Public Principles of Public Debt: A Defense and Restatement
An overwhelming proportion of the outstanding public debts of all nations represents debt which was originally created to finance public expenditure during periods of war. It will be useful to examine this war-created debt as a special case.
The classical model which has already been thoroughly developed is fully applicable for that portion of war borrowing which represents a wholly voluntary transfer of economic resources from private to public uses. No additional discussion of this type of war debt is needed. Government sale of securities to individuals or to nonbanking institutions during the war periods is no different, in economic effect, from the sale of such securities during periods of substantially full employment. If, during the initial phases of a war, some unemployment exists, war borrowing might take on some elements of the Keynesian as well as the classical model. But no essentially new elements appear which warrant further discussion.
The classical model is clearly inappropriate for the portion of war borrowing which implements or facilitates the transfer of real resources from the private to the public economy involuntarily. Government sale of securities to the banking system introduces a unique operation which seems to combine elements of the classical and the unemployment models. Some further discussion of this sort of war debt is necessary.
Periods of deep depression are not the only ones in which government may employ public debt issue as a means of accomplishing the effects of direct currency creation. Government borrowing from the banking system during war periods is of this nature. If the government borrows from the system in such a way that the borrowing-expenditure process itself creates sufficient excess reserves to finance fully the banks' purchases of bonds, the results are substantially identical to an outright operation of the printing presses with the same subsequent expenditure of the funds.
A significant share of the outstanding national debt of the United States was created in this way during World War II. In December, 1939, the banking system held $18.4 billion in federal government securities. This total had increased to $115 billion by December, 1945. The increase was divided between the commercial or member banks and the Federal Reserve Banks in an approximately four-to-one ratio. By the process of borrowing one fourth of the additional $100 billion from the central banks, sufficient excess reserves were created (when the funds were spent) to allow the member banks to purchase the other three fourths. By this operation, the excess reserves of commercial banks were continually increased, and with these excess reserves they were able, and encouraged, to purchase government securities yielding a low, but positive, interest return. No liquidity was sacrificed in the operation, and, as Paul Samuelson once suggested, it could have been a 1 per cent war as easily as a 2 per cent one.*90 This manner of financing a large part of World War II expenditures was the major cause of the postwar difficulties in preventing and controlling inflation effectively.
The effects upon the general level of prices stemming from this type of debt creation are equivalent to the direct creation of currency. In this the operation is akin to depression-born debt which is financed out of private hoards and excess bank reserves. The only difference is that, in the latter case, individuals and financial institutions do sacrifice some liquidity in exchange for the claims on future income which is transferred to them in the form of debt instruments.
On the spending side, however, this war borrowing operation takes on classical overtones. When the funds secured in exchange for debt instruments are expended, resources are purchased which would have otherwise been used by private people. A real transfer of resources from the private to the public sector of the economy does take place. This transfer is, however, only superficially similar to that which takes place in the classical model. The wartime transfer is not effected by any reduction in private disposition over economic resources when the debt is issued. No private person suffers current sacrifice of goods and services until and unless the funds are spent. The borrowing side involves no part of the real transfer. This real transfer is wholly effected on the expenditure side by the incremental purchasing power which the newly acquired funds make available. The expenditure of the debt-financed funds creates inflation, and the inflation facilitates the real transfer. Thus, completely unlike the classical case in which bond purchasers voluntarily abstain from current usage of economic resources in exchange for future income claims, private people during the period of expenditure give up current command over resources involuntarily. With their given money incomes, individuals find that they are unable to purchase as many real goods and services as before the combined debt-expenditure operation takes place. These effects occur whether the inflation takes the open or the repressed form. If there is open inflation, individuals are confronted with higher prices. If the inflation is repressed by some sort of direct controls, they cannot freely purchase as many goods and services as they should like at established prices. In either case, they give up real goods and services which the government acquires. And there is no quid pro quo transaction involved here at all. Individuals give up current real income with no promise of future income. Inflation is equivalent to a tax which coercively imposes upon individuals some sacrifice of current resources.
This type of government borrowing, coupled with the subsequent expenditure of the funds, is important for our purposes because it allows us to locate other sources of support for the basic propositions of the new orthodoxy. Much of the interest in the theory of public debts, and consequently much of the discussion, has arisen from wartime borrowing. And here the failure to distinguish between the effects of the securing of the funds and the spending of them has been evident. The idea that the primary real burden of public debt must be borne by the generation living during the period of the initial borrowing operation seems to have stemmed from an implicit consideration of this particular model of war borrowing, to the exclusion of all other possible cases. If the two sides of the combined borrowing-expenditure operation are not separated, the primary real burden may, somewhat legitimately, appear to fall on the currently living and not on future generations.
But here the error is identical to that which is involved when depression-created debt is considered. If the borrowing alone is considered, a primary real burden is present which is shifted forward in time to the shoulders of future taxpayers. So long as the debt instruments carry with them some obligation to transfer to bondholders some future incomes, future taxpayers bear a real-debt burden. This is the cost side of the fiscal operation.
This burden of debt, or real cost of securing the funds, may be as low as desirable, since debt issue here is not a genuine "market" transaction. The burden will depend upon the amount of subsidy paid to the banking system. And this burden may be, and normally will be, dwarfed in significance by the burden of inflation which the debt operation facilitates. This burden of inflation must be borne largely by the generation living at the time that the resources are given up. It cannot be shifted easily. But this burden of inflation is not due to the debt per se. It arises solely from the fact that the government has not withdrawn, through taxation or through "real" borrowing, sufficient funds from the private economy to finance its purchases without a distortion of the general price structure. (This point was discussed briefly in Chapter 7.)
From all this it follows that the question "Who pays for the war?" is not equivalent to the question "Who bears the burden of the war debt?" The answer to the first question may be the one which the orthodox theory would suggest. If the war is financed largely by taxation and by inflation, whether the latter is directly implemented by currency creation or indirectly through the disguise of war borrowing from the banking system, individuals of the generation living during the war will pay most of the real cost. They will do so to the degree that taxation and inflation impose upon them, coercively, some current sacrifice of real goods and services. Inflation becomes a true tax, regardless of the way that it is financed, and the real burden of taxation does rest on the current generation.
Insofar as the inflation is made possible by the issue of debt, there is added to the current cost a supplemental real burden which is shifted forward to future generations of taxpayers. There is a burden of the debt in addition to a burden of tax (inflation). But this burden of debt may be insignificant in comparison to the larger burden of taxation (inflation).
But is there a "net" burden of debt in this model? Future taxpayers are, of course, worse off with the debt than they would be with its alternative, currency creation, which would have accomplished the same real purposes during the war period. We have stated that these future taxpayers bear the burden of war debt. But have we not overlooked the fact that this sort of war debt also involves beneficiaries who are better off with the debt? It is true that this sort of war borrowing, which provides a subsidy to the banking system, makes certain groups in society better off than they would be under currency inflation. Quite apart from the spending side of the operation, the issue of this type of debt benefits certain groups. No longer is debt issue a market transaction in which individuals or institutions exchange current command over resources or liquidity for future claims on income. The banking system essentially gets "something for nothing" in this operation. And the benefits from the debt accruing to individuals connected with the banking system may be offset against the real burden imposed on future taxpayers.
The recognition that, in this particular case in which debt issue does not represent a "market" transaction, beneficiaries of debt issue per se do exist does not modify the basic conclusions reached in the "general" theory of public debt developed in this book. First of all, the benefits which the banking system receives will tend to be fully capitalized at the moment of issue. Once this capitalization takes place, future receipts of debt interest cannot be classified as differentially beneficial to the holders of debt instruments. As we have argued earlier, future tax payments will not normally be fully discounted. Therefore, during any particular period subsequent to that of the initial debt issue, the social group, as a unit, will experience a "net" burden of debt.
Even if this capitalization aspect is neglected, however, the location of the burden of debt on future taxpayers is still legitimate. Let us assume that bondholders (commercial banks) do not capitalize future interest subsidies. (This assumption would amount to saying they do not carry government securities in their portfolios at capital values.) In future periods these institutions receive interest receipts equivalent to interest outpayments by taxpayers. These two items are offsetting in the aggregate, and no "net" burden could be defined. But surely here it is both appropriate and legitimate to separate the burden or cost of the debt issue from the benefits of the debt. There exists a beneficiary group, quite apart from any gains from exchange which may be present in the other cases. We have shown that the relevant alternative to debt issue here is currency creation which would accomplish the same real purposes. The question becomes that of determining the differential effects of debt issue. These effects may be separated into two parts: benefits and costs; and these two may be compared and evaluated. This is precisely equivalent to the procedure in the classical model.
One relevant alternative to debt issue is a failure to undertake the public expenditure. It is necessary to evaluate carefully the differential effects of the combined debt-expenditure operation. To do so we isolate the cost side, which is represented by taxes borne by future taxpayers, from the benefit side, which is represented by the social real income estimated to accrue to future citizens from the public project financed. Nothing in the classical model necessitates that there must be a "net" real burden when both sides are taken into account. In the classical model another alternative to debt issue is current taxation to finance the same expenditure. Here the differential effects of debt issue are taxes on future generations on the one hand and a lightened tax load on the current generation on the other. The social decision must always involve a comparison. To return to our example of war borrowing from the banking system, in which those who bear the cost may be from among the same generation as those who secure the benefits, it is necessary to separate the two sides of the account and to compare the costs with the advantages. Here the comparison will be based exclusively on distributional considerations. The debt is not necessary to finance the war; this can be accomplished through currency creation. The debt issue is justified here only if the distributional benefits more than offset the distributional costs. Note here that we cannot a priori make any such comparison as this. We cannot say that war borrowing from the banking system is socially undesirable. In the majority of cases, this sort of borrowing may lead to distributional effects which may be considered undesirable by many people. But nothing definite can be stated on this score.
There may, of course, be peacetime periods for which this war-borrowing model is applicable. For example, political pressures against higher interest rates during a period of prosperity and full employment may force the Treasury to refinance maturing long-term issues held by the public by the sale of new securities to the banking system, in part to the Federal Reserve Banks. In this case, the discussion of this chapter almost fully applies. The borrowing operation itself does nothing to reduce private purchasing power; and when the spending of the funds takes place (in this case retiring bonds held by the public) inflation must be the result.
Elements of this war-borrowing model may also be present in depressions. If the government borrows from the Federal Reserve Banks at a positive rate of interest, and subsequently from the newly created excess reserves of commercial banks, there is no sacrifice of liquidity in the purchase of the government securities. The minimum real cost of securing the fuller resource utilization remains zero, but the real burden of the debt must be offset here against the real benefits accruing to the interest-recipient beneficiaries.
The discussion of this and the preceding chapter reveals that much of the fuzziness of the new orthodoxy stems from a consideration of the depression borrowing from idle hoards and war borrowing from the banking system. As we have shown, neither of these forms involves "real" borrowing, that is, no reduction in private disposition over resources is required on the borrowing side of the fiscal operation. The transfer of resources in the depression case can be secured more efficiently through money creation since some resources are previously unemployed. In the war case, the transfer from the private to the public sector is facilitated solely by the incremental addition to purchasing power. This, too, could have been achieved by direct currency creation.
In either of these cases, if the debt alternative is adopted, some burden is imposed on future taxpayers. But this cost need not be large since bond purchasers only give up liquidity in the one case and nothing in the other. This cost of debt will tend to be insignificant when considered relative to the benefits from public expenditure in the depression case and relative to the burden of inflation in the war.
Notes for this chapter
Paul A. Samuelson, "The Effect of Interest Rate Increases on the Banking System," American Economic Review, XXXV (March, 1945), 16-27.
End of Notes
Return to top