Public Principles of Public Debt: A Defense and Restatement
By James M. Buchanan
- Ch. 1, The Economists and Vulgar Opinion
- Ch. 2, The New Orthodoxy
- Ch. 3, The Methodology of Debt Theory
- Ch. 4, Concerning Future Generations
- Ch. 5, The Analogy: True or False
- Ch. 6, Internal and External Public Loans
- Ch. 7, Consumption Spending, the Rate of Interest, Relative and Absolute Prices
- Ch. 8, A Review of Pre-Keynesian Debt Theory
- Ch. 9, Public Debt and Depression
- Ch. 10, War Borrowing
- Ch. 11, Public Debt and Inflation
- Ch. 12, When Should Government Borrow
- Ch. 13, Should Public Debt Be Retired
- Ch. 14, Debt Retirement and Economic Stabilization
- Appendix, A Suggested Conceptual Revaluation of the National Debt
Concerning Future Generations
The new orthodoxy of the public debt is based upon three propositions. If these propositions can be shown to be false, the modern conception of public debt must be radically revised. If these propositions can be shown to be true in reverse, the conception must be completely discarded. I shall attempt, in this and the following chapters, to accomplish this reversal. I shall try to prove that, in the most general case:
1. The primary real burden of a public debt is shifted to future generations.
2. The analogy between public debt and private debt is fundamentally correct.
3. The external debt and the internal debt are fundamentally equivalent.
The Analytical Framework
Initially I shall discuss public debt in what may be called its “classical” form. The existence of substantially full employment of resources is assumed. Secondly, I shall assume that the debt is to be created for real purposes, not to prevent or to promote inflation. The government desires to secure command over a larger share of economic resources in order to put such resources to use. This assumption suggests that debt instruments are purchased through a transfer of existing monetary units to the government. Thirdly, I shall assume that the public expenditure in question is of a reasonably limited size relative to both the total income and investment of the community, and, consequently, that the effects of the sale of government securities on the interest rate and the price structure are negligible. Fourthly, I shall assume that the funds used to purchase government securities are drawn wholly from private capital formation. I shall also assume that competitive conditions prevail throughout the economy. Finally, I shall make no specific assumption concerning the purpose of the expenditure financed. I shall show that this purpose is not relevant to the problem at this stage of the analysis.
These assumptions may appear at first glance to be unduly severe. They will, of course, be relaxed at later stages in the argument, but it is perhaps worthwhile to point out that these assumptions are largely applicable to the debt problem as it has been, and is being, faced in the 1950’s. They apply, by and large, to the highway financing proposals advanced by the Clay Committee in early 1955. They apply, even more fully, to the debt problems facing state and local units of government, which alone borrowed more than five and one-half billions of dollars in 1956.
By contrast, the assumptions do not accurately reflect the conditions under which the greater part of currently outstanding public debt has been created. This qualification may appear to reduce somewhat the generality of the conclusions reached. Such is, however, not the case. The initial restriction of the analysis to public debt in the “classical” form allows the characteristic features of real debt to be examined; other forms of public debt are less “pure,” and it is appropriate that they be introduced only at a second stage of analysis. When this is done in later chapters, the conclusions reached from the initial analysis will be found generally applicable, and the apparently contradictory conclusions stemming from the new orthodoxy will be explained on the basis of the methodological confusion discussed in Chapter 3.
The first of the three basic propositions will now be examined in the light of the specific assumptions stated above.
The Shifting of the Burden to Future Generations
Before we can proceed to discuss the question of the possible shifting of the debt burden, we must first define “future generations.” I shall define a “future generation” as any set of individuals living in any time period following that in which the debt is created. The actual length of the time periods may be arbitrarily designated, and the analysis may be conducted in terms of weeks, months, years, decades, or centuries. The length of the period per se is not relevant. If we choose an ordinary accounting period of one year and if we further call the year in which the borrowing operation takes place, t0, then individuals living in any one of the years, t1, t2, t3, … tn, are defined as living in future “generations.” An individual living in the year, t0, will normally be living in the year, t1, but he is a different individual in the two time periods, and, for our purposes, he may be considered as such. In other words, I shall not be concerned as to whether a public debt burden is transferred to our children or grandchildren as such. I shall be concerned with whether or not the debt burden can be postponed. The real question involves the possible shiftability or nonshiftability of the debt burden in time, not among “future generations” in the literal sense. Since, however, the “future generation” terminology has been used widely in the various discussions of the subject, I shall continue to employ it, although the particular definition here given should be kept in mind.
What, specifically, do the advocates of the new approach mean when they suggest that none of the primary real burden of the public debt can be shifted to future generations? Perhaps the best clue is provided in a statement from Brownlee and Allen: “The public project is paid for while it is being constructed in the sense that other alternative uses for these resources must be sacrificed during this period.”*39 (Italics mine.) The resources which are to be employed by the government must be withdrawn from private employments during the period, t0, not during any subsequent period.
This last statement is obviously true, but the error lies in a misunderstanding of precisely what is implied. The mere shifting of resources from private to public employment does not carry with it any implication of sacrifice or payment. If the shift takes place through the voluntary actions of private people, it is meaningless to speak of any sacrifice having taken place. An elemental recognition of the mutuality of advantage from trade is sufficient to show this. If an individual freely chooses to purchase a government bond, he is, presumably, moving to a preferred position on his utility surface by so doing. He has improved, not worsened, his lot by the transaction. This must be true for each bond purchaser, the only individual who actually gives up a current command over economic resources. Other individuals in the economy are presumably unaffected, leaving aside for the moment the effects of the public spending. Therefore, it is impossible to add up a series of zeroes and/or positive values and arrive at a negative total. The economy, considered as the sum of the individual economic units within it, undergoes no sacrifice or burden when debt is created.
This simple point has surely been obvious to everyone. If so, in what sense has the idea of burden been normally employed? The answer might run as follows: To be sure no single individual undergoes any sacrifice of utility in the public borrowing process because he subscribes to a voluntary loan. But in terms of the whole economy, that is, in a macro-economic model, the resources are withdrawn from private employment in the period of debt creation, not at some subsequent time. Therefore, if this sort of model is to be used, the economy must be treated as a unit, and we may speak of a sacrifice of resources during the initial time period. In the macro-economic model we are not concerned with individual utilities, but with macro-economic variables.
It is perhaps not surprising to find this essentially organic conception of the economy or the state incorporated in the debt theory of Adolf Wagner,*40 but it is rather strange that it could have found its way so readily into the fiscal theory of those countries presumably embodying democratic governmental institutions and whose social philosophy lies in the individualistic and utilitarian tradition. The explanation arises, of course, out of the almost complete absence of political sophistication on the part of those scholars who have been concerned with fiscal problems. With rare exceptions, no attention at all has been given to the political structure and to the possibility of inconsistency between the policy implications of fiscal analysis and the political forms existent. Thus we find that, in explicit works of political theory, English-language scholars have consistently eschewed the image of the monolithic and organic state. At the same time, however, scholars working in fiscal analysis have developed constructions which become meaningful only upon some acceptance of an organic conception of the social group.*41
In an individualistic society which governs itself through the use of democratic political forms, the idea of the “group” or the “whole” as a sentient being is contrary to the fundamental principle of social organization. The individual or the family is, and must be, the basic philosophical entity in this society. This being true, it is misleading to speak of group sacrifice or burden or payment or benefit unless such aggregates can be broken down into component parts which may be conceptually or actually imputed to the individual or family units in the group. This elemental and necessary step cannot be taken with respect to the primary real burden of the public debt. The fact that economic resources are given up when the public expenditure is made does not, in any way, demonstrate the existence of a sacrifice or burden on individual members of the social group.
The error which is made in attributing a sacrifice to the individual who purchases a security, be it publicly or privately issued, has time-honored status. One of its sources, for there must be several, may lie in the classical doctrine of pain cost. Nassau Senior is generally credited with having popularized, among economists, the notion of abstinence. This concept was introduced in order to provide some philosophical explanation and justification for profits or returns to capital investment. The individual, in abstaining from consuming current income, undergoes the pain of abstinence which is comparable to that suffered by the laborer. Abstinence makes the receipt of profits, in an ethical sense, equally legitimate with wages in the distributive system of the late classical economists.
Traces of this real- or pain-cost doctrine are still with us, notably in certain treatments of international trade theory, but neoclassical economic theory has, by and large, replaced this doctrine with the opportunity cost concept. Here the works of Wicksteed and Knight generally and of Ohlin in particular must be noted. In the neoclassical view, resources command a price not due to any pain suffered by their owners, but because these resources are able to produce alternative goods and services. Resources may be used in more than one line of endeavor. A price, that is, a payment to the resource owner, is necessary in order to secure the resource service. Its magnitude is determined by the marginal productivity of the resource in alternative uses.
This shift of emphasis from the real-cost to the opportunity-cost conception has profound implications, some of which have not yet been fully understood. The real-cost doctrine suggests, for example, that a man is paid because he works, while the opportunity-cost doctrine reverses this and suggests that a man works because he is paid. The emphasis is placed on the individual choice or decision, and the gain or benefit side of individual exchange is incorporated into the theory of market price. The classical economists did not clearly view the distributive share as a price and the distribution of real income as a pricing problem.*42 Neoclassical theory does interpret the distributive share as a price, and the factor market is subjected to standard supply and demand analysis. The mutuality of gain from trade becomes as real in this market as in any other.
It becomes irrelevant whether the individual undergoes “pain” as measured by some arbitrary calculus when he works. If he works voluntarily, he is revealing that his work, when coupled with its reward, enables him to move to a preferred position. The individual is in no sense considered to be paying for the output which he cooperates in producing, merely because his productive services enter into its production. I am not, in my capacity as a member of the faculty of the University of Virginia, paying for the education of young men merely because my time is spent in classroom instruction, time which I could spend alternatively in other productive pursuits. Clearly the only meaningful paying for is done by those parents, donors, and taxpayers, who purchase my services as a teacher. What I am paying for when I teach is the income which I earn and by means of this the real goods and services which I subsequently purchase. Only if a part of my income so earned is devoted to expenditure for education can I be considered to be paying for education.
All of this is only too obvious when carefully considered. It is a very elementary discussion of the wheel of income which every sophomore in economics learns, or should learn, on the first day of class. If sacrifice or payment is to be used to refer both to the producer and the final consumer of goods and services, we are double counting in the grossest of ways; we are paying double for each unit of real income. We are denying the existence of the circular flow of real income in an organized market economy.
It is not difficult to see, however, that this error is precisely equivalent to that committed by those who claim that the real payment or sacrifice of resources must be made by those living in the period of public debt creation. The purchaser of a government security does not sacrifice resources for the public project; that is, he does not pay for the project any more than I pay for the education of young men in Virginia. He pays for real income in some future time period; he exchanges current command over resources for future command over resources. No payment or sacrifice is involved in any direct sense. The public project is purchased, and paid for, by those individuals who will be forced to give up resources in the future just as those who give up resources to pay my salary at the University of Virginia pay for education. It is not the bond purchaser who sacrifices any real economic resources anywhere in the process. He makes a presumably favorable exchange by shifting the time shape of his income stream. This is not one bit different from the ordinary individual who presumably makes favorable exchanges by shifting the structure of his real asset pattern within a single unit of time.
All of this may be made quite clear by asking the simple question: Who suffers if the public borrowing is unwise and the public expenditure wasteful? Surely if we can isolate the group who will be worse off in this case we shall have located the bearers of the primary real burden of the debt. But clearly the bondholder as such is not concerned as to the use of his funds once he has received the bond in exchange. He is guaranteed his income in the future, assuming of course that the government will not default on its obligations or impose differentially high taxes upon him through currency inflation. The taxpayer in period t0 does not sacrifice anything since he has paid no tax for the wasteful project. The burden must rest, therefore, on the taxpayer in future time periods and on no one else. He now must reduce his real income to transfer funds to the bondholder, and he has no productive asset in the form of a public project to offset his genuine sacrifice. Thus, the taxpayer in future time periods, that is, the future generation, bears the full primary real burden of the public debt. If the debt is created for productive public expenditure, the benefits to the future taxpayer must, of course, be compared with the burden so that, on balance, he may suffer a net benefit or a net burden. But a normal procedure is to separate the two sides of the account and to oppose a burden against a benefit, and this future taxpayer is the only one to whom such a burden may be attributed.
Widespread intellectual errors are hard to trace to their source. We have indicated that the pain-cost doctrine may have been responsible for some of the confusion which has surrounded public debt theory. But there are other possible, and perhaps more likely, sources for the future burden error. One of the most important of these is the careless use of national income accounting which has grown up in the new economics. Attention is focused on the national or community balance sheet rather than on individual or family balance sheets. In relation to debt theory, this creates confusion when future time periods are taken into account. There is no net change in the aggregative totals which make up the national balance sheet because the group includes both bondholders and taxpayers. The debits match the credits, so no net burden in the primary sense is possible. “Future generations” cannot be forced to pay for the resources which have already been used in past periods.
This simple sort of reasoning makes two errors. First, the effect on the national balance sheet is operationally irrelevant. As pointed out above, the nation or community is not a sentient being, and decisions are not made in any superindividual or organic way. Individuals and families are the entities whose balance sheets must be examined if the effects on social decisions are to be determined. The presumed canceling out on the national balance sheet is important if, and only if, this is accompanied by a canceling out among the individual and family balance sheets.
A moment’s consideration will suggest that genuine canceling in the latter sense does not take place. The balance sheet of the bondholder will include an estimated present value for the bond, a value which is calculated on the certain expectation that the interest payments will be made and the bond amortized when due. These interest payments represent the “future” income which the bondholder or his forbears paid for by the sacrifice of resources in the initial period of debt creation. These payments are the quo part of his quid pro quo. They are presumably met out of tax revenues, and taxpayers give up command over the use of resources. This sacrifice of income has no direct quid pro quo implication; it is a sacrifice imposed compulsorily on the taxpayer by the decision makers living at some time in the past. To be sure, as pointed out above, if the public expenditure is “productive” and is rationally made, the taxpayer may be better off with the debt than without it. His share of the differential real income generated by the public project may exceed his share of the tax. But the productivity or unproductivity of the project is unimportant in itself. In either case, the taxpayer is the one who pays, who sacrifices real resources. He is the final “purchaser” of the public goods and services whether he is a party to the decision or not. His is the only sacrifice which is offset, if at all, by the income yielded by the public investment of resources made possible by the debt.
From this analysis it is easy to see that much of the recent discussion on the burden of transfer misses the point entirely. One senses as he reads the discussion, notably that of Ratchford, that underlying the argument for the existence of a transfer burden there is an implicit recognition that the primary real burden does fall on taxpayers of future generations.*43 But the transfer burden advocates were unable to escape the real sirens of the new orthodoxy, the national balance sheet and the false analogy. Their deep and correct conviction that all was not happy in the conceptual underpinnings of the newly rediscovered edifice properly led them to re-examine the theory; but they accepted entirely too much before they started. They gave away their case, and their efforts resulted in little more than a slight modification of the theory. They did little more than to force a new emphasis on the secondary burden of making transfers.
The transfer burden analysis suffers the same methodological shortcomings as the more general approach of the new orthodoxy. If public debt issue is analyzed in terms of the whole set of relevant alternatives, in this case notably that of taxation, the burden associated with the making of an interest transfer cannot fail to be viewed as a primary one akin to that which is imposed through taxation. The failure to consider the position of the individual bondholder under each of the alternative situations led the “transfer burden” analysts to accept the fundamental premise of the new orthodoxy, that interest payments do, in a differential sense, represent a net “transfer” among individuals within the economy.
The Ricardian Approach
We may now discuss the unique set of assumptions under which the primary real burden of the debt cannot be shifted forward in time. And surprisingly enough, this is not mentioned in the new orthodoxy at all. It is propounded by the classical economist, David Ricardo. Ricardo enunciated the proposition that the public loan and the extraordinary tax exert equivalent effects on the economy. His argument is as follows:
When, for the expenses of a year’s war, twenty millions are raised by means of a loan, it is the twenty millions which are withdrawn from the productive capital of the nation…. Government might at once have required the twenty millions in the shape of taxes; in which case it would not have been necessary to raise annual taxes to the amount of a million. This, however, would not have changed the nature of the transaction. An individual instead of being called upon to pay 100£ per annum, might have been obliged to pay 2000£ once and for all.*44
Under these Ricardian assumptions, the full burden of payment for the public project, whether it be a war or a royal ball, will be borne by the generation which lives at the time of the expenditure. But this is true only because Ricardo assumes that the creation of the debt, with its corresponding obligation to meet the service charges from future tax revenues, causes individuals to write down the present values of their future income streams. The tax reduces the individual’s current assets directly; both the gross and net income streams over future time periods are reduced, these being equivalent in this case. The loan does not affect the gross income stream, but it does impose a differential between this and the net income stream. Capitalized values will be figured on the basis of income streams net of tax. Therefore, present values of assets will be immediately reduced by the present value of the tax obligations created by the future service charges. Present values will be identical in the two cases.
This Ricardian reasoning is correct within the framework of his assumptions. But it should be noted first that this is not at all the reasoning of the new orthodoxy. Ricardo places the primary burden of the public loan on the taxpayer not the bond purchaser. The primary burden is placed on the taxpayer because he writes down the value of his capital assets in anticipation of his obligation to pay future taxes to service the debt.
This Ricardian proposition has been much discussed in the Italian works on fiscal theory. This Italian contribution will be treated in some detail in the Appendix to Chapter 8. It will be sufficient at this point to indicate in general terms the deficiencies in the Ricardian proposition. The major objection which has been raised to the proposition is that individuals do not fully discount future taxes. While full discounting may take place for those individuals who own income-earning assets, this reasoning cannot be extended to individuals who own no assets. For the individual owning capital, it is possible that he will write down the value of his assets and transmit them to his heirs at the reduced value. In this case the burden of the tax required to service the debt can be said to be borne entirely by the individual in the initial period. The necessity of transferring income to bondholders will not reduce the present value of expected utilities for future taxpayers because this would have already been discounted for in the past.
For the individual who owns no capital assets, however, this analysis cannot fully apply. Since slavery is not an acceptable institution in the modern world, individuals are not treated as capital assets and traded in accordance with capitalized values. The individual human being as either a capital asset or a liability disappears at death, and his heirs inherit directly neither his asset characteristics nor his liabilities. For this reason, the individual who owns no capital assets will not fully capitalize the future tax burden involved in the interest charges. He will capitalize them, if at all, only within the limits imposed by his effective planning horizon, that is, only to the degree which he, individually, conceives that he will be a future taxpayer. And, since human life is short, much of the debt burden must remain uncapitalized. Therefore, even granting all of the other Ricardian assumptions, which are extremely restrictive, the burden must rest on “future generations,” at least to some degree.
Insofar as other individuals who do own capital assets do not plan to submit these intact to their heirs, that is, insofar as family relationships do not make individuals act as if they will live forever, the Ricardian proposition is further weakened. And when the possibility of individual irrationality in discounting future tax payments is introduced, as Ricardo himself recognized, the location of the debt burden on future taxpayers is even more clear.
It must be concluded, therefore, that this Ricardian analysis, as amended, introduces only a slight modification of the conclusions earlier attained. The primary real burden of a public debt is borne by members of the current generation only insofar as they correctly anticipate their own or their heirs’ roles as future taxpayers, and take action to discount future tax payments into reductions of present capital values. Insofar as the time horizons of individuals are not infinite, that is, insofar as future individuals are considered to be separated conceptually from present individuals, there must be some shifting of the primary real burden to future generations.
It has been shown in this chapter that the primary real burden of a public debt does rest largely with future generations, and that the creation of debt does involve the shifting of the burden to individuals living in time periods subsequent to that of debt issue. This conclusion is diametrically opposed to the fundamental principle of the new orthodoxy which states that such a shifting or location of the primary real burden is impossible. We have examined the reasons for the widespread acceptance of the nonshiftability argument. We have isolated at least some of the roots of the fallacy. Among these are the pain-cost doctrine and the use of national rather than individual balance sheets.
The primary real burden of the debt, in the only sense in which this concept can be meaningful, must rest with future generations at least in large part. These are the individuals who suffer the consequences of wasteful government expenditure and who reap the benefits of useful government expenditure. All other parties to the debt transactions are acting in accordance with ordinary economic motivations.