Public Principles of Public Debt: A Defense and Restatement
Should Public Debt Be Retired?
As of July, 1957, the federal government debt of the United States amounted to $272.5 billion. State and local units of government owe an additional $55 billion (estimated). The annual cost of servicing the national debt for fiscal 1958 is estimated to be $7.8 billion; state and local debt service charges will require perhaps an additional $1.7 billion.
These are significant totals even in the most prosperous economy that the world has ever known. Public debt cannot be relegated to the position of an unimportant or minor aspect of the nation's fiscal structure. And even should the question discussed in the last chapter never arise, the manifold problems arising from managing the existing public debt are continuous and difficult.
This book will not include a specific discussion of debt management. The management of the national debt introduces issues and questions of complexity and detail which are not essential to the main argument of this book.*99 For my purposes it will be useful to consider only those broader aspects of debt management which appear to be modified in some way by the reversal of conceptual approach.
The central issue in debt management concerns the rate at which public debt is to be retired. Should an active and vigorous attempt be made to reduce the interest-bearing public debt? There are several parts to this question, and each must be considered separately.
First of all, it is necessary to distinguish debt retirement as such from the means of financing such retirement. There are two ways in which net debt retirement may be financed, taxation and money creation. Only the first of these is open to subordinate units of government. Both are open to the central government which possesses money-creating powers. In the full employment model, however, money creation has effects which are equivalent to taxation. This makes it useful to consider this model separately. In this chapter I shall assume that the economy is characterized by full employment and a stable price level. Only in the following chapter shall I introduce the stabilization aspects of debt retirement operations.
Public debt instruments take several forms. Therefore, it is also necessary to specify what part of the debt is to be retired. The retirement of debt held by the commercial banks and the central banks has different effects from the retirement of debt held by individuals or by nonbanking institutions. In this chapter I shall consider only the retirement of public debt held by these latter groups. The retirement of bank-held debt cannot be discussed apart from the stabilization aspects; these will be fully explored in the following chapter. I shall assume here that the question is one of retiring debt held by individuals, the retirement to be financed by taxation.
This is not to suggest that the tax-financed retirement of privately held debt will be without effects upon the level of absolute prices or upon employment. The deflationary effect of the taxation must be compared with the inflationary effect of the repayment. The combined operation may be inflationary or deflationary in net terms, the result being dependent upon the relative propensities of taxpayers and bondholders to spend out of ordinary income, and the degree to which debt repayment does not represent an ordinary income payment for bondholders. Assuming roughly equivalent behavior patterns for taxpayers and bondholders, the combined operation is probably deflationary due to the fact that debt instruments possess some "moneyness." But there is no a priori way in which this effect can be established. It seems best, therefore, that this sort of taxation-retirement operation be examined for effects other than those upon the absolute price level.
We may first review briefly the implications for this question to be drawn from the currently orthodox theory. This theory makes a rather sharp conceptual distinction between the federal debt and the state and local debt. The federal debt is an internal debt, while state and local debts are external to the borrowing jurisdiction in large part. For the latter, the new orthodoxy would suggest a reasonably rapid rate of repayment in order to reduce the annual interest drainage out of domestic tax revenues. For the federal government debt, on the other hand, the new orthodoxy would suggest that the rate of retirement is not very important. This debt, being internal, involves a mere transfer of purchasing power from the taxpayers to the bondholders. While some secondary burden of the interest transfer is acknowledged, this is conceded to be relatively minor.
The approach of this book views debt retirement in a different light. It has been demonstrated that there is really little difference between internal and external debt. The fact that interest payments on public debt represent transfers of purchasing power from taxpayers to bondholders becomes more important. The sacrifice imposed on the taxpayer is a very real cost which is measured in terms of the alternative goods and services which he could otherwise enjoy were the debt service eliminated.
The implication of this approach is not that debt retirement must be attempted during all normal periods of economic activity. In the full employment model which we are considering, the alternatives to an active policy of debt retirement are tax reduction and public expenditure expansion. In reaching a collective decision, an attempt must be made to compare the benefits to be secured from each of these three alternatives. The benefits from current retirement of public debt will be enjoyed in large part by future taxpayers. They will be relieved of the necessity of having to finance the required interest transfers. On the other hand, the benefits from a reduced rate of current taxation (or a differentially lower rate) or from an expansion of current public expenditure of the "consumption" variety will be enjoyed primarily by those individuals living at the time in which the retirement decision must be made. If a society faced with an outstanding public debt makes a decision to undertake substantial retirement, it is deliberately providing future benefits at the expense of current enjoyments. From the analysis of the collective decision-making process which was briefly developed in the preceding chapter it may be readily inferred that decisions are likely to be biased against retirement. Experience in democratic countries seems to bear this out. Rarely have nations deliberately retired significant portions of their debts. In effect, a public debt, once created, will tend to be permanent unless some repayment mechanism is built into the original decision and this commitment is honored. The necessity for this built-in device for repayment seems to have been vaguely recognized by those early advocates for the establishment of sinking funds, although they became somewhat overenthusiastic over the effects of compound interest.
But at this point we are interested in the normative question "Should debt be repaid?" Whether or not the people acting collectively will choose to repay need not delay us here.
There are essentially three separate arguments which may be employed. Two of these will be found to lend support to a policy which considers the debt permanent and attempts no significant retirement. The third argument strongly suggests that an active policy of debt retirement should be attempted.
The Equity Argument
The first argument is a highly tenuous one in that it introduces again the whole question of equity as among separate generations of taxpayers. As we have indicated earlier, the benefit principle of taxation is no more applicable for allocating the costs of public expenditure among separate generations than it is for allocating costs among separate individuals or groups within a single time span. In a purely negative sort of way, however, this principle may be useful. The taxpayers living in 1958, for example, would be required to give up a substantial sum if the national debt were to be liquidated, once-and-for-all, during this calendar year. The argument that such liquidation should take place does not seem acceptable on any grounds. A part of the federal debt, even in a pure or nonmonetized sense, was created to finance World War II. Although this was perhaps a highly "productive" expenditure of funds, the incremental real income stemming from "victory" may be said to be permanent. The "benefits" from this war expenditure have become a part of the national capital. The 1958 taxpayer receives no greater share of this benefit than will the 1975 or the 2000 taxpayer. And neither had any voice in choosing debt creation over wartime taxation. It does not seem unreasonable to suggest on equity grounds, which should command wide acceptance, that the portion of the "pure" national debt created during war periods should be considered permanent and that no real effort should be made to retire it.*100
A similar conclusion may be reached with regard to that portion of real debt which was created for purposes which involve no directly measurable public investment. Debt issue which was used to finance purely wasteful expenditure must fall in this category, as must debt which was issued unnecessarily due to the failure of government to secure funds efficiently through money creation. This, combined with the war debt considered above, makes up a good part of the national debt, and this argument indicates that no serious effort should be made toward retirement. The refunding of this type debt into consols bearing no fixed maturity date would seem a reasonable management decision.
An extension of this argument may be applied to public debts which are created to finance specific capital improvement projects having a limited life. Only a small portion of the federal debt falls under this heading, but the greatest part of state and local debt fits in this category. Debts created to finance such projects as highways, parks, schools, and river basin developments are a few of the many examples which may be mentioned. The effective life of any specific investment project is limited, that is, the project will yield real income only over a finite period of time. The equity argument would suggest that the practices followed by private business firms be adopted for public debt of this nature. Debt should be fully amortized within the period of the effective life of the capital asset. Especially as applied to state and local units of government, this argument has additional merit in that such practice tends to preserve the credit rating of the borrowing unit. Debt retirement policy cannot be divorced from the effects on the securities market, although, for the national government, the connection may be quite remote.
Diminishing Marginal Utility
The second argument that may be used in support of a no-retirement policy represents an application of the conception which provides the theoretical support for progressive taxation. The whole basis of the argument is very shaky, and the more sophisticated students of public finance have long since discarded it, but it is still found in many modern treatments. The normative proposition runs as follows: In allocating the tax burden among individuals, the principle of minimizing the aggregate sacrifice on the whole group should be employed; aggregate sacrifice may be minimized by taxing the rich since the marginal utility of income declines as more income is received.
As applied to debt retirement out of current taxation, this argument would suggest that, in a growing economy in which individual incomes are expected to increase over time, no retirement should take place. Future generations of taxpayers, enjoying a higher income than present taxpayers, can service the debt at a net sacrifice which discounts to a smaller present value than that which would be imposed upon present taxpayers with current redemption of debt.
In its strictest form, this argument requires not only that interpersonal comparisons of utility can be made within a single time period but also that some comparisons can be made between separate time periods. In spite of the fact that the impossibility of making such comparisons in any objective manner has been recognized for a long time, the argument cannot be discarded out of hand. It retains some value if the utilities introduced are conceived to be, not those of the individual taxpayers, but rather those of the individual taxpayers as they are interpreted by the individual as a decision maker. If the diminishing marginal utility of income is widely acknowledged, and if social policy based on the assumption of a diminishing marginal utility of income is deemed to be reasonable, the collective choice process may produce results equivalent to those which would have been forthcoming from the assumption of full measurability and comparability.
Although it was not posed in utility terms, the implications of the Domar contribution to debt theory are essentially equivalent to those implicit in the utility argument. Domar showed that the tax rates necessary to service a given public debt must diminish with rising national income. From this the inference is clear that the higher the income the lower the marginal utility of the fixed absolute transfer. He clearly suggests that debt should not be retired. The solution to the debt problem lies not in efforts toward retirement, but in "trying to find ways of achieving a growing national income."*101
Debt Retirement and Economic Growth
The two arguments discussed above tend to support a policy which calls for the servicing of the outstanding public debt of the national government but which makes no positive effort to retire debt. There is a third approach to the problem which leads to the opposite results.
If economic growth is widely accepted, as it seems to be, as a desirable attribute of a well-functioning economic system, public policies aimed at increasing the rate of growth may seem advisable. Implicit in any public action to stimulate growth is the assumption that the rate of growth produced by individual choices within the institutional complex of social, economic, and political forces is less than a "desired" or "optimum" rate. It is, of course, impossible to define an "optimum" growth rate, and it is almost equally difficult to defend the result of private decision making as providing any approach toward an optimum. In setting aside income for savings and in making investment decisions, individuals will tend to adjust their subjective rates of time preference to the rate of return on investment. But the individual schedules of time preference may tend to "underestimate" the value of future income, underestimate being defined in accordance with some indefinable criterion.
Some public action toward stimulating capital formation, an admittedly essential element in rapid economic growth, may be held to be desirable even though the impossibility of determining the "optimum" growth rate is fully acknowledged. There are, of course, many ways in which society could change its public policy in order to stimulate capital formation. Most of these need not concern us here.
One means which is often proposed is that of direct public investment. If this investment (for example, in highways or in river basin development) is debt financed, it will stimulate public capital formation largely at the expense of private capital formation. There will be some net stimulant to economic growth only if the rate of return on the public projects exceeds that which is sacrificed on the private projects which would otherwise have been carried forward. This clarity in conclusion is achieved by begging all the questions concerning the difficulty of defining economic growth in the first place.*102 To the extent that the proceeds from the sale of public debt instruments are drawn from consumption uses rather than private capital formation, the rate of growth is apparently increased. But it cannot be stated without qualification that a reduction in consumption does not also deter economic growth.
A second scheme is that of financing public investment projects directly from current tax revenues. This will clearly shift the cost to current taxpayers, and it will reduce both the current rate of private consumption as well as private capital formation. Since consumption will be reduced more than in the preceding case, this policy may provide a considerably greater stimulus to economic growth than the financing of similar investment through the medium of loans unless the incentive effects of taxation are overwhelming. There are limitations on any policy of direct public investment, however, in that those areas of economic activity suitable for public investment, and of the type which stimulate economic growth, may be severely restricted.
For a country faced with a large public debt held by private individuals or nonbanking institutions, there exists a third alternative which seems to be preferable in some respects to the two mentioned above. If the community desires collectively to provide a stimulus to more rapid growth, and private capital formation is believed to be important for growth, it can adopt an active policy of debt retirement, financing this retirement out of current tax revenues. This operation will, on the tax side, affect the rate of capital formation only to the same extent as the second alternative mentioned above. And, on the repayment side, a substantial stimulus to new capital formation will be provided. Individuals and institutions securing liquid funds in exchange for government securities will tend to channel a large share of these funds into the market for private securities. Security prices will go up, interest rates will fall, and investment demand will be increased. A relatively small proportion of the liquid funds returned to individuals and institutions in exchange for government securities will be devoted to spending on current consumption items.
This debt retirement method has the advantage of stimulating growth in private capital formation as opposed to public investment. The market can be allowed to serve its normal function in distributing the new investment funds among competing projects in such a way that some approach to an "efficient" utilization is achieved. This means of allocating investment funds does not exist for public investment projects, and competing uses for newly available funds must survive the test imposed by pressure-group political jockeying. There is little assurance that the public investment funds are distributed in accordance with any acceptable criterion of "efficiency."
The suggestion that debt retirement out of current tax revenues provides, in some respects, a superior means of stimulating rapid economic progress does not imply that direct public investment should not be undertaken at the same time. If there exist peculiarly public projects which are closely related to economic development, and which are characterized by underinvestment, direct public outlay is indicated. It has been claimed, for example, that the national investment in an interstate highway network will be highly productive and that such investment will greatly enhance the possibilities of growth in the United States. Any over-all policy program which is directed primarily toward promoting economic growth will probably include some such direct public investment financed preferably out of tax revenues, but also perhaps from new debt issue. The program should also include some debt retirement financed out of current tax revenues. Thus, returning to Domar's argument for a moment, we find that one of the best ways of achieving a growing national income lies in debt retirement.
Notes for this chapter
For a current attempt to construct an economic theory of debt management see Earl Rolph, "Principles of Debt Management," American Economic Review, XLVII (June, 1957), 302-20. For a general discussion of debt management in the immediate postwar period, see Charles C. Abbott, Management of the Federal Debt (New York, 1946).
As indicated in Chapter 10, a large part of nominal debt issued during World War II was not "pure" debt at all, but instead was disguised currency creation. Insofar as this is true, the burden of debt per se is relatively small and the question of retirement becomes relatively unimportant.
Evsey D. Domar, "The 'Burden of the Debt' and the National Income," American Economic Review, XXXIV (December, 1944). Reprinted in Readings in Fiscal Policy (Homewood, Ill., 1955), pp. 479-501, especially p. 500.
Cf. G. Warren Nutter, "On Measuring Economic Growth," Journal of Political Economy, LXV (February, 1957), 51-63.
End of Notes
Return to top