Public Finance in Democratic Process: Fiscal Institutions and Individual Choice
By James M. Buchanan
- Ch. 1, Introduction
- Ch. 2, Individual Demand for Public Goods
- Ch. 3, Tax Institutions and Individual Fiscal Choice
- Ch. 4, Tax Institutions and Individual Fiscal Choice
- Ch. 5, Existing Institutions and Change
- Ch. 6, Earmarking Versus General-Fund Financing
- Ch. 7, The Bridge Between Tax and Expenditure in the Fiscal Decision Process
- Ch. 8, Fiscal Policy and Fiscal Choice
- Ch. 9, Individual Choice and the Indivisibility of Public Goods
- Ch. 10, The Fiscal Illusion
- Ch. 11, Simple Collective Decision Models
- Ch. 12, From Theory to the Real World
- Ch. 13, Some Preliminary Research Results
- Ch. 14, The Levels of Fiscal Choice
- Ch. 15, Income-Tax Progression
- Ch. 16, Specific Excise Taxation
- Ch. 17, The Institution of Public Debt
- Ch. 18, Fiscal Policy Constitutionally Considered
- Ch. 19, Fiscal Nihilism and Beyond
Tax Institutions and Individual Fiscal Choice
Recall the characteristics that describe the restrictive model of tax-price invariance:
1. the tax is newly imposed;
2. the revenues from the tax are clearly earmarked for the financing of a single public good or service;
3. the benefits from this good or service are currently enjoyed;
4. the amount of tax, per unit of the public good, to the individual, is independent of his own, or others’, behavior in collective choice;
5. the amount of the tax, per unit of public good, is independent of his own, or others’, behavior in market choice;
6. the amount of the individual’s total tax bill depends strictly on the quantity of public good that the community chooses to supply.
As the analysis showed the fifth characteristic is violated when the individual is allowed to vary the tax base through his own behavior; the sixth condition is modified in consequence. In addition, under progressive rate structures, the fourth condition is not likely to be met.
Each of the institutions examined in Chapter 3 is normally classified in the category of “direct” taxation. The person upon whom the fiscal obligation is levied is presumed to be the person that the collectivity
intends as the final payer of the tax. The fact that the individual may be able, within limits, to vary his own liability for the tax is not, presumably, taken into account directly in the decision concerning the distribution of the total tax load among individuals and groups. By contrast, under an “indirect” tax, the person or entity that is legally obligated to pay is not, presumably, selected with the deliberate understanding or prediction that final payment will be borne. The tax is imposed on the basis of some more or less definite predictions about the behavioral responses of such directly obligated taxpayers, responses that are aimed at shifting or transferring the final burden onto others in the community. Final incidence of the tax is supposed, therefore, to rest with individuals who are only
indirectly affected by the fisc; that is, with those whose net tax obligation stems from modifications in the market behavior of others in the directly assessed group. The members of the latter group serve, in a very real sense, as the set of tax collectors for the treasury. The tax liability of any individual depends
directly on the behavior of these intermediary entities. From this it follows that, even in the absence of all behavioral response on the part of the final taxpayer, our fifth descriptive characteristic would have to be modified so that,
5a. the amount of tax per unit of the public good, to the individual, is directly dependent on the behavior of other members of the community in market choice, even if independent of his own behavior in market choice and also of others’ indirect market responses.
We know, of course, that the latter half of this condition is not fulfilled under familiar indirect tax institutions. Nonetheless, the condition is useful as a benchmark for comparison. A specific excise tax levied on a consumption item that has zero price elasticity of demand for all individuals in the group would approximately meet 5a.
Initially, we shall concentrate on the first half of condition 5a in order to show how the additional element of interdependence that is introduced tends to increase the individual’s range of uncertainty. The very indirectness of payment insures this result. The individual, as bearer of the real costs of those goods and services supplied publicly, is not assessed directly or personally. The sensation of paying funds directly to the fisc in “exchange” for the availability of public goods is absent. To reach a mental state comparable to that in which the direct taxpayer finds himself, a translation of sorts must be completed. Normally the individual will be partially conscious of the fact that conditions under which he makes market choices are modified by the tax. But to compute his own tax liability he must make a set of calculations over and above all of those required under comparable direct tax institutions. He must first distinguish between the conditions of market choice before and after the tax. Having done this, he must effect a translation of the differences into a cost- or tax-price equivalent. Even if we assume that he is rationally motivated, the individual may find it almost impossible to act on the basis of any reasonably accurate evaluation of alternatives. This preliminary conclusion will become evident as we examine various tax institutions more fully.
Corporate Income Taxation
The taxation of business income is an important means of meeting revenue needs in modern fiscal systems. If the business corporation is considered as a person, as it is in strictly legal terms, this tax belongs in the category of direct taxation, and scholars in public finance have often classified it in this way. For purposes of this analysis, however, the corporation tax cannot be treated in this fashion. Recall that individual behavior in collective fiscal choice processes is the subject of inquiry. And corporations, as such, do not participate directly. Corporations do not vote, although at a secondary level of consideration corporate interests may influence political decisions. There remains, nonetheless, a fundamental distinction between corporate behavior in the private sector and in the public sector of the economy. Corporations, as such, do “vote” with their dollars in the market alongside private individuals and families. They do not “vote” directly in public choice. For purposes of this analysis, the tax on corporate income must be classified as indirect.
As before, we want to isolate if possible the variables that enter into the decision calculus of a single voter-taxpayer-beneficiary. Again let us remain within the restrictive confines of the earlier models. Revenues from a newly imposed tax are earmarked for spending on a single public good, units of which are to be available to all members of the group. Some previously agreed “constitutional” decision is presumed to have selected the tax on corporate income as the revenue-raising instrument. Initially, we want to avoid the problem of distributional differences among separate persons, and to concentrate on individual choice for the public good. To do so, we may assume that the individual, whose calculus we examine, is genuinely “representative,” and his private economy can be described in terms of shares of ownership in corporate stock and dollars of purchases from the corporate sector.
How will such a person behave in “demanding” public goods? How will he go about making an estimate of the “price” at which such goods are made available to him?
Initially, let us assume away the whole complex of issues concerning the short-run incidence of the tax. Assume that the tax rests exclusively on the owners of shares in corporate enterprise, and that this is known. The tax does not affect corporate output. In other words, let us assume that the tax is an “ideal” one, levied on pure economic profit, which all corporations try to maximize.
If the rate of tax is predetermined, the representative shareholder can estimate, within some limits, his own share in corporate-tax liability under these highly restricted conditions. As we have shown in the discussion of the earlier models, however, the rate of tax cannot be determined independently of the decision on the quantity of public goods to be supplied. If we think of the group as voting or deciding in some other fashion on various proposals for spending on public goods, we must allow the rate of tax to be adjusted. Or, alternatively, if we think of the group as “voting” on the rate of tax to be levied, we must allow the quantity of public goods to remain dependent on the tax-rate decision. In either of these cases, the individual must make some estimate as to the size of the aggregate tax base. In this extreme model, where the tax is levied on pure economic profit, there is no direct behavioral response on the part of the corporation. However, even here, the independent variability of the tax base introduces major uncertainty into the choice problem faced by the representative individual. The situation is roughly comparable to that faced under the personal income tax when the individual has no control over the amount of income that he receives. The uncertainty is greater under the corporate tax, however, due to the greater volatility in aggregate corporate profits.
Once we modify the model to allow for some behavioral response on the part of the corporation, additional elements of uncertainty are introduced, similar to those examined under the personal income tax. Both individual and aggregate base variability are increased, and with this, uncertainty in any fiscal choice that the individual must make. And, in each instance, the individual must make the translation from corporate to personal liability. He must compute a personal liability from the predicted workings of a nonpersonal tax.
This distinguishing feature of all indirect taxes may be illustrated by an elementary comparison between the corporate and the personal income tax. Under the latter, the individual varies his own tax liability, in tax-price terms, by varying the amount of taxable income that he earns. His own ability to do so implies also that the tax-price he faces is indirectly dependent on the behavior of others who can react similarly. In acting to reduce the tax base, each taxpayer imposes an external diseconomy on all fellow taxpayers. To an extent, this sort of personal interdependence remains under the corporate tax. Individuals may, within limits, reduce the tax-price per unit of public goods through withdrawing resources from corporate investment. And any one person’s final tax obligation becomes reciprocally dependent on the activity of all other persons in making such allocative adjustments. In this respect, the two taxes differ only in the degrees of response. The additional factor that the corporate tax necessarily introduces is the “bridge” between the individual and the corporate entity. To become liable for tax, it is the corporation that must earn taxable income, not the individual. And to reduce its liability for the tax directly, the corporation must reduce taxable income. In order to estimate his own share, therefore, even apart from his own influence over aggregate investment in the corporate sector, the individual must predict how the corporation itself will behave in response to the tax. In other words, an additional decision-making entity is introduced between the individual and the fisc. A whole set of new predictions must be made concerning the decision-making processes of this in-between institution, the corporation, processes themselves involving most of the problems of group rather than individual decision-making.
As in other models, the central features are clarified by posing specifically a choice situation. Suppose that the individual must decide how to vote on a public spending proposal, with revenues to be raised exclusively from a tax on corporate net income. For now, assume away net resource shifts into and out of the corporate sector. For expositional simplicity, think of the proposal as one aimed at expanding national park facilities from the proceeds of the tax, with the rate of tax, which is proportional, to be residually determined after the community decision on budget is made. Should our reference person support or oppose this proposed extension in public-goods supply? He must, obviously, make some sort of estimate as to the tax-cost that such a proposal will involve. This estimate will depend on the amount of net income that he predicts for the corporation (or set of corporations) that he owns, whether or not the corporate responses to the tax are considered. Hence he must predict the behavior of others than himself, even if the necessary interdependence among separate taxpaying units is left wholly out of account.
If his investment in corporate ownership is relatively favorable, the individual will find himself paying a relatively high tax-price per unit of public goods that are available to him. If, by contrast, his investment is relatively unsuccessful, he will find that he obtains collectively supplied goods at “bargain tax-prices.” The discrimination among individuals in actual tax-prices paid for public goods will vary directly with the rate of yield of their corporate portfolios. This relationship should yield testable hypotheses concerning individual behavior in demanding public goods and services. If such goods are characterized by positive income elasticities, which seems empirically descriptive, the individual should,
ceteris paribus, demand a larger quantity of public goods under the corporate income tax than under the comparable personal income tax, which is, of course, levied directly on less residual components of income. Actual testing of such an hypothesis would, of course, be extremely difficult, due to the necessity for cutting through the maze of information and uncertainty differences confronted in the two institutions. A second conceptually testable hypothesis is that, given the institution of proportional taxation of corporate income, individuals whose portfolios embody relatively greater “riskiness” will tend,
ceteris paribus, to demand a somewhat larger public-goods outlay than those whose portfolios exhibit less “riskiness.” Both of the hypotheses here are derived from an analysis similar to that which was first made familiar to fiscal scholars by Domar and Musgrave,
*13 in their discussion of corporate income taxation and risk-taking. In the context of this study, the proportional tax on corporate income makes the individual shareholder’s “purchase” of public goods into a risky venture.
To this point we have left distributional considerations out of account. This is wholly unrealistic in regard to corporate income taxation since one of the essential features of this tax is its lack of generality. The tax is necessarily discriminatory, and this in turn implies that the position of the individual in the economic process must be considered before his reactions to fiscal choice proposals can be predicted. This, in turn, requires that we develop more specific models for corporate tax incidence. If, as was assumed above, the final incidence rests largely with stockholders, nonholders will tend to approve all extensions of spending so long as the incremental benefits are expected to be positive. Different assumptions as to final incidence, or, more importantly, as to standard attitudes about incidence, will produce different results. Beyond this, once distributional considerations are raised, models for group choice are required for any predictions about choice behavior. While these extensions are required to make the analysis complete, at the elementary and exploratory level of this study they will be left aside. The problems of fiscal choice confronted by the voter-taxpayer-beneficiary under corporate income taxation, even in the simplest of “representative” man models, are sufficiently difficult to suggest those that might arise in still more complex settings.
General Sales Taxation
As a second major institution of indirect taxation, we shall examine
general sales taxes. Specifically, let us look at a flat-rate, or proportional, tax levied on the value of all goods and services sold at retail in private markets. The model could, of course, be readily modified to allow for specific exemptions or for imposition at different stages. We retain the essential features of the previous tax models discussed. We consider the tax to be newly imposed, and to finance a single good.
Similar to the tax on corporate income, some of the difficulties that any individual voter-taxpayer must face as he tries to decide rationally on the quantity of public goods that he prefers are illustrated by the disagreements, even among the experts, as to the actual incidence of this tax. If fiscal economists, who have specialized in the theory of incidence, are not agreed on just who does, in fact, “pay for” the public goods that are purchased with revenues from taxes levied on general sales, how can individual choice be made under anything other than gross uncertainty?
The individual should recognize that the tax drives a wedge between consumer-goods prices and productive-service prices. Relative to final product price levels, factor prices must fall, and, consequently, incomes earned from the sale of factors will be reduced in real purchasing power, regardless of monetary adjustments. In a perfectly working competitive economy, the effects of the general sales tax should not differ greatly from those of a proportional tax on personal incomes or on personal consumption expenditures, depending on whether or not investment goods are included or excluded from the tax base. The prospective taxpayer may even recognize all this in some proximate way. But it is useful to recall that the competitive model of market process is designed for explaining
general patterns of effect. The model is not especially helpful to the individual (even he who understands it) who lives in the real-world economy, and who must decide how to vote on spending proposals, given sales-tax financing. For this choice, the tax is not similar to either the proportional income or the expenditure tax. Under either of the latter, the individual can estimate with reasonable accuracy the base upon which his tax liability will be computed. Also, since these taxes are personal, he can make his own decisions concerning adjustments to their imposition. These steps become immensely more complex under indirect tax institutions.
He will recognize that he will not, personally, be required to pay out funds to the fisc in “exchange” for public goods. Revenues are collected only from sellers. Only if the individual should serve in some functional capacity as a retailer will he be conscious of the direct fiscal transfer. The individual who does not serve in such capacity must try to estimate the differences in his market opportunities before and after the tax. As suggested, he may accept the hypothesis that factor prices will fall relative to product prices. However, this general effect of the tax will never be uniform over all markets either functionally, spatially, or temporally. Recognizing this, the individual must try, as best he can, to predict the effects of the tax on his own income shares, in real value terms. These effects will depend upon the particular supply conditions characterizing the markets for his own productive services and upon the organization of the industry utilizing these services, among many other things. What the individual must predict here are the behavioral responses of many decision-making units in the economy, other than himself. The interpersonal interdependence, the externality, that was shown to be significant even under proportional income taxation, becomes enormously complex under general sales taxation. The behavior of other persons and firms, not only in earning income, but in apportioning resources, in pricing products and services, in purchasing final output, in adjusting to price changes, is necessarily relevant to the effects of the tax on the individual. At best, predictions amount to no more than rather inaccurate “guesses.” Investment in knowledge must surely stop far short of even the economist’s level of prediction. The range of uncertainty that must face the individual when he makes a final fiscal decision must be extremely wide, and most persons are likely to rely on very crude rules-of-thumb, perhaps made available to them through press media and stated in very simple averages.
Despite all of the difficulties involved, the individual must, nonetheless, choose (or acquiesce in the choices made for him by others). A demand or marginal evaluation schedule for the public good may be derived in a reasonably straightforward manner, since presumably the individual can make some rough estimate of the benefits he secures. On the tax or cost side, however, he may either grossly underestimate or grossly overestimate the tax-price that this institution imposes on him. No particular direction of bias seems indicated by this analysis. Relative to the model of invariant tax-price, the individual under sales taxation may choose more or less public spending. At a later stage of discussion, when the possibility of fiscal illusion is introduced, this conclusion will be re-examined.
Specific Excise Taxation
The remaining important institution of indirect taxation is that of partial or discriminatory excise taxation. Many real-world systems of excise taxes, which levy charges on the sale, use, or consumption of several products or groups of products, combine elements of general sales taxation, considered above, and specific or partial excise taxation. For present purposes, it is sufficient to consider only the polar models.
We examine here the behavior of the single utility-maximizing individual as he confronts the financing of a new public good from the proceeds of a tax to be levied on one commodity only. How will he estimate the tax-price that collective supply of the public good will impose upon him? Under this model of clearly discriminatory taxation, it is more difficult to leave aside differential impacts on separate persons and groups, but we may commence the analysis by neglecting this aspect, even here. We may do so by supposing, initially, that all members of the group purchase and consume the single product that is to be taxed, say, whiskey or tobacco, and that the differential patterns of consumption are not significant enough to generate widely different patterns of response.
In such a model, as in each of the indirect tax models especially, the tax-price that the individual must pay for a unit of the public goods depends directly on the behavior of other persons in an exceedingly complex chain of economic interdependence. To estimate this tax-price, the individual must predict the reactions of those whom the legislature makes initially responsible for payment. The behavior of retailing firms must be predicted, along with the responses of resource suppliers and product demanders in the aggregate. The difficulties in making accurate predictions are evident, but it should also be noted that, precisely because of the selectivity of the tax, these difficulties are not so great as those encountered under either of the two institutions previously examined in this chapter. Textbook economics makes this point. The primary adjustments to be predicted take place via increases in the prices of the taxed commodity. Adjustments in factor prices generally, while predictable to a degree, normally assume quite secondary significance. Naïve predictions made by the potential taxpayer to the effect that commodity prices will increase by the amount of the expected tax per unit will not be wildly in error if markets are reasonably competitive, if resources are not highly specialized, and if time is allowed for supply adjustments. This naïve prediction enables the potential taxpayer to make some predictions of his own about responding to the tax. The interdependence among all taxpayers with regard to the aggregate base of tax remains, but insofar as all persons are predicted to act similarly, tax-price can be estimated with some accuracy, at least as compared with alternative tax institutions.
When differential responses among individuals and groups are anticipated, the estimation of individualized tax-cost is subject to significantly greater uncertainty. Here the individual must examine his own demand for the taxed commodity relative to that of his fellows. It becomes obvious that nonconsumers, along with consumers who can themselves respond most effectively to the tax-induced price increase, will tend to secure “bargains.”
The concentration of attention on individual behavior in public or collective choice as opposed to individual behavior in private or market choice should again be emphasized. Our concern is with the quantity of public goods to be supplied. In the model where partial excise taxation is the financing device, it becomes especially tempting to say that the individual’s behavior in market choice is a part of his “collective” decision. This would suggest that, when he purchases a unit of the taxed commodity, say, a bottle of whiskey, he does so with the knowledge that he is buying a package that includes two components, the whiskey that is directly utilized along with the public goods that are to be financed with the proceeds of the tax. Such a tie-in model is misleading, however, since the individual will extend his purchases of the privately consumed commodity, whiskey in this example, to the point where his marginal evaluation of this alone equals the marginal price, including tax. The fact that the public goods financed by the tax are also valued by the individual has no effect on his margin of choice for the private good. There is no way in which the individual can adjust the margin of provision of the public good through his market behavior. This choice arises only when the individual participates, not as an independently acting purchaser-consumer, but as a voter-taxpayer-beneficiary.
Summary and Conclusion
In this chapter and the one preceding, some of the familiar tax institutions have been examined in an attempt to determine their relative effects on the information-uncertainty elements that must enter into any individual’s efforts to estimate the costs of public goods. The institutions have not been analyzed in detail, and the many sub-models that might be introduced under each broad category have not been explored, although some of these may prove sufficiently unique to warrant special treatment. Several conclusions may, however, be drawn even from the limited analysis.
The model of tax-price invariance assumes a position all its own, as does its familiar analogue, the lump-sum tax, in the more orthodox tax theory based on the usage of Pareto-efficiency criteria. The approach of this study is, of course, closely related to the welfare analysis of tax institutions, but the differences should be kept in mind. No attempt is made here to array tax institutions in terms of economic efficiency, as such. The invariant tax-price is unique for our approach, not because it exerts no influence on individual behavior in market choice, the traditional requirement for the absence of an “excess burden,” but because only the absence of such influence enables the individual to choose fiscally on the basis of a well-informed comparison of alternatives.
By and large, those tax institutions that have been shown by the traditional welfare analysis to generate relatively less “excess burden” will be the same institutions that allow the individual to choose relatively more rationally as a participant in collective choice processes. There are exceptions to this rule, however, as is evidenced by the partial or discriminatory excise levy. Traditional welfare analysis suggests that this tax tends to distort the choice pattern of the consumer of private goods to a greater extent than a more general excise tax. As the above analysis has indicated, however, the individual may be able to choose a preferred quantity of public goods upon a more rational consideration of alternatives here than under a more general tax. He may be able to do so precisely because the discriminatory nature of the tax makes the effects and incidence more certain than those of the more general levy. Hence, “efficiency” in fiscal choice, which depends on the prospects for informed decisions by individual participants, may require greater distortions in market choices if the result is greater predictability. The tax on corporate income provides an even more dramatic illustration. If, in fact, this tax could be levied on pure economic profit, there are no short-run effects on market behavior of individuals or firms. The necessary conditions for Pareto optimality are not modified by the tax. However, the analysis has shown that even such a tax would introduce major elements of uncertainty in the fiscal choice problem confronted by the individual, and because of this the tax surely generates “inefficiency” in the final selection of some most preferred mix between private goods and public goods.
One significant difference between the results derived from the fiscal choice approach and those derived from orthodox welfare analysis involves the theory of the second-best. In its various forms, this latter theory states that it is not possible to judge a single distortion as nonoptimal, on Pareto-efficiency grounds, until and unless there is some assurance that there exist no other violations of the necessary marginal conditions for optimality. Hence, even the lump-sum tax (or, in our models, tax-price invariance) cannot necessarily be predicted to generate greater over-all efficiency than other taxes of comparable magnitude. This theorem is correct, within certain limitations, when a global view of “efficiency” is taken. In the analysis of this study, by contrast, the invariant tax-price unequivocally allows for a more “efficient” fiscal choice than comparable institutions. Only under this institution can the individual participant in collective choice predict the results of group action on his own economic position with any degree of accuracy.
The discussion of tax institutions in Chapters 3 and 4 has as a central feature the interpersonal interdependence that the two-sidedness of the fiscal system necessarily introduces. On several occasions reference has been made explicitly to the “externalities” inherent in individual responses to tax imposition. This suggests that a more formal analysis could be developed within the “externality” terminology that is familiar to theoretical welfare economists.
Quarterly Journal of Economics, LVIII (May, 1944). Reprinted in
Readings in the Economics of Taxation, ed. R. A. Musgrave and C. Shoup (Homewood: Richard D. Irwin, 1959), pp. 493-524.
Southern Economic Journal, XXIII (July, 1966), 35-42.