"Fiscal Policy" and Fiscal Choice:
The Effects of Unbalanced Budgets
In making a decision as to whether he should support or oppose a proposed expansion in public spending, the individual must in some manner construct the bridge between the benefits and the tax-cost. The preceding chapter examined the ways in which the institutions of choice might affect the construction of such a bridge under the over-all constraint of budget-balance. This constraint, if it is known to the individual participant, facilitates comparison of benefits and costs. The next step in the analysis allows budgets to be unbalanced, and we seek to trace the effects that this single institutional change can exert on the individual's calculus in making tax decisions on the one hand and spending decisions on the other. The initial and elementary conclusion is that any comparison of costs and benefits becomes more difficult here than under a regime of over-all balance between revenues and outlay.
Unbalanced budgets are almost always possible in real-world fiscal systems. Any strict balanced-budget restriction must, at best, be considered as being imposed only on an overly simplified model, preliminary to the more general model that allows for unbalance. The earlier discussion is helpful, however, in that it facilitates concentration on the effects of unbalance, as such, on individual behavior.
The apparent splitting of the fiscal process into two parts was shown to produce potential gaps between preferred spending on public goods and services and preferred levels of taxation. Until and unless these gaps are eliminated, budget deficits tend to emerge from democratic decision processes. In our previous model, the deficits remained potential because of the imposed restriction of balance. Once we drop this restriction, emergent deficits can become actual deficits. To the extent that political decision-making institutions split the fiscal choice into apparently separated tax and expenditure choices, the potential deficit predicted as a part of an individual's choice calculus will tend to be transformed through the interaction of all individuals into an aggregate result.
If deficits are allowed to arise, they must be financed. And the manner in which they are financed may itself exert important influences on the individual's ability to make a reasoned comparison of public benefits and public costs. Broadly speaking, there are only two means of financing budget deficits. One is by borrowing; that is, by issuing interest-bearing debt obligations in exchange for current command over purchasing power. The second is by printing money; that is, by issuing or creating non-interest-bearing money or currency which becomes acceptable directly as purchasing power. The effects of unbalanced budgets on fiscal choice behavior must be examined under each of these two methods of financing. The second of these methods, currency creation, can only be exercised by governmental units that possess effective money-creating powers; normally, national or central governments. But national governments also assume some responsibility for the level and the movement of the aggregative economic magnitudes in the economy: income, employment, price levels. This fact further complicates the analysis here since it is evident that the real bridge between tax and spending decisions depends critically on the state of the aggregate economy. It will be necessary to examine how changes in the aggregate, macro-economic changes, affect the individual's own bridge between the two sides of the fiscal account, and through this, his own behavior in making fiscal choices.
Real Debt and Budget Unbalance
Even for the single individual or family, the private budget need not be balanced in each and every accounting period. "Going into debt" or borrowing either from internal or external sources almost always provides a means of resolving conflicts between income and outlay. Nevertheless, for the single family, debt issue provides at best a temporary reconciliation, a breathing space, until more permanent measures for correction may be taken. The effects of the existence of borrowing, as a temporary means of covering deficits in private budgets, are probably important in some cases, but these do not warrant further consideration here.
For governmental units, the borrowing alternative to taxation as a means of financing public expenditures is almost always available, within limits. Our question is: How will the knowledge that debt can be issued to cover deficits affect the choice behavior of the individual citizen? Let us again reduce this problem to the simplest possible model. Suppose that spending is to provide only one public good or service, and that only one tax is to be utilized. The individual confronts an apparent dual decision, one on the amount of public outlay, the other on the rate of tax. However, once we allow for the possibility of debt issue, there need be no real bridge between an expenditure and a tax decision, even for the whole community, in the sense that resources devoted to public spending need not be withdrawn in comparable magnitude from private spending, as payment for the public goods and services provided. If government borrowing provides the means of covering residual differences between preferred levels of spending and preferred levels of taxation, these two primary choices become independent, in any current-period sense, not only to the individual as he participates in collective choice, but also in real terms for the whole community. Under such conditions, a collective decision to spend does not imply a collective decision to tax currently, and a collective decision to tax does not imply a collective decision to spend currently in the same amount.
The presence of the debt alternative tends, therefore, to widen possible divergencies that may arise between the weighing of costs and of benefits in the two sides of the fiscal decision. When faced with a proposal for expanding public expenditures, the individual will tend to include a lower value for the opportunity cost of choosing favorably than he would in the balanced-budget model, other things equal. For several reasons, he will not treat the discounted value of future taxes that debt issue embodies as equivalent to nondiscounted current taxes. In other words, the individual can be predicted to "vote for" somewhat larger extensions in public outlay here than he would support in the "no-debt" model. On the other side of the budgetary process, when he is faced with a tax choice, he will not associate directly collective revenue shortfalls with curtailments in public services because of the availability of borrowing. He will be somewhat more reluctant to approve tax increases than he would be in the no-debt model. These predictions de-rived from models of individual behavior will tend to be transformed into collective-community outcomes.
It should be emphasized that these conclusions as to the effects of public borrowing on fiscal choice are wholly independent of the normative question as to whether or not borrowing "should" be an alternative to tax financing. The conclusions reached are both restricted and intuitively obvious; borrowing makes individuals more reluctant to levy current taxes upon themselves and others, and less reluctant to expand public spending programs.
Some vague recognition of this proclivity probably explains the origin and widespread use of constitutional debt limits that are imposed on governmental units, at all levels. In the absence of such limits, say, on a local government, the workings of democratic choice process might well produce debt issue beyond limits of "capacity," although increasing costs of credit might impose barriers to over-extension. With the imposition of limits, however, further possibilities for decision conflict arise. If a local community should approve spending of X dollars, taxation that would bring in revenues of Y dollars, but is allowed to borrow only Z dollars, when Z is less than the difference between X and Y, there emerges a conflict that must be resolved. And major inefficiencies can arise when constitutional debt limits, designed to minimize excessive debt issue, serve to inhibit what is essentially "productive" borrowing.
In one sense, local government borrowing is analogous to family or private borrowing in that it can provide only for temporary and extraordinary deficits. Without recourse to money creation, local governments must look to their own credit worthiness. There remains a fundamental difference between local government borrowing and family borrowing that should not be overlooked. In the latter, there is normally a single, responsible decision-making unit. In a democratically organized political group, by contrast, the individual participant is aware that he is not, individually or personally, responsible for group or collective decisions. He participates in these decisions, he expresses his own preferences, and he recognizes, more or less accurately, that collective choices influence his own well-being. He will not, however, feel the same sense of what might be called "unit responsibility" that he will feel in the private family decision process. In other words, precisely because he is "individual," he will not wholly identify his own interest and responsibility with that of the political group of which he is a part. His membership in the political community allows him, so to speak, to act under a system of limited fiscal responsibility. If the democratic processes of his local government should expand debt issue to the point of default, the individual is under no personal obligation to make good on the community debt. He is in a position, for purposes of decision, much like that of a shareholder in a limited liability enterprise, but without the latter's interest in "efficiency."
Any sense of fiscal identification that the individual might possibly feel as a member of a local government unit becomes less pronounced as the number of citizens in the group increases. At the national government level, there is essentially no feeling of private fiscal responsibility on the part of the individual citizen. This makes the suggested influence of public borrowing on fiscal choice more significant at the national government level, even if we remain within the individual calculus and ignore the much enhanced prospect that both tax and spending decisions at this level will involve important considerations of intergroup conflict. To demonstrate this point, examine the individual citizen's role in the 1963-64 discussion of tax reduction in the United States. What did he write his congressman, or what would he have said if he had written, and what were the underlying elements in his choice? Tax reduction was, as is usual, discussed independently of public spending. The individual was confronted with reasonably adequate measures of the additional private funds that he might secure under the proposed reduction schemes. What were the costs, to him, of favorable action on tax reduction? Clearly, he knew that these costs would not take the form of any reduction in current levels of public service provision. There were, in effect, no apparent costs that he could offset against the benefits of tax reduction promised him. He recognized that national debt (and/or currency) would be issued to finance any deficit that might have been increased as a result of the political decision process in which he participated.
Similar behavior can be predicted on the spending side of the account. If the individual citizen were asked, in mid-1963, his opinions on proposed expansions in the federal space program, he could, roughly and in some fashion, measure benefits in terms of sport, national prestige, adventure, technological fallout, etc. But what were the costs? He would not have translated the costs of the space program into increased taxes. And for a very simple reason: the individual knew that he would not have to pay such taxes. The predictable result of a democratic choice process is the generation of budget deficits when borrowing is available as an alternative to taxation unless deficit creation is not somehow restrained by constitutional limitations.
This result is, of course, reinforced when the emergence of budget deficits is rationalized and justified on "fiscal policy" grounds. The Keynesian and neo-Keynesian arguments in support of deficits tend to accentuate and to legitimatize the proclivity toward deficit creation that democratic governments inherently possess for the reasons developed. If this tendency is so pervasive, however, the question may be asked as to why deficit creation had not got out of hand even before the appearance of the Keynesian apologetics. The answer lies, not in the presence of genuine fiscal responsibility on the part of the individual, as citizen, or through him, on the part of his legislative representative, in the making of everyday decisions on taxes and spending. The answer lies, instead, in the fact that "constitutional" restrictions on debt issue (and/or currency creation) have been present, even at the national government level. Although it is not written down as such, the "balanced-budget rule" has been an integral part of the broader unwritten fiscal constitution of the United States. It seems probable that it is only the strength of this restriction, in part based on traditional ethical considerations, that has kept deficits within bounds of reasonable propriety in past periods. So long as the individual citizen accepts the "mythology" of budget-balance, this unwritten constitutional rule will continue to exert a limiting force on deficit creation. The effect of the Keynesian and post-Keynesian arguments is to undermine the "constitutional" status of this rule.
Once again it is necessary to state that the analysis here is not concerned with whether or not such results are or are not desirable. Nor does the conclusion about the bias toward deficit creation carry with it any particular implication about levels of public spending and of taxation. To say that, as it operates, democratic procedure tends to generate budgetary deficits is not the same as saying that public spending programs are "too large." This latter implication would hold only if there should exist agreement that public debt "should" not be issued. Clearly, there is no basis for such normative agreement. Hence, all that is implied is that public spending probably tends to be larger than and taxation less than they would be in the absence of the debt (and/or currency creation) alternative.
This is a very simple conclusion that amounts to saying nothing more than that national debt, as an institutional alternative to taxation, tends to produce budget deficits. This might appear as tautological in that, in the absence of debt, deficits would not be possible, ignoring for the moment the resort to money creation. But the conclusion has more content than this version of it suggests. The introduction of the debt alternative to taxation makes the bridge between cost and benefit more difficult for the individual to construct. This is a positive conclusion and should allow derivative hypotheses to be empirically tested. The analysis does not suggest that resort to the borrowing alternative is not desirable in many situations, for reasons that can readily be developed. Nor does the analysis imply that the Keynesian destruction, or attempted destruction, of the effective "constitutional" rule of budget-balance may not have been independently desirable.
The issue of public debt should never be confused with the issue of currency. Nothing has plagued modern economic policy analysis more than the persistent refusal of economists to make this distinction clearly. Public debt embodies an obligation to make interest payments in periods of time subsequent to issue. Currency involves no such obligation and, for this reason, its issue becomes a distinctly different fiscal operation. For governments that possess the authority to create money, there are two, not one, means of financing deficits that may be produced by emergent gaps between preferred spending rates and preferred tax rates. The question that is now relevant is how currency creation differs from borrowing in its influence on the fiscal choice behavior of the individual.
To answer this question it is necessary to make some assumptions about the state of the economy at the time of the operation. If resources are fully employed, or employed to the extent that increased aggregate demand will produce price-level increases without output increases, the issue of new currency is equivalent to the levy of an indirect tax on the users and the holders of cash. In this limiting case, therefore, despite the initial gap between approved public spending and approved levels of taxation, no real "deficit" in any genuine sense appears when the new-currency "tax" is employed as the residual financing device, the balancer. Insofar as the individual recognizes this and, in his choice calculus, cuts through the apparent illusion that currency issues create, he may be more reluctant to approve new spending projects than he would be in situations where genuine public debt is the balancing device. Insofar as he is able to make the proper bridge between benefits and costs, these costs are measurable under currency creation in this model in terms of current income units. In the debt-creation case, by contrast, the comparable measure of costs must be computed in terms of present values of future income units. Insofar as future taxes are not wholly capitalized, there will be some tendency for residual borrowing to generate larger budget deficits than residual currency creation, other things equal. This conclusion depends on the assumption that the individual is able to dispel the illusion that money creation necessarily introduces. This is, of course, an important proviso, and the effects of this illusion may overwhelm those emphasized here.
Currency Creation and Unemployment
The analysis must be modified when we shift out of the full-employment model. Consider now the opposite extreme; assume that there exist unemployed resources to the extent that aggregate employment and output can be expended without generating price inflation. In this situation, the financing of budget deficits by currency creation does not impose a current indirect tax on the holders and users of cash. The recognition of the possible real-world existence of this limiting case was the essential novelty of the Keynesian "revolution" in thinking about economic policy. In such a situation of deep depression, which did seem to characterize the 1930's, a decision to expand public spending does not imply an offsetting real cost to the individual, as a voter-taxpayer-beneficiary, either currently or in future periods. Professor Abba Lerner was basically correct in his early insistence that, in such situations, there is no underlying real cost of public spending, provided that it is financed by pure currency creation. The real bridge does not exist here, in either community or individual terms, and there is no logical economic basis for the imposition of taxes. The financing of budget deficits by currency creation becomes the logical translation of economic reality into meaningful decisions as these are confronted by individuals in the group, through their legislative assemblies. In fact, the "ideal" structure in such situations is one in which only spending decisions are proposed. Failing this, the complete divorce of spending decisions from taxing decisions is desirable, and ideally rational behavior would involve the approval of expenditure expansion and tax reduction simultaneously until the growth in aggregate economic activity requires the acknowledgment of the real bridge between the two sides of the fiscal account. It was precisely to facilitate such a genuine splitting of the fiscal decision that the Keynesian and the neo-Keynesian attack on the budget-balance rule was launched. The difficulty is, of course, that "constitutional" rules may be helpful in constraining choice behavior in certain situations but may become undesirable in other situations, and vice versa. If the ultimate effects of the Keynesian attack are to undermine the budget-balance rule, fiscal choice during periods of deep depression will, without doubt, be "improved" since the institution of balance in such periods serves to distort individual choice. However, this may well be accomplished at the expense of "worsening" the results of fiscal choice during periods of high income and employment, when the rule of budget-balance does assist the ultimate choosers in making the proper bridge between the two sides of the account.
In the normal order of events, the economy will be neither in "full employment," in the sense described above, nor in "unemployment," in the contrary sense. At almost any time, an increase in aggregate demand will both generate some additional employment and output and some inflation in prices. The mix between the employment-output effects and the inflationary effects will, of course, vary over the phases of the so-called "business cycle," but both effects will normally be present to some degree. Let us then examine the fiscal choice behavior of the individual as he might confront federal tax proposals and federal appropriations measures. Suppose that he assumes that deficits emerging from revenue-expenditure combinations will be financed solely by currency creation. How will he construct the necessary bridge between benefits and costs? He will, probably, tend to approve spending projects that require more revenue than he approves in taxes, over all phases of the cycle combined. The general bias toward deficit creation remains. But the real cost of government spending projects will vary over the different stages of economic activity, and, ideally, choice behavior should embody some recognition of this variance. It is clear, however, that this variance adds yet another element of uncertainty to the individual's decision calculus. When he writes to his Congressman approving, say, a proposed expanded program for anti-missile missiles, how much will he expect this expansion to cost him, individually and privately? Given the tax structure as it exists, and assuming that revenues were just equal to total expenditures prior to the fiscal decision under consideration, the individual knows roughly what his total tax liability is. But he now proposes to expand the rate of public spending without, at the same time, changing tax rates. How will he estimate his costs? If the program is approved, and the deficit created, price inflation and/or greater national output will result. The real costs suffered by the individual will vary greatly depending on the precise breakdown between the price and output effects of the deficit-money-creation operation. If the deficit expands total output, the additional missile defense is secured at little or no cost. If, on the other hand, price inflation results, the additional defense is provided only at a real cost imposed on the individual through his holding and his usage of cash. To make any reasonably accurate translation of a spending proposal into tax-costs, the individual must predict the movement of the aggregative variables in the whole national economy. This movement, in turn, depends on the behavior of all of the economic units in the system, as well as the external variables. It is difficult to think of a situation where the interdependence involved in an individual choice calculus could be greater.
Debt Creation, Currency Creation, and Uncertainty
The analysis suggests that under either of the two extreme or limiting models, that of full employment or that of unemployment, currency creation as the residual financing institution should be more conducive to rational individual behavior in fiscal choice than debt issue. If employment is effectively full, currency issue becomes equivalent to a current-period tax, one that is somewhat more likely to be correctly weighted than the future-period taxes embodied in debt issue. In such situations, the deficits produced in democratic-decision processes are likely to be somewhat smaller under currency creation than under debt. On the other hand, in the Keynesian unemployment model, currency creation embodies no real cost and clearly this should produce larger deficits than debt issue, which does embody some real cost, even if not fully sensed by individual participants. The conclusion must be that in either of these two models, debt issue is a second-best residual financing device or institution.
Currency creation remains relatively more efficient as the residual financing institution for those models which allow some combination of employment-output and price effects so long as the mix between these is known with certainty. The individual will always be able, in such circumstances, to make a better comparison between benefits and costs if he knows that potential deficits are to be financed by currency creation rather than by the issue of interest-bearing debt. This conclusion must be modified, however, if uncertainty as to the mix between employment-output and price effects is present. Here second-best or relatively inefficient institutions may be supported on logical grounds, and resort to debt issue as the residual financing device may be justified. Consider a setting roughly equivalent to that faced in the United States in the early 1960's where unemployment was quite high but where controversy raged as to whether this was attributable to deficient aggregate demand or to structural factors. We may compare the two institutions for residual financing in this setting. The individual who placed most weight on the deficiency in aggregate demand would tend to assess the real costs of spending programs somewhat higher under the public debt alternative than he would under the currency creation alternative. On the other hand, the individual who placed the higher weight on structural elements in unemployment would tend to assess the real costs of spending somewhat lower under the public debt alternative than under the currency alternative. The effect of the public debt alternative is that of bringing the two assessments more closely into agreement, but providing some built-in offset to error in each. Under certain configurations of possible error, more rational choice behavior might well be produced by reliance on the debt alternative.
A Regime of "Functional Finance"
After the early enthusiasm for Keynesian ideas, during which policy proposals were often advanced with little regard for the structure of political decision-making institutions, a more realistic discussion of budget unbalancing, of "fiscal policy," has taken place and is continuing. How can the budget of the national government be utilized as a tool in an over-all policy for maintaining desired values for the macro-economic variables within an effectively democratic process? It came to be widely acknowledged that "functional finance," the deliberate manipulation of the budget for macro-economic policy purposes, could hardly be expected to work well when both revenues and expenditures remain within the control of representative legislative assemblies. The inherent bias toward deficit creation that we have discussed came to be recognized, along with other structural defects with fiscal policy weapons. The hope that functional finance might lead to symmetry over the whole cycle vanished. The widespread acceptance of these facts led advocates of fiscal policy to advocate modifications in the basic institutional structure. Proposals were made to shift the authority over decisions concerning both tax rates and spending rates to the executive and to remove these from direct legislative control. The executive, who is presumably under less direct fiscal pressure from the electorate, the individual citizens with whom this study is primarily concerned, was presumed able to operate more "efficiently." Tax rates and spending rates could, presumably, be moved up and down more freely so as to promote macro-economic objectives, and the inherent conflicts of democratic decision processes largely eliminated. The Commission on Money and Credit in 1961 proposed that a step be taken in this direction by granting to the President discretionary power to move first-bracket rates of the personal income tax up and down to facilitate fiscal policy action.
Let us assume that such discretionary power, additional to that now possessed, is transferred to the executive by the legislative body in some quasi-constitutional delegation. The executive would then be empowered to expand or to cut back spending projects and to reduce or to increase tax rates. The effects of such a change on individual fiscal choice, which is our center of attention, are clear if we assume that this choice is exercised primarily through the representative legislative assembly. The shifting of additional power to the executive removes effective control over ultimate fiscal decisions another step away from the individual citizen and creates for him still further uncertainty concerning the relationship between the benefits that he secures from governmental programs and the costs that he must suffer through the payment of taxes. It is obviously impossible to delegate to the executive additional "functional finance" powers without, at the same time, granting to it additional powers over the basic fiscal decision itself, that is to say, over the ultimate mix between private goods and public goods, and over the composition of the latter.
This power is already possessed by the executive to a significant degree in the current American institutional structure. It is possible for spending rates to be speeded up or slowed down, especially in the defense sector of the budget, and certain discretionary powers are also present on the tax side, notably in connection with rules for depreciation. In addition, the executive has the formal responsibility for preparing the expenditure budget, as a plan for the whole governmental fiscal operation. It would, however, be possible to shift power further to the executive, as the various proposals suggest. This change would lessen the individual's control over decisions, and to the extent that the executive remains insensitive, or less sensitive, to pressures from the electorate, the biases analyzed in this chapter and the preceding one might be reduced. In such an executive-power system of decision-making, the whole analysis developed in this study is changed in character. In such a system, where the decisions made by individual citizens are largely confined to "choosing the choosers," perhaps the traditional models of public finance theory, those which implicitly assume the presence of decision-makers divorced from the citizenry, would become more suitable.
A Regime of Rules
The recognition of the usefulness of the balanced-budget constraint on democratic decision-processes as well as the need to allow for budget-unbalance as a weapon in macro-economic policy led to various proposals for alternative budgetary-fiscal policy rules. One of the most important of these was that for "budget-balance at full employment," which was proposed in the 1940's, and widely accepted during the 1950's as the norm for policy. This rule was replaced, to an extent, by that of "budget-balance at potential GNP" in the 1960's. These are defensible rules for policy, but they are rules for the sophisticated, for the expert, and they cannot be expected to inform the consciousness of the individual potential taxpayer-beneficiary, or even that of his legislative representative. Such rules as these cannot be expected to mitigate significantly the biases in democratic processes that the abandonment of the strict budget-balance rule produces.
Alternative proposals have been made for the introduction of more definite rules concerning the increase in the supply of money over time. Such a proposal could, if desired, require that new currency be issued only through the budget. In this way, budget deficits would be a permanent feature of the growing national economy, but such deficits would be held strictly within check by the monetary-growth rule. While this proposal for a monetary-growth rule, associated with Professor Milton Friedman, would be less adaptable to the day-to-day adjustments required for macro-economic objectives, somewhat more informed consideration of the benefits and costs of public programs might be possible for the individual citizen than would be the case under the more sophisticated alternatives.
The effects of budget unbalance on the individual's ability to make the appropriate comparisons between public service benefits and tax-costs in a regime of effectively full employment seem clearly undesirable. The knowledge that residual gaps between preferred levels of spending and preferred levels of taxation will be financed either through public debt issue or through currency creation will surely make the individual less willing, and less interested, to construct the bridge between the two sides of the fiscal account that any fully informed fiscal decision would require.
The strict requirement of budget-balance will, however, during periods of unemployment also distort the individual choice calculus. Under the balance constraint, the individual will necessarily overestimate the real costs that public expenditure programs involve in such circumstances. In the essentially mixed post-Keynesian world, where elements of both the full-employment and the unemployment model are likely to be present, the relaxation of the budget-balance rule along with the absence of an agreed-on alternative makes any reasoned comparison of benefits and costs almost impossible. To the extent that governmental budgets are used to achieve what are essentially macro-economic objectives without the constraints of predictable rules, the scope for individual control over the size and composition of budgets, through ordinary democratic procedures, must be progressively reduced. How is it possible for the individual to answer the question, How much "public goods" should I "purchase"? if, because of uncertainty concerning the relationship between the two sides of the budget account, the real "price" to the individual, of these public goods, is continuously and unpredictably changing?