Public Principles of Public Debt: A Defense and Restatement
Public Debt and Depression
The three fundamental propositions of the currently accepted theory or conception of public debt have been demonstrated to be false and insupportable within the framework of the classical assumptions. The discussion should prove sufficient to destroy any claims that the "transfer payment" approach provides a "general" theory. But is the new orthodoxy not acceptable when the classical assumptions are dropped? And, by implication, does not the classical reformulation fail to apply in conditions of economic depression? Is there one theory of public debt which is appropriate under the classical assumptions and another which is applicable when "Keynesian" conditions prevail?
In this chapter I shall show that there is only one "correct" theory of public debt, and that the reformulation of classical theory developed in this book does provide a "general" theory which is applicable to all possible situations, and under all sets of assumptions. There is no need for two contrasting theories of public debt.
The important difference between the classical and the Keynesian models lies, of course, in the assumption concerning the level of employment. The full employment assumption will now be relaxed, and it is here assumed explicitly that there exist in the economy some economic resources which are not employed at the time of debt creation.
The importance of the two methodological principles discussed in Chapter 3 is worth re-emphasizing. First of all, the necessity of introducing and examining the necessary offsetting or compensating changes in the other fiscal variables cannot be neglected or assumed not to take place. Relevant alternatives must be compared. Secondly, although the compensating changes must be examined as an integral part of the analysis, the two sides of the transaction must be conceptually distinct. Debt theory has been characterized throughout its history by an oversight of both of these principles.
In the classical model it is necessary to distinguish between the real and the monetary aspects of a government borrowing operation only in a secondary sense. The existence of a money economy does not seriously modify the real nature of the transaction involved. Roughly similar conclusions would be produced from an analysis which assumes a pure barter economy. The government borrows resources, either directly or indirectly, through the medium of some generally accepted item classified as money which, in turn, gives to the government command over the disposition of resources. The resources which are transferred to government, via the debt operation, would have been used by private people had not the transfer taken place. In return for these resources, the government promises to transfer a certain amount of resources to its creditors during future time periods. Without debt creation, the government could have secured the usage of resources currently only through taxation. Inflation of the currency as a means of financing public expenditure in this model becomes a form of taxation and should be analyzed as such.
The introduction of unemployment suggests that a different sort of operation must take place when the government borrows. Regardless of the manner in which the government secures the funds, the public expenditure for resources does not necessarily involve any transfer of real resources from current private employments. If the government purchases the services of unemployed resources, no real transfer need take place; no private person need give up any disposition of economic goods and services. In one sense, therefore, no "real" borrowing from the private economy need take place at all. In the extreme case in which the government exclusively purchases the services of previously unemployed resources, the real costs of such resources need not be other than zero. The public expenditure undertaken may involve zero real cost.
This situation which allows resources to be secured at zero real cost may be achieved, however, only if the government is an efficient purchaser. Since no private person gives up any current disposition over resources, and no real transfer takes place, the government need make no promise to private people to pay them real income in future periods. The actual private disposition over current resources need not be reduced. To provide the monetary means for its actual purchases of resource services, the government may create money quite readily. The efficient means of purchasing the services of unemployed resources is through inflation of the currency. Individuals owning the resources concerned are made better off, and no one in the economy is made worse off.
In this currency inflation-public expenditure operation no public debt is created. Neither the present generation nor future generations undergo any sacrifice of utility or of real goods and services in order to secure the benefits accruing from the public projects constructed with previously unemployed resources.
Governments are not, however, always rational or efficient purchasers. Experience, especially that taken from the depression period of the 1930's, provides ample evidence that governments will refrain from creating money during depressions, and that they will, instead, create interest-bearing debt. And, even now, there is perhaps only a rather naïve hope that governments in the future would be any wiser and would improve substantially on this record.
If governments do choose deliberately to be inefficient purchasers, that is, if they sell interest-bearing securities instead of issuing noninterest-bearing money, no longer can the unemployed resources be put to work at a zero cost. The zero real cost is a minimum which can only be reached by rational policy. If interest-bearing debt is issued, some unnecessary real cost is introduced into the borrowing side of the fiscal operation, and even though the resources will otherwise be idle, the transference of these resources to the government must involve some future sacrifice of individual utilities.*88
If, in its sale of securities, the government withdraws funds from either consumption or investment spending, the subsequent utilization of these funds does nothing toward increasing the level of total employment. Debt creation under such circumstances is in all respects similar to the classical model already discussed. When debt creation during depression is discussed separately as a distinct fiscal operation, presumably the funds are drawn from idle balances and, therefore, when these are expended, idle resources are put to work. But if the securities sold are interest bearing, regardless of the amount or degree of unemployment existing in the economy, some real-debt burden is created. This debt burden is the real cost of putting the idle resources to work. And this real cost will be shifted forward to future taxpayers. If the government borrows to finance the building of a post office in a deep depression, the real cost of the post office is represented by the goods and services which taxpayers in the future could purchase if they were not obligated to finance the interest payments.
The interest-bearing, government securities are purchased by individuals and financial institutions from idle funds, and the real cost or primary debt burden is entirely unnecessary and could, and should, be avoided by direct money issue. At this point we may raise the question: Why is it necessary that the government pay interest on bonds sold to individuals or institutions who purchase these bonds from idle funds? What do these individuals or institutions give up in exchange for the claims to future incomes represented by the bonds? This question is answered when it is recognized that money as an asset yields some utility income to its holder in an uncertain world. Some sacrifice of current liquidity is involved in the exchange of idle cash for a government bond. The claim on future income may be considered a payment for this sacrifice of liquidity. But this payment is unnecessary, because the government does not add to its own liquidity in the process. It secures no additional liquidity from private people in the exchange. The government is always infinitely liquid so long as it possesses money-creating powers.
If the individuals and institutions do not purchase the bonds, they retain a more liquid cash position which, at least in their calculations made in the initial period, can be expected to yield them a utility income over future time periods. This alternative utility yield is comparable to the interest yield secured from the bonds. Bond purchasers are better off in future periods only insofar as these two expected yields differ.
The taxpayer, on the other hand, finds his utility in future periods reduced by the necessity of having to finance the interest payments. And it is reduced by the utility which he could have secured from the usage of the full amount of the transfer. The burden of debt, which is the real cost of the expenditure, rests on the future taxpayer just as in the classical model.
It seems clear that the sort of internal public debt discussed here is the model which many of the advocates of the new orthodoxy have used in constructing the debt theory which is currently espoused in our textbooks. Internal debt, considered as a substitute for money creation, rather than internal debt as a means for facilitating the real transfer of resources from the private to the public economy, is the model for the post-Keynesian construction. The failure of the new orthodoxy to see that its model of debt did not really overthrow the traditional theory lies in the confusion and the intermingling of the effects of securing the funds and the effects of spending them, the second methodological error mentioned. When unemployment does exist, the return to be expected from public expenditure which puts the unemployed resources to work is represented by the whole amount of the additional real income which these resources produce. On the payments side, the real costs which are present are unnecessary, and, if incurred at all, represent exchanges of claims on future income for current sacrifices of liquidity. Hence, the two sides of the fiscal operation embody quite different orders of magnitude, something which is not normally present in the classical model. In the latter, the debt-expenditure decision may be genuinely marginal; that is, the benefits and the real costs may be in roughly comparable dimensions. In the unemployment model, the benefits overwhelm the real costs in any absolute sense. This discrepancy explains the relative overemphasis on the benefits from debt-financed expenditure, but it does not remove from existence the real cost, and hence a real primary debt burden, albeit wholly an unnecessary one.
If the government is wise in such situations, it will issue currency outright. If it does this, the rate of return on its outlay will be infinite in real terms, or at least potentially so in the extreme case in which only unemployed resources are put to work, and no effects on prices take place. (This is the familiar assumption that there exist no bottlenecks in the economy to cause any price inflation until a full employment position is attained.) The cost of creating money is zero or nominal and the return is measured in the full amount of additional income generated. However, even if the government chooses to ignore this most efficient manner of financing expenditure during the depressions, and instead decides to issue interest-bearing debt as a means of securing idle cash from private hoards and from excess reserves of banks, the rate of return over its outlay is still likely to be tremendous.
This situation is not likely to be found often in the case of the private individual borrower, and, in this respect, public debt in depression is different from private debt. But the fundamental similarity is not modified since the private borrower could conceptually find himself in such a situation. The government in a depression is analogous to the individual who has secretly located a rich gold mine, and who needs only a shovel to begin work. The rate of return over the outlay on the shovel promises to be extremely high, and if the individual borrows to finance the purchase, he will certainly find himself far better off than had he remained idle. But this does not mean that he should consider the shovel to be free. To make the analogy more complete, however, we should have to allow the individual to have an old shovel around which he overlooks, one which would have done the job equally well. No borrowing is really required so his rate of return over cost could be infinite (labor cost neglected). It is not that the basic analogy between the public and the private economy becomes false during depression situations, but rather that the public economy is faced with a rare opportunity to make a tremendous return.
Similar conclusions may be reached in regard to the contrast made by the new orthodoxy between external and internal debt. When depression conditions prevail, the internal debt form is always indicated by purely market criteria. If fully rational the government will print money, that is, "borrow" at a zero rate of interest. If less than fully rational, it may choose to use internal debt as a substitute for money creation, paying a nominal return to reward people for the sacrifice of current liquidity. Assume, however, that liquidity preferences are extremely high, and that internal debt creation, even though the funds are to be drawn solely from private hoards, involves a high rate of interest, higher than the external borrowing rate. In this case the external rather than the internal loan is indicated, provided that the same task can be accomplished with each. The purpose of the borrowing must be that of securing funds which, when used, will bring idle resources into production. This suggests that funds must be in units of domestic currency and that they must be expended in the domestic economy. In many cases, the external borrowing operation may not make this possible. In these, the rate on external relative to internal loans is not relevant. The internal loan exists as the only possible alternative to money creation. The external borrowing operation may necessarily be real in the sense that the borrowing economy secures the current usage of resources drawn from outside its geographical limits.
If, on the other hand, external borrowing allows a country to secure funds of domestic currency, or of an international currency convertible readily into domestic currency, the external loan will be the efficient form if the loan rate is lower than that required for the internal loan. Of course, it must be kept in mind that neither loan form is efficient when compared to money creation.
It may be concluded that the difference between the external and the internal loan is not relevant so long as they both apply to the same task. But if the one involves real borrowing of resources from abroad while the other implies merely the putting to work of previously unemployed resources, no comparison is possible. We are talking about two kinds of fiscal operation here, not about two kinds of loans.
From the above discussion it is not difficult to understand why and how the basic propositions of the new orthodoxy were developed. The difference in the alternatives faced by government during periods of deep depression and those normally faced by private people appeared to suggest a fundamental distinction between public and private debt, between internal and external debt, and between the possibility of shifting the burden in the one case and the other. As we have shown, these fundamental distinctions do not exist.
The basic confusion involves a comparison of public debt, as a form of or substitute for money creation, with private debt which is normally real in the sense that the borrowing operation does require the withdrawal of economic resources from current usage by some other economic units, within or without the economy.*89 It does not seem likely that this confusion would have developed if governments in the Great Depression had adopted rational monetary policy and created money in the direct manner. Instead they chose to sell securities internally to finance deficits, at low or nominal rates of interest. Some of these securities were drawn from the excess reserves of the banking system, some from private hoards. In either case, much of the depression-born debt was equivalent in effect to new money creation. But the operation appeared to be a debt operation, and was discussed as such, since debt instruments were transferred to private people, instruments which represented an obligation of the government to pay an interest when due.
The funds so raised did not, when used, cause the government necessarily to bid away resources from private employments. Some resources previously unemployed were put to work. Debt issue when coupled with such expenditure appeared always to be desirable, something which could not apparently be claimed for private debt. In this way, the whole new approach to the public debt was built up, an approach which is applicable only to the narrowest of irrelevant comparisons. The advocates of the new orthodoxy were really comparing the issue of public debt, combined with the expenditure of the funds raised, with the policy of no action on the part of government. In this comparison, the first alternative is clearly to be preferred during periods of deep depression. No resources are given up currently, and even though they must pay interest in the future, that is, bear the primary debt burden, taxpayers are still likely to be much better off as a result of the combined borrowing-expenditure operation.
On the other hand, when they considered private debt, the advocates of the new orthodoxy were thinking of a real borrowing operation in isolation from the expenditure. Only this can explain the idea that private debt is necessarily burdensome, and that the carrying of private debt is oppressive. Had the expenditure side of the private debt been added into the picture, as it was with the public debt, the oppression would have been seen to arise only if the original combined borrowing-expenditure operation was unwise. A wisely chosen, debt-financed private undertaking may, similar to the public debt in deep depression, create much more income than it costs over future time periods. And, when compared with the no-debt alternative, the combined operation may actually be highly beneficial to the debtor. The new orthodoxy of the public debt and its central propositions were thus built up on the basis of an asymmetrical analysis of public and private debt. If we limit the comparison to the borrowing side, as this essay has done in preceding chapters, there is no distinction to be drawn. But if we include both the borrowing and the expenditure side, there is no distinction either. And the whole question of whether or not future generations of taxpayers in the public debt case and private debtors in the other are better or worse off depends on a comparison of the productivity of the project financed with its real cost.
Notes for this chapter
The oversight of this rather obvious point in the discussion of debt theory may be based on a slightly distorted meaning which is given to real cost in much careless textbook discussion. The opportunity or alternative cost of a good is often defined as those goods and services which could have been produced if the resources employed had been used in a different fashion. From this simple definition, the real cost of depression-built post offices, for example, has often been held to be zero, or nearly so, because the resources employed were not diverted from alternative employments, actually or potentially. This line of reasoning overlooks the all-important point that the absence of alternative opportunities for resource employment sets a minimum of zero to the real cost of a good or service. There is nothing to insure that this minimum is always attained. It can be attained only with fully efficient purchasing. Individual analogies are easy to draw here. If I choose deliberately to pay a price for a good in excess of that set by the market forces, the real costs to me are in excess of the goods and services which the resources could have produced in alternative employments. The alternative or opportunity cost defined in the usual manner represents the minimum real cost, not the actual real cost.
During periods of depression, the creation of private debt may also serve as a reasonably good substitute for money creation. Insofar as money borrowed to put resources to work comes from hoards, no real transfer of resources takes place. Cf. Emerson P. Schmidt, "Private versus Public Debt," American Economic Review, XXXIII (March, 1943), 119-21. When liquidity positions are considered, the issue of any debt, public or private, must serve as a partial substitute for money creation. That is to say, any debt, even if real, will introduce additional "moneyness" into the economy, at least to some degree. This is obvious when public debt is considered; it is less obvious for private debt. But, as McKean has shown, the illiquidity or "non-moneyness" of the debt obligation for the debtor cannot fully offset the liquidity or "moneyness" of the debt claim for the creditor. (See R. N. McKean, "Liquidity and a National Balance Sheet," Journal of Political Economy, LVII , 506-22. Reprinted in Readings in Monetary Theory [New York, 1951], pp. 63-88.)
End of Notes
Return to top