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
Fiscal Nihilism and Beyond
This book is an attempt to develop, in a preliminary fashion, parts of a
theory of fiscal choice. The central presumption is that individuals do make fiscal choices through their participation in political process. If the potential taxpayer-beneficiary has no part in choosing either the private goods-public goods mix or the institutions through which he pays for and enjoys public goods, there is little purpose served in any analysis of the feedback effects of such institutions on his behavior. The traditionalist moves from analysis to prescription without necessary recourse to individual preferences. He sets up criteria for fiscal reform without asking how individuals themselves make fiscal choices. Since he must presume that individuals have little or no power of ultimate choice, resort to extra-individual, external norms becomes acceptable, indeed essential, if anything at all is to be said.
This is intended to be an indictment of orthodox scholarship in public finance, but not to be an undue criticism of practicing scholars. Within the tradition, effective research has been accomplished, and the frontiers of knowledge have been pushed into continuing retreat. But what is most urgently required is precisely a shift out of this tradition, out of the mainstream. Essentially the orthodox tradition is nondemocratic, with no emotive significance intended. Decisions for the polity must be made exogenously to the individual citizen and coercively imposed upon him.
If political order is presumed to be workably democratic, individuals must be presumed to participate variously in the making of fiscal choices. They may, of course, do so quite indirectly and at times almost unconsciously, but their behavior becomes a proper subject of scientific inquiry. The awesome gap in our knowledge is apparent here. We need to know much more about how individuals behave in collective decision processes, and we need to know more about the workings of those institutions that transmit and translate individual preferences into collective outcomes.
But what about norms? Where are “principles” to be found? What are the criteria for fiscal reform? Should A or B be chosen? Does the model of individualistic fiscal choice simply ignore such questions, or does it point to its own prescriptions? Is fiscal nihilism the ultimate outcome? Does the approach produce effective criticism of long-established norms while replacing these with none of its own making? The traditional objectives of equity and efficiency may be shown wanting, but they have provided a frame for discussion. What is proposed or implied in their stead?
The institutional-choice analysis has suggested a partial answer to such questions. Hopefully, such an analysis of fiscal choice processes can provide a basis for laying down criteria for reform. But what will these criteria be like? If individuals are presumed to choose for themselves, how can analysis do other than examine choice behavior and attempt to predict the outcome? To “improve” choices here must the specialist become a moralist who preaches a new choice ethic? Perhaps the answer is implied in the question. Improvements in individual choice behavior can result from positive analysis. Fiscal theorizing at this level has, as its ultimate purpose, objectives that are analogous to those that guide “consumers’ research,” “operations research,” or “systems analysis.” The ultimate choice-makers, whoever these may be, can make “better” decisions to the extent that they are made more fully informed as to the alternatives which they confront. Analysis has as its purpose the clarification of the various alternatives, the prediction of the consequences of the separate lines of action.
The Theory of Incidence
Properly interpreted, the whole of the theory of fiscal incidence can be incorporated in the fiscal choice approach. Surely it is equally appropriate for the theorist to assist, ultimately, in the choice-making of individual citizens and in the presumed choice-making of some ruling authority. For the bulk of the work on incidence theory, the underlying political framework remains essentially unimportant.
The student of fiscal incidence and effects does not inquire about and is not concerned about the origin of or the selection among the alternatives that he analyzes. His task is that of predicting the comparative effects of different fiscal devices, real or imagined. He examines individual market responses to imposed fiscal phenomena, and he traces the primary, secondary, and tertiary stages of such responses to a point where final patterns of effects can be isolated.
Even within incidence theory, there remain gaps in the traditional analysis that have gone largely unnoticed because of the underlying political framework. The specific objective of incidence analysis has been that of predicting the
real effects of alternative fiscal devices, of locating the
real pattern of final burden of taxes and benefits from public spending. Who does pay the taxes? Who does enjoy the benefits? These are important and relevant questions that should interest the fiscal decision-maker, whoever he may be. There are, however, less apparent but nonetheless significant questions that should also be asked, and, if possible, answered. Who
thinks that he pays the taxes? Who
thinks that he enjoys the benefits?
Incidence theory has largely ignored these latter questions. To an extent, this neglect is explained by the fact that scholars have been economists, not psychologists. And as economists they have properly concentrated on real, rather than apparent or illusory, values. This apart, however, they have been uninterested in individuals’ attitudes toward fiscal devices or instruments, as such. The emphasis has been, on the one hand, in predicting the allocational responses to fiscal changes, and, on the other, in determining the real pattern of final effects. The policy objectives that have been implicit in traditional scholarship, those of economic or allocational efficiency and distributional equity, have in this way exerted feedback effects on even the most positivistic elements in incidence analysis. The theorist who has operated within the orthodox allocational framework has been interested in predicting how an individual will respond in the marketplace when the retail price of a final product is increased due to the imposition of an excise tax. He has been unconcerned, or relatively so, about whether, in making this response, the individual attributes the price change to the tax or to any of the many other possible causal factors. The same theorist, who may have had implications for distributional equity in mind, has also been interested in imputing directly to the individual consumer a final share, in either relative or absolute terms, of the net burden that a tax embodies. How much does the individual really pay, absolutely or in proportion to some income-wealth base? Implicit here lies the presumption that the “social welfare function,” the preference function for the “chooser” for the group, incorporates somehow the real pattern of incidence rather than any apparent or consciously realized pattern.
It is evident that questions about the individual’s attitude toward the fisc, toward taxes and benefits, become important either in an explicitly defined ruling-class, elitist model of politics or in an individual-choice, democratic model. How conscious are taxpayers of the burdens involved in the costs of public services? How conscious are beneficiaries of the values of public goods? Such questions as these become vital in any model that presumes that individuals make their own fiscal choices, directly or indirectly. The whole problem of fiscal consciousness is relevant for fiscal choice, and, in one sense, real burdens and real benefits become important only to the extent that they are effectively translated by individuals into “felt” or “consciously realized” burdens and benefits.
If viewed in this perspective, the discussion contained in Part I of this book can be treated as an extension of incidence theory. Analysis there was aimed at predicting the effects of various fiscal instruments on individual choice behavior in political processes.
The Theory of Public Goods
Traditional public finance theory has been concerned primarily with individual choices in response to imposed fiscal conditions. In this book, we have discussed two additional levels of individual choice behavior which, combined, provide the elements of specifically
fiscal choice. There is what we have variously called day-to-day, in-period, ordinary, operational, or budgetary fiscal choices. By these descriptive terms we have meant simply that, under any institutional setting, individuals will exercise their powers of decision and select somehow among alternative possible outcomes. Given any conceivable tax structure, and given any conceivable rule for amalgamating separate individual choices into a group decision, a specific set of public goods and services will be financed, purchased, and supplied. Apart from this level of choice, and in one sense “superior” to it, there is the stage or level where the institutional structure itself is selected. This level of choice has been the subject of attention in Part II where it has been suggested that many fiscal instruments can best be analyzed institutionally.
The operational level or stage of fiscal choice has been examined only indirectly in this whole book. The analysis of Part I was aimed at developing certain predictions about the influence of various institutions on this choice behavior of individuals, but the discussion did not contain the process of choice itself. Quite apart from the universal problem of space and time limits, there are reasons for this relative neglect. In the first place, the formal theory is quite complex, and many elements remain to be perfected. Secondly, and more importantly, the modern theory of public expenditure, which is surely the most exciting recent work in public finance literature, can be brought within the over-all framework of this study without difficulty. Shorn of its occasional “social welfare function” overtones, which become both unnecessary and impossible in an individualistic model of political order, this modern theoretical construction may be interpreted in such a way as to allow predictions to be made about the outcomes that will tend to emerge from the complex interplay of individual preferences as these are expressed through collective decision-making processes.
In its standard formulation,
*104 this theory of public-goods supply is explicitly normative. It purports to lay down the necessary marginal conditions that should be met if economic resources are to be allocated optimally in the public sector. Optimality or efficiency in resource use is defined in the Paretian sense, and a single optimum point or position (any one from among an infinite number of such points or positions) is defined as one from which no change can be made without harming at least one person in the relevant group. The necessary marginal conditions that must characterize such a position are defined without reference to nonindividual norms and also without reference to the political or institutional processes that might produce such an outcome. The standard discussion stresses that such optimal outcomes cannot, in fact, be predicted to emerge from the private or independent behavior of individuals, analogous to that pressure toward optimality which does characterize behavior in market interactions. Individuals will rationally behave as “free riders” in trying to enjoy public goods and services; as a result they will tend to find themselves caught in a “many-person prisoners’ dilemma.”
Given this widely accepted and explicitly normative version of the theory of public goods, how may it be transformed so as to enable us to predict the characteristics of the outcomes that will emerge from actual political processes? To construct this bridge between the formally correct and abstract normative theory of public-goods supply (which is derived from the theoretical welfare economics that owes its origins to Vilfredo Pareto), it is necessary to go back to one of Pareto’s own contemporaries, Knut Wicksell. In any over-all evaluation of the history of fiscal thought, Wicksell alone commands the heights of genius. He worked independently from Pareto, of course, and his own discussion of the “principles” for fiscal organization seem, at first glance, quite different from the formal statements of necessary marginal conditions that we associate with Paretian welfare theory. Wicksell was equally the armchair theorist, but he framed his whole discussion of fiscal choice in terms of political institutions, in terms of the processes through which individual preferences are translated into collective or group decisions.
Wicksell suggested that the unanimous consent of all parties should be the criterion for decisions on fiscal matters. Although it was developed independently, it is evident that this criterion is the political counterpart of the Pareto criterion for optimality. If, from a given position, no change can be made through general agreement among all parties, the initial position may be classified as one belonging to the optimal or efficient set. On the other hand, if a change is proposed and all members of the group agree to this change, the initial position is nonoptimal. Wicksell’s discussion contains specific institutional suggestions for implementing the rule of unanimity in the reaching of fiscal decisions.
In this Wicksell variant, the theory does become a theory of fiscal choice in a positive sense. If an institutional rule is imposed to the effect that all fiscal decisions, all taxing-spending decisions, must be made only after the unanimous agreement among all parties, the necessary conditions for optimality, defined in the Paretian sense, will characterize the outcomes that tend to emerge from the collective choice process. The only qualification that need be placed on this general proposition is that choices must be made marginally or in small steps. The theory of fiscal choice, so interpreted, does not, of course, allow us to predict
what outcomes will tend to emerge. The Pareto surface contains an infinite number of optimal positions or points, and, at each stage of the journey toward this surface, the division of the “gains from trade” among persons will tend to restrict the size of the finally attainable set. The theory enables us only to define the characteristics of the solution, not to specify the elements contained within it. In this sense, the theory of fiscal choice is wholly analogous to the “theory of consumer’s choice” which is a standard part of the economist’s equipment.
While Wicksell does provide us with a bridge between the normative theory of “optimal resource allocation” and the positive predictions that may be desired in an individualistic model, the severe restrictions that his institutional constraints impose on individual behavior in collective choice must be acknowledged. Under a genuine rule of unanimity, individuals will be led to invest resources in strategic bargaining, investment which will, in the net, prove wasteful to the group as a whole. This type of individual behavior is not the same as the “free rider” sort which would characterize individual attitudes toward voluntary contributions for public goods. Under unanimity, some agreement might ultimately be reached at each stage on the way to a final outcome, but serious resource wastage might occur, the most important element of which would be measured in the costs of delaying agreement. Decision-making in groups, bargaining, is a costly process at best, and costs may become prohibitively high under a rule of unanimity, despite the acknowledged relevance of this rule, and this rule alone, for guaranteeing that action taken is, indeed, of net value for the group.
Wicksell sensed the problem here in his expressed willingness to allow for some relaxation of the institutional rule of unanimity, and in his specific proposal for a qualified legislative majority—although he left the precise size of his majority ambiguous. If the rule of unanimity is relaxed, the single participant in group choice cannot proceed on the assumption that his own agreement is required for collective decision. He will be much less inclined to invest resources in bargaining tactics. Decision-making costs are reduced dramatically. At the same time, however, any departure from the strict unanimity requirement means that inefficient or nonoptimal outcomes may emerge. The final result of the collective decision process need not be Pareto optimal; the necessary marginal conditions need not be satisfied.
What is suggested is some balancing off of the two sides of the account, some comparison of the costs of inefficient or nonoptimal outcomes with the reductions in costs (benefits) that are expected to arise from the facilitation of decision-making. This is essentially the comparison that Gordon Tullock and I discussed at some length in
The Calculus of Consent, although the analysis there was not confined to fiscal choice.
Once this step is taken, the theory moves beyond the operational choice level into considerations of institutional-constitutional choice, the level or stage discussed in Part II of this book. Through some calculus of comparing costs, it becomes possible to discuss optimal rules and institutions within which choices are to be made, choices which are, themselves, predicted to produce outcomes or solutions that are not always located on the standard Pareto surface. What becomes conceptually predictable under this theory is not the characteristics of particular outcomes, but, instead, the general features of a whole probability distribution of outcomes. We shall return to a further discussion of this theory of institutional choice in a later section. Before this, however, it will be useful to return to the level of in-period budgetary choices. The discussion of the possibilities of developing positive theories of fiscal outcomes has not yet been exhausted.
Fiscal Choice Under Fixed Institutions
At any moment in time, some political “constitution” exists that specifies the manner in which collective decisions, including fiscal decisions, shall be reached. This structure may be described in detail only by the complex rules and procedures governing the whole set of political institutions. This very complexity makes it incumbent on the theorist to abstract the essential elements of the structure, to simplify, and to construct models of political choice-making. With these models, he may then try to predict the characteristics of the outcomes that will emerge. Any realistic model will, of course, incorporate a political decision rule that requires the assent of less than all members of the group. One such model is simple majority voting, a model that was introduced and discussed in Chapter 11. Under such an operative rule for reaching group decisions, what characteristics of final outcomes can be specified? Some analytically meaningful results can result from attempts to answer this question; the literature on the solutions to majority-rule games and on majority-coalition formation is relevant and important. Somewhat more restriction may be placed on the analysis of fiscal outcomes if additional constitutional constraints are imposed on the models. In addition to the majority-voting rule for making political choices, it is possible to fix the institution under which taxes are to be paid, through which public goods and services must be purchased. Through this dual set of institutional-constitutional restrictions, the outcomes of the fiscal choice process may be somewhat more narrowly circumscribed and the analysis made somewhat less general than in unconstrained majority-rule models. Only in Chapter 11 has this sort of theorizing been attempted in this book. The exploratory efforts there are presented more or less as lead-ins for further possible research. Both the rules for making political decisions and the institutions through which fiscal outcomes are produced are subject to wide variations, even within the framework of any existing political-fiscal order.
What results are to be expected from such theorizing? No model that allows for genuine individual choice can predict the precise outcomes that will emerge from a decision process, whether this be the private choice of a single person or the collective choice of a group of persons. The economist, the theorist of consumer’s choice, cannot predict the mix of goods that a particular housewife will purchase in the market. Similarly, the fiscal theorist cannot predict the particular mix of public goods that will be chosen by a community of persons. But it may be useful to extend this comparison with the theory of consumer’s choice somewhat further here. As suggested above, only under the somewhat rarified institutional assumptions imposed in the Wicksellian model can the outcome of the fiscal choice process be described by the familiar equalities among marginal rates of substitution. Under almost any remotely relevant institutional restrictions, the outcomes will tend to be nonoptimal in the Pareto sense. If the limits of theory are exhausted with the classification of particular outcomes into nonoptimal and optimal sets, there would be little purpose in the analysis of differing institutional structures. Something more than mere classification of outcomes within the nonoptimal set can be made. The various institutional combinations can be arrayed in terms of the predicted degree of “nonoptimality” of the outcomes that they are expected to produce over a whole sequence of separate decisions.
Our attention in this book is concentrated primarily on fiscal institutions, not on the institutions of political decision-making. The procedure suggested, therefore, is that of attempting to array alternative fiscal arrangements under each possible political decision structure. As an illustration, refer to the models introduced in Chapter 11. Assume the presence of simple majority voting for reaching all political decisions. The next step is that of comparing predicted outcomes under separate and alternative fiscal institutions. Compare, for example, the outcomes to be predicted under a head tax with those under a proportional income tax. Which of these series of outcomes seems to be “preferred” on efficiency criteria? The Pareto criteria can serve as the benchmarks from which possible departures are measured.
It is useful to recall that the choice of a tax institution can serve as a substitute for a decision-making institution and vice versa. Conceptually, in decisions on the appropriate quantity of a single public good, there will always exist some tax institution which will produce “optimal” outcomes, under
any and all rules for reaching a collective decision. The more “efficient” the tax institution is in this sense, the less “inefficient” will be any given departure from unanimity in the political decision structure. This point was illustrated in some of the models developed and discussed in Chapter 11. If the tax institution should be such that each person is obligated to pay for public goods so that tax-prices equal the schedule of marginal benefits, any conceivable decision rule will yield the Pareto-optimal quantity of public goods. The fact that such a tax institution always exists conceptually does not, of course, imply that it can be determined independently of the revealed choices of individuals themselves. If an omniscient observer should be present, and if he were asked to “read” all individual preference maps, he could then describe the “optimal” structure of tax-prices. Failing this, there is no means of ascertaining with any degree of accuracy the “efficient” tax structure or institution.
If the tax institution is not the “efficient” one, either because its selection cannot be made independently, or because nonefficiency criteria are also relevant, then the political decision rule can be important in determining the degree of efficiency in the outcomes that emerge. For example, if the tax rule states that all persons must pay equal taxes, then the delegation of political decision-making power to a single person produces less inefficiency than such a delegation would produce under no such tax restriction. The dictator’s possible exploitation of his fellows is reduced by the requirement that he, like his fellows, must pay a share of the total tax load. Since, in the normal order of events, the tax institutions in existence will not approximate those that are “efficient,” analysis must consider carefully the effects upon outcomes produced by various political decision rules. In this analysis no simple conclusions can be reached by trying to array alternative political institutions under separate fiscal arrangements. For example, suppose that the constitution dictates that all public goods shall be financed by equal taxes on all persons. It does not follow at all that the “efficient” decision rule, that of unanimity, will produce “optimal” results for any and all tax allocations. It seems obvious that unanimity in this case of equal taxes might be one of the worst of rules for reaching group decisions. There may exist some “efficient” decision rule in a regime of poll taxes, which an omniscient observer could specify, but it becomes extremely difficult to think of meaningful procedures through which such a rule could be independently discovered. For this reason, it seems preferable to consider the political decision rules as being, somehow, less subject to deliberate variation than the fiscal institutions.
The suggestion was made above that various fiscal institutions could be arrayed or ranked in terms of their predicted ability to produce “efficient” outcomes, these being defined in the standard Pareto fashion. This raises the whole question of norms once more. Is the suggested procedure not equivalent to reintroducing the economist’s normative standard? If so, what has become of the model in which individuals are simply observed to choose what they will?
There is no paradox here when the proper relationship between the criteria of efficiency and individual choices is recognized. What does the economist mean by an “optimum allocation of resources”? He really means that allocation which is produced by the uninhibited interplay of private individual choices and nothing more. The extension to the supply of public goods is straightforward. An “efficient” public-goods provision is that which would tend to emerge from the “ideal” institutions of individual-collective choice. It becomes appropriate, therefore, to discuss various institutions in terms of their predicted tendencies to promote or to prevent the attainment of such outcomes. No external ethical norms concerning the actual shapes of these outcomes need be introduced at all in order that some institutions may be called “better” than others, by efficiency criteria. An analogy may be helpful. We may say that a clear windshield is “better” for driving an automobile than a dirty windshield, without any reference to where the driver wants to go. Given any route, he will drive “more efficiently” if he is able to see where he is going. Similarly, we may say that certain institutions are “better” than others, quite independently of the outcomes that will be produced. Whatever these may be, they are reached more efficiently under some institutions than under others.
The incorporation of the traditional equity norms into the individualistic model cannot be accomplished so readily. Some of these issues will be discussed in a later section. First, however, there is more to be said concerning efficiency.
The Choice Among Fiscal Institutions
For what purpose does the analyst array the various fiscal institutions in the procedure suggested above? Ultimately at some higher-stage or higher-level “constitutional” choice, individuals themselves must select the set of fiscal institutions, rules, and regulations under which in-period budgetary choices shall be made. The vital distinction between fiscal choice under specified and preselected institutions and the choice among such institutions themselves cannot be overstressed, and a simple example may prove helpful even at the expense of repetition with earlier discussion. Consider a group of persons organized as a political community, and for simplicity think of the political constitution as having been fixed. But no fiscal constitution exists. The opportunity, or the necessity as the case may be, arises for some group outlay on a public good or service, say, defense against external enemies. The community, acting as a unit, must decide on how much of this public good to supply, and it must decide how the costs shall be distributed among the citizens. These two separate decisions are interdependent under any political choice rule. An individual’s behavior in voting for public goods will be influenced by his predictions as to the tax allocation. As we have previously noted, it is in this setting that the “free rider” problem emerges to complicate and to confound fiscal decision-making.
One means of sharply reducing the investment in strategy and of generating directness in individual response is for the group to reach some agreement on how the tax-costs shall be distributed among persons in advance of and independently of the decision on public-goods quantity. In our example, the community may approve a “fiscal constitution” even before any need or opportunity for spending on external defense is anticipated. This institutional or constitutional choice implies, of course, that some inefficiency in any final outcome as measured in public-goods supply must be predicted as highly probable. In the example here, if the fiscal constitution agreed upon dictates that all public goods, including defense, must be financed from head taxes imposed equally on all citizens, the specific supply of defense in any particular period may diverge considerably from that which would be “optimal.” Given sufficient investment in bargaining and discussion, some rearrangements of tax shares might be worked out that would enable the community to shift somewhat closer to the Paretian welfare surface in almost every particular case. When, however, it is recognized in advance that such rearrangements would have to be worked out for each separate public good or service supplied and in each separate time period, the costs of securing reasonably efficient outcomes may become prohibitive. Some structural adjustment in the direction of selecting tax allocations that determine individual cost shares (individual tax-prices) over a large basket of public goods and services and over a whole series of time periods may be individually and collectively rational.
Once this is recognized, the whole notion of “efficiency” is necessarily modified. An institution may well be “efficient,” even though it is recognized that “inefficient” outcomes will be produced through its operation. The central question in institutional choice is that of selecting the most “efficient” institution. What scheme or rule for collecting taxes is relatively most efficient when the public goods to be supplied from the tax revenues remain wholly unspecified? As noted in Chapter 14, this shift to institutional choice seems partially to rehabilitate traditional fiscal theory and to justify a consideration of “tax principles” independently from the expenditure side of the fiscal account. Efficiency here must be discussed in terms of the probability distribution of outcomes that a tax institution is predicted to produce over a series of separate time periods and over a series of different benefit imputations.
A New Approach to Fiscal Justice
The methodology for the analysis of institutional efficiency is drawn from several sources: the modern theory of statistical inference, the theory of games, the theory of political constitutions, and, also, recent philosophical discussions of “justice.”
*106 This latter discussion is especially relevant since it allows us to relate the institutional-choice approach to the traditional discussion of justice or equity that has occupied so much of the fiscal literature. The methodology that embodies as its characteristic feature a sharp differentiation between the outcomes of a choice process and the rules or institutions that generate such outcomes is, of course, wholly different from the traditional approach in fiscal theory.
How will an individual choose among the alternative institutions of taxation? In an idealized position of choice here, the individual is uncertain both as to his own share of the benefits that might accrue from the spending programs that may be adopted and as to his own economic position (upon which taxes would presumably be based). In such a situation, he must try, as best he can, to choose an institution that will work tolerably well under almost any set of circumstances. The analogy with the choosing of rules for an ordinary game of poker is a close one. The individual will try to select rules that seem to be “fair.” At this level of consideration, “fairness” and “efficiency” merge and come to mean the same thing. It seems also appropriate to use the word “justice” here, as Rawls has done in his discussion of ethical norms. In terms of some normative personal ethic, the individual “should” choose
as if he is in such an idealized position, even if he is not, and the criteria for his decision can be summarized as those of “justice.”
Our central concern is not, however, with the ethics of individual behavior, but rather with the prediction of behavior in institutional-choice situations. Normally individuals will not find themselves in the idealized conditions. A person will probably have some idea as to the pattern of benefit imputations over time, and, even more probably, he will have some idea as to his own economic position in future periods. Nevertheless, it is not unreasonable to suggest that uncertainty elements in both respects loom relatively important in his decision calculus, and to the extent that they do so, it is appropriate to examine the notions of both “efficiency” and “justice” in the framework discussed here.
Theoretical welfare economics enables us to define the necessary marginal conditions that must be satisfied for an allocation of economic resources to be efficient. Straightforward extension of this analysis to “theoretical institutional economics” should enable us to define a similar set of conditions that would have to be met if an institutional arrangement or rule is to be classified as “efficient.” It now seems quite possible that future developments will in fact allow for general statements of such conditions. At this time, however, we must be content with more ambiguous and less rigorous definitions. Analysis remains at the stage of examining various institutions under varying sets of assumptions, with criteria for efficiency being largely derived from introspection.
It was noted above that the criteria for “efficiency” and “justice” merge and become identical under the institutional-choice approach, at least in its idealized form. This is, of course, sharply at variance with public finance orthodoxy, where these two objectives are distinct. “Equity” or “justice” has been traditionally held to require the introduction of external ethical norms. A long-standing principle of normative public finance theory has been that “equals should be treated equally,” the principle that has been called one of “horizontal equity” by R. A. Musgrave.
*107 Corollary to this, there has been the principle of vertical equity: “Unequals should be treated unequally.” But to what degree? This has remained the central issue in normative tax theory, and it has been resolved only upon the introduction of some external value scale, some “social welfare function” that is defined by the observer. Since individuals have no part in the formation of this scale, except as the observer chooses to take their preferences into account, the conceptual task of the analyst is simply that of “reading off” the solution that best achieves the indicated equity objective.
In its most modern formulation, represented in Musgrave’s treatise, fiscal theory contains a paradox. The allocative function of the fiscal mechanism is sharply differentiated from the distributive function. In the former, individual choices are allowed to serve as the basic determinants of outcomes, at least in some normatively idealized sense. In the latter, however, resort to an external value scale is necessary. “Efficiency” criteria are derived from individual preferences; “equity” criteria are derived from external sources. Hence, efficiency and equity not only represent different and often conflicting objectives; they are also different philosophically, being derived ultimately from two quite distinct sets of values.
In the modified approach proposed here, these two sets of criteria become one, and both are derived from individual preferences. A fiscal institution that is efficient is also just, and vice versa, since these terms cannot be distinguished in the individual institutional-choice context. The individual who is presumed to be making a choice among alternative fiscal institutions does so on the grounds of his own utility-maximization. This insures that he will tend to select that institution or rule that he considers most efficient privately considered. But this institution will also tend to be that which is considered just for the simple reason that the individual cannot predict with accuracy his own position under the subsequent operation of the institution. He will be led to choose an institution that will treat him “fairly” or “justly” wherever he might find himself located.
To the extent that the individual’s actual choice position is not that defined in the idealized model, the efficient fiscal instrument for him will not be that which would be observed to fulfill criteria for “justice.” To the extent that the individual can predict with accuracy the future imputation of public benefits and/or his own income-wealth status, his utility-maximization will lead him to select institutions that will provide differential advantages. As previously noted, and as will be discussed again below, it is the recognition that this conflict will arise which makes the importance of the conditions for institutional choice so important and points the way toward proposals for reform aimed at shifting these conditions.
The Redistributive Function
The two approaches, the orthodox or traditional one and that which has been partially developed in this book, may be compared in their treatment of the income-wealth distribution problem and its relationship to the fiscal mechanism. As suggested, the standard treatment here explicitly invokes external norms. For example, Henry Simons accepted “greater income equality” among individuals as a social objective that the fiscal system “should” be organized to promote. Modern works call on some “social welfare function” to determine the single most-desired point from among the infinite set of the Pareto welfare surface.
This resort to external norms is eschewed in the individualistic model. Is it then possible to say anything at all about the redistributive function of the fiscal mechanism? There are two separate levels of response to this question which must be kept distinct, again illustrating the relevance and importance of the two stages of fiscal choice that have been emphasized. If an individual’s economic position is clearly identifiable, along with those of his fellows, and if a single one-period choice is confronted, he may, of course, choose to tax himself for the purpose of transferring income to those less fortunate than himself. In this sense, redistribution is a public good in the classical form, and there are evident externalities to be internalized by collectivizing the “consumption” of this good. Therefore, even in the purely individualistic model that is confined to single-period choice, some net redistribution would tend to be carried out by the fiscal system. The limits of this income transfer would be quite confined under normally expected circumstances, however, and this transfer would depend on the fact that individuals include the utility of others than themselves as arguments in their own utility functions.
Redistributional elements become much more important at the second or institutional level of fiscal choice. Under the idealized conditions, the individual cannot identify his own income-wealth prospects over time with accuracy. Hence the choice-maker should, rationally, act as if he confronts a probability distribution of possible income-wealth positions, and he should select that fiscal structure that maximizes expected utility. It is easy to see that the individual might under these conditions build in important elements of net income-wealth transfer, not because he pays any attention at all to the utility of his fellows, but simply because he wants to insure for himself a satisfactory post-fisc income level. There will surely exist some “optimal” degree of net redistribution, and this will tend to be considered in the individual’s choice of a fiscal constitution under almost any political decision rule. It is to be expected, therefore, that net redistribution will characterize the operations of the fiscal constitution over time.
*108 It is reasonable to suggest that this sort of calculus is helpful in “explaining” redistributional elements that are found in modern fiscal structures, perhaps even more helpful than the vaguely asserted notions of equalization presumed in the standard treatments.
Directions for Reform in Fiscal Structure
The analysis of this book has been basically positive. The purpose has been that of predicting the effects of specific institutions upon fiscal choice and of predicting the types of institutions that might be selected. The discussion of the underlying efficiency norm was presented through its derivation from individual preferences rather than in the more usual shorthand conception which seems to imply that the criterion is independently discovered. The analysis does point toward general normative conclusions, however, and it is appropriate that some of these be outlined briefly in this section.
Given the complexities of modern budgets and the large numbers of individuals who hold membership in most governmental jurisdictions, it takes little or no theorizing to suggest that any attempt to attain efficiency in the supplying of each and every public good and service in each and every fiscal period would be economic as well as political folly. Public goods must be supplied within the context of a fiscal constitution, which is described as a quasi-permanent and quite complex set of institutions and rules that specify what tax instruments are to be employed and when and how, when public debt is to be issued, how budgets are to be made, etc. What must be sought for, realistically, in any reform, are “improvements” in this fiscal constitution. How can those elements that seem to produce inefficient results be eliminated and more efficient elements substituted?
The standard procedure is for the expert to place himself in the role of constitution-maker and to discuss the drawing up of the “ideal fiscal structure.” This procedure remains possible, and the interpretation of justice-as-efficiency sketched above suggests that considerable material of interest might be developed in this fashion. Such an attempt would, however, probably be foredoomed to failure. No effectively democratic society would be disposed, nor should it be disposed, to turn over the remaking of its fiscal or political constitution to a single expert or to a body of experts.
Recognizing this, the student of the fiscal process can begin a less exciting but more productive task. He can try to suggest specific changes in the
ways in which individuals make constitutional choices rather than changes in the choices that “should” be made. He can suggest modifications in the structure of choice itself that may lead individuals, through the political decision procedures, to choose more efficiently among alternative fiscal instruments. As noted, individuals will tend to equate efficiency with justice, properly interpreted, if they are confronted with institutional-constitutional choice in its idealized setting. It is possible in many circumstances to suggest changes which have the effects of placing individuals closer to such situations. The first requirement is that all genuinely constitutional changes (whether they be called this explicitly is largely irrelevant) should be treated by individuals as quasi-permanent or long-run changes. The most important single improvement in the fiscal system might well be the introduction of specific lags between decision and implementation along with the requirement that decisions, once made, must remain in force over some minimal period of time.
An illustrative example is provided in inheritance and estate taxation. In the standard discussion of fiscal reforms, merely to raise issues concerning this tax is to choose sides. And the choosing is not difficult. Those persons who identify themselves with favorable asset positions tend to argue persuasively against increases in and for reductions in such taxes. Those other persons who cannot or do not make this identification argue, with equal persuasiveness, that these taxes should be made confiscatory. The collective decision process becomes strictly analogous to a zero-sum game, and no reasoned discussion of an efficient or optimal scheme or asset-transfer taxation can possibly take place.
How may this state of affairs be improved? Surely not by the various experts posing as authorities and invoking time-honored principles to support their own personal preferences. The inference to be drawn is that changes in asset-transfer taxation cannot be discussed dispassionately so long as these are discussed as
current changes. Therefore, the implication is that modifications in the structure of such taxes should, ideally, be discussed only with significant time lags between decision and action. Reasoned, and reasonable, discussion should be possible on the most efficient structure of asset-transfer taxation that would come into effect in, say, a quarter or a half-century after decision. Individuals who participate in the discussion on this basis will be unable to identify their own positions so clearly; their self-interest will be long-term. This proposal seems farfetched only because it has not been explicitly examined, although the familiar Rignano plan can be interpreted as a vague normative statement of the same idea.
*109 This is presented here only as a simple and single example of the sort of reforms that might be expected to improve the choice among fiscal institutions. No exhaustive discussion of such reforms will be made here, and none is found in this book.
Time is, of course, the element that converts an interclass, intergroup decision into a reasoned one on which general agreement becomes possible. Different time lags may be appropriate for different institutions. Time has rarely been treated as a variable by economists, yet it seems evident that the temporal characteristics of a decision can have a major impact on the manner in which individual decision-makers evaluate and choose among alternatives. Everything is variable in the long run, including the individual’s own economic position and the pattern of benefit imputation from public spending programs. The individual who may be quite eager to support a temporary one-year tax on new automobile purchases to finance a one-year subsidy to a world’s fair in his home city, may be quite reluctant to support a permanent tax on new car purchases to finance annual world’s fairs in a series of cities.
The Constitutional Attitude
The effective operation of democratic government, in its fiscal as well as its nonfiscal aspects, requires the adherence of its citizens to what may be called the “constitutional attitude.” Given the high cost of making collective decisions, government can function properly only if a large proportion of its day-to-day operations take place within a quasi-permanent constitutional structure. Individuals, and groups, must recognize the importance of constitutional-institutional continuity, and the dependence of democratic process on firm adherence to such continuity. If this is not recognized, and if individuals come to consider governmental processes as nothing more than available means through which separate coalitions can exploit each other, democracy cannot, and should not, survive. Fiscal institutions are a part of the political constitution, broadly considered, and especially in the sense noted here. Changes in the fiscal constitution must be treated as quasi-permanent and long-lasting features of the social structure. If individuals, and groups, including politicians, come to consider seriously the possibility of manipulating basic fiscal institutions for the accomplishment of short-run-purposes, bargaining elements will quickly swamp all efficiency considerations.
In the final analysis, “justice,” “efficiency,” “fairness,” whichever term is employed, can be expected with a reasonable degree of certainty only when individuals (or their representatives) are placed in the position of choosing for themselves, not as instant, momentary beings, but as a whole complex probability distribution of potentialities. To expect the poker player with a pat hand to agree to a new deal is to place entirely too much dependence on human ethics. The rules of the game, political or otherwise, may properly be drawn up only in advance of play, and by the players themselves. And, as play proceeds, rules should be changed to apply only to later rounds of play.
In fiscal theory, as in politics generally, scholars need to pay more attention to the working out of rules or institutions through which final outcomes emerge and less attention to the shape of these outcomes themselves although these must, of course, be relevant to an evaluation of the institutions. Improved allocations, or outcomes, can be achieved only through improvements in the institutions that generate them, and improvements in such institutions, in turn, can be achieved only if their proper role in the whole structure of democratic process is appreciated and understood. Perhaps more than their fellows, scholars themselves need to acquire a “constitutional attitude.”
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), 350-55; and Musgrave,
The Theory of Public Finance (New York: McGraw-Hill, 1959).
Finanztheoretische Untersuchungen (Jena: Gustav Fischer, 1896). The major portions of this work are translated as “A New Principle of Just Taxation,” in
Classics in the Theory of Public Finance, ed. R. A. Musgrave and A. T. Peacock (London: Macmillan, 1958), pp. 72-118.
Philosophical Review, LXVII (April, 1958), 164-94; “Constitutional Liberty and the Concept of Justice,”
Nomos VI, ed. C. Friedrich and J. Chapman (New York: Atherton Press, 1963); and, somewhat earlier, “Two Concepts of Rules,”
Philosophical Review, LXIV (January, 1955), 3-32.
The Theory of Public Finance.
The Calculus of Consent (Ann Arbor: University of Michigan Press, 1962), especially Chapter 13.
The Social Significance of the Inheritance Tax, trans. W. J. Shultz (New York: Alfred A. Knopf, 1924).