The Demand and Supply of Public Goods
By James M. Buchanan
First Pub. Date
Indianapolis, IN: Liberty Fund, Inc.
First published in 1968 by Rand McNally & Company. Foreword by Geoffrey Brennan.
The text of this edition is copyright ©: 1999 Liberty Fund, Inc. Picture of James M. Buchanan: File photo detail, courtesy Liberty Fund, Inc. James M. Buchanan, Charlottesville, Virginia, 1964.
- Ch. 1, A Methodological Introduction
- Ch. 2, Simple Exchange in a World of Equals
- Ch. 3, Simple Exchange in a World of Unequals
- Ch. 4, Pure and Impure Public Goods
- Ch. 5, Many Private Goods, Many Persons
- Ch. 7, The Publicness of Political Decisions
- Ch. 8, The Institutions of Fiscal Choice
- Ch. 9, Which Goods Should Be Public
- Ch. 10, Toward a Positive Theory of Public Finance
- Supplementary Reading Materials
A Methodological Introduction
People are observed to demand and to supply certain goods and services through market institutions. They are observed to demand and to supply other goods and services through political institutions. The first are called private goods; the second are called public goods.
Neoclassical economics provides a theory of the demand for and the supply of private goods. But what does “theory” mean in this context? This question can best be answered by examining the things that theory allows us to do. Explanation is the primary function of theory, here as everywhere else. For the private-goods world, economic theory enables us to take up the familiar questions: What goods and services shall be produced? How shall resources be organized to produce them? How shall final goods and services be distributed? Note, however, that theory here does not provide the basis for specific forecasts. Instead, it allows us to develop an explanation of the structure of the system, the inherent logical structure of the decision processes. With its help we understand and explain how such decisions get made, not what particular pattern of outcome is specifically chosen.
This process of explanation involves several stages. There is first a set of conjectural predictions, a set of basic behavioral hypotheses, or laws. These may be wholly conjectural, requiring the mental feat of constructing the pound of
ceteris paribus. On occasion, hypotheses may be derived that involve empirically testable implications, and when data can be assembled properly evidence may be adduced in corroboration or refutation. This strictly positive content of economic theory has, perhaps, been somewhat overemphasized in recent years to the partial neglect of theory’s more basic function. This is the development of the logical structure of an economy through the making of what may be called inferential predictions. The trained economist can predict the general shape or pattern which tends to emerge from the exchange or market process. These predictions are not of the conditional “if
B” variety, at least not in any directly analogous sense. Instead, these generalized predictions take the form, ”
A tends to equal
B.” The distinction here between elementary conditional predictions and inferential predictions has not been fully appreciated, perhaps because both are present in the central body of economic theory.
Conditional predictions take the form: If price falls, quantity demanded increases; if price increases, quantity supplied increases. All such conditional predictions, whether empirically verifiable or not, are combined to generate a logical structure for the whole system of behavioral interactions that we call the economy. To the extent that the conditional predictions in the set are valid, inferences may be drawn concerning the general characteristics of the outcomes that will emerge. These inferences are also predictions, and they are essentially descriptive in nature. They provide information about the relationships among variables: Prices will equal costs; wage rates for similar workers will be equalized; factors of production will earn their marginal product.
A vital link in the logical chain between conditional and inferential prediction has been deliberately omitted in the above sketch. Assume that the conditional hypotheses of the economist are valid. That is to say, the predicted behavioral responses are correct. Individuals will buy more goods when prices fall; firms will supply more goods when prices rise, etc. It is impossible to move from this knowledge directly to the statement that “prices will tend to equal costs,” until and unless we postulate something about the institutional-organizational structure within which individuals are allowed to make choices. Orthodox procedure in this respect has been that of explicitly or implicitly postulating competitive organization. Once this missing step is added, the inferences about results or outcomes follow logically from the set of conditional hypotheses. The descriptive characteristics of the results can be indicated.
In their most sophisticated form, these characteristics are presented as the familiar statements for the necessary marginal conditions for efficiency or optimality, the presumed domain of theoretical welfare economics. It is important to note that these conditions are inferential predictions and that they are positive in content, given that competition is postulated as the organizational structure. These conditions become conceptually refutable predictions about the descriptive characteristics of the results of the market interaction process. No normative elements need be introduced.
The weak step in this methodological procedure is the assumption that must be made about institutional-organizational structure. Only to the extent that this assumption is relevant will inferences be corroborated. As an example, consider the economist faced with predicting the effects of the 1965 excise tax reductions in a particular industry. Assume he predicts that prices will fall to a certain degree; his predictions are, we shall say, refuted by events. Does this refute the underlying conditional hypothesis that firms in the industry are profit-maximizers, or does it, instead, refute the hypothesis that the industry is competitively organized? Clearly, it may do either, or neither if still other relevant variables have changed. The standard procedure of assuming competitive order when this seems convenient is not acceptable. Appropriately thorough analysis should include an examination of the institutional structure itself in a predictive explanatory sense. The economist should not be content with postulating models and then working within such models. His task includes the derivation of the institutional order itself from the set of elementary behavioral hypotheses with which he commences. In this manner, genuine institutional economics becomes a significant and an important part of fundamental economic theory.
If human interaction is limited to voluntary exchange conceived in its broadest sense, a theory of institutional structure can be derived, yielding something closely akin to the standard model of competitive order as the end or equilibrium product. In other words, a somewhat loosely defined competitive economic organization can be predicted to emerge from the play of human interaction so long as this interaction is limited to voluntary exchange. Using nothing more than his standard tools, the economist can predict, first, the emergence of this structure, and, secondly, the characteristics of the outcomes that such a structure will tend to produce.
Only after this stage is reached can the economist begin to talk about the relationship between competition as an organizational structure, and efficiency. No criteria can be externally introduced. Efficiency becomes a descriptive term that is used to specify the existence of certain relationships among variables and among institutions which are produced through the process of voluntary exchange. The satisfaction of the necessary marginal conditions for efficiency, viewed in this light, becomes a prediction of results that will tend to emerge from the exchange process, not a criterion for telling us what should be present in order to further some externally derived value norm. The derivation of these necessary conditions, and of the institutional structures that will cause them to be satisfied from the choice processes of individuals engaging mutually in trade, is the central task of economic theory. When observed results appear to counter those predicted, either in terms of specific characteristics of outcomes or in terms of institutional structure, explanation of divergence becomes a supplementary and proper task. And analysis, here as elsewhere, must proceed simultaneously at several levels.
Theory of Public Economy
The extended methodological digression on the function of orthodox economic theory in application to the private economy is designed to provide some assistance in discussing the analogous role of theory as extended to the public economy, to the demand for and the supply of public as opposed to private goods. At base, the economist must begin from the same set of conditional hypotheses. He deals with the same individuals as decision-making units in both public and private choice, and, initially at least, he should proceed on the assumption that their fundamental laws of behavior are the same under the two sets of institutions. If he predicts that the average or representative person will purchase a greater quantity of private good
A when the relative price of
A is reduced, he should also predict that the same person will “purchase” a greater quantity of public good
B when the relative “price” of
B is lowered. This step in itself represents a significant departure from orthodoxy in public finance. Individual behavior patterns in demanding public goods, in participating in political decision processes, in voting, have not been examined in detail by economists (or by anyone else). A body of theory devoted to individual participation in voting processes is only now emerging. And even here, the individual’s behavior in demanding public goods, as some functional relationship between quantity demanded and the “tax-price” that he pays, has not been studied either analytically or empirically. Even more dramatic departures from public-finance orthodoxy are required, however, when inferences as to results are drawn. There is nothing analogous here to the competitive model, the use of which so greatly facilitates our elementary textbook predictions concerning the outcomes produced under voluntary exchange processes in the private-goods sector.
As suggested above, many economists have more or less jumped over the step of institutional theorizing in their analysis of markets, perhaps without fully realizing that they have done so. They are able to do this because the competitive-model assumptions yield predictions about outcomes that are not dramatically at variance with observation, tending thereby to corroborate both the assumptions and the conditional hypotheses. Despite all of the discussion about the unrealism of these assumptions, they remain paradigmatic for economists. Decisions on the demand-supply of public goods are made through political, not market, institutions, and there is no analogue to competitive order that eases the analytical task.
There are two possible ways along which the analyst might proceed. First, a specific political decision structure can be postulated and inferences made concerning the pattern of results that will emerge. Alternative models can be tried, and various differences in predictions noted. This approach has much to recommend it. However, nothing can be said about efficiency in this framework.
The second approach is that of making an attempt to derive the institutional structure from the broadly conceived exchange process. The economist can try to predict, as best he can, what sort of political decision structure will tend to emerge from the voluntary “political exchanges” that may be entered into by rational persons. Once this decision structure is derived, he may be able to characterize outcomes of actual processes in a manner that is analogous to his treatment of the private-goods sector. To a limited extent, the term “efficiency” may be introduced to describe certain outcomes, with this term having essentially the same meaning as that which applies in the private-goods world.
There must remain, however, an important difference in the degree of relevance that theory has in the two sectors. As Wicksell so perceptively noted, outcomes or results of individuals’ choices for public goods in discrete instances can only be classified unequivocally as efficient or optimal by some external observer if group decisions are made under some effectively operating rule of unanimity. For discrete allocations, political-choice institutions embodying decision by unanimity become the analogue to market-choice institutions that are described as perfectly competitive. In both cases, we are dealing with idealizations. For the latter, however, observed interactions seem to produce proximate realization, and the ideal commonly becomes, in one sense, the accepted norm for policy changes. That is to say, the institutions of the competitive market economy have been widely accepted to be desirable, over and above their place in the analysis which suggests that these describe the structure that would tend to emerge, ideally, from the free workings of voluntary exchange processes. Presumably the costs of achieving some approximation to the ideal here are not considered sufficiently high to warrant significant modifications in the norm, although some of the discussions of workable competition may be so interpreted. It is for this reason that efficiency conditions applicable to the private-goods economy have been widely understood as carrying important normative implications. And the very use of the emotive words “efficiency” and “optimality” tends, of course, to reinforce this interpretation.
I have suggested above that the familiar conditions need not embody such normative implications. Basically, they represent nothing more than inferences drawn from the set of hypotheses that make up economic theory, inferences that describe certain results that will tend to emerge from the interaction of many separate persons in voluntary exchange processes, including the institutions themselves as variables subject to choice. The drawing of such inferences, which are themselves predictions, remains within the scope of positive economic theory, and hence within the professional competence of the economist. He can, and should, say nothing whatever concerning the desirability of such outcomes or such institutions as might generate these outcomes.
The barrier between positive theory and normative advice must always be vigilantly maintained. It is difficult to accomplish this separation even in the strict private-goods world, as the discussion here suggests. Theoretical welfare economics, as a subdiscipline, is considered by many economists, perhaps by most, to involve necessarily normative elements. As I have tried to indicate, however, the fundamental content of this subdiscipline can be incorporated into positive theory with no normative overtones.
The same barrier between positive and normative theory is much more difficult to maintain when the demand-supply of public goods is introduced. Here the role of theory seems much more limited, and the analysis much less relevant to the observed world. The theoretical idealization analogous to the competitive order, that represented by Wicksell’s unanimity rule for making group choices, is sufficiently removed from real-world experience so that it rarely serves even as a norm for policy action. Presumably, by contrast with the private-goods sector, the costs of attempting to approximate the ideal here are considered to be so great that wholly different norms must be introduced.
Properly conceived, however, theory can do precisely what it can do in the private-goods world. It can describe, and at several levels, the outcomes that will tend to emerge from the process of voluntary exchanges among individuals. It can do no more than this, and the economist has no role in pushing further. By the nature of the different universe that he confronts, the limits of theoretical relevance for the economist seem to be reached much earlier here. In a genuine sense, all discussions of political-decision rules can be interpreted as treating of “workable unanimity,” but the distance between the ideal and the alternatives that seem plausibly possible is so great as to cause the ideal itself to lose apparent relevance.
The reason is not difficult to find. A community of individuals decides to demand goods and services publicly through governmental-political processes, rather than privately, precisely because the bilateral exchanges facilitated by market arrangements are insufficiently inclusive. External effects are exerted on parties other than those directly entering into the market exchange, and these effects are considered to be relevant and important. “Exchanges,” trades, agreements among all members of the community are deemed more efficient by these members. Multilateral agreements are, however, far more costly to negotiate than bilateral ones. In addition, the incentive for initiating negotiation leading toward agreement in such cases may be absent. These facts are evident to such an extent that it often appears as folly to make any attempt to examine the outcomes that genuinely voluntary exchange processes would produce in the theoretical idealization described by the unanimity rule. The limits of the voluntary exchange theory of the demand for and the supply of public goods are indeed narrow, especially when compared with its analogue, the theory of perfectly competitive markets.
The exercise is nonetheless useful, and it does provide the only available “pure theory” of public finance, upon which all derivative theoretical constructions rest. By first ignoring the costs of negotiating
n-person agreements, by ignoring the absence of individual incentive to organize agreements in the
n-person case, the theorist can proceed with his description of the results of idealized political process. These descriptions are wholly analogous to those made about results of market processes that are characterized by perfectly competitive conditions. The statements of the necessary conditions for efficiency are closely similar in the two cases, and in neither is normative content necessary. The satisfaction of the necessary marginal conditions may or may not represent desirable social objectives, and it is not the role of the economist to make such a determination.
One of the primary purposes of this book is that of stating these conditions and examining their implications. The theory is intended to be positive, and its extremely limited relevance is recognized and acknowledged. It is the “pure voluntary exchange theory of public finance” and is presented for the simple reason that this theory must first be developed rigorously before we can begin to examine more relevant models. Again the theory is on all fours with that of perfectly competitive markets; only after the latter was fully worked out could more refined analysis begin. In specific terms, the theory presented in the early part of this book describes the results that the political process would produce if a general rule of unanimity should be operative. The treatment here is in the strict Wicksellian tradition, and is, in fact, Wicksell revisited or modernized.
Initially, the costs of negotiating
n-person agreements are largely ignored. In a broader framework, and at a later stage, these costs must be introduced since they are essential to an understanding of the public economy. Analysis at this second stage must incorporate the costs of reaching agreements, or making collective decisions, and an economic theory of political constitutions developed. The individual’s own recognition that, in the public-goods world, he is likely to be caught in an
n-person analogue to the prisoners’ dilemma will prompt him to agree to “workable unanimity” rules. He will trade off some efficiency (as measured by the standard criteria) in exchange for more efficient decision-making. The whole theory of political order becomes directly relevant to the demand and the supply of public goods, inclusively considered.
The analysis is developed progressively from the simplest models to complex ones. Chapter 2 examines the demand-supply of a single pure public good in the highly restricted two-good, two-person, world-of-equals model. Only the world-of-equals assumption is dropped in Chapter 3. The purity of the public good is abandoned in Chapter 4, and the analysis is extended to a many-person group in Chapter 5. The novel world where all goods are public is treated in Chapter 6. The problems presented by the publicness of any political decision are introduced in Chapter 7, and the specific institutions of fiscal choice are considered in Chapter 8. The interesting and important question that has been assumed to be central in much of the modern theory, Which goods should be public? is examined in Chapter 9. Suggestions for a positive theory of public finance are advanced in the concluding chapter.
For those students and scholars who do not fully share the methodological approach that I have suggested, and whose interests lie primarily in the derivation of the necessary conditions for Pareto efficiency or optimality in the public-goods sector, most of the analysis is applicable and relatively straightforward. To an extent, my treatment can be interpreted within this framework as an alternative version of the normative theory of the public sector in the Samuelson-Musgrave tradition.
One additional and final point should be made in this introductory chapter. The demand for and the supply of public goods are discussed throughout the book under the assumption that the community contains a specific number of persons. I shall neglect in this book the important set of issues that is introduced when attempts are made to determine efficient or optimal sizes of membership in sharing groups. I hope to develop some of the analysis of these issues in a later work.