Public Finance in Democratic Process: Fiscal Institutions and Individual Choice
Individual Demand for Public Goods
The analysis is designed to contribute to the derivation of a "theory of demand for public goods and services." Difficulties arise at the very outset, however, when we begin to think about public goods and services in such terms. How are public goods demanded by individuals? What are public goods?
For our purposes, any good or service that the group or the community of individuals decides, for any reason, to provide through collective organization will be defined as public. The inclusive category may include some goods that Samuelson and others have designated as "purely collective," but it may also include other goods and services, with the degree of "publicness" ranging from zero to 100 per cent. The inclusive definition is suitable because our purpose is that of analyzing the organization of public-goods provision, and not that of determining the proper classification of particular goods and services independently of organization. Our purpose is not that of answering the question: What goods should be public?
Purely Collective Goods
It is, however, precisely because goods and services that are provided governmentally are rarely, if ever, wholly collective that problems arise in discussing demand. Recall the initial Samuelson definition of a purely public good;*6 one that must be consumed equally by all members of the collective group. If a unit is available to any one member of the collectivity, a unit must be, by definition, also available to each other member of the group. The benefits are wholly indivisible with respect to the shares of the separate individuals. Only in such a polar case can the "quantity" of a public good be defined unambiguously. In this polar model, the individual compares potential costs and benefits of a public good which he knows to be available equally to all. It is possible to discuss the demand of the individual in this case without introducing the complexities that are involved when partial divisibility of benefits among separate individuals is present.
To facilitate analysis, think of public expenditure decisions as being made marginally or incrementally. The group decides on the number of units to be provided one at a time, and serially. The choice is not all-or-none. It is now possible to construct, conceptually, for each individual in the community a schedule or curve of marginal evaluation. Assume that only one public good is considered. If we neglect income effects, this schedule or curve is fully analogous to the demand schedule or curve for an ordinary private good or service. Figure 2.1 illustrates. On the vertical axis is measured the individual's marginal evaluation of the public good in dollars. On the horizontal axis is measured the quantity of the good, per time period, potentially available to the individual, and equally available to each other individual in the group.
In this simple construction, which wholly neglects income effects, we can think of this demand or marginal evaluation curve as indicating the quantities of the public good that would be optimally desired by the reference individual at each of a series of different "tax-prices per unit" confronting him. For simplicity, we may assume that the "tax-price per unit" in each case remains constant over quantity. In other words, the "marginal tax" remains equal to the "average tax," per unit, as these are faced by the choosing individual. Through this convention, we may think of the individual as being confronted with a series of horizontal "supply curves" for the public good analogous to the standard derivation of demand curves for private goods. For most of the analysis of this chapter more complex models need not be introduced. For those who are concerned about the neglect of income effects, however, it should be noted that, through specifying a single schedule of tax-prices, that is, through fixing a single "supply curve" for the public good, as faced by the individual, we may fix a unique curve of marginal evaluation. The circularities in this procedure need not be damaging at this stage.
In Figure 2.1, Da represents the single person's demand for the single public good under one specifically designated institution of payment, that which is identical to payment in the private economy. At a tax-price per unit of 0T, the individual would, if his own private desires are fulfilled, "purchase" an amount, 0X, of the public good. There is, of course, no assurance that the individual can be at or even near to his own preferred position, his private "equilibrium," in public-goods purchases. The final quantity of goods provided must be chosen by the whole community, acting through complex institutional processes. The demand curve for a single public good allows us to think more clearly about the individual's participation in such decision processes, even though we recognize that he will not normally confront fiscal alternatives in such an abstracted setting.
Assume that the individual knows that the public good, to be made equally available to all persons, will be financed by a specific tax institution that will be expected to impose upon him, personally and privately, a per-unit charge shown at 0T. The individual will tend to approve all spending proposals for extending the provision of the good up to an amount 0X. Similarly, he will tend to "vote against" all proposals for providing more than 0X, keeping in mind our assumption that all-or-none choices are not presented. It should be noted that the individual calculus depicted in this demand-curve construction is straightforward. The individual has no incentive to "conceal" his true preferences for the public good or service. This aspect of behavior is absent because we have postulated that a specific tax scheme is pre-selected, externally to the chooser, and that his own action cannot modify the tax-price per unit at which the public good is made available to him. The methodological defense of this approach is postponed until Chapter 9.
Suppose that the collective good is Polaris submarine defense, and that the quantity is measured by the number of submarines in commission. The same number of submarines is available to each and every citizen. Suppose further that the provision of this particular defense is to be financed through the imposition of an equal-per-head tax and that the marginal cost of supplying additional submarines is equal to average cost. The individual taxpayer-beneficiary can, in this abstracted model, estimate his own "private" or "individualized" share in the collective benefits provided at each level of submarine defense and also his own "private" share in the tax-cost that this defense embodies. Now assume that the decision on quantity to be supplied is made by a referendum process, where the individual is allowed to vote on successive spending proposals, commencing with one unit and increasing until some group decision is attained. He will tend to vote in favor of proposals for spending up to a level of 0X submarines, and he will vote against all proposals for extensions beyond this quantity. As emphasized, the individual person will not be allowed to make an independent quantity adjustment; he cannot, by definition, consume a quantity different from anyone else. For this reason alone, and even in this highly abstract referendum model, it will be unlikely that his preferences will be fully reflected in the group decision outcome. But this outcome, both in simple and complex models, can only be determined through some analysis of the choices of the individuals who participate.
We seek now to examine the effects on an individual's choice behavior that might be exerted by the institutions of taxation, by the methods through which the individual is required to meet his financial obligations for the public good. Suppose that the position shown in Figure 2.1 reflects the individual's predicted response under a head tax. Let us now modify the taxing institution. Assume that the same collective good, submarine defense, is to be financed, not through the levy of a head tax, but through a tax on the net income of corporate enterprises. How will this change in the institution of payment affect the private decision of the single voter-taxpayer-beneficiary? Three possible effects may be distinguished. First of all, the new means of payment may be more or less "convenient" to the individual, quite apart from considerations of uncertainty or ignorance about shares in the aggregate tax liability. In other words, any individual will have some preference ranking for the various means of meeting financial obligations to government, even if he must pay the same net tax in each case. This scale of preference is rarely noted in connection with ordinary market choices because the individual is considered to be free to make payments in any manner that he chooses. He is not required to discharge obligations in any particular way, as he normally is in the fiscal process.
To isolate this first effect on individual choice behavior, suppose that, as an owner of a specific number of corporate stocks and as a consumer of a specific quantity of commodities produced in the corporate sector, the individual estimates his own personal tax liability to be the same as that predicted under the head tax scheme depicted in Figure 2.1. Suppose, however, that he would "prefer" to pay this sum under the corporate tax arrangement rather than through the head tax. For purposes of his decision calculus, this would have the effect of shifting the supply curve, or tax-price line, downward, even though the net tax paid is the same in the two cases. Due to the nonpecuniary advantages and disadvantages of the various methods of payment, the individual may act differently in separate institutional situations, even with equivalent net taxes. The effects on his choices will be identical to those stemming from a tax-price reduction. This nonpecuniary or convenience effect does not seem likely to be of major importance in influencing the individual's behavior in demanding public goods. It is introduced briefly here primarily for analytical completeness.
The second, and much more important, effect of the suggested shift in institutions arises, not out of preferences for or against particular means of payment, but out of the differential effects on the uncertainty and ignorance concerning the individual's own share in the aggregate tax liability. The taxpayer may know that he owns a specified number of corporate shares and he may also know how many commodities he purchases from the corporate sector. But he may have no idea at all about how much his own share in the cost of an additional Polaris submarine will be under the institution of the corporation income tax. The contrast in this respect with the head tax, under which we have assumed the individual able to predict his own tax liability with reasonable accuracy, seems dramatic. His estimate of the "private opportunity cost" of the public good is likely to be grossly in error. This introductory example indicates that the influences of fiscal institutions on the information pattern of the individual warrant careful examination.
The third possible effect of the change in fiscal institutions on the behavior of the individual stems from the fact that he is able, under most tax schemes, to influence the tax-price, the terms of trade, with the fisc. The individual may, through modifying his pattern of private earning or spending, or through participating in collective decisions, change the net tax-price per unit of public good that he confronts. This third influence is not present under the equal-per-head tax, used here as a benchmark. But it will be present under most other institutions. If the rate structure of the tax is chosen independently of the decision on the amount of the public good supplied, we may still think of the individual as being faced with a horizontal "supply" curve. His behavior as a direct participant in the collective choosing process cannot modify the tax-price per unit, although it can, of course, modify the total tax bill. However, if the tax base involves any relationship to his behavior in the private economy, he will be able, by changing this behavior, to modify this tax base, and, through this, the tax-price per unit at which he, along with others, may "purchase" the public good. For example, under the personal income tax, the individual, through not earning taxable income, may slightly increase the tax-price at which units of public good are made available to every other member of the community while at the same time lowering the tax-price for himself. His own estimate of the tax-price that he faces will depend, therefore, on some prediction as to the private behavior of others as well as his own. Clearly, additional uncertainty is introduced in an individual's decision calculus.
How do these separate effects combine in influencing the behavior of an individual in demanding a single public good? Figure 2.2 illustrates. Instead of a uniquely determinate "equilibrium" position, depicting the most preferred position of the individual, we get, at best, a whole range of indeterminacy. As we allow the "tastes" of the individual to vary over a relatively narrow range we can think of minor shifts in the effective supply curve. But as the information pattern of the individual is modified under the various fiscal institutions, we can think of this shifting taking on major proportions. The supply curve, upon which the individual actually makes a choice, may fall anywhere within the shaded range drawn in Figure 2.2 as the tax institutions change. Clearly, the behavior of the individual as he participates in collective decision processes will depend, and significantly so, on the way in which his tax bill is presented to him.
As shown in Figure 2.2, the range of choice is wide, even in this highly rarified model. The individual would approve all spending programs, regardless of the tax institution, up to 0X. He would disapprove all programs beyond 0X'. Within the broad range, 0X-0X', he might approve or reject proposals to finance extensions or contractions, depending on the particular institution of payment.
This preliminary discussion is designed to make one elementary point. The effects of the institutions of payment on individual choice behavior are more important in fiscal choice than they are in market choice. Part I of this study may be summarized as an attempt to make some rudimentary predictions concerning these effects. Will the individual choose to spend more publicly under income taxation or expenditure taxation? Before this question can be directly answered, what variables must we analyze?
The analysis of an individual's demand for public goods is relatively simple only in the polar case of the purely collective good. We know, of course, that such goods rarely exist, in any descriptively realistic sense, and that governmental units provide goods and services with widely varying degrees of benefit divisibility among separate persons. Individual shares cannot normally be treated as equal, and units of the public good available to one individual need not be equally available to all others. Insofar as any divisibility of benefits among separate persons is introduced, the consumption of units by one person must decrease the availability of units for others in the group. To use Musgrave's terminology,*7 exclusion rather than nonexclusion applies, at least to a degree, for most publicly supplied goods and services.
This fact of partial divisibility makes it necessary to introduce, for almost all publicly provided goods and services, two distinct demand elements. The first is the private demand for the good or service, as this is exhaustively treated in basic economic theory. The demand for divisible components of the good is an inverse function of the direct user price that is charged, other things equal. The familiar propositions in the standard theory of consumer's choice apply; the fact that the supplier happens to be the collectivity has little relevance. The individual is a quantity-adjuster, or potentially so, and different individuals may consume different amounts.
This individual or private demand for excludable or divisible components of goods provided collectively is not, however, the individual demand with which this study is concerned. By definition, divisibility implies that separate units may be demanded and consumed, individually and privately. This behavior is outside our field of reference, however, since we are concentrating attention on individual choice and behavior in collective decisions to purchase and to consume public goods. The demand for "private" or divisible components is absent from individual behavior here. Individual responses in demanding quantities of a public good as a participant in political choice processes, and as related to tax-prices, not user prices, become the relevant behavioral elements.
An illustration will prove helpful. Consider a good that is characterized by both "collective" and "private" components, say, the services of a municipal park. Decisions as to the amount and quality of park services must be taken collectively through some set of institutional rules for reaching political-administrative decisions. Individuals participate in these decisions, and it is this participation that is the object of our researches. Behavior in this process may, however, depend upon the private demand for the services of the park, and this demand may, in turn, be functionally related to the direct user charge that is to be placed on usage of the facility. However, insofar as the collectivity, in making decisions, does not respond directly and automatically to privately expressed demands and nothing more, that is, insofar as pure market criteria are not utilized, other demand considerations enter into the individual's decision calculus. These reflect, or should reflect, the genuinely indivisible components that the facility embodies, and the demand for these collective elements can be treated in the same manner as in the analysis of the purely collective good, discussed above.
Let us say that a decision is made to charge direct users twenty-five cents for each visit to the park. This privilege of strolling or sitting in the park, at twenty-five cents per visit, is equally available to all citizens, or rather it is considered to be potentially available at the moment of collective decision. The individual, as a member of the municipal community, considers the availability of park services, at twenty-five cents, to be genuinely collective in the polar sense. Through this convention of incorporating direct user pricing into the institution itself, we convert, as it were, all quasi-collective goods and services into purely collective goods for purposes of the analysis. In effect, the procedure amounts to breaking down the mixed public-private good into its two component elements. Once we have specified a user price for divisible components, we may proceed to construct a demand schedule or curve for the collective components. The fact that a person knows that he will be charged a user price of twenty-five cents will, of course, affect his estimate of the tax-price that incremental additions to the facility will involve. While he may value collective components of a zero-user price facility, at the same level of usage, higher than he would value components of a positive-price facility, he will also recognize that the revenues from user pricing in the latter case will reduce the share of the facility that must be financed from tax-prices. Direct user prices will be treated as a partial substitute for tax-prices.
We may illustrate this in Figure 2.3. Da represents the individual's demand for the park services, as a collective good, on the assumption that these services are to be made available at zero-user prices. Db represents the same person's demand curve for services on the assumption that a twenty-five cent user charge will be levied. As drawn, this curve lies below Da throughout the range. This relationship need not hold universally. If congestion is sufficiently serious, the individual may value park services higher with than without user prices. The estimated tax-price that he will be required to pay is clearly lower in the second case than in the first. If 0T is the predicted tax-price when park services are made available free of direct user charges, the position of private "equilibrium" will be at E, with the most preferred quantity (size of facility) at 0X. At the lower estimated tax-price, 0T', which accompanies the demand curve, Db, the equilibrium position is shifted to E', with preferred quantity 0X'. The geometry of Figure 2.3 shows that the position of individual "equilibrium" may move in either direction along the abscissa with the introduction of direct user charges. Even without undue congestion, collective decision processes may well generate more total investment in public facilities that provide quasi-collective services with the charging of direct user prices than without. Public parks may be larger in municipalities that charge user prices than in those that do not.
This analysis may be generalized for any degree of "publicness," from pure collective goods to pure private goods. Consider a good that is supplied through ordinary market institutions, or through a public enterprise that operates solely in accordance with profitability criteria. Even in this case, it is possible to conceive of an individual demand for the good, as a collective, not a private, good, quite apart from demand for the specifically divisible components. Under market institutions, the tax-price that the individual expects to pay is, of course, zero since direct user charges, market prices, are expected to finance the whole supply. Nevertheless, an individual demand schedule for the good, as a collective good, may exist, and this schedule may be employed to indicate various tax-prices, over and above market prices, that the individual would be prepared to pay for the availability of the good, at market prices. The right to purchase unlimited quantities at going market rates becomes, in this model, a purely collective good for the purposes of the analysis. Since the "supply" curve facing the individual is, however, expected to be that reflecting zero tax-price, the individually preferred quantities are those generated through the institutions of the market.
The point is that the elements of demand for any good, whether this be classified as wholly, partially, or not at all "public" by the standard criteria, may be factored down into private and collective aspects.*8 Recognizing this, it is possible to analyze individual demand for collective benefits provided by any good or service in terms of the Samuelson polar model. We can ask the same questions posed in the preceding section. How do the institutions of payment, the form in which taxes are imposed, affect the demand of the individual for those components of goods and services supplied publicly that must be equally available to all members of the group?
Notes for this chapter
Paul A. Samuelson, "The Pure Theory of Public Expenditure," Review of Economics and Statistics, XXXVI (November, 1954), 387-89; "Diagrammatic Exposition of a Theory of Public Expenditure," Review of Economics and Statistics, XXXVII (November, 1955), 350-56.
Cf. R. A. Musgrave, The Theory of Public Finance (New York: McGraw-Hill, 1959).
This is noted, in a somewhat different way, by Burton A. Weisbrod in his article "Collective-Consumption Services of Individual-Consumption Goods," Quarterly Journal of Economics, LXXVIII (August, 1964), 471-77.
End of Notes
Return to top