Public Finance in Democratic Process: Fiscal Institutions and Individual Choice
The Institution of Public Debt
When should governments borrow rather than tax? This is a classic question in the theory of public finance, along with those discussed in the two preceding chapters. Can the institutional-choice approach to fiscal systems be applied to this question?
We are concerned with public debt as a fiscal institution through which a collectivity may finance public goods and services and with the individual's evaluation of this institution. Governments borrow as an alternative to taxing, and it is appropriate to consider borrowing as an addition to the revenue-raising alternatives listed at the beginning of Chapter 16. Are there any conditions that may cause the utility-maximizing individual to select, at the moment of constitutional-institutional choice, the public loan over any of the tax alternatives? Recall the characteristics of the situation that we have presumed to confront the individual with at this moment of choice. He is not choosing between debt issue and taxation for the financing of a specific public good or service in a specific time period. The individual recognizes that the fiscal instrument to be chosen will remain in force over a whole series of time periods, and that it will be employed to raise funds for a stream of public goods and services, with the precise nature, the range, and the extent of these goods and services to be determined from period to period, and with the benefits from this stream of services wholly unpredictable at the moment when the revenue-raising institution is to be selected. It is in this situation that we ask the question: Will the individual find it desirable for the collectivity to resort to debt issue?
It will be helpful to employ the same device that was introduced in Chapter 15. Assume that a given individual is assigned a specific share in some aggregate community liability and that he is able to define this share in terms of a definite present value at the moment of constitutional-institutional choice. For simplicity, we shall say that this present-value liability is set at $1000. This liability must be recognized as such under any of the institutional alternatives that the individual confronts. Hence, the public debt instrument, as one of these alternatives, must be defined in such a way that the individual is required to meet the obligation over a time span that is within his own planning horizon. That is to say, public debt must be considered for periods of sufficiently short maturity to insure that the individual making the choice shall recognize that he must amortize his own share in the community liability during his own planning period. If this constraint is not imposed, there would be no way of making a present-value liability under debt equivalent to those under various tax institutions. Therefore, we shall simply postulate here that the debt will, if issued, be amortized over a period of, say, ten years. In this restricted model, the individual can, through his community's resort to public debt, postpone current payment for public services for a maximum of ten years. The question becomes: Are there conditions under which he would wish to select the institution which facilitates such a postponement? Will he prefer that the government issue bonds as a means of raising revenue to finance public services for each of the first nine years? Or, will he choose to have his government rely on one or several of the standard tax instruments?
As in some of the previous models, it will be helpful to assume initially that the individual knows with certainty the pattern of his income receipts and his private spending over the relevant time period. It is not necessary to specify any particular pattern of either of these streams. When we allow the public debt alternative to be considered, we reach quickly what appears to be an unorthodox or startling conclusion. Given such an opportunity as that posed, the rational individual will always choose that all public goods and services be financed through public debt issue. This result seems striking at first glance, and it seems to be so much at odds with accepted principles of fiscal practice that one searches for the fallacy that must be hidden somewhere.*94
There is, however, no such fallacy lurking in the underbrush, and within the limits of the model examined here the conclusion holds. Why should the individual select the debt alternative? He will do so because this alternative is the only one that allows him full freedom of choice in adjusting his income-spending pattern over time. Public debt, as an institution, effectively allows the individual to meet his assigned liability "optimally," and it is the only revenue-raising alternative that accomplishes this, given the operation of a capital market that requires private borrowers to pay something over the government borrowing rate. As the collectivity borrows to finance currently supplied public goods and services in each period, the individual is placed in the position of borrowing, one stage removed, at the government rate. In effect, through issuing debt, the government is borrowing for the individual. If, therefore, the individual's pattern of net income receipts or spending over time is such that he desires to postpone meeting his fiscal obligation, the public debt alternative enables him to do this at no net cost. On the other hand, if his pattern of income-spending flows is such that he chooses to discharge his obligation early during the time sequence, he can always do so by purchasing government securities and holding these until the time of debt retirement-taxation, when his accumulated assets will just offset his accumulated tax obligations. In effect, the public debt allows the individual both to borrow and to lend at government rates, and hence to remove any temporal distortion from his spending pattern.*95
The analysis suggests that the isolated individual should rationally select public debt as the means for financing public goods and services. Why has this alternative not commanded more respect in the institutional structure of real-world fiscal systems? The underlying assumptions of the model require more careful consideration.
The individual's share in the aggregate community liability has been assumed to be preassigned in some present-value sense, at least insofar as this informs his own choices. But can such a share really be assigned in advance? The difficulties that arise here are not the same as those we have discussed previously in the analysis of progressive income taxation. The difficulties here stem from what we may call the "contingent liability" that public debt must embody under normal political circumstances.
Let us assume, as before, that, provisionally, some share in an aggregate community liability has been assigned to the reference person, and that similar shares have been parceled out to all members of the political group. Next, assume that no distributional problems explicitly arise here; the individual proceeds as if he will fully meet his own assigned share over the period in his own optimally selected manner. On the basis of some such calculation as that outlined above, he opts for the debt alternative; others in the group agree, and all public services are initially financed by public loans. The reference person then carries out his plans as projected, meeting his fiscal obligation as these plans dictate.
Now suppose that the final accounting period arrives; all issues of debt must be retired. The individual in question has accumulated, through his purchases of bonds over the period, sufficient assets to meet precisely the share of the liability during the final period that his plans dictated. All seems well; he seems to have chosen the ideal fiscal arrangement.
All is not well; and herein lies the rub. Suppose that a second person, Mr. B, likewise made optimal spending plans when the time sequence commenced and the fiscal alternative was selected. However, suppose that B has failed, over time, to live up to such plans as he had initially laid down. The final period arrives, and he has not accumulated sufficient assets to offset the meeting of his fiscal obligation. He simply cannot "pay off debt" or "pay taxes" in this final period and discharge his assigned multiperiod liability. This failure of B to live up to his rationally projected plans need not bother our first person, A, except to the extent that he understands that B's plight imposes a clear contingent liability on him. The funds have all been spent in the separate periods in financing the public goods and services. The aggregate liability for the whole collectivity must be paid, assuming that the community does not choose to default on its loan. B has, however, behaved either irrationally or irresponsibly over the period and he cannot meet the share that he implicitly agreed to meet. Consequently, it falls to the remaining members of the political group to bear the liability that was initially assigned to B. Others will find themselves paying for B's profligacy or deceit.
Note that A himself, the individual whose calculus of choice we are considering, will behave here precisely in accordance with his plans. He will, nonetheless, be unduly burdened at the end of the period to the extent that B's behavior runs contrary to B's projected plans at the start of the time sequence. The reference person, A, will tend to recognize this contingent liability aspect that the public debt instrument may embody. When he does so, he will tend to reject the debt alternative, and to select instead a tax institution, despite the acknowledged superiority of debt in terms of efficiency criteria in an isolated individual income-spending pattern.
No irrationality has been introduced in the analysis here. The reference individual need not fear for his own ability to meet targets that he lays down at the time of institutional choice. He will tend to reject the generalized usage of public credit not because he fears that he cannot live up to the model of behavior that he sets himself, but, instead, because he fears that some among his fellow citizens may fail to live up to their own targets of behavior. The acceptability of public debt requires, then, that the individual not only predict his own rational behavior, but, also, that he can predict with reasonable certainty that all other members of the group, or at least a sufficient portion of them, will likewise behave rationally and responsibly. This requirement becomes extremely restrictive and seems likely, in most cases, to rule out general approval of the debt alternative.
Why do these same fears concerning the rational and responsible behavior of other persons not arise in the considerations among the various tax alternatives? The analysis of the preceding chapters has shown that the efficiency advantages of both progressive income taxation and specific commodity taxation may stem from the fact that these institutions allow the individual to shift his fiscal liability through time in such a way as to reduce temporal distortions in his spending pattern. Why will not irrational or irresponsible behavior on the part of others than himself here too affect the individual's own liability?
The essential difference between these tax institutions and that of public debt stems from the fact that, under any tax institution that allows for fluctuations in individual liability over time, separate persons in the group tend to offset each other. Periods when one person's income is relatively low, or private spending relatively high, may be periods when another person's income is relatively high, or private spending relatively low. If the individual assumes that over-all income and spending in the economy will remain roughly stable or rise steadily, this result is assured. Public debt, by contrast, involves no such offsetting through the fiscal structure. All public goods and services are financed by debt in the initial periods (in the general model considered here). Fiscal bills pile up; no one is required currently to pay for the stream of public services. Current-period adjustments, if they occur, must take place within the private accounting systems of individual citizens. The fact that rationally behaving individuals may be acting in accordance with optimal plans is not externally revealed to the observer, nor is this behavior required in any way.
The reference individual may, of course, also harbor some doubts about his own rationality in following out some predetermined plan of spending-saving. To the extent that he does so, he will reject the debt alternative, quite apart from the contingent liability effects here emphasized.
The above analysis demonstrates that even if the individual knows with certainty his income prospects and his private spending plans over the relevant time span, and even if he is confident that he can carry out some predetermined plan of behavior, he may still reject the public debt as his most preferred general financing institution because he cannot predict that others will behave responsibly and rationally. As we have done in previous chapters, the unreal assumptions as to certain prospects must now be relaxed. Assume now that the reference individual's income prospects and/or private spending needs are uncertain and subject to some fluctuations over time. How will this change the conclusions about his attitudes toward the debt institution?
It is best to examine the individual's choice calculus on the assumption that total income in the whole community remains uniform over time or grows at some predictable rate. This allows us to concentrate on fluctuations in individual income receipts or in private spending needs independently of over-all aggregate fluctuations. As suggested earlier, with uncertainty as to receipts or outlays over time, the very notion of some optimally planned pattern of spending through time scarcely exists. The individual's decisions will be informed largely by current or in-period comparisons of income receipts and outlays. He should be able, however, to distinguish among periods of relative affluence and relative penury, not necessarily in advance but as events materialize. In the former, he will tend to put aside some income as savings to protect his economic position contingently over possibly lean periods. In the latter, he will tend to resort to the loan market, either borrowing from himself out of accumulated savings or externally from others. Previous chapters have demonstrated how the progressive income tax on the one hand and specific commodity taxation on the other may allow the fiscal structure to facilitate the individual's temporal adjustments. The question here is whether or not public debt can accomplish similar purposes?
The answer is negative. Public debt does not provide a means of bunching fiscal liability during periods of relative affluence comparable with the other two institutions discussed. For the whole group, public debt allows for a postponing of fiscal liability through time, but in the final accounting period, when debt must be retired, some taxpayers will be affluent, some will be penurious. Not knowing to which group he may belong, our reference individual will rationally reject the debt alternative as the general financing instrument. Public debt allows such a bunching in time only in the certainty model, and only in a highly restricted form of this.
In the real world some mixture of the certainty and the uncertainty models normally is descriptive. The individual can make some reasonable predictions as to his income prospects, and he can project within limits his needs for private spending over time. There is some sense in his attempts to frame optimal saving-spending plans. The important barrier that prevents his selecting resort to public credit as the general financing device lies in the contingent liability aspect discussed with respect to the certainty model above.
Public Debt and Individual Planning Horizons
In all of the models discussed to this point, the individual has been assumed to evaluate public debt as a means of financing general public goods and services. As compared with its alternatives, debt issue has been assumed to impose comparable liabilities, computed on some present-value basis, on the individual who is making the choice among the fiscal institutions. If human beings should live eternal lives, no problem need arise. But since they do not, it was necessary to postulate that debt must be issued in such a way that amortization occurs within the planning horizon, the life span, of the individual decision-maker. If this restriction is not placed on the models, there is no way in which public debt can be made genuinely comparable, in our terms of reference, with tax alternatives except through some quite arbitrary assumptions about human behavior. If individuals, despite the limitations on human life, treat their heirs as lineal extensions of their own lives, which was the assumption always made by Ricardo, no problem arises. In this case, individuals act as if they live forever. But individuals may not behave in such a fashion, and if they do not, public debt, which allows them to postpone fiscal liability, may provide a means of redistributing the net fiscal load intertemporally. If, at the time of constitutional-institutional choice, the individual considers public debt as a means of shifting the final fiscal liability forward in time to "future generations," he will, of course, tend to select this instrument on the basis of utility-maximizing considerations.*96
The rejection of the debt alternative in this limited-time-horizon situation must be based on the individual's acceptance of some ethical principle of intergeneration equity. If he makes plans on the basis of a limited time horizon and does not fully incorporate the interests of his descendents in his own, the individual will tend to select debt as the means of financing public goods and services unless he is deterred by some such ethical norm. Of course, if general acceptance of debt issue should become widespread, reflecting an absence of the effectiveness of this norm, the likelihood that future generations would, in fact, default on inherited debt obligations would quickly become an economic deterrent to this institution. For these and other reasons, it is appropriate that the analysis here be restricted to the model where debt is amortized within the planning horizons of the decision-makers.
The Public Debt Illusion and Its Converse
To this point the analysis has been limited to an evaluation of public debt as a general source for raising governmental revenues. It has been argued that the individual's probable rejection of this alternative stems, at least in large part, from his distrust of fellow taxpayers' ability or willingness to carry out optimal spending-saving plans. One element of this mistrust may arise out of the recognition that public debt may generate a fiscal illusion. Although fiscal liabilities are created at the moment that debt is issued, individuals may not act as if such liabilities exist. They may not fully capitalize the future taxes that the debt must embody, in service and amortization charges, and if they do not, they will not behave rationally in making plans to discharge their own shares in such aggregate liability.
The debt illusion has its converse, however, and when this is also recognized, public debt again assumes a limited but legitimate place in the acceptable array of fiscal instruments. To this point, as noted, revenue-raising alternatives, including debt, have been considered a means of raising general revenues for the financing of all public goods and services. Although the assumption was not explicitly made, the results derived are wholly appropriate only if the benefits from the provision of public goods are concentrated during the periods when the public outlays are actually made. This does not characterize all public outlay; some takes the form of capital investment which yields benefits over a whole series of time periods.
Let us now examine this sort of public outlay independently. Assume that the individual is faced with selecting the appropriate fiscal instrument or institution for financing only quasi-permanent public goods, the benefits from which will be fully realized only over the long run. As in the more general model, we assume that the individual does not know precisely the pattern of these capital projects, and he has no way of predicting whether or not he will personally benefit from such projects in particular periods of time. His task is that of selecting the fiscal means of financing public capital projects, and these only, with the actual decisions on the form and extent of these projects to be made in subsequent periods.
Here the choosing person may recognize that the temporal distortion between the receipt of benefits from public capital projects and the impact of the payment institution may tend to bias in-period decisions against such outlays under the standard taxing instruments. In other words, for projects that involve benefits which accrue over time, there may exist some "asset illusion." The individual may not fully capitalize the future benefits that such quasi-permanent outlays will yield. If he does not, he will not make "proper" decisions concerning the amount of taxes to be levied or the "proper" allocation of funds within a limited revenue budget. Budgetary decisions will tend to be biased in favor of short-term and against long-term public projects.
If the individual, at the level of institutional choice, recognizes that this sort of illusion is likely to occur, he may prefer that public debt be authorized as the revenue-raising device for such projects. Here, the individual who makes in-period operational budgetary choices may suffer both a public debt illusion and an asset illusion. He may fail to capitalize both the liability that the debt side embodies and the benefit stream that the asset embodies. But these two illusions become offsetting here, and the individual may predict that more rational in-period budgetary choices will emerge under such a structure than under one that is limited to tax financing for all outlays.
Note also that if it is limited to financing only capital projects, the public debt alternative need not involve the contingent liability element to the same extent as the more general model. Operating under the debt illusion, individuals may not make adequate plans to meet fiscal liabilities when these are due. However, insofar as the projects financed are genuinely chosen so as to yield benefits over time, presumably the ability of individuals to meet postponed liabilities is enhanced by these public service benefits, which are, in one sense, translatable into real incomes. The contingent liability element cannot be wholly eliminated, even for debt issue limited to the financing of long-term capital projects, because the accrual of benefits need not be distributionally equivalent to the optimally projected allocation of fiscal liabilities. It should perhaps also be noted that the ethical principle against the issue of debt which embodies some transfer of net fiscal liability to future generations of taxpayers does not fully apply when debt is limited to financing genuinely long-term projects. In this case, future generations enjoy the benefits as well as inherit the liability.
The analysis suggests, therefore, that public debt issue may be chosen as an appropriate part of the over-all "constitution" of a fiscal structure, provided that limitations are imposed to insure that debt financing be restricted to projects that yield benefits over time. "Capital budgeting" can be rationalized on the basis of the individual decision calculus here introduced. These conclusions are similar to those that were developed in the traditional or classical theory of public debt, and they have been incorporated into responsible fiscal practice. This correspondence itself, along with other instances noted in this book, tends to corroborate the efficacy of the general approach adopted.
Notes for this chapter
For several years this result has been discussed among economists at the University of Virginia as "Tullock's fallacy," since it owes its local origin to my colleague, Gordon Tullock. In a recently published paper, E. J. Mishan has indirectly noted the same point. See his "How to Make a Burden of the Public Debt," Journal of Political Economy, LXXI (1963), 529-42, especially note 5. The point is, of course, implicit in the traditional Ricardian notion that the public loan and the extraordinary tax are fundamentally equivalent for the individual, since this argument assumes that the individual is able both to borrow and to lend at the government borrowing rate.
The analysis of de Viti de Marco, although itself incomplete, is suggestive of the approach to debt theory developed in this section. See Antonio de Viti de Marco, First Principles of Public Finance, trans. E. P. Marget (New York: Harcourt-Brace, 1935), pp. 377-98.
This is not the place to repeat the analysis that demonstrates that public debt does, in fact, involve a postponing or shifting forward in time of fiscal liability. On this, see my Public Principles of Public Debt (Homewood: Richard D. Irwin, 1958). The discussion among scholars on this subject since 1958 is collected in James M. Ferguson (ed.), Public Debt and Future Generations (Chapel Hill: The University of North Carolina Press, 1964).
End of Notes
Return to top