To me it is far more pleasant to agree than to differ; but it is impossible that one who has any regard for truth can long avoid protesting against doctrines which seem to him to be erroneous. There is ever a tendency of the most hurtful kind to allow opinions to crystallize into creeds … A despotic calm is usually the triumph of error. In the republic of the sciences sedition and even anarchy are beneficial in the long run to the greatest happiness of the greatest number.
W. Stanley Jevons (1970, p. 260)
Scientific progress is a process of creative destruction. What is destroyed is the intellectual capital of other scientists whose resistance to accepting new contributions is not only understandable, but desirable; it is only by overcoming this resistance that the few genuine contributions can be separated from the more numerous invalid proposals.
Melvin W. Reder (1982, p. 20)
At the beginning of this book, we suggested that what separates schools in science is the questions posed, not the answers given. Certainly in the history of economics, major innovations have occurred when theorists observing the same phenomena asked new questions about it. And these innovations have been opposed precisely because different questions were being asked. In the eighteenth century, students of government sought to devise policies that would achieve a pre-ordained social pattern and lead to an increase in national wealth. Adam Smith altered the nature of the inquiry by asking a different question. He speculated on how a nation of autonomous individuals, each seeking to increase his own wealth and well-being, could produce an overall order that was no part of anyone’s intention. The answers to this question spawned what we now recognize as economic science.
A century later, Ricardianism had transformed economics into an arid set of laws of distribution and production, conjoined with a dubious population theory. Jevons, Menger and, to a lesser extent, Walras posed a new question: What governs the choices of consumers? The resulting neoclassical analysis transformed economics into the science of choice. It also transformed economics from a predominantly macro- into a predominantly microanalytic discipline.
In this century, Keynes changed economics by inquiring into the determination of aggregate demand and national income. In a sense, economics had now come full circle back to macroanalysis and the study of distributional rather than allocational questions. The subsequent dilution of Keynesianism in the neoclassical synthesis is well known but does not change our interpretation of the Keynesian message.
Subjectivists do not perform constrained maximization problems differently than neoclassical theorists. Nor are subjectivists privy to a special mathematics that yields different solutions to familiar equations. Subjectivists do, however, ask different questions. To relate our philosophy of science to our concise history of economics, subjectivists try consistently to analyze economic phenomena within the framework first sketched out by Menger in the last century. This involves, inter alia, a focus on the goals and plans of individuals, not on the objects or instruments of these plans; the reconstruction of observed phenomena in terms of individual choices; and the analysis of overall outcomes as the unintended results of individual interactions.
We have attempted to steer a course, as it were, between Charybdis and Scylla. On the one hand, we have sought to avoid unnecessary controversy. We are certainly aware that, within the neoclassical edifice, many of the individual points that we have made in this book are recognized and incorporated in the work of at least some major figures. Indeed, wherever possible we have cited the orthodox literature to emphasize points of convergence or, alternatively, the subjectivist elements in neoclassical economics. Where we are critical, our purpose in citing specific individuals is not to foment controversy but to be precise in our criticism. In every instance, we have attempted to cite an author who is representative of the orthodox position on a particular point.
On the other hand, we have been cognizant of the necessity of clearly stating the differences that do exist between subjectivists and their neoclassical brethren. It should not be surprising if there were substantial overlap between Austrian or subjectivist economics and neoclassical orthodoxy. The Austrians were one of the three founding schools of modern economics; and, in fact, there are more Austrian elements in neoclassical economics than are recognized in textbooks (including, via Lord Robbins, the very conception of economics as the study of the allocation of scarce means among competing ends). The reader may at times wonder why we have not emphasized this overlap more. The simple reason is that the subjectivist analysis would then have been obscured. Our purpose in writing this book was to restate and advance subjectivist economics. This goal would scarcely have been promoted by presenting an exhaustive list of neoclassical propositions with which we agree.
There is a “sponginess” to neoclassical economics that enables it to absorb divergent elements around it without ever emphasizing their main points. These fringe ideas become footnotes to which theorists can refer as evidence that they have taken the ideas into account. For instance, Keynes’ liquidity preference concept is incorporated in neoclassical economics as an interest-elastic demand for money. The idea has been formally incorporated, yet Keynes’ main point has been lost: the instability of liquidity preference. Nor are the effects of an unstable liquidity preference on interest rates and capital markets adequately treated in orthodox analysis. In terms of our own ideas, we have chosen to compare them to the dominant themes of neoclassical economics and not to unrepresentative footnote presentations.
We have also attempted to strike a balance between, on the one hand, presenting theoretical and methodological abstractions and, on the other hand, offering examples or applications. We did not wish to overwhelm the reader with too much of the former, or to confuse him with too much of the latter. If, however, we were to characterize the subjectivist literature generally, we would have to admit that it has been relatively long on methodological prescriptions and short on applications. It is time to do the difficult work of applying subjectivist ideas to actual problems. We are aware that most of our applications involve only a sketchy analysis of a particular problem. In almost every case, more work needs to be done. We believe, however, that the time is ripe for applying subjectivist insights to current economic problems. Accordingly, we outline here some areas where we think the returns would be particularly high.
First, we would mention the whole area of law and economics. We do so not only because we have each written in this area, but because it is a field that virtually demands the theoretical-institutional approach advocated by subjectivists. If any area involves a blending of theoretical and institutional analysis, it is law and economics. Nowhere else is an institutionless analysis less fruitful and more destructive. For instance, a good bit of ink has been spilled on the question of the efficiency of common law. This view, first of all, treats an undesigned order as if it were the creation of individuals; and, second, it puts the theoretical cart before the horse. What is efficient depends on the institutional environment. One can inquire whether a particular institution is consistent with a whole set of institutions (as when one analyzes the consistency of a common law doctrine with all others). It is quite meaningless, however, to ask whether an entire set of institutions – like the property rights structure – is efficient. This question treats as “givens” conditions that change mutatis mutandis as property rights evolve (Rizzo, 1980b; O’Driscoll, 1980). Of course, having made this point, we have only opened up a research program, not completed one. Needless to say, an important aspect of that program involves a theory of these institutions that does not treat them as the outcome of a collective constrained-maximization problem. The next item on our subjectivist research agenda is a suggestion for a start in this direction.
The analysis of money as an evolved social institution is a quintessential subjectivist topic. Viewing money as the unintended outcome of a market process involves thinking about money in a fundamentally different way (Frankel, 1977). Money then becomes not something that individuals can shape or control, but an institution or order to which they must accommodate themselves. In this process of accommodation, individuals will inadvertently affect the monetary order. The resulting monetary “innovations” will not, however, be changes that individuals have deliberately adopted: rather, these changes will represent evolutionary adaptation. Even when dramatic choices are made, as when a gold standard is substituted for a fiat money system, or fixed exchange rates are substituted for flexible exchange rates, policy-makers have not created new monetary institutions so much as they have moved the country from one given environment or regime to another. Policy-makers do not necessarily have a great deal of explicit control over the characteristics of either environment. They may surely take actions that affect their environment, but this does not imply that they can control the outcome. If we may suggest an analogy, we each have options with respect to which kind of climate we live in. We choose by moving from one climate to another, however, not by altering the climate of the area in which we presently live. The sum total of individual actions may affect the climatic environment, but even collective action will not enable individuals to effect a desired outcome.
The monetary approach suggested here has policy implications, of which two can be mentioned. The first concerns the nature of monetary rules. Every monetary system has or can be described by rules, including rules governing the growth of the supply of money. If a monetary system is an evolved economic order, then it is difficult to imagine how these rules could be externally imposed or chosen exogenously. The rules themselves, including the money growth rules, must be the product of an evolutionary process – a process of economic adaptation. Macroeconomic planning for a fixed rate of growth of the money supply has all the allure of microeconomic planning – and all the pitfalls, if recent history is any guide.
The second implication is more specific, though it is related to the first one. If the monetary institutions and money itself are part of the market system, then it is difficult to imagine how the quantity of money could be exogenous or independent of individual choices. Empirical evidence aside, there are strong a priori reasons to suspect the endogeneity of the money supply. Of course, empirically we may recognize that monetary questions have historically become political questions, and that political intervention leads to exogenous influences on the quantity or rate of growth of the money supply. This recognition puts the question of the exogeneity of the money supply in quite a different light, however, than that produced by, say, monetarist analysis. In the evolutionary view, money is a market institution and its supply is endogenously determined. There may be exogenous factors, as there are exogenous factors like fortuitous inventions in economic growth. An unexpected discovery of new sources of base money would be an example of an exogenous monetary change. For the supply of money to be strictly exogenous, however, it must be transformed from an economic or market institution into one subject to rigid political control. And market forces will constantly be at work offsetting the effects of these controls.
We think that an evolutionary approach to money has more explanatory power for analyzing recent financial market innovations, including the multiplication of close money substitutes (if not money themselves), than a positive theory of exogenous money determination.
Finally, we turn once again to the topic of competition. Economists analyzing competition as a process need to develop a theory of nonprice information structures. Neoclassical formulations rely almost exclusively on prices to convey information. This is untenable as soon as one incorporates real time and expectations in the analysis. Spot-price movements no longer convey unambiguous signals. Even were futures markets to be “complete,” the passage of time would upset any temporary intertemporal equilibrium. Not only do producers require information about trade-offs, they need to know how other producers will react to these prices. No matter how rich the array of futures markets producers confront Keynes’ beauty contest problem.
Entrepreneurs are not, of course, paralyzed by this informational void. Their ability to plan in a tolerably stable environment indicates the existence of other, nonprice, sources of information. There must be market institutions and rules of thumb that “fill in” some of the information not provided by prices. This realization requires, however, that we rethink our attitude toward familiar institutions and practices. For some of the required information will be about product characteristics, production plans, and even prices charged by competitors. Competition may, then, depend on the existence of institutions and practices presently regarded as non-competitive or anti-competitive.
In Chapter 5 we discussed endogenous uncertainty in the guise of Keynes’ beauty contest and the Holmes-Moriarty story. We saw there that, as long as independent decision-making remains independent, the search for more knowledge does not reduce uncertainty. It merely heightens the level of guessing and counterguessing. In this sense, then, endogenous uncertainty is ineradicable. On the other hand, as we have previously intimated, endogenous uncertainty can be reduced or eliminated if actors (agree to) follow “arbitrary conventions.” Independent firms, for example, face endogenous uncertainty in the form of trying to guess the price decisions of their rivals. In “oligopolistic” markets, the price that any given firm should charge will depend, in part, on the prices that he expects other firms to charge. Their decisions will in turn depend on what he is expected to decide. Thus, the expected price distributions faced by firms are endogenous; that is, they arise from the very operation of the market process. One “arbitrary convention” or institution that can reduce this form of uncertainty is the exchange of reliable price information among the rival firms.
Firms are subject to both exogenous and endogenous uncertainty. While exogenous risk may be transformed, by definition nothing can be done to eliminate it. Endogenous uncertainty arises from the very operation of the market. At least in principle, some actions can be taken to lessen this type of uncertainty. Consider uncertainty about what price to charge. At any given moment, firms observe a distribution of prices charged by their competitors, a distribution that need by no means be normal. Firm owners are aware, moreover, that they do not have complete information about prices. The variance of prices reflects, inter alia, temporary cost advantages, special discounting and product differentiation.
Each firm owner desires information that will improve his planning ability. This would include information about the mean and variance of the price distribution, product characteristics, amounts sold at these prices, etc. All of this information is either given or superfluous in perfectly competitive models, but is available only at some cost in competitive markets. Though he would never describe it in this way, the entrepreneur is seeking information about equilibrium conditions in his market. Though he may have quite different uses in mind for this information, a competitive firm owner requires much of the same information needed by a cartel.
By reducing price uncertainty, information-sharing arrangements may reduce the variance of prices. If firms are risk-averse, this may also reduce the mean of the price distribution. This latter point is emphasized in Dewey (1979), while Rizzo (1984) presents the more general argument.
None of this sanctions other aspects of collusive arrangements, especially not legal barriers to entry. Indeed, the existence or absence of legal blockages to competition would surely be crucial in ascertaining empirically whether an exchange of information were facilitating collusion or abetting competition.
We are not arguing that the exchange of price information would be generally characteristic of competitive industries, though some of it probably occurs in every industry (including among economists). We are suggesting, however, that competitive markets will generally rely on institutions and practices to generate information assumed in neoclassical analysis to be given by price signals alone. Sharing of price information may be one of these practices, which may also be institutionalized. A theory leaving no room for these nonprice sources of information is an impoverished – indeed, a biased – analysis of competition.
In conclusion, we have attempted in this book to offer a sample of research topics whose development would especially benefit from a consistent application of subjectivist economics. This sample is certainly not meant to be exhaustive, reflecting, as it does, the constraints both of space and of the authors’ own interests. We will count this book a success if readers begin adding their own items to the research agenda.