 1933 inclusive and the subsequent recovery to 1937 show
a much greater sensitivity. However, beginning with 1935 the figures
have been appreciably affected by changes in rates.

The cigarette figures, also in Table 28, must be taken with some reserve because they reflect not only resistance to cyclical influences, but a shift from other forms of tobacco. From 1927 to 1930 the yield of the tax on large cigars dropped from $24,000,000 to $21,000,000. It had fallen to $11,000,000 by 1933, and it recovered only to $12,000,000 in 1936. The tax rates did not change

                              TABLE 28


                      (In Millions of Dollars)

                               Customs           Small-
        Fiscal Year             Duties       Cigarette Tax

           1923                  562              183
           1924                  546              204
           1925                  548              225
           1926                  579              255
           1927                  605              279
           1928                  569              302
           1929                  602              342
           1930                  587              360
           1931                  378 /b/          359
           1932                  329              317
           1933                  252              328
           1934                  315              350
           1935                  345 /c/          385
           1936                  387 /c/          425
           1937 /d/              402 /c/          ---

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

/b/ Sharp changes in rates made by Tariff Act of 1930.

/c/ Influenced by changes made under reciprocal trade pacts under the Trade Agreements Act of 1934. The few changes of this kind made before the end of the fiscal year 1934 apparently did not influence the 1934 total appreciably.

/d/ Estimated. See Table 12, p. 88.

[pages 326 and 327]



[Table omitted]

[page 328]

during this period. The tax on manufactured tobacco and snuff dropped
from $76,000,000 in 1923 to $62,000,000 in 1936, with rates

In order to compare readily the degree of fluctuation shown by the several taxes, some index of fluctuation is needed. This has been made by fitting a trend to the adjusted yield and measuring the standard deviation from the trend. The results are shown in Table 29. /4/ Another index, not so useful except for the income taxes, is also shown in Table 29 -- namely, the standard deviation from the average, or mean. A considerable spread between the two figures, as in the tax on sales of produce futures and the taxes on cigarettes and cigars, indicates of course a pronounced trend. This feature is also evident in the gasoline tax figures. The deviation from the trend, for total gasoline taxes, is only 8 per cent, while the deviation from the mean is 27 per cent. Even the latter figure, however, is much less than the deviation shown by the stock transfer tax during approximately the same period -- 52 per cent (from the trend) and 57 per cent (from the mean). In between are the federal estate and corporation income taxes, with a 37 per cent-40 per cent range.

The personal income tax indices are somewhat misleading because of the shortness of the period covered and because this period happened to include the last of a pronounced upswing and practically the whole of the subsequent decline. The trend shown is therefore spurious, and the figures for standard deviation from the trend -- ranging between 16 per cent and 69 per cent for the various income groups -- are too small to show the real situation. The standard deviation from the mean -- ranging from 35 per cent to 70 per cent -- gives a better indication of the degree of fluctuation. The greatest fluctuation is of course found in the top income group. In

/4/ For a more detailed description of Table 29, see Note 3, Cyclical Stability of Certain Taxes, p. 563.

[page 329]

a depression the taxpayers fall down through the brackets, so to
speak. Many in the middle and lower brackets disappear into the sub-
bracket cellar where they pay no tax, but their places are taken, in
part, by the former dwellers in the higher brackets. In contrast, no
source of replacement for the highest bracket exists.


No forecast is made here about the probability of a period of rapidly rising prices in the United States in the near future, but the current interest in the subject is recognized. Therefore an analysis is made of some of the effects on the tax system that might be expected if prices were to rise fairly steadily for five or ten years up to a new plateau several times as high as the starting point. In place of a generalization about the whole tax system, two major taxes are examined -- the real estate tax and the personal income tax. Attention is restricted to revenue aspects, and the problems of justice are therefore not considered. /5/

Real Estate Tax

If prices rise and government costs rise with them, and if the yield of the real estate tax fails to keep pace, a slashing of local services or a demand on the federal and state governments for aid may result.

The problem is not one of maintaining a given dollar yield, but of maintaining an increase in dollar yield; for, with continually rising prices, government costs will rise. The yield will increase if the tax base -- that is, the assessed value of the real estate -- increases. The tax base will increase if the real or market value increases, and if assessors recognize the increase by raising assessed values.

/5/ The material in the remainder of this chapter is drawn from memoranda by J. Wilner Sundelson and Paul Strayer on the property tax and the income tax, respectively, in periods of rapidly rising prices. See also the analysis made by Jacob Viner, "Taxation and Changes in Price Levels," Journal of Political Economy, XXXI, No. 4 (August 1923), 494-520.

[page 330]

Rise in Tax Base

Market values of real estate, on the average, might rise as fast as government costs, during a period of generally rising prices. The rise would be geographically uneven, however, and some of the local government units might soon find themselves in acute fiscal distress. Rent control, a common feature of European inflations, would restrict to a significant degree the increase in urban values. The value of land under long-term lease at fixed rentals would also lag. The growing value of the privilege held by the tenant under long-term leases would not be reflected in any part of the real estate tax base. This value is, rather, an intangible property right. As such, it is not subject to the real estate tax.

The homestead exemptions recently enacted in several states make the tax base more responsive to price changes, although, because of exemptions, the starting point is at a lower level. /6/ The exemptions are commonly in terms of absolute amounts -- $5,000 in Florida, for instance. A rise in value of homestead real estate will therefore result in a more than proportionate rise in the tax base. Thus, if a property is worth $6,000 and is exempt to the extent of $5,000, leaving $1,000 taxable, a doubling of the property's value would make the tax base seven times as high. If all properties above $5,000 are taxed in full, a doubling of the value of homesteads would still more than double the tax base.

Even if market values rise as fast as governmental costs, assessed values will lag behind market values and probably behind governmental costs, under prevailing methods of assessment. The most frequent period of reassessment of real estate in the states is one year, so that a price rise can be well under way before assessed values begin to move.

For only about half the states, however, is property

/6/ See p. 295.

[page 331]

reassessed at this one-year interval. In nine states -- almost all of
them farming states with few large urban centers -- the reassessment
is made once every two years. Pennsylvania has a three-year interval;
nine states, again chiefly farming states, have a five-year interval;
Maryland and Ohio have a six-year interval and Connecticut has a ten-
year interval, but both Ohio and Connecticut officials are empowered
to assess annually.

The failure of assessments to rise might be offset by increases in the rate of tax, but states have imposed limits on local rates and, through constitutional amendments, on their own rates. In a few states constitutional limits have been imposed on the aggregate of state and local property taxes. /7/

Delay in Payment

Even if the tax base rose rapidly enough to match the increase in governmental costs, the customary delay between assessment and payment would cause trouble. Time must be allowed for review and equalization, and the task of original assessment alone must occupy a considerable period. Even after the liability has been fixed, common practice allows a brief period during which no penalty applies and then a long period, ranging from several months to more than a year, during which the penalty for non-payment is only a fairly high rate of interest. Taxpayers who realize the advantages of borrowing during an inflation are likely to borrow in this way from the government. Even interest rates of 10 per cent and 12 per cent a year, which represent the maximum charge in most of the states, would not deter them.

If the difficulty became acute, the states and localities might be driven to use index numbers to revise assessments in a blanket fashion and to revise amounts of tax due.

/7/ See p. 297.

[page 332]

Personal Income Tax

During the first year of the price rise the current revenue from the federal personal income tax would decrease steadily in purchasing power. The dollar amount would not be increasing, since it would be based on the period before prices started to rise.

After the first year, however, not only the government's dollar income but also its real income -- income measured in purchasing power -- would be increased. This gain would be the result of three factors: (1) the real value -- purchasing power value -- of the personal exemptions and the credit for dependents would decrease, (2) the average rate of tax would increase as more and more taxpayers found themselves in the higher-rate brackets, and (3) capital gains representing dollar gains but not purchasing power gains would be taxed.


If prices and incomes doubled, the real value of the $2,500 exemption given to married couples and heads of families would be worth only as much as a $1,250 exemption would be under the old price level, and the $1,000 exemption given to single persons would drop in real value to $500. So many people would have such increased dollar incomes -- though no increase in real incomes -- that millions more of them would be paying an income tax.

Effective Rates

The effective tax rate may be defined as the ratio found by dividing the tax due by the net income before deduction of personal exemptions or earned income credit. Thus, a married taxpayer with two children under eighteen, receiving a salary of $4,000, gets a personal exemption and credit for dependents of $3,300, under the Rev-

[page 333]

enue Act of 1936. He would, therefore, have a surtax base of $700. By
further deducting an earned income credit of $400, a normal tax base
of $300 is reached. There is no surtax, since the first $4,000 of
surtax base is exempt. The normal tax is $12, and the effective rate
is $12 divided by $4,000, or 0.3 per cent.

If, in this illustration, the taxpayer's dollar income doubles, keeping pace with a doubling of prices in general, he has a surtax base of $8,000 minus $3,300, or $4,700, and a normal tax base of $3,900 (the earned income credit increases to $800). This results in a surtax of $28 and a normal tax of $156, a total of $184. Since the net income is $8,000, the effective rate is 2.3 per cent. The net in- come doubles, but the tax increases to fifteen times and the effective rate to nearly eight times the former amount, because of the operation of the exemption provisions.

Suppose that another 100 per cent increase in prices in general brings with it another 100 per cent increase in the taxpayer's salary, to $16,000. The surtax base is $12,700, the normal tax base is $11,300 (the earned income credit increases to $1,400). The surtax is $496, the normal tax is $452, and the total tax is $948. The income doubles, but, owing to the operation of the exemptions, and, this time, the progressivity of the surtax rates also, the tax increases to more than five times the former amount, and the effective rate (now 5.9 per cent) increases to about two and one-half times the former rate.

The effect is much less severe on taxpayers with very high incomes. They start from so high a level that they experience relatively little increase in rates when their dollar incomes increase. The rate of progression tapers off at the upper end and finally stops completely, since all income in excess of $5,000,000 is taxed at a flat rate of 79 per cent. Thus, if prices and incomes triple, the taxpayer who starts out with $4,000 (and ends with $12,000) will have his tax, measured in dollars, multiplied by

[page 334]

43.5, while the taxpayer who starts with $100,000 (and ends with
$300,000) will pay only 5.1 times as large a tax as before. The tax
in terms of purchasing power -- the "real" tax -- will be 14.5 times
as great as before for the first taxpayer and only 1.7 times as great
for the second taxpayer.

Capital Gains and Losses

If prices in general double, the taxpayer who sells for $10,000 some property that he bought at the start of the price rise for $5,000 makes no gain in real purchasing power. However, he pays a tax on the dollar gain, which is $5,000. Moreover, losses in real purchasing power -- for instance, if the asset rises only to $9,000 -- are not deductible even in part. In fact the net result may be, as in this case, a taxable capital gain unless the real loss is so great as to be also a loss in terms of dollars. Even then the deduction is severely limited. /8/

In effect, then, the income tax in an inflationary period reaches every transfer of property holdings that shows a dollar gain, regardless of whether there is a real gain or not.

Loss to Government

The absolute amount of increase in revenue that the government can obtain from the personal income tax with assumed increases in prices and tax bases cannot be estimated here. Even in real revenue, it is probably large. If the rise in prices continues very long, however, or becomes very rapid, the government may lose on balance because of one factor: the time interval between the income-year on which the tax is based and the payment of the tax. The real value of the government's claim diminishes steadily as prices rise. Over a period of years,

/8/ See p. 302.

[page 335]

the real value received by the government may be less than if no rise
in prices and incomes had occurred at all.

If prices and incomes rise at such an extreme rate as 10 to 1 each year, the tax burden in terms of purchasing power actually decreases for a taxpayer starting from the $20,000 income level. These calculations assume prompt payment. In practice, many taxpayers would withhold payments, thus borrowing from the government. Stiff penalties would be needed to prevent this action. If prices are rising 100 per cent each year, and if the real rate of interest is 6 per cent, the penalty rate necessary to prevent delinquency from being profitable to the taxpayer is 112 per cent.


In any general rise of prices the course of events will of course be more complex than has been indicated above. For one thing, not all incomes will rise in the same proportion. Thus, any attempt to solve the problems of real revenue, to say nothing of the problems of justice and administration, must await detailed study. However, the probable correctives would include some index of prices to deflate money incomes and to inflate tax amounts due. Also, the time between the tax-base period and the date of payment might be shortened.

[page 336]

Section C. Ease of Administration

Chapter 22


As will be shown below, the data on cost of tax administration in the United States reflect both efficiency and undue economy in enforcing the tax laws. They also reflect the use of business firms as unofficial tax collectors. Instead of dealing directly with millions of individuals, the authorities often collect, from a relatively small number of business firms, taxes that can be passed on to customers. These firms, like other taxpayers, undergo certain expenses in complying with the tax law. The cost of obtaining expert advice, the cost of maintaining special bookkeeping routines for tax purposes, and similar expenses represent the "cost of compliance." It is not included in the figures on cost of administration.

Evasion seems to be serious only in a few taxes. Almost every tax has a fringe of evaders who, though numerous, account for only a small part of the total revenue payable. Evasion is of course illegal, in contrast with avoidance and the minimizing of a tax, which consist merely in taking advantage of whatever avenues of escape the legislator has left.

Existing data on cost of administration are fragmentary, /1/ and only one set of reliable figures is available on cost of compliance. Even when the information is accurate and precise it is usually difficult to interpret. Moreover, the differences in the cost figures for the various taxes are not striking. Little evidence is at hand, consequently, for preferring one tax to another because

/1/ See the memorandum by Thomas J. Reynolds on cost of administration in the forthcoming volume, Studies in Current Tax Problems.

[page 337]

of differences in cost of collection or cost of compliance. In
contrast, differences in the extent of evasion and the possibilities
of checking it are in several instances obvious enough to supply a
basis for choice.


The cost of administering a tax is usually compared with the revenue obtained from it, in order to judge whether the cost is high or low. For example, the New York State tax commission reports that in the year when it collected a record amount from the personal income tax ($82,234,368, in the calendar year 1929) the cost of administration was $675,187, or 4/5 of 1 per cent of the yield. In contrast, the cost of administering the gasoline tax for about the same period amounted to only 1/5 of 1 per cent. It does not follow, however, that the personal income tax was four times as difficult to administer as the gasoline tax. Percentage-cost figures never give a complete and accurate picture of relative administrative expense.

Ambiguity of Percentage Costs

In the first place, precise costs for each tax cannot be known. A tax-administering body usually has large overhead expenses that cannot be allocated accurately to the various individual taxes. Occasionally taxes are so closely linked in the process of collection that any segregation of their costs, while conceivable, is not practicable.

Some kinds of governmental costs are traceable to certain taxes but are never considered a part of tax administration costs. For example, salaries paid to judges who listen to tax litigants and costs of the Department of Justice incurred in prosecuting tax evaders are not figured as tax costs. Business firms acting as collection agents are sometimes recompensed by a share in the proceeds, but the published figures of tax revenue and expense usually fail to include this amount.

[page 338]

Furthermore, the rate of a tax must be known in order to interpret the percentage cost of its administration. Up to a limit not yet reached by most of the taxes in the United States, a higher rate of tax results in a lower percentage cost -- under increasing rates, the total revenue usually grows faster than the total cost of administration. Similarly, when the yield of the tax fluctuates with business conditions, a high yield is usually accompanied by a low percentage cost.

Finally, and most important of all, the meaning of the percentage-cost figure varies with the degree of enforcement. The tax administrator is usually supposed to allow no evasion at all. He never reaches this goal, but the nearer he comes to it, the higher the average cost of administration is. Moreover, the "last" or hardest dollars of tax extracted from the least co-operative taxpayers may cost 100 per cent, or even more, while the average percentage for the tax as a whole is still not strikingly high. Any tax-administering body will show a low percentage cost if it collects only the quasi-voluntary contributions that will come in under almost all tax laws without any pressure.

Sometimes, as under a few of the state sales taxes, the law sets a limit to the percentage cost. In these cases the recorded percentage will not rise above a certain figure, no matter how great the evasion.

Despite all these limitations, percentage-cost figures must be used for want of a better standard of comparison. Properly qualified, they at least serve to correct major errors in judging administrative costs. They may also indicate when the administration of the system as a whole needs thorough reform.

Lack of Data for Federal Taxes

The information available on costs of administration is largely confined to state and local taxes. The federal government gives only one over-all figure for its in-

[page 339]

ternal revenue costs -- except for the agricultural adjustment taxes
while they were in force. /2/ Similarly, it gives one over-all figure
for customs. The 1935 figure for internal revenue other than
agricultural adjustment taxes was 1.54 per cent. The processing tax
figure was 0.72 per cent. The customs duties figure was 5.65 per

It would be particularly interesting to know the cost of administering the federal spirits tax (to be distinguished from the beer tax). This seems to be the one important place in the federal tax system where a significant amount of revenue is being lost through evasion. /3/ Although cost data for the other major federal taxes would probably require no substantial adjustment for evasion, some reservation might be made in connection with the personal income tax and the gift tax. A considerable number of low-income taxpayers may be evading the former tax, and the latter is still too new to be judged.

State and Local Costs

Gasoline Tax

Among the major state and local taxes, the gasoline tax has the best record. The cost is less than 1/2 of 1 per cent in about half the states and less than 1 per cent in practically all of them.

The figures are understatements to some extent, but they reflect the economy of collecting high-rate taxes chiefly from a few taxpayers -- in this case, usually the refineries or other wholesalers. They also reflect an unwillingness to spend money to check an appreciable amount of evasion. Extreme estimates put evasion as high as $100,000,000 a year; actually, the amount is probably much less. Total state and local collections were slightly more than $500,000,000 a year when this

/2/ See Note 1, Federal Administration Costs, p. 565.

/3/ See Note 2, Liquor Tax Evasion, p. 566.

[page 340]

extreme estimate was made. However, even allowing for the lowering in
cost that results from tolerating evasion, the gasoline tax at
present rates seems to have an inherently low percentage cost. /4/

Motor Vehicle Tax

The reported percentage cost for the motor vehicle taxes is high, averaging about 6 per cent. Part of the cost is due to the inclusion, at least in some instances, of amounts spent for traffic regulation. In any event, the motor vehicle tax is more costly to collect than the gasoline tax.

Income Tax, Death Duties, and Sales Tax

The other important state and local taxes usually cost from 2 per cent to 4 per cent. Here and there a slightly higher figure is found for the personal income tax, particularly in rural areas. The tendency in many states, however, to neglect the enforcement of the income tax law robs the reported figures of much of their meaning. /5/ On the other hand, the income tax rates are usually so low that it would be difficult for any administrator to show an exceptionally low percentage cost.

In the state of New York, which accounts for nearly half of the personal income tax revenue of the states, and which probably enforces the law as strictly as any state, the cost has ranged from 0.70 per cent, in 1930, to 2.85 per cent, in 1921 when personal exemptions were low. Figures for the corporation income tax are usually somewhat lower than those for the personal income tax -- though comparable data are few -- and the sales tax costs are usually somewhat higher. Death duty costs show considerable variation, but in few states is the cost above 4 per cent.

/4/ See Note 3, Gasoline Tax Collection Costs, p. 567.

/5/ See Note 4, Administration of State Income Taxes, p. 568.

[page 341]

Property Tax

Practically no data exist on the cost of administering the real estate tax. Scattered evidence indicates that it usually ranges between 2 per cent and 4 per cent -- an unfavorable showing in view of the high rate of the tax. Cost figures on that part of the property tax levied on personalty are of little value without a complicated adjustment for the large amount of evasion, especially in those states that try to tax intangible property at the same high rate as real estate. /6/

In general, then, the concentration of wealth, income, and business activity, coupled with an unwillingness to extract the last legitimate cent from the smallest taxpayers regardless of cost, insures that any major source of revenue in the United States will show a cost of tax administration of less than 6 per cent.


The amount of time and money that the taxpayer spends in complying with a tax law is a cost of administration that does not appear in the official statistics. This "cost of compliance," in its broadest sense, includes items not ordinarily associated with mere compliance, such as expenditures made by honest taxpayers to check evasion by dishonest competitors, and expenses of litigation. In any sense, it includes such costs as fees to accountants for making up returns and expenses resulting from the necessity of special accounts for tax purposes.

A recent study of costs of compliance reported by a group of corporations includes the following figures: state corporation income tax, 9.5 per cent (median for a group of 76 corporations); federal corporation income tax, 4.7 per cent (median for a group of 95 corporations); state sales taxes, 3.7 per cent (arithmetical average for a

/6/ See Note 5, Definition of Cost under the Property Tax, p. 569.

[page 342]

group of 91 corporations); property tax, 1.04 per cent (arithmetical
average for a group of 122 corporations) or 1.2 per cent (median for
the same group). The figure for all taxes, federal, state, and local,
was 2.3 per cent (arithmetical average for a group of 163
corporations). The data refer to the year 1934. /7/

An inverse relation between public costs of administration and private costs of tax compliance is suggested by these figures. This may be explained in part by the fact that under self-assessed taxes, such as the income tax, the taxpayer has to assemble all the data and even calculate the amount of tax due; under a government-assessed tax, such as the real estate tax in most states, all this work is done by the government.

The highest compliance figures are usually those of firms that are engaged in business in many states. They are bothered by a great number of tax forms and intricate problems of allocation of the tax base among the states.

A high cost of compliance is not always an unmitigated evil. The figure may simply indicate that the tax has forced on business men accounting methods that are desirable, regardless of the tax. The pressure of the federal corporation income tax, for example, has greatly improved accounting practice.


The taxpayer becomes, in effect, an unofficial revenue officer when he is made subject to a tax whose burden he is supposed to pass on to someone else. For example, no one supposes that the federal tax on cigarettes is paid wholly out of the profits of the cigarette manufacturers, from whom it is collected. Part of the tax -- indeed, probably almost all of it -- is passed on, through wholesaler and retailer, to the cigarette smoker in the form of higher prices for cigarettes.

/7/ See Note 6, Cost of Compliance Study, p. 570.

[page 343]

Conceivably, the legislator might collect the tax directly from the person whom it is intended to burden. As a practical matter, however, the government cannot ask each cigarette smoker to pay a certain number of cents for each package of cigarettes that he has smoked during a certain period. Levying the tax on the manufacturer instead of on the consumer can therefore be regarded in large part as an administrative device. It obviates inordinately high costs of collection, intolerable evasion, and inconvenience to the consumer.

Disadvantages of Relying on Business Firms

Not only the taxpayer (the manufacturer) but also the others in the chain -- the wholesalers and retailers -- must use their distributive mechanism for tax purposes if the burden is to rest finally upon the consumer. Although the use of a private mechanism such as this is often unavoidable, it has some serious disadvantages.

Failure to Place Burden Where Intended

The government may be deceiving itself in thinking that the burden is being passed on to the consumer. It may instead be carried by the producer, or the wholesaler, or the retailer, or shifted "backward" by a reduction in wages paid, or rent paid. /8/ This point is irrelevant if the tax is laid blindly, without thought of the burden, as many undoubtedly are.

Cost Imposed on Business Firms

A second disadvantage lies, of course, in the cost imposed on the unofficial collectors. The amounts involved have been indicated in the discussion above on cost of compliance. The question may be raised, however, of the fairness of imposing such a cost.

Most complaints on this issue seem to be merged in-

/8/ See pp. 360-61.

[page 344]

distinguishably with complaints about cost of compliance in general
-- that is, the complaints are made regardless of whether the
immediate taxpayer is also supposed to be the ultimate burden-bearer.
If he is not supposed to be the ultimate burden-bearer, his costs of
compliance may be called "agency costs." They represent costs that he
incurs in acting as an unofficial government agent. Occasionally,
such a tax as the recent short-lived federal tax on checks raises a
clear-cut protest over agency costs. In this case, the complaint came
from the banks.

The government rarely offers any compensation to its unofficial agents. Sometimes, however, the business firm is allowed to keep a certain share of the tax money. For example, wholesalers are allowed to keep 10 per cent of the cigarette tax imposed on them in Alabama. Reasonable doubt may exist whether this represents a real compensation. A reduction of the tax rate by 10 per cent ought to have the same effect. Indeed, no necessity for reimbursing the taxpayer exists if he can shift the cost, like the tax itself, forward to consumers in higher prices. Whether any given taxpayer can or cannot do this is a question requiring special study. It would be especially useful to discover whether the given taxpayer's "agency costs" are disproportionate to those of his competitors. A uniform allowance to all will not apportion reimbursements equitably among the taxpayers.

Another cost must sometimes be borne by business firms intermediate between the taxpayer and the ultimate burden-bearer. The shifting of the tax results in higher prices and the higher prices necessitate added working capital, on which interest must be paid. This cost, however, like the tax itself, may be shifted to others under certain conditions. Moreover, it is not a social cost. By incurring it, the business firms save the government interest on money it would have to borrow

[page 345]

if it were to wait and collect the tax when the article was consumed.
Apparently no tax laws in the United States allow compensation to
intermediaries (as distinguished from immediate taxpayers) for this
or any other cost traceable to the tax.

Extent of Dependence on Business Firms

Only a minor part -- apparently about 30 per cent -- of federal tax revenue in the United States comes from taxes that are obviously not supposed to be passed on in business transactions. This includes the personal income tax, the excess profits tax, the estate tax, and the gift tax. Probably much of the state and local property tax and motor vehicle license tax also belongs in this group.

The corporation income tax, which produces about one-fifth of the federal tax revenue, is probably intended to burden not consumers but corporate stockholders. Thus the tax is not supposed to be passed on in business transactions, but the corporation acts as unofficial tax agent for its stockholders.

Almost all the other important kinds of tax -- federal, state, or local -- either represent a clear intent to use the business firm as unofficial agent through business transactions or give little indication of the lawmaker's purpose. In the former group are to be found the gasoline tax, most of the customs duties, the sales tax, the tobacco tax, and the liquor tax, with a number of similar commodity and service taxes of minor importance. In the latter group may be placed a large part of the property tax and the motor vehicle license tax, and the general corporation taxes except the federal tax and most of the state taxes on net income. This grouping can only be approximate since interpretation of the legislators' attitude is not easy, and the attitude differs from one jurisdiction to another.

[page 346]


A tax administered at a low percentage cost, with practically no evasion, and a low cost of compliance would approach the ideal from the purely administrative point of view. The taxes in the United States, as has been shown above, fall short of these objectives by an appreciable -- though, in general, not an alarming -- distance. The probable causes for this relative failure will now be examined.

Amount of Money Spent on Administration

Economy that is unintelligent, at least from an administrative point of view, is one of the most widespread causes. The amount of tax revenue brought in as a result of recent studies of tax enforcement carried on by white-collar relief workers has been illuminating. Two of the largest investigations -- one on some federal excise taxes and another on the property tax in Texas -- are said to have more than paid for themselves. The study of minor federal excises resulted in collections and assessments totalling eight times the cost. /9/ Probably few taxes in the country are being collected so vigorously that the cost of the last or marginal tax dollar received is 100 per cent of the revenue.

The taxes that would respond most favorably to increased expenditures on administration are not necessarily the taxes that are being evaded the most. Only tax administrators, and only those administrators who are especially familiar with many taxes, are qualified to hazard a guess on where an additional dollar of administrative expense would bring in the most revenue.

Centralized Administration

A proper balance between administrative centralization and decentralization is important not so much in

/9/ See Note 7, Added Revenues Obtained from Added Expenditures, p. 570.

[page 347]

changing the percentage cost of collection as in checking evasion and
in helping the taxpayer to lower his cost of compliance. In the
administration of the real estate tax, for example, additional
centralization has been called for by most students of the problem,
at least to the extent of having full-time county assessors replace a
larger number of part-time, poorly trained, town or village

Whether administration would be further improved if the state were to select and control the assessors of all property still seems debatable. The experiment has not yet been tried. Certain classes of property, however, are well adapted to highly centralized assessment. Almost every state has central assessment for a large part of the property of public service corporations -- railroads, light and power plants, etc. In Ohio, New Hampshire, and Tennessee, a special state- wide tax on income from certain intangibles has replaced part of or all the ad valorem tax on such property. To this extent, centralized administration has replaced local administration. Large industrial plants that extend over more than one local taxing district and intangible property, even under an ad valorem tax, offer the most attractive fields for the further extension of centralized assessment in the immediate future.

Improvement would come chiefly in two forms. Evasion, including a natural failure to protest against a relative underassessment, would be checked, and the injustices that arise through overassessment would be lessened.

Decentralization, on the other hand, has been urged for the administration of the federal income tax. The present system of referring practically every serious dispute to the central authorities necessitates expensive trips to Washington and, in general, results in long delays and a baffling feeling of remoteness. The officials outside of Washington are granted little authority to

[page 348]

settle matters on behalf of the administration. Decentralization
similar to that under the property tax has been suggested. Each local
administrator, though a federal employee, would settle practically
all questions without appealing to the central authorities in
Washington. Nation-wide uniformity in interpretation of the tax law
would be risked to gain speed in settlement and to decrease both
costs of compliance and costs of administration. Obviously, this plan
would work intolerable injustices if the rank-and-file personnel were
not of a high quality. /10/

An interesting type of centralization, midway between state administration and federal administration, is effected through the co-operative action of several states. This is of marked importance in the gasoline tax. Regional associations of gasoline tax administrators have grouped themselves into the North American Gasoline Tax Conference. The result appears to have been an appreciable decrease in evasion by clandestine transportation of motor fuel in trucks and barges across state lines. On a less ambitious scale, regional associations of administrators of state motor vehicle taxes, liquor taxes, property taxes, sales taxes, and tobacco taxes have developed within the past five years. /11/

Fair Treatment by the Government

For most taxes in the United States at the present time, greater cultivation of a spirit of fairness toward the taxpayer would probably decrease evasion. Adequate tax administration depends upon a certain amount of co-operation by taxpayers. This will be lacking if the taxpayers feel unjustly treated.

Rightly or wrongly, a sense of unfairness seems to have become disturbingly acute among those who pay

/10/ See Note 8, Centralization versus Decentralization, p. 571.

/11/ See Note 9, Associations of Tax Administrators, p. 572.

[page 349]

the federal income tax. The present writer has personal knowledge of
a number of instances where the internal revenue agents were branded
as arrogant and stupid. The complaints are disquieting because they
come from persons who are not contentious or fault-finding, and who
are eager to play the part of good citizens. A more fundamental and a
better-established complaint is that the Bureau of Internal Revenue
considers itself primarily an agent of the federal Treasury, and, as
such, bound to decide all doubtful points in favor of the government.
This complaint raises a basic issue of the status of a government
department. The Treasury Department does at least take the part of
the taxpayer by giving a refund, even when the taxpayer fails to
request it, of any overpayment that is discovered in the process of
checking the returns.

In large part, the sense of unfair treatment results from certain legislative provisions defining net income, over which the administrator has no control. Examples are: the taxation of capital gains with severe limitations on the deduction of capital losses, the prohibition of consolidated returns, and the refusal to allow certain expenses incurred in one year to offset income gained in another year.

Whatever the cause or the remedy may be, the federal income tax is in danger, because a self-assessed tax such as the income tax is particularly dependent upon the active co-operation of almost all the taxpayers -- in contrast, for example, with the customs duties, the tobacco tax, and the death duties. On the other hand, no evidence exists that the federal income tax is in any immediate peril of falling to the low level of the intangibles taxes levied at high rates by some states and localities, where the unfairness of both law and administration and the consequent resentment of taxpayers have admittedly made a joke of the tax.

[page 350]

Unfortunately, the spirit of fairness may under certain circumstances lead to an increase in both cost of administration and cost of compliance. In the complicated economic and social structure of the modern world a taxing instrument must often be correspondingly complicated if it is to meet current demands for justice. The personal income tax, again, offers a good example. It is capable of a relatively close adjustment to individual ability to pay, but only at the price of added administrative difficulty. A case in point is the intricate part of the federal law dealing with reorganizations and other transactions where tax liability or tax reduction is deferred until a final settlement of the investment. Again, a motor fuel tax whose proceeds are dedicated to highway construction and maintenance is more easily administered if no refunds or exemptions are allowed to purchasers of gasoline for use, not on roads, but in farm tractors or for dry cleaning.

Height of Rate

Finally, the incentive to evasion offered simply by high rates is of great importance. The federal spirits tax at $2.00 a gallon is a comparative failure in an administrative sense, despite the care taken in the law and regulations to prescribe in minute detail the manner in which the taxpayer's business shall be conducted. At a much lower rate evasion might almost disappear. The gasoline tax rate is high enough to induce certain expensive and partially successful attempts to evade it. For instance, non-taxable substances, such as kerosene, are known to be blended with taxed gasoline and sold to consumers as pure and wholly taxable gasoline.

There is no uniform level for all taxes at which evasion can be expected to increase markedly. The federal cigarette tax, at $3.00 a thousand cigarettes (6 cents for a package of twenty), probably represents a higher per-

[page 351]

centage of the manufacturer's sales price than the federal spirits
tax. The high cigarette tax rate, however, has caused no evasion of
consequence. Where high rates do not cause evasion, it is usually
because of the concentration of the business in the hands of a few
large concerns that are easily supervised. If the business is one in
which a small amount of capital is sufficient to start a productive
unit, a high rate of tax will usually create a difficult problem of
enforcement. In this respect, the manufacture of cigars contrasts
with the manufacture of cigarettes, and the distilling of spirits
contrasts with the brewing of beer.

Aside from the spirits tax, the gasoline tax, and possibly the intangible property tax, few, if any, major taxes in the United States would show an appreciable decrease in evasion with a mere lowering of the rates unaccompanied by more vigorous enforcement.

[page 352]

Section D. Tax Consciousness

Chapter 23


Various taxes differ in the number of taxpayers that they make aware of the tax burden and in the degree of tax consciousness that they arouse in each taxpayer. Thus the same amount of total tax revenue can be raised with widely varying patterns of tax consciousness.


Several factors determine how conscious of taxation the populace will be. The most important, of course, are the number of people who pay taxes and the amount that they pay directly to the government. When a person pays a tax as a separate transaction between himself and the government, he is far more aware of it than when he pays it as a hidden charge -- for example, in rent -- or as part of another transaction, as in the purchase of goods to which a separately charged retail sales tax applies. For the purposes of this discussion, a corporation is considered as but one taxpayer, while a tax on a partnership is considered as a tax paid by each of the partners. The term "family" follows the Census usage of including not only normal families but also persons living alone. Direct taxpayers -- those who pay the tax money directly to the government -- are considered first, and in this discussion they include both those who can and those who cannot shift the burden of the tax to others.

Motor Vehicle Tax

The tax that is paid directly by the largest number of people is that on motor vehicle licenses. The total num-

[page 353]

ber of motor vehicles registered for license taxation in the United
States on December 31, 1935, was 26,221,052. Of this total,
22,565,347 were passenger automobiles. /1/ No data have been found
that indicate how many of the passenger automobiles are used by
families with more than one automobile or by business firms. Perhaps
from 15,000,000 to 20,000,000 families own at least one passenger car
each -- that is, from one-half to two-thirds of the 30,000,000
families in the United States.

Not every automobile used by a business firm represents an additional direct taxpayer. Some firms own more than one automobile, and in some cases the taxpayer is already included in the 15,000,000 or 20,000,000 families.

In any event, the number of individuals who pay an automobile license tax is large enough to create an impressive amount of tax consciousness. The tax probably falls in the range of $3.00-$15.00 a year for most of the families. It is ordinarily not payable in installments. /2/ Some automobile owners, however, who are unaware of the great spread between the size of the tax and the cost of the license plate, may possibly think that the license tax is not a true tax but a payment for the license plate itself. Also the acuteness of tax consciousness is probably less if a business firm thinks that it is shifting the tax.

Property Tax

The real estate tax is close to the motor vehicle tax in the number of direct taxpayers involved. By far the largest body of real estate taxpayers consists of home owners. The United States contains 14,000,000 families that own their homes -- between 3,000,000 and 4,000,000 of them farm homes. In all the states except Florida, which has extensive homestead exemption, practically every home owner pays a real estate tax directly to the

/1/ Automobile Manufacturers Association, Automobile Facts and Figures (New York, 1936).

/2/ Tax Systems of the World, pp. 152-57.

[page 354]

local government, and sometimes an additional tax to the state
government. Therefore almost half the total families in the United
States pay a real estate tax directly. /3/ The amounts are
substantial. At 2 per cent of true value, which is not an unusually
high or low rate for most of the country, a home owner with a $5,000
property pays $100 a year. Moreover, home owners in general doubtless
realize that they are not able to shift the tax to anyone else.
Payment can usually be made in not more than four installments.

The owners of the 11,000,000 rented dwellings in the United States likewise pay a real estate tax. Their number cannot be determined, but it must be far less than 11,000,000. Because of the possibility of shifting part of the tax to tenants, however, many of the owners must feel less concern about the tax than those who live in their own houses.

Probably most of the dwelling owners also pay an automobile license tax. The property tax on dwellings, added to the automobile tax, therefore intensifies tax consciousness instead of increasing its spread.

In some states the personal property tax on household goods, clothing, etc., and on intangibles causes widespread tax consciousness. In general, however, the personal property tax may, for present purposes, be considered chiefly a tax on business firms (other than dwelling owners).

Personal Income Tax

After the property tax, the personal income tax supplies the greatest number of direct taxpayers. It is probable that few of them, however, pay directly neither the automobile tax nor the property tax on dwellings. Adding the income tax to the other two taxes, therefore, brings in a few additional taxpayers, intensifies, as a second tax, the tax consciousness of a considerable num-

/3/ See Note 1, Ownership of Homes, p. 573.

[page 355]

ber of automobile taxpayers or property taxpayers, and acts as a
third tax for a number of persons who own both dwellings and

Number of Returns under Present and Past Exemptions

Under the federal personal income tax, 4,094,420 individual returns were filed for 1934, but less than half -- 1,795,920 -- were taxable. The tax paid equalled 6.1 per cent of the net income of the taxable returns. Husband and wife together were represented in 657,795 returns, hence almost 2,500,000 taxpayers, practically all of them adults, were represented directly by the taxable federal returns. In addition the taxable returns showed 827,084 dependents. The preliminary report for 1935 incomes filed to August 31, 1936, shows 4,473,426 individual returns, of which 2,067,736 were taxable. /4/

The number of adults represented by the 1934 returns -- 2,500,000 -- may be compared with the total number of persons of twenty-one years of age or over in the United States, in 1934, estimated at 75,000,000 out of a total population of 126,000,000. /5/

The figure, 2,500,000, loses some of its significance when the size of the tax is considered. Of the taxable single returns included in this figure, about 650,000 were for incomes of less than $2,000. The federal tax on a single person's income of $2,000, assuming that it is all earned income, is $32. About 270,000 married couples living together and 20,000 single persons, heads of families, showed taxable incomes of less than $4,000. The federal tax on a married couple, or a single taxpayer who is head of a family, with an earned net income of $4,000 and assuming one dependent, is $28. Consequently, for at least 1,200,000 out of the 2,500,000 taxpaying indi-

/4/ Statistics of Income, 1934, Tables 2 and 5, pp. 60-61 and 66-67; Treasury Department, Bureau of Internal Revenue, Statistics of Income for 1935: Preliminary Report of Individual Income Tax Returns Filed to August 31, 1936, p. 1.

/5/ Derived from Department of Commerce, Bureau of Foreign and Domestic Commerce, Statistical Abstract of the United States: 1934, Fifty-sixth Number, 73 Cong., 2 sess., House Document No. 368, Tables 11 and 27, pp. 9 and 34.

[page 356]

viduals the federal tax payable per return was not more than about
$30, on 1934 incomes. Moreover, since the tax can be paid in four
installments, any one payment need not have been more than $7.00 or

Although the amounts computed above seem small, they are not negligible -- particularly since most taxpayers entertain no hope of shifting their income tax to someone else. Some doubt may exist, however, whether the tax at these levels is high enough to induce much concern over the amount of governmental expenditures.

In the fairly prosperous year of 1924 the federal exemptions were the same as those reflected in the 1934 statistics, but the form of the earned income credit did not tend to reduce the number of taxable returns, as the 1934 provision did. In 1924, 4,490,000 taxable returns were filed -- two and a half times as many as in 1934. This number was exceeded only in 1920, when, with considerably lower exemptions, 5,518,000 taxable returns were filed -- three times as many as in 1934. /6/

In some of the states, where lower personal exemptions are in force, the persons subject to the income tax are relatively more numerous. For example, Kansas, where the exemptions are $1,500 for married persons, $750 for single persons, and $200 for dependents, received for 1934 incomes 71,500 returns, of which 48,700 were taxable. /7/ These taxable returns represented a total of 92,000 persons, out of a population of 1,848,000. Thus about 5 per cent of the populace is taxed, as against 3 per cent under 1934 federal returns. About twenty-five states have exemption levels lower than those of the federal government, partly because of the effect of the earned income credit in the federal tax. /8/

It must be recalled, however, that more than one-third

/6/ Statistics of Income, 1920, and Statistics of Income, 1924.

/7/ Kansas State Tax Commission, Income Tax Law and Regulations 35 . . . Statistics for 1933 and 1934, pp. 113-14.

/8/ See p. 301.

[page 357]

of the populace lives in states that do not impose personal income

Number of Returns under Hypothetical Exemptions

If the personal exemptions and the credit for dependents of the federal personal income tax were lowered by about 50 per cent, the number of taxable returns would be greatly increased. For incomes of less than $5,000 the number would rise from 1,000,000 to 8,100,000, giving a total of 8,400,000 taxable returns. Both figures are for 1933 incomes, and they would of course be much larger than that in a prosperous year. /9/ Very few new taxpayers with incomes of more than $5,000 would be discovered by this change.

The present exemptions are $2,500 for a married couple or head of family and $1,000 for a single person, and the credit for dependents is $400 for each dependent. However, the effect of the earned income allowance is that all single persons with an income of $1,111 or less and all married couples (or heads of family) with an income of $2,778 or less are exempt.

The 8,100,000 figure assumes a lowering of the exemptions to $1,000 and $500, and the credit to $200. If the reductions were only to $1,500, $800, and $300, respectively, the total number of new taxable returns under 1933 conditions would be, not 7,100,000, but 2,400,000, giving a total of 3,700,000 taxable returns. Both of these hypothetical situations assume continuance of the 10 per cent earned income credit.

The total number of new returns, taxable and nontaxable, would, of course, be much greater. In 1933, the number of taxable and non- taxable returns filed from all income groups was 3,700,000. With the lower of the two sets of exemptions and credit noted above, the total returns would have been nearly 13,000,000 -- that is, 12,-

/9/ The 1934 figure to compare with 1,000,000 is 1,380,000. The 1934 data were not, however, available in time to compute the increase under 1934 conditions.

[page 358]

200,000 returns with a net income above the exemption limits, /10/
plus perhaps 500,000 other returns. With the higher exemptions and
credit, total returns would have been between 7,000,000 and
8,000,000. /11/ These totals, so far as they reflect non-taxable
returns, are of course much more useful in indicating the
administrative task involved than in showing possible tax

Business Taxes Generally

The other taxes are of little importance in a discussion of direct taxpayers. The business tax that accounts for the largest number of direct taxpayers is the sales tax. In the United States the sales tax is restricted usually to sales by retailers. In 1933 there were 650,000 retailers in the twenty-two states that had a sales tax at the end of 1936, and another 100,000 or 150,000 can be added for New York City and New Orleans, bringing the total to 750,000 or 800,000. /12/ Only a few of this group are exempt from the sales tax.

Whatever the precise figures may be, however, the number of individuals who are conscious direct taxpayers is far smaller than for the automobile tax and the others discussed above. Many consumers bear the burden of the sales tax, and even pay it in a legal sense, but they do not make the payment themselves directly to the government. The purchasers' consciousness is, therefore, not considered here but in a subsequent section. In view of the relatively small number of direct payers involved in the sales tax, it is not necessary to consider the other business taxes, which reach an even smaller number of taxpayers directly.

/10/ See Note 2, Increases in Returns from Changes in Exemptions and Earned Income Credit, p. 573.

/11/ Ibid.

/12/ Department of Commerce, Bureau of the Census, United States Summary of the Retail Census for 1933, Press Release, December 3, 1934, pp. 9-10; Carl Shoup, "The Experience of Retailers under New York City's Sales Tax," Bulletin of the National Tax Association, XXI, No. 4 (January 1936), 98.

[page 359]

Death Taxes and Gift Taxes

The death taxes and gift taxes reach relatively few taxpayers. Only 12,724 returns were filed under the federal estate tax in the calendar year 1935, and 1,614 of these were for non-resident decedents. Of the remainder, 8,655 were taxable, but 4,483 of these paid a total tax of only $1,797,000 -- about $400 each. /13/ While the lower exemptions of many of the state death taxes increase the number of taxpayers considerably, the total number affected is still relatively small.

In contrast with the income tax, almost all those -- decedents and beneficiaries -- who bear the burden of the death taxes are, in any one year, a new group who have not paid the tax before, or at least not for many years. For a given number of returns and a given amount of tax over a period of years, therefore, the death tax creates more tax consciousness. On the other hand, because of the general remoteness and indefinite time of payment the intensity is not so great. The taxpayer probably does not worry so much about possible future payment as he does with the income tax and other regularly recurring taxes.

Payroll Taxes

The taxes levied in the Social Security Act are of uncertain importance as instruments of tax consciousness. Sufficient data are not yet available to make a comparison with other taxes, but 26,000,000 employees will probably contribute to the old-age pensions when the plan is in full operation. /14/ According to the definition of direct taxpayer used in the present discussion, however, the only direct taxpayers under the act will be business firms.

/13/ Statistics of Income, 1934, Tables 1-5, pp. 36-47.

/14/ John G. Winant, Chairman, Social Security Board, New York Times, May 8, 10, 1936.

[page 360]

Number of Direct Taxpayers

Apparently about two-thirds of the families in the United States are made tax conscious in the sense that they pay a substantial amount of money in taxes each year directly to some governmental unit. Without the automobile tax, only about half the families would be direct taxpayers. Practically every business firm pays a tax of some kind. The unincorporated firms, however, consist of "families," most of whom are probably included in the direct taxpayers, just noted, under the automobile tax, the dwelling tax, and the income tax. On a strictly numerical basis, corporations are insignificant. Through their owners, creditors, and others, their tax burdens exercise a significant degree of tax consciousness; it is, however, indirect.


In many states the gasoline tax and the retail sales tax create a large number of conscious indirect taxpayers. Because the amount of the tax is usually made known to all purchasers, they realize that some definite part of the purchase price is destined eventually to reach the government. They may also believe, if they stop to think of it, that the indirect burden is a real one -- that if the tax were removed the price that they would have to pay would be correspondingly lower. Whether the burden is in fact a real one is not relevant. A merchant who sells an article for 98 cents plus 2 cents sales tax might possibly have sold it for $1.00 even if no sales tax had been in force. For all practical purposes, however, the consumer does not know it and he may, in effect, be a conscious indirect taxpayer even though the tax is not shifted to him. Sometimes, the consumer does know it, as in the New York City restaurant that posts a sign reading: "All wines and liquors reduced 1 cent from prices

[page 361]

posted; added 1 for tax." The result is a special kind of "tax

Sales Taxes

Fifteen of the twenty-two sales-taxing states, and New York City and New Orleans require the retailer to state the tax to the consumer as a separate item instead of hiding it in the price. /15/ This requirement has been sponsored chiefly by the retailers themselves in the expectation that it would facilitate shifting and would also, by keeping consumers aware of the tax, improve the chances of repeal. The seventeen jurisdictions that practice separate charging comprise about one-third of the total population of the country. In a few of the other jurisdictions the practice is more or less widespread on a semi-official or unofficial basis.

In all twenty-four jurisdictions the sales tax is so broad that, wherever a separate charging plan is used, few, if any, residents can be unaware of the tax. The kind and degree of tax consciousness that is caused, however, is difficult to ascertain. The day-by-day payment of extra pennies or the handling of metallic tokens, cardboard tax receipts, or prepaid tax coupons may develop an intelligent interest in government. They may, on the other hand, create a pin-prick irritation against taxes in general, or simply an almost unconscious response whose real meaning is lost.

Other Taxes

The possibilities of the social security taxes have been mentioned above. It will be of interest to discover whether a long- continued payroll deduction results in the employees' finally forgetting that the cause of the deduction is a tax. Perhaps the form in which they are

/15/ William Paul Walker and Everett C. Weitzell, The Retail Sales Tax, with Particular Reference to Administrative Problems of Its Collection (College Park, Md., February 1936), p. 13, and American Retail Federation, Sales Taxes: A Digest of Each State Law (Washington, November 18, 1936).

[page 362]

notified of the deduction will prevent a loss of tax consciousness.

The consumer's attention is drawn to some of the state commodity taxes by tax stamps that show the amount of tax paid by the merchant. The federal government, however, has not adopted this policy. The federal tax stamp on packages of twenty cigarettes, instead of informing the taxpayer that the manufacturer has paid a tax equivalent to 6 cents, merely states the administrative classification, "Class A." Similarly, the amount of the federal internal revenue tax on spirits, $2.00 per proof gallon, is not stated on the tax stamp.

Landlords have on occasion considered the idea of presenting their tenants with rental bills broken into two parts: rent proper and property tax paid by the landlord. Allocation of the tax among the apartments of a building is difficult, however, and no landlord could know exactly how much of the tax he was in fact shifting. The latter problem exists for all taxes that are not borne entirely by the direct payer, but it is particularly difficult in connection with the real estate tax -- partly because of the distinction that must be drawn between land and buildings. /16/

Wherever consciousness of indirect taxation is found, two or more strata of tax consciousness exist, since the person who pays the money directly to the government is sure to be aware of the tax.


A number of miscellaneous factors may increase tax consciousness.

A tax may reach such a stage of unpopularity that a large proportion of the taxpayers evade it successfully. This sort of tax consciousness, however, is scarcely considered desirable. Evasion of this degree exists in the

/16/ See pp. 296-97.

[page 363]

states that try to impose the ordinary property tax on such
intangible property as stocks, bonds, etc. In some places, however,
the evasion becomes so habitual and so complete that practically all
consciousness of the tax is lost.

Variation in the amount of tax payable from year to year probably increases tax consciousness -- compare, for example, the income tax with the real estate tax. The effect is particularly great when the variation seems to have a close relation to variations in the volume of governmental expenditures. Moreover, the taxpayer may feel that he has discounted some of the steadier taxes in his purchase price, as when he buys real estate.

Publicity of the amounts paid by income-taxpayers encourages curiosity in one part of the general public and may cause resentment in the taxpayer. Wisconsin has allowed publicity of income tax returns, and the federal government has experimented with it in 1924 and 1934 and again, with respect to salaries only, under the 1936 act.

Administration of a tax in an officious and generally tactless manner is another way to heighten consciousness into resentment. /17/

Well-organized publicity by private groups has at times created passing waves of tax consciousness. In some states, as in Ohio, the organized movement has aroused the taxpayer to such a pitch that drastic tax limits have been put in the constitution. It was claimed that 265,000 signatures were obtained on a petition for this constitutional amendment. /18/ Corporations sometimes make their shareholders aware of the burden of taxation through comments by officials in their reports or on other occasions. /19/

/17/ See p. 349.

/18/ Adam Schantz, "Real Property Tax Limitation," Proceedings of the . . . National Tax Association, 1934, p. 43.

/19/ See pp. 343-44.

[page 364]

The income taxes usually allow deduction of certain taxes paid, in computing net income. Many taxpayers have been made conscious of these taxes -- with even a momentary feeling of gratitude -- when they have calculated the amount that could be deducted because of taxes paid on admissions, gasoline, retail purchases, income, etc.

Finally, the cost of compliance has been an important stimulant of tax consciousness -- especially in certain business circles where the problems of federal and state income taxation have required elaborate record-keeping. Among individuals this cost is often a psychic cost rather than a monetary cost. Generally, the more complex the tax, the greater the consciousness of it will be.

[page 365]

Section E. Intergovernmental Co-ordination

Chapter 24


Many of the aims discussed in preceding chapters can be most nearly achieved under a centralized tax system. The taxes generally conceded to be most adaptable to the personal status of the taxpayer -- notably the personal income tax -- can usually be administered with the least evasion by the government that has the widest jurisdiction. Moreover, a single tax on income or business may be cheaper to administer, and its incidence can be better controlled from the viewpoint of justice, than independent federal and state taxes. And, except where special benefit is involved, taxes that are uniform in rate and base for the whole national economy are less apt to exert an undesired influence on industrial developments than taxes that vary from one jurisdiction to the next. /1/

A decentralized government requires independent taxing powers, however, and independent taxing powers cannot be reconciled with a uniform and centralized tax system. Consequently, a tax that would be discarded in a centralized system as relatively inequitable, or costly to administer, might be given an important place in a decentralized system because it was one of the few taxes available to a small jurisdiction. Many of the faults of existing tax systems in the United States are directly attributable to decentralized government, and some of them cannot be remedied without sacrificing an important degree of state or local independence. In consequence, the choice lies, to some extent, between justice and efficiency

/1/ Special benefit is used here to cover any advantage that a local government may offer an industry even though the cost of the superior services given is covered by a general tax levy rather than a special benefit tax or charge.

[page 366]

in taxation on the one hand and local self-government on the other.

In practice, the attempt to couple the virtues of centralized taxes with the virtues of decentralized government has led to a more rapid centralization of taxation than of other functions. This is not, however, a complete solution of the problem. It is important, therefore, to consider how much the tax system can be centralized without sacrificing local responsibility. In the absence of such responsibility the merits of local self-government are largely illusory.


Pressure for an integrated tax system has two sources: (1) increasing government expenditures and (2) increasing integration of the economic system of the country as a whole. Such a tax system can yield more revenue with less unwanted interference with business than independent and varying systems. This point is important to taxpayers and government alike.


With the creation of the national government in this country, revenue sources were divided in the usual manner of federal states. The customs duties were assigned to the national government, and all direct taxes were left to the state governments. In practice there was complete separation of sources, but the tax powers of the two jurisdictions overlapped. Both governments might levy excises, and the national government had the power to levy direct taxes, although only in proportion to the population of the various states. /2/ Actually, the national government rarely attempted to make such per capita levies.

The first permanent invasion by the federal govern-

/2/ See Note 1, Uniformity Requirement for Federal Taxes, p. 512.

[page 367]

ment of the field of direct taxation was the income tax, permitted by
the Sixteenth Amendment. The states had previously made little effort
to develop this form of direct taxation. Most of them, in their
earlier history, found property and poll taxes adequate for their
purposes. The states also made little use of the excise taxes that
have played so important a part in national finances beginning with
the Civil War. Consequently, until the past quarter century
practically no duplication occurred in federal and state taxes.
During the past twenty-five years, however, the situation has
altered. The growing demand for revenues has led to an increasing
number of duplications, and investigators of the problem now number
them in hundreds.

Two taxes on the same base are not in themselves harmful if they are properly integrated and the combined rates are not excessive. State and local governments shared the same taxes -- chiefly the property tax -- without conflict as long as rates remained low. When combined rates are high, however, an increase in the rate levied by one jurisdiction makes increases in the rates levied by other jurisdictions more difficult, and conflict arises. The conflict lies deeper than the duplication of particular taxes. Fundamentally these overlapping jurisdictions are competing for the same economic resources, whatever the form of the taxes by which the different shares are obtained.

Co-ordinate Governments

Competition is not limited to federal and state governments and state and local governments. Counties compete with towns, and towns compete with school districts. The competition is not even limited to jurisdictions that cover the same geographic area. One city may not tax real estate situated in another city; but it may, by granting tax exemption for a limited period, induce

[page 368]

a manufacturing concern to build new plants within its boundaries
instead of in a neighboring city. In this way it may ultimately add
to its taxable real estate at the expense of its neighbor. /3/

States may compete with one another in much the same fashion. For interstate business, state competition may take another form. One state may obtain its maximum revenue from a corporation income tax by allocating all interstate income, for purposes of taxation, to the state in which gross receipts arise. Another state may obtain its maximum by allocating such income to the state where the tangible property of the corporation is located. If each state pursues its own immediate interest, some corporations may pay taxes in both states on their interstate income while their competitors pay taxes in neither. These facts are so well known that it is hardly necessary to enlarge on them here in order to demonstrate the importance of co-ordinating tax systems.


Constitutional Limitations

The obstacles to co-ordination are legal and political rather than economic. Federal and state constitutional limitations set the framework within which tax systems may be constructed.

Specific restrictions are few. The federal Constitution is generous in its grant of tax powers to the national government, and yet it does not seriously limit the tax powers of the states. It forbids the federal government to levy a direct tax except in proportion to the population of the states. Court interpretation and the Sixteenth Amendment have restricted this prohibition, for all practical purposes, to a federal property tax levy. The states, on their side, may not impose customs duties; nor may they tax interstate commerce, or property that has been held

/3/ See pp. 148-50.

[page 369]

to be beyond their jurisdiction. /4/ Neither government may levy
export duties.

Within these restrictions, the whole field of taxation is open to both federal and state governments. The consequence is extensive duplication of taxes. If the federal Constitution had made a sharp division between federal and state taxing powers, the resulting separation of sources would have limited the area of conflict. This solution of the problem would bring certain difficulties, however. /5/

Effective co-ordination of tax systems does not demand complete separation of tax sources. In the absence of such a separation, however, it does demand a larger measure of control in the hands of the central government than is feasible in a federal state. This centralization of control is necessary to resolve the conflicts between national and state governments and between states.

The most significant provision of the federal Constitution, from the point of view of interstate tax relations, is the prohibition of taxation of interstate commerce. Without this provision state tax relations might be even more chaotic than they are, but it has brought problems of its own. Not only has it failed to prevent some forms of double taxation, but, with the growth of interstate business, it has offered an increased opportunity for tax avoidance, which no amount of state co-operation can abolish. The problem is particularly serious for corporation income and gross sales taxes.

The control of local taxation that is granted to state governments by their constitutions is sufficient to prevent conflicts between the state and local jurisdictions and also between one local jurisdiction and another. However, a majority of the constitutions require that

/4/ See Note 1, Supreme Court Decisions, p. 575.

/5/ See pp. 374-76, 383-84.

[page 370]

taxes should be equal and uniform. This requirement, as interpreted
by the courts, has frequently barred the use of such taxes as the
progressive income tax. As a result, the states have had to rely so
heavily on the general property tax that its availability for local
use has been seriously impaired. Co-ordination between the federal
and the state governments and between the various state governments
is outside the sphere of state constitutional control.

Limitations Inherent in Decentralization

Constitutional restrictions do not offer an insurmountable barrier to co-ordination. Constitutions can be amended as need arises, however slow and cumbersome the process. The real obstacle lies in the nature of decentralized government itself, not in the constitutional provisions that have created and preserved it. A uniform tax system requires a centralization of tax power, and probably also of tax administration, but both kinds of centralization are more consonant with a unitary than with a federal state.

Complete centralization of the tax system in the hands of the national government would leave state and local governments entirely at the mercy of the central government. Whatever other rights and privileges they might have would be in name only if they were denied any measure of independent financial power. The central government might, of course, grant the underlying governments substantial freedom, together with the means of financing their activities. But, whether support were provided in the form of direct subsidies or through delegation of certain tax powers, such freedom would be on sufferance, at best. If the federal form of government and local self-government are to be maintained, complete centralization of the tax system is impossible.

[page 371]



Co-ordination of the tax systems of jurisdictions that occupy the same geographical area does not necessarily imply complete centralization. The more serious conflicts resulting from independent, overlapping, and to some extent competitive tax systems may be avoided in other ways. Five specific methods will be briefly reviewed.

(1) The first is to adopt a complete separation of sources (using "sources" in its commonly accepted meaning of types of tax) between federal and state or state and local governments.

(2) A second method is to share the same sources, with due regard for one another's interests and the interests of the taxpayers. Independent federal and state taxes on the same base are apt to be wasteful from an administrative point of view, but they can be made practically uniform in base, reasonable in rate, and -- except for the additional cost -- no more injurious to the taxpayer than a single tax that produces an equal amount of revenue.

(3) A third method is to establish a uniform base for each tax and a single administration, each jurisdiction remaining free to fix its own rates. This obtains most of the desired unity without diminishing the responsibility of the rate-fixing jurisdiction. It does not, of course, resolve the conflict arising from varying rates among governmental units of the same rank. Moreover, it is probably a solution for state and local, rather than federal and state, authorities, since it practically assumes an element of compulsion. States might, of course, agree voluntarily to use a federal tax base and to avail themselves of federal administrative machinery. They would, however, sacrifice a measure of their independence in doing

[page 372]

so, no matter how willingly the sacrifice might be made. Actually the
use of a uniform tax base, with a single administration but varying
rates, resembles closely the manner in which some states share the
property tax with their subordinate governments, except that
administration in the latter case is local. In many states, however,
the base used by state and local governments varies, and, even where
it does not, local administration introduces inequalities not
contemplated by law and not apparent in the nominal rates.

(4) A fourth method of co-ordinating the taxes of the different layers of government is the crediting device adopted by the federal government in the estate tax and the payroll tax for unemployment insurance. /6/ Lacking the power to force the states directly to adopt uniform taxes, the federal government has bought uniformity in this way. In both these instances, however, the end in view was not so much the avoidance of federal-state competition as the avoidance of competition among states. The states have such extensive control of their local governments that they are hardly likely to use this indirect method. It has been recommended, however, for a motor passenger carrier franchise tax in New York State. /7/

(5) A fifth -- and complete -- solution is central administration of all taxes with distribution of part of the yield to underlying governments. At one extreme the distribution may take the form of fixed percentages of specified taxes returned to the locality of origin. Here, the central government serves as a collecting agent, and little more, although the decision on the form of tax is in its hands. At the other extreme, the distribution may take the form of a grant-in-aid distributed in accordance with need. Here, the subordinate government usually

/6/ See pp. 26-27.

/7/ Report of the New York State Commission for the Revision of the Tax Laws, 1932, p. 168.

[page 373]

surrenders a large measure of control in spending the money and in
addition loses all control of the source and amount of its income. To
the extent that the tax conflict is solved by this means, local self-
government and states' rights are relinquished.

Co-ordinate Governments

The co-ordination of the tax systems of local jurisdictions of equal rank will depend on the power of the superior government. Any state government can obtain complete harmony of local taxes within its boundaries. If the state fixes uniform bases for all local taxes and adopts uniform rules for the allocation of bases when the situs is not clear, the injustice of double taxation is avoided. If, in addition, it fixes rates in order to attain complete uniformity, the revenues lose their adaptability to local conditions, and the local officials lose responsibility. With these losses the chief merits of the locally administered tax disappear. If central administration is adopted, the shared tax or the grant-in-aid, or both, will be resorted to as long as local authorities are left with spending powers.

Federal control over states does not extend to requiring any marked degree of uniformity among them in tax bases or rates. The Fourteenth Amendment, as it is interpreted by the Supreme Court, prevents the taxation of (1) real or tangible personal property outside the boundaries of the taxing state, and (2) intangible personal property belonging to anyone not domiciled in the state (except possibly intangibles used in a business). But such control as the federal government can exercise must come, for the most part, through indirect channels. The crediting device, noted above, is the outstanding example, and it has been effective as far as it has been used.

The federal government might, however, buy uni-

[page 374]

formity in other ways. It might, for instance, use its own powers of
taxation to levy a tax on a base that is also available to the
states, and all or a substantial share of the proceeds might then be
distributed to the states that refrained from levying taxes on the
same base. If the states' shares were large enough, they would
doubtless adopt such a plan. The distribution of the proceeds would
not need to be in accordance with their origin. It might even take
the form of a grant-in-aid for a specific purpose.

In the absence of federal control, co-operation may be used to avoid conflicts between state and state although genuinely conflicting interests will limit its application. The reciprocal provisions of state inheritance taxes are a step in this direction.



Recent developments in the United States are important, both as they indicate trends and as they demonstrate which methods are, and which are not, successful in co-ordinating revenue systems under a decentralized form of government.

Separation of Sources

As noted above, the separation of sources of federal and state revenues, which prevailed in our earlier history, has been definitely, and probably permanently, abandoned. This development is not peculiar to the United States. One by one, particularly since the war, other federal states have given up the separation of sources characteristic of their earlier history.

The only important federal tax forbidden the states is the customs duties, which brought the federal government only 10 per cent of its tax revenues in 1936. The only important state tax forbidden the federal government is the property tax, which brought the states only

[page 375]

6 per cent of their tax revenues in 1936. /8/ The federal government
does not levy a motor vehicle license tax or a general sales tax, but
it competes to some extent with its automobile excises and its many
specific sales taxes. In the important fields of personal income,
corporation, alcoholic beverage, gasoline, and tobacco taxes, both
federal and state governments have taxes on practically the same

The states might conceivably give up their tobacco taxes and the federal government, in exchange, might give up its gasoline tax. It is hardly to be expected, however, that either will give up income, corporation, and alcoholic beverage taxes. These three taxes contributed more than half the federal tax revenues in 1936 and also more than half the tax revenues of some of the states. Other important overlapping taxes are death taxes and (in a few states) stock transfer taxes. According to estimates, in 1934, 90 per cent of federal tax revenue and 57 per cent of state tax revenue came from bases subject to both federal and state taxes. /9/ In 1936, 80 per cent of federal tax revenue and 55 per cent of state tax revenue came from such bases. These figures include the entire proceeds of certain federal taxes that are subject to state taxation in only a part of the states.

The trend in state and local tax duplication is the reverse of that for federal-state tax duplication. Separation of sources is increasing. In 1902 the general property tax, which supplied local governments with more than 90 per cent of locally administered tax revenue, also contributed more than half of state tax revenue. In 1925 it continued to supply the local governments with much the larger part of their tax income, but contributed less than one-third of state tax revenue. And in 1935 state tax revenue from this source had dropped to 7 per cent.

/8/ See Table J, pp. 530-33.

/9/ J.W. Martin, "Trends in Federal-State Taxation Relationships," The Southern Economic Journal (July 1936), pp. 71-72.

[page 376]

Such large states as New York, Pennsylvania, Illinois, Ohio,
Michigan, California, Massachusetts, and Iowa levied no state
property tax in 1936. /10/

Localities have not achieved financial independence, however. On the contrary, they are more and more dependent on the states for support. But both state and local governments are depending less on the locally administered general property tax and more on state- administered sources. Only in the technical sense of changes in amounts of revenue from given sources is separation of sources increasing.

Federal and local taxes overlap very little. Isolated instances may be found, such as the county income tax in New Castle County, Delaware, the recently repealed New York City estate tax, and occasional local gasoline and beverage taxes. Little reason exists to expect any important increase in this overlapping, however. Those taxes that local governments might administer with some success -- gasoline and beverage taxes -- are so firmly entrenched in state tax systems that there is little probability that local governments will be permitted to invade the field.

Co-operation in Utilizing Independent Taxes on the Same Base

The federal and state governments have not levied independent taxes on the same base without taking account of the taxes already in force in the various jurisdictions. Many of the taxes give evidence on this point. For example, the provisions of the state income tax laws have obviously been influenced by federal tax provisions. Differences in federal and state practice have been more often stressed than similarities, but actually the similarities are striking. Also, the federal tax provides for deduction of state taxes in computing taxable net in-

/10/ See the memorandum by Thomas J. Reynolds on state property tax rates in the forthcoming volume, Studies in Current Tax Problems.

[page 377]

come, and most state taxes permit deduction of the federal tax.
Moreover, it is widely accepted that the high federal surtax rates
preclude high state rates. Similar evidence of a consideration of the
taxes imposed by other jurisdictions can be found in other taxes.

Formal co-operation is relatively new. The appointment of a Subcommittee on Double Taxation of the Committee on Ways and Means in 1932, however, gave official recognition to the problem. The first meeting of the Tax Revision Council in 1935, with representatives of federal, state, and local governments, also gives promise for the future.

Uniform Tax Bases with Varying Rates

The common European practice of permitting local governments to levy their own rates on tax bases fixed by the state, with state collection of both state and local taxes, has never gained a foothold in the United States. However, the administration of the property tax in some states, particularly the tax levied on public utilities, closely resembles this form of taxation. As state interference with local property tax administration increases, the European practice may well be adopted. This form of co-operation is open to federal and state governments only if the states are willing to sacrifice an important measure of independence.

Crediting Devices

The use of the crediting device to co-ordinate federal and state taxes, introduced in 1924 for death taxes, has been fairly successful. Its application to the income tax and to other taxes has frequently been suggested. It is used in the payroll tax for unemployment insurance for a somewhat different purpose, but it probably also assures a freedom from conflict in this new field. This indirect form of coercion is not used by the states in their local

[page 378]

relations, since they have more direct methods of control at their

Central Administration and Redistribution of Revenue

Strangely enough, the method of co-ordinating tax systems that is least compatible with decentralized government -- centralization of taxes -- is the one gaining the greatest headway in the United States at the present time. Taxes have not, of course, been transferred from local to state and from state to national governments. However, centrally administered taxes have increased in proportion to the total more rapidly than the taxes of the underlying jurisdictions. In the short space of three years, 1932-35, federal tax revenues increased from 23 per cent to 37 per cent of the total tax revenues, state tax revenues increased by a fraction of 1 per cent of the total, and local tax revenues declined from 57 per cent to 44 per cent of the total. /11/

The rapid growth of federal tax revenues is to some extent an emergency and a cyclical phenomenon. The same development, to an even greater degree, occurred during the war period. But the proportion of federal taxes failed to return after the war to its pre-war level, and there is good reason to believe that it may not drop even to its pre-depression level in the future. The increase of state taxes compared with local taxes is not a depression phenomenon and may be expected to continue.

This increase of centrally administered taxes has been accompanied by a much smaller increase in expenditures on centrally administered functions. A large part of the adjustment has come, instead, through the increased sharing of state-administered taxes with local governments and through federal and state grants-in-aid. In 1932 local governments obtained 14.5 per cent of their

/11/ See p. 107; Financial Statistics of State and Local Governments: 1932; Report of the Secretary of the Treasury, 1932.

[page 379]

tax revenue from state-administered taxes -- 9.9 per cent in the form
of grants-in-aid and 4.6 per cent in the form of shared taxes. The
remainder of their tax revenue, 85.5 per cent, came from locally
administered taxes. In 1935 local governments obtained 16.5 per cent
of their tax revenue from state-administered taxes -- 12.0 per cent
from grants-in-aid and 4.5 per cent from shared taxes. /12/ This
difference is not striking. In addition, however, the local
governments received 24.4 per cent of their tax revenues from federal
grants-in-aid for welfare, leaving only 59.0 per cent of tax revenue
from locally administered taxes.

Federal aid to localities is an innovation. The federal government has for many years been developing a policy of grants-in- aid to induce the states to develop activities that the federal government had no power to force upon them. But local governments have hitherto been left entirely to state control and aid. The federal aid has passed through state hands, of course, and states have supervised its expenditure. The actual administration has been essentially local, however.

The federal government has likewise contributed substantial sums to state support through grants-in-aid for relief, highway construction, and other purposes. In 1932 such grants supplied the states with 12.5 per cent of their tax revenues, the remaining 87.5 per cent coming from state-administered taxes. In 1935 the corresponding percentages were 22.2 and 77.8.

As noted above, this extensive federal aid is essentially a depression phenomenon. But the sources of revenue that are most sensitive to business conditions are the federal and state taxes. It seems probable, in consequence, that while local revenues rise gradually with the slow recovery of property values (increased tax limitations are an additional handicapping factor), state and federal tax

/12/ See Note 2, Shared Taxes and Grants-in-Aid, p. 575.

[page 380]

revenues may experience a rapid rise. Federal budget estimates for
1937 contemplate a 55 per cent increase in tax revenues over 1935. No
comparable increase in local taxes can be expected.

This increased income may be applied to debt reduction rather than aid to state and local governments, but it will probably encourage a continuance of the grant policy on a somewhat larger scale than in the pre-depression years. The provisions of the Social Security Act give even more convincing evidence that in the long run federal grants can be expected to increase. Further development of state grants likewise seems probable. These have a longer history than federal grants, and, although they have grown more slowly, their steadier development makes their permanent use even more probable.

Grants-in-aid signify a loss of independence by the receiving jurisdiction. They are invariably accompanied by restrictions on their use. In addition, they frequently demand of the receiving jurisdiction a quality of service and a predetermined amount of financial participation that leaves little discretion to the authorities who receive the aid.

The practice of sharing state taxes with local governments has grown rapidly in recent years, largely as a result of the growth of highway expenditures and their support from motor vehicle taxes. More than half the state tax revenue that was shared with local governments in 1935 was from motor vehicle taxes. This form of assistance may prove less permanent than the grant.

Shared Taxes versus Grants-in-Aid

The essential difference between the shared tax and the grant- in-aid is not that the latter is usually earmarked for a specific purpose. It is, rather, that the amount of the shared tax usually varies from year to year with the yield of the tax whereas the amount of a grant is meas-

[page 381]

ured by the need of the receiving authority. Smaller jurisdictions
probably cannot be asked to share changes of fortune over the
business cycle, for example, equally with the larger jurisdiction.
Local authorities are experiencing increasing difficulty in adjusting
their own revenues to needs. Moreover, the revenue system of the
larger jurisdiction has a much greater adaptability to economic

The grant is more readily accompanied by central control than
is the shared tax, and this control will probably increase. The
temptation to dictate is very great, and the distribution of
increasing sums of money unaccompanied by control fosters local

Between 1932 and 1935 shared taxes declined slightly in relative importance, while grants-in-aid increased -- in spite of the general tendency to share the new liquor taxes and, occasionally, the gross sales taxes. The explanation is, of course, that, since grants are based on need, they are more apt to increase than decrease in time of depression, whereas shared taxes diminish with declining yields. Shared taxes may again rise rapidly as tax yields rise.

The federal government has yet to experiment with the sharing of tax yields, although plans for sharing many important taxes have been proposed. The Graves-Edmonds plan, for example, proposes federal administration and federal and state sharing of a general manufacturers' excise, and liquor, gasoline, and tobacco taxes.

Co-ordinate Governments

Conflicts between local governments are practically limited to the property tax. Even here the conflict is not so much among co- ordinate as among overlapping jurisdictions. Tax limits have protected the taxpayer to some extent, but they have only aggravated the conflict among the jurisdictions themselves. Since real estate clearly can

[page 382]

be taxed only by the jurisdiction in which it is located, serious
conflict between county and county and city and city has been

The conflict between state and state increases, however, as the amount and variety of state taxes increase. Since the adoption of the crediting device for death taxes in 1924, the federal government has made little effort to exercise such limited control as it might have in order to reduce this conflict. It did not even extend the credit to the additional estate tax, levied in 1932 and increased in 1934 and 1935. The credit for the payroll tax is for quite a different purpose. Supreme Court decisions in death taxes have helped in some instances -- notably the line of decisions starting in 1929 in which it has been held that intangibles, except possibly those used in a business, may be taxed only by the state where the owner resides. /13/ Unfortunately, however, some problems of double taxation of intangibles remain. The recent Dorrance case demonstrates that, at least until another Supreme Court decision, two states may tax the same property simply by claiming to be the state of residence of the owner. The use of the shared tax would bring uniformity between state and state, as well as between federal and state governments, but the federal government, as noted above, has yet to experiment with this device.

Co-operation among the states has progressed more rapidly in recent years than federal control. Reciprocal agreements for inheritance taxes eliminated much serious overlapping before the series of Supreme Court decisions just noted. Recognition of the interests of both the taxpayer and other states is to be found in the provision of the New York personal income tax law for reciprocal tax credits. By this provision, New York, in general, exempts residents of a state that in turn exempts New York residents. The tendency to adopt the "Massachu-

/13/ See Note 1, Supreme Court Decisions, p. 575.

[page 383]

setts formula" in preference to any other in allocating corporation
income for state taxation is eliminating some double taxation in that

States tend to model their new tax laws on the laws in neighboring states, probably because the latter are known to be administratively feasible. Even though uniformity is not the purpose here, it is a result. Uniform tax laws do not necessarily, however, avoid double taxation, since the laws may provide for the taxation of both resident and non-resident income.

Organized efforts to achieve co-operation are relatively new, but they have aroused a great deal of interest and an enthusiastic response from state officials. The Interstate Commission on Conflicting Taxation has made important facts available and has brought the problem to the attention of those concerned. The recommendations of this committee deal with the conflict between federal and state governments rather than the conflict among the states, but the two problems are closely related, and to some extent they can be solved by the same means.


In judging the relative merits of the different adjustments that have been listed above, two factors must be kept clearly in mind. (1) Will the adjustment maintain the desired degree of local responsibility? And (2) can it adapt revenues to needs?

Separation of source meets the first test, but not the second under existing governmental organization. The largest share of governmental functions is still performed by local authorities and the only important tax that most local governments can administer successfully is the property tax, which is not always equal to the task. Even though local expenditures as a whole could be met from a reasonable tax rate levied on property as a whole,

[page 384]

separation would not offer a satisfactory solution. The fundamental
difficulty is that wealth is very unequally distributed, and the
localities that are poorly endowed are the ones with the greatest
needs. Inequalities are in fact so great that even states are not
always large enough to be self-supporting under all conditions.

Independently administered taxes on the same base, and centrally administered taxes with a uniform base and varying rates, likewise meet the test of responsibility but fail in flexibility. It is not enough to have the rates determined by the authority who receives the proceeds when the base itself is inadequate. The crediting device fails to meet either test, although this failure does not condemn its limited application.

The shared tax likewise fails to meet either test if the distribution is on the basis of origin of the tax revenue. Even when the shared tax is distributed on a per capita basis, or on some more direct measure of need, it fails in a time of emergency. Local governments have not the same powers of adjustment to falling tax yields that larger jurisdictions have. The grant-in-aid, properly distributed, meets the test of adaptability to needs but fails to meet the test of maintaining local responsibility. It can, of course, be safeguarded with adequate controls, but to the extent that controls are applied local independence is circumscribed.

The problem of tax co-ordination cannot be solved as long as a large part of our business is interstate and much of it nation-wide. There are, of course, sectional differences in industrial development, and a substantial part of our business activity is essentially local in character. Unfortunately these sectional differences do not follow state lines; and even a local business may be interstate. There is, thus, an irreconcilable conflict between our industrial unity and our political disunity. This can be alleviated, but not abolished, by financial measures.

[page 385]

The extensive redistribution of wealth between geographical areas through taxation and grants may be defended on economic grounds. Although it is true that an unwise state-aid policy may result in the perpetuation of submarginal communities that might better be depopulated, some communities that are too poor in tax resources to support themselves are a vital part of our national economy. The wealth that they create, in cooperation with others, simply appears in taxable form outside their legal jurisdiction. values are rarely the result of purely local forces. Even the real estate values in large cities are not entirely local in character. They are created in the first instance by the dense local population. But large cities are the least self-sufficient communities that exist. Their taxable wealth would melt very rapidly if they were cut off completely from the surrounding territory. It is not enough, however, to justify the redistribution of revenues on these or any other grounds; some workable method of redistribution must be found.

If enough tax revenues were redistributed today, through grants- in-aid or any other means, to enable local governments to meet the standards of expenditure set by the state itself, some communities would receive more than nine-tenths of their income from the central government. And to establish the necessary safeguards to insure wise spending under these conditions would mean the practical abandonment of local self-government.

The problem is not purely a financial one. Other powerful forces are impelling centralization of government. And some centralization is apparent. The taking over of roads and schools in North Carolina recently and Virginia's state support and administration of all rural roads are isolated instances of a widespread movement. The mileage of state highways grows year by year. Welfare is more and more accepted as a national obligation.

[page 386]

The movement is slow, however, and it might be unfortunate to hasten it for purely fiscal reasons. Many of the best achievements of government in this country have been achievements of local government. Local administration is more adaptable to local needs and wishes. Consequently, it is important to consider what financial measures are most suitable for the present degree of centralization or decentralization.

The greatest degree of local independence could be obtained if local governments were left in full enjoyment of the tax on real estate. State governments have other sources at their disposal, and probably few would find its complete abandonment a serious hardship. Some of the poorer agricultural states would, of course, find it difficult to fill the gap. Most communities could support a substantial measure of local self-government from the proceeds of a moderate tax rate. As a result of this shift more taxes might be levied on real estate than on personal income, though the latter is in general, of course, a more satisfactory measure of ability to pay. However, this disadvantage might well be offset by the gain in local independence. This tax would of course have to be supplemented by state assistance. Grants-in-aid are more adaptable to both local need and central control than other forms of state assistance. It might be necessary, in addition, to sacrifice a large measure of local government in very poor communities.

The federal-state and interstate conflicts are complicated by legal barriers that are not present in the state-local problem. Fortunately, however, the inequality in resources is not so great among states as among local governments. Furthermore there is no reason to suppose that state governments will have to give up any large measure of independence because they are unable to support themselves. Present conditions point to a continued use of the same tax bases by both jurisdictions,

[page 387]

although a greater separation of sources is probably feasible. The
two interests might be reconciled to some extent both through
cooperative effort and through such indirect pressures as the federal
government can bring to bear. /14/ The sharing of federally
administered taxes should not weaken state responsibility or result
in a maladjustment of revenues if the device is used moderately. If
it is used to force a completely uniform and unified tax system,
however, both these results are inevitable.

An irreconcilable conflict admits of no clear-cut solution. Complete co-ordination of the tax system cannot be achieved under a federal government. The most that can be hoped for is a series of compromises. These will not remove the fundamental conflict, but they may remove some of the consequences.

/14/ in order to further cooperative effort the establishment of an "impartial, disinterested commission of the highest possible dignity and standing, to represent neither the states, nor the Treasury, nor the President, nor the Congress, but the people of the United States" has been recommended. This commission would be for the purpose of formulating policies. Address by Robert M. Haig at the opening session of the second Interstate Assembly, February 28, 1935, in The Book of the States . . . I (3rd ed.; Chicago, 1935), 312-13.

[page 388]