VI Tax Exempt Securities

Should the tax exemption of Government securities by abolished? If so, what differentiations should be made with respect to future issues and outstanding issues authorized and sold under various terms and conditions? The problem is: first, what is fiscally and socially desirable; secondly, what is constitutionally and politically feasible.

Constitutional Question. -- With respect to the constitutional question, if a layman might venture an opinion it would be that the Supreme Court, with a different personnel or under other conditions than those which it has actually confronted, might have built up a very different line of decisions, but in fact it has established precedents that make invalid Federal and State taxation of each other's outstanding securities, at least, without the consent of the authorizing Government, and except in extreme emergencies, for example, such as might justify such extreme action as nullification of the gold clause in Government issues. Of course, the personnel of the Court and the pressure of events and of public opinion upon the Court change from time to time; there have been important reversals in the past, there may be others in the future.

But with the constitutional lawyers in hopeless disagreement and with no desire to make hazardous forecasts, we shall leave the question of constitutionality to the Supreme Court, which will not pass on it unless a law is enacted and a real case is brought before it.

Fiscal And Social Questions. -- The fiscal and social aspects of this problem of exemption are in almost as much of a snarl as are the legal aspects. Opponents of exemption have been most vocal in recent years, and they claim that it causes substantial loss of revenue, the nullification of the progressive or ability-to-pay principle of taxation, and the consequent maintenance of a privileged class. Examination of available evidence indicates that there is some truth in these claims but not so much as is usually believed, though it must be admitted that there are many gaps in the evidence and most of the analyses are not conclusive with respect to certain important points. Inasmuch as the present writer has had insufficient time and assistance to carry this analysis appreciably further this summer, the best that is practicable at this time is to summarize the data that are available, including the few additional that have been secured; to indicate the gaps that need to be filled; and to suggest the conclusions that are indicated, even though a number of them are insufficiently substantiated.

Outline Of Problems. -- The main issues, other than legal, may be suggested in outline form, somewhat as follows:

What would be the effects of abolition of the tax exemption of
Government securities

     If the Federal and Sate Government reciprocate

          With respect to income taxes only?
          With respect to property and certain other
          taxes possibly?

     If Federal and State Governments do not reciprocate,
     or do so only partially?

Important effects to consider are

     Fiscal effects

          On tax yields                         )
          On Government interest payments       )  of Federal,
                                                )  State,
                    Because of additional tax   )  and
                                                )  local
                    Because of uncertainty of   )  Governments
                    future taxes                )
               On extent of public expenditures )
               (extravagance)                   )

     Effects on private capital market and business

     Social effects

          Because of graduated rates

          Because of privileged classes, etc.

Amounts Of Securities Outstanding. -- On June 30, 1934 there were outstanding Federal interest-bearing securities in excess of $26,000,000,000, besides nearly $3,000,000,000 in Federal Farm and Home Owners' Loans, or a total of about $29,000,000,000. The gross State and local debt on December 31, 1933 is estimated at over $19,000,000,000, (net debt at $17,000,000,000). Thus, the grand total of interest-bearing Federal, State, and local issues amounts to from $46,000,000,000 to $49,000,000,000. Of the Federal issues nearly half are wholly tax exempt, but interest on the remainder is subject to the surtax. Of the State authorized issues all are exempt from Federal income taxes. Most State issues are exempt from their own State and local taxes but not from taxation by other States when in the hands of owners subject to the jurisdiction of the State levying the tax. Inheritance and estate taxes are commonly considered taxes on transfers or privilege and are not subject to the usual limitations on property taxes and on some income taxes.

The annual interest on Federal issues amount to about $950,000,000 (about 3 1/3 per cent) on State and local issues probably to about $1,000,000,000 (about 5 1/2 per cent).

Government Securities Outstanding on June 30, 1934.

United States Government                        $26,084,249,000
United States Possessions                           160,000,000
State and Local                                  19,429,000,000 /1/
Federal Farm and Home Owners' Loan                2,811,108,975

Total                                           $48,484,447,975

/1/ This figure, presented by the Division of Financial and Economic Research of the Treasury, is an estimate as of December 31, 1933.

A large part of the Federal and other Government bonds are held by banks and other corporations not subject to Federal surtaxes, hence the Government fails to collect taxes on interest from most of its bonds that are supposed to be taxable, though ultimately some of this is reached in the hands of dividend receivers. There are many gaps in the information respecting the amounts of Government bonds held by various income and other classes of owners, hence it is impossible to determine with accuracy how much extra in taxes the Government would get if none were tax exempt. /12/

Revenue Losses. -- The August 30, 1934 Treasury estimate /13/ of additional revenue that would be gained by subjecting the interest from Federal, State and local Government issues to Federal income taxes is as follows:

Corporation income tax                                $34 million
Individual income tax                                 148.5 million

Total                                                $182.5 million

The corresponding estimate of Mr. L. H. Parker, Chief of Staff of the Joint Committee on Internal Revenue Taxation, nearly a year ago, before the 1934 Act was passed and when passed and when the public debt was smaller than now, was that extra revenue "would not exceed $160,000,000 annually". Various estimates have been made by others.

Effects Of Taxes On Interest Rates. -- There is even more uncertainty about how much extra interest the Federal Government and others would have to pay on their debts if tax exemption were eliminated. Mr. McLeod, Government Actuary, estimates 1/2 of 1 per cent for Treasury bonds and 5/8 -- 3/4 of 1 per cent for certificates of indebtedness as the probable differential effect of Federal taxes.

The accompanying graph (Graph 11) supplied by Mr. McLeod shown the differences in yields from Federal bonds subject to surtaxes only and from others wholly exempt.

Table 14 gives similar data in tabular form for the First Liberty Loan 3 1/2's (wholly tax exempt) and the First Converted 4 1/4's (subject to surtaxes). The average differences shown over a period of 14 years is .727 per cent. If the very abnormal figures for 1920 were eliminated, the average would be slightly lower.

On a Federal debt of $30,000,000,000, 1/2 of 1 per cent amounts to $150,000,000 annually; 7/10 of 1 per cent amounts to $210,000,000. If State securities were taxed, the effect upon them would have to be included also.

The estimates of revenue losses above include that from Federal taxes on State and local, as well as Federal, issues. If these estimates are correct, the elimination of tax exemption would appear to indicate little or no fiscal gain to the Federal Government and a substantial loss to the States and localities, assuming Federal taxes only. State taxation of State and Federal bonds might mean little or no gain to the States, but certainly much loss to the Federal Government.

In 1924 the Federal Trade Commission estimated the loss of Federal taxes due to tax exemption of securities to be between $181,000,000 and $100,000,000 in 1923, with the latter figure probably more nearly correct. /14/ It suggested that taxing such securities might raise the interest rate very little or not at all because very all because very large public debts might so saturate the market that buyers subject to surtaxes would have to pay no more than the marginal buyer. This is a common argument, which appears to overlook the fact that all buyers are marginal with respect to the last units purchased by each. The Commission did recognize, however, that the interest rate might be raised by taxation.

Table 14. -- Prices and Yields on First Liberty Loan 3 1/2 Per Cent Bonds and First Liberty Loan Converted to 4 1/4 Per Cent Bonds, June 30, 1920, to June 30, 1934 (a) (Yields calculated at maturity date)

Year                       First 3 1/2's (b)
and Day
                    Prices                       Yield
                                               Per Cent

June 30, 1920       $91.00 -- $91.08             4.049
Dec. 31, 1920        89.90 --  89.92             4.130
June 30, 1921        86.54 --  86.60             4.369
Dec. 31, 1921        94.84 --  94.94             3.816
June 30, 1922       100.04 -- 100.10             3.497
Dec. 31, 1922       100.96 -- 101.00             3.441
June 30, 1923       100 14/32 -- 100 17/32       3.470
Dec. 31, 1923        99  5/32 --  99  6/32       3.553
June 30, 1924       101 17/32 -- 101 19/32       3.401
Dec. 31, 1924       100 30/32 -- 100 31/32       3.439
June 30, 1925       100 31/32 -- 101  1/32       3.435
Dec. 31, 1925        99 17/32 --  99 20/32       3.528
June 30, 1926       101 11/32 -- 101 15/32       3.406
Dec. 31, 1926       101  5/32 -- 101  9/32       3.417
June 30, 1927       101       -- 101  2/32       3.428
Dec. 31, 1927       101 20/32 -- 101 26/32       3.379
June 30, 1928       100  2/32 -- 100  5/32       3.492
Dec. 31, 1928        99 22/32 --  99 27/32       3.517
June 30, 1929        96 10/32 --  96 13/32       3.781
Dec. 31, 1929        99 15/32 --  99 16/32       3.540
June 30, 1930       100 31/32 -- 101  1/32       3.422
Dec. 31, 1930       101 26/32 -- 101 29/32       3.352
June 30, 1931       102 16/32 -- 102 19/32       3.294
Dec. 31, 1931        97 27/32 --  97 28/32       3.683
June 30, 1932        101 4/32 -- 101  6/32       3.401
Dec. 31, 1932        102 9/32 -- 102 11/32       3.297
June 30, 1933       102 17/32 -- 102 19/32       3.270
Dec. 31, 1933       100.18 (Closing)             3.447
June 30, 1934       104.04 (Closing)             3.111

Year                       First 4 1/4's (c)
and Day
                    Prices                      Yield     Difference
                                               Per Cent   in Yields
                                                           Per Cent

June 30, 1920       $85.60 -- $85.70             5.251        1.202
Dec. 31, 1920        85.10 --  85.20             5.300        1.270
June 30, 1921        87.23 --  87.32             5.141         .772
Dec. 31, 1921        97.08 --  97.18             4.440         .624
June 30, 1922       100.14 -- 100.18             4.240         .743
Dec. 31, 1922        99.06 --  99.10             4.311         .870
June 30, 1923           98 --      98 4/32       4.382         .912
Dec. 31, 1923         98 9/32 --  98 11/32       4.366         .813
June 30, 1924        102 4/32 -- 102  7/32       4.103         .702
Dec. 31, 1924       101 16/32 -- 101 18/32       4.145         .706
June 30, 1925       102 21/32 -- 102 25/32       4.062         .627
Dec. 31, 1925       101 20/32 -- 101 24/32       4.131         .603
June 30, 1926       102 15/32 -- 102 16/32       4.073         .667
Dec. 31, 1926       102 30/32 -- 103  2/32       4.033         .616
June 30, 1927       102 29/32 -- 102 31/32       4.034         .606
Dec. 31, 1927       103 12/32 -- 103 16/32       3.994         .615
June 30, 1928       101 18/32 -- 101 22/32       4.126         .634
Dec. 31, 1928        100 6/32 --  100 9/32       4.231         .714
June 30, 1929         99 1/32 --   99 3/32       4.325         .544
Dec. 31, 1929        101 7/32 -- 101 10/32       4.147         .607
June 30, 1930        102 3/32 --  102 8/32       4.071         .649
Dec. 31, 1930       102 28/32 -- 102 30/32       4.007         .655
June 30, 1931       103 11/32 -- 103 13/32       3.963         .669
Dec. 31, 1931        99 10/32 --  99 24/32       4.292         .609
June 30, 1932       101 20/32 -- 101 22/32       4.105         .704
Dec. 31, 1932        102 8/32 -- 102 11/32       4.039         .742
June 30, 1933        102 5/32 -- 102 10/32       4.039         .769
Dec. 31, 1933        101.18 (Closing)            4.097         .650
June 30, 1934        103.13 (Closing)            3.912         .801

                                               Average         .727

(a) Compiled from monthly reports of the Treasury Department.
See statement of Government Actuary -- Form A.

(b) This issue is exempt, both as to principal and interest,
from all taxation, except estate or inheritance taxes, imposed
by the United States, its possessions, or by any State or local
taxing authority.

(c) This issue is exempt, both as to principal and interest,
from all taxation except (1) estate or inheritance taxes, and
(2) graduated additional income taxes, commonly known as
surtaxes, and excess profits and war profits taxes. The
interest on an amount of such bonds the principal of which does
not exceed in the aggregate $5,000 owned by any individual,
partnership, association, or corporation, is exempt from the
taxes provided for in subdivision (2) of this section.

Of course, if both Federal and State Governments tax each other's bonds, the difference in interest rates by virtue of this taxation might be less than if only Federal or only State bonds were exempt. On the other hand, if the bonds of either or both jurisdictions are made taxable, an extra element of uncertainty is introduced, no one can know what the tax rates will be a year hence, particularly 20 or 40 years hence, consequently the normal interest rate will have to take account of this uncertainty as well as of the tax itself. The uncertainty of the effect of a definite tax and the additional uncertainty of the effects of unpredictable taxes raise doubts of the desirability of abolition of exemption in the minds of those intimately connected with the Government bond market, especially now when large refundings and new flotations are being considered. But to make the change after all important issues are out would be like locking the barn door after the horse is stolen.

Dr. C. O. Hardy, who has made the latest published comprehensive study /15/ of this subject, comes to the conclusion that losses by virtue of tax exemption of Government bonds are commonly much overestimated; that with respect to wholly tax exempt Federal securities, the Treasury probably losses about the same in taxes as it gains in interest, but that it sustains a net loss with respect to Federal Farm Loan bonds; that the Federal Government loses because it cannot tax State and municipal securities, but that this loss is no greater than the issuing Governments gain; that most other arguments advanced in favor of the abolition of tax exemption are unsound or of insufficient importance to justify the proposed change, but, on the other hand, that most arguments generally advanced in defense of tax exemption are also of no weight. He found no evidence that high surtaxes interfered seriously with business activity or added to the cost of living. Moreover, he thought that the tax disparity caused by the very high surtaxes as compared with taxes on those with small incomes could be justified only as a means of equalizing incomes or of offsetting the regressivity of other taxes.

As stated earlier, the data bearing upon the various questions involved in this problem are so incomplete and elusive that the analyses based on them are inconclusive, so much so that those who have strong feelings against tax exemption certainly will hesitate to accept either the indications of the estimates given above or Dr. Hardy's conclusions.

Comprehensive Study Recommended. -- The completion of the study of tax exempt securities -- so well begun by the Couzens Committee, the Federal Trade Commission and Dr. Hardy -- is a very large task, but the importance of the issues involved appears to justify exploring further the possibilities in this connection. The task could probably be performed by the Treasury Section of Financial and Economic Research if it does not have too much other work, but the task is too large to be taken on incidentally as a small extra.

It would appear more appropriate for the Treasury than for almost any one else to make this very important study, but, if this is not feasible, one might consider making unofficial inquiries to see if Dr. Hardy and the Brookings Institution or the National Bureau of Economic Research would undertake the task. /16/ This should be done at once to avoid delay in legislation before too many more bonds are issued. Without some such extensive study, about all one can do now is to make the best estimates and guesses possible on the basis of studies and facts at hand. The issue is rather important, it is being pressed by business organizations and by Congressmen more and more each year. The Treasury and Congress are practically forced to take some position on it.

Tentative Conclusions. -- The writer's tentative and insufficiently supported opinion is that the Federal Government loses comparatively little fiscally through tax exemption of its own securities, that the States and local units of Government probably lose still less on the whole -- in fact, they may gain --; but that exemption may nullify graduation in part and foster certain privilege unduly, although it is popularly thought to foster even more privilege than it in fact does.

On the whole the writer is quite uncertain as to what is wise for the Government to do at this time about an elimination of tax exemption of Government securities, though he is inclined to favor it if the cost is not too great. He does not at present believe, however, that much, if any, net revenue is to be gained in this way. But if a more exhaustive study of the subject shows that the Governments concerned would lose no substantial amount of net revenue, if any, by the abolition, he would be in favor of it because of its effect upon progressive taxation, social justice, and class feeling.

Some argue that tax exemption never had the constitutional support commonly supposed, and that it has even less since the adoption of the sixteenth amendment. One way to test this would be for Congress to put the question up to the Supreme Court by the enactment of a law. It is very doubtful if the States will approve most of the proposals for a constitutional amendment permitting the taxation of these securities. It would be a breach of faith to tax all present outstanding Government securities under the guise of an occupational excise tax as is proposed by some governmental authorities, even if this method is technically constitutional as some proponents claim. It would not be so indefensible, however, to use such a method to reach future issues.

Recommendations. -- If the study suggested above could be started promptly with an adequate staff, recommendations could probably be made before the end of the next session of Congress, in time for that body to submit an amendment based on the finding. If there is no prospect that Congress will soon have any more information than is now available, the writer would make the following recommendations, though with much difference for reasons already indicated:

1. Submit a Constitutional Amendment (Oliver plan) to permit the Federal Government to levy non-discriminatory income taxes upon future issues of securities authorized by the states, with provision for returning proceeds ratably to the States. Or,

2. Submit Constitutional Amendment to permit both Federal and State Government to levy non-discriminatory income taxes upon each other's future issues of securities. (Government compensation might be included. Very doubtful if three-fourths of States would ratify this, and it is not urged strongly.)

3. If neither of the above is politically feasible, enact statute to permit non-discriminatory Federal and State levies of general excise taxes, measured by income, including income from tax-exempt Federal and State authorized Government securities, upon future issues of such securities, (Not urged.)

Any one of these plans, if adopted, would settle, or lead to settlement of, the most important constitutional questions now involved, though any plan involving a constitutional amendment would probably require several years for ratification.

The Oliver plan has much more chance of ratification by 36 States than the second plan. It would permit a trial of the effects of taxing Government securities and might result in additional revenue, possibly to both State and Federal Governments, unless borrowing rates were raised too much. It would enable the Federal Government to get a little more in taxes than it turned back to the States even though it distributed to the States the amounts of tax collected at average rates on State securities, because it would throw more income into surtax brackets where the rates are above the average. This extra amount is relatively insignificant, however. If it worked well, this plan would probably lead the States to tax their own securities, but it would not permit the taxation of Federal securities by 48 different States and their numerous local governmental units with all varieties of methods and rates. Even though such taxes were non-discriminatory, they might have serious effects, due to uncertainties as well as to terms of taxes per se, as noted a few pages above.

The second plan is not very attractive on account of the objections noted in preceding paragraphs; moreover, it has little change of ratification. If no better plan is evolved, however, it might be tried.

The third plan is a rather devious one; direct methods would be preferred if feasible. If applied to future issues only, however, as any plan should be, bond buyers would be put on notice and no breach of faith would be involved.

Note. -- In order to give a better idea of the problem summarized above and also of the gaps in data at hand, there is in the Appendix to this section an outline or brief statement of the study we planned to undertake but had to give up because of the difficulty in getting the desired data with the help and time available. (Pages 221-223a)