Economics as a Coordination Problem: The Contributions of Friedrich A. Hayek
In contrast to the majority of economists, [Austrian economists] talk and act like people who are doing extraordinary science. They produce relatively more books and contribute fewer articles to established journals. They do not write text books; their students learn directly from the masters. They are very much concerned with methodological and philosophical fundamentals and what makes the label extraordinary most applicable to their work is that they share a conviction that orthodox economics is at the point of breakdown, that it is unable to provide a coherent and intelligible analysis of the present-day economic world (Edwin G. Dolan, "Austrian Economics as Extraordinary Science," in The Foundations of Modern Austrian Economics [Kansas City: Sheed & Ward, 1976]).
The crises are being proclaimed everywhere by pundits and serious economists alike. When Sir John Hicks lends support to this chorus of woes in the very title of his recent book,The Crisis in Keynesian Economics, then truly something is amiss with economics. Moreover, a plethora of cures has been offered from various sources. But the patient must be excused if he hesitates before ingesting any of these elixirs. For many of these appear oddly familiar. In fact, some appear very similar to all-but-forgotten diseases from which our existing regimen was intended to protect us. Ricardian value theory, with its objective value theory and its absence of a theory of demand, scarcely represents an attractive alternative to general equilibrium theory. In many respects, of course, neo-Walrasian general equilibrium theory and Ricardian classical political economy share common faults; the tendency in both research programs to assume perfect foresight is one striking example.*22 At least neoclassical economics, in its dominant neo-Walrasian variant, provides the basis for a satisfactory theory of demand. Ricardianism and its intellectual offshoots surely offer no such hope. One need only examine Mr. Sraffa's tour de force, Production of Commodities by Means of Commodities,*23 which in many ways represents the logical culmination of Ricardian economics: general equilibrium theory without demand.*24 This cannot represent an acceptable alternative to modern economic theory, however flawed the latter may be. It would truly be a cure worse than the disease.
Orthodox economists are justified, then, in their steadfast adherence to existing economic theory, given the visible alternatives. Nor does "radical economics" offer a very attractive substitute research program. Insofar as radical political economy is Marxist, and insofar as Marxist political economy is Ricardian, then it surely suffers many of the same faults as the other neo-Ricardian alternatives.*25 In addition, the metaphysical underpinnings scarcely recommend the Marxist alternative. In a profound sense, then, all neo-Ricardian research programs, far from being "radical," are reactionary intellectual developments.
None of the above is intended to deny that there is a general crisis situation in economic theory, of which various particular crises, such as that in Keynesian economics, are but manifestations of a more general intellectual malaise. Nor am I entirely out of sympathy with particular "radical" criticisms of orthodox models. I am simply arguing that the crisis has simply not been diagnosed, though its symptoms are now well known. They have been observed in numerous recent works, including Hicks's and in previous chapters of this book. But no satisfactory diagnosis having been made, it is no wonder that the putative cures are misconceived.
The analysis of the previous chapter provides the basis for a diagnosis. Ricardian economics is an unsatisfactory theory because of its inability to explain demand or short-run pricing. Even its long-run value is marred, dependent as it is on a materialist conception of costs. In the Ricardian analysis, costs depend on "the ultimate conditions on which nature yields her stores."*26 It was from this materialist conception of costs that subjectivist neoclassical economics was to emancipate us. This Ricardian conception is not entirely consistent with a thorough-going subjectivist cost theory, in which a cost is a forgone utility.*27 Once again, the forgoing in no way subtracts from Ricardo's many valuable contributions, not the least of which is the concept of the margin.
Neoclassical economics represents a demonstrable improvement over Ricardian analysis, insofar as the former advanced our understanding of those phenomena least understandable in the Ricardian approach.*28 The so-called Marginalist Revolution provided a basis for a solution to Ricardian lacunae. All three variants of neoclassicism, the Walrasian, Austrian, and Jevonian, focused on utility as an explanation of demand and made use of the older concept of the margin to explain pricing. The Austrians, particularly Wieser, developed the idea that costs are forgone opportunities, not coefficients of production dictated by physical laws of production. Their consistent methodological subjectivism enabled the later Austrians, particularly Mises and Hayek, to perceive that the only relevant forgone opportunities are those perceived by the individual decision maker. The pure logic of the economic short run and that of the long run were seen to be the same and to have no necessary connection with any laws of production that so occupied the Ricardians.
But Ricardian elements remained embedded in the emerging neoclassical synthesis, particularly its Marshallian variant. Though graduate school price theory today is basically Walrasian, intermediate textbooks abound with confusing Marshallian and Ricardian concepts intermingled with a neo-Walrasian analysis. The classical short run is superimposed on models that preclude imperfect and incomplete adjustment. The U-shaped cost curve—indeed, the whole cost curve apparatus—is a distracting Marshallian intrusion. No wonder undergraduates are confounded by all this!
Yet professional economists have learned their early intellectual gymnastics in this environment. Probably few discard entirely the more classical elements in orthodox economics. As a consequence, Ricardian analysis persists even in the "higher level" theoretical output of journals.
My suggestion, then, is that the emergent theoretical edifice in the twentieth century—which in the English-speaking world was predominantly Marshallian at first and later predominantly Walrasian—contained inner contradictions that had to be resolved before a satisfactory theoretical edifice could be constructed. It happened that the problems first became manifest in two distinct areas in the first decades of this century: monetary theory and the theory of costs, or supply.
Monetary theory occupied an unsatisfactory position so long as it remained separate and distinct from the general theory of value. Monetary theorists, who in this period were likely to be accomplished value theorists as well, discovered in effect that much of monetary theory was a holdover from Ricardian political economy. As was noted in chapter 3 of this book, Wicksell took the initial steps toward an integration of monetary and value theory. But Wicksell himself was incapable of solving the root theoretical conundrum: how can one apply the marginalist calculus to the demand for money? It was Ludwig von Mises who finally solved this problem, by recognizing the necessity of subjectivizing monetary theory in what surely remains one of the more neglected works of monetary theory.*29 Curiously, though, he is accused of perpetuating the very problem that he solved.*30
By building on the foundations laid by Mises, and indirectly by Wicksell, Hayek provided a microfoundation for monetary theory. He did so by showing that monetary theory could be viewed as an extension of the barter theory of price, rather than approached macroeconomically in the Ricardian tradition. For Hayek, the quaesitum of monetary theory is not the determination of the value of money, but the determination of the effects of monetary disturbances on (relative) prices and production. Moreover, Hayek's microfoundations and his general theoretical approach are independent of his specific empirical hypothesis about precisely how monetary disturbances typically operate. That is, the approach can be readily adopted to changing institutional arrangements (for example, the increasing role of government expenditures) and thus take account of changing consequences for the real sector of monetary disturbances.
Keynes was a comparatively late convert to these changes in monetary theory. Perhaps the reason was that he could never slough off entirely his Marshallian epidermis. Or, perhaps, as Mrs. Robinson now claims, Keynes did not at all times fully comprehend the point of his own revolution.*31 In any case, he was never fully successful in articulating the centrality of relative prices in any monetary explanation of cyclical fluctuations.*32
Hayek worked out the fundamentals of the problem in much more detail. This is most obvious in the case of the analysis of investment demand, but it is even more true in his analysis of prices as signaling devices and the role of changes in the array of relative prices on entrepreneurial expectations. In this regard, Hayek always treated changes in expectations as endogenous, whereas Keynes saw them more as an exogenous element in the market system.
The second area of concern was cost and supply theory. This area involved a number of problems, related by the common element of an ill-defined cost theory. The literature is too vast and diverse to cite here, but it covered such problems as the relation between costs and supply and the disappearing supply curve in monopoly. The aspect of this general problem that interests us here is the question whether opportunity costs are subjective or objective in nature. Hayek's interest in this issue can be seen as one with his interest in monetary questions. He wanted to develop a consistently methodological individualist and subjectivist theory of the coordination of economic activities. It would not do to leave cost theory hanging in the air, as it were, borrowed from Ricardian economics. The problems with inherited cost theory become apparent in the Socialist-calculation debate. Costs are not data for the individual, given to him as they are to the economist as ideal observer who constructs the models of decision making. Different transactors have different perceptions of the data and acquire different bits of knowledge about relevant opportunities. Moreover, even if all individuals, mirabile dictu, had the same experiences, they would interpret these differently. It is as if each of us saw part of some production function. Even if some central allocator costlessly and instantaneously collected all our perceptions as they occurred, he would discover that they did not make up one common production function. If perceived opportunities differ, the criterion of allocating according to costs—as though costs were a unique and measurable magnitude—is seen as non-operational, illusory, and, hence, irrelevant.*33
All of these issues were actively discussed and debated in the thirties. Though surely no part of his intention, what Keynes did in The General Theory was to direct attention away from these central theoretical questions. Whatever The General Theory may have taught us, it answered none of these questions. Rather, the Keynesian revolution diverted attention away from these questions and effectively closed off paths of inquiry that at the time evidenced every sign of leading to a solution.
My argument, then, is that the various crises in economics are manifestations of inconsistencies in the neoclassical research program. These inconsistencies are present because of the curious admixture of Ricardianism and neoclassicism present in the modern research program. The inconsistencies were never resolved, though their resolution was in sight in the thirties when it was aborted by the Keynesian revolution. The current situation is the result of forty years of repressed debate over the very fundamental questions that occupied economists in the earlier period. We are now condemned to relive the debates unless we succeed in using the earlier discussion as starting points.
In particular, Mises and Hayek went a long way toward solving a theoretical problem that besets economists today—the integration of monetary and value theory. Whatever else can be said about Hayek's monetary economics, his work does consist of a monetary theory erected upon a consistent microfoundation. Whatever else can be said about almost any other monetary theory of economic fluctuations, it lacks that consistency. Since an avowed purpose of macro and monetary theory today is to provide a microfoundation for the analysis of economic fluctuations, this alone recommends a reconsideration of Hayek's work. It is true that an acceptance of his approach involves abandoning some of what is presently taken for granted (for example, monetary theory is the determination of the value of money). But once one accepts the fact that something is not quite right with the grand neoclassical synthesis, one is virtually committed to abandoning something substantial in the orthodox research program. The only question is what.
It would be to go far beyond the scope of this book to outline in detail a new Austrian or Hayekian research program. Such a program is really only in the first stage of development.*34 This is partly because Hayek's attention turned elsewhere at a crucial moment, and because the Keynesian revolution swept aside many of the interesting questions that would surely be addressed in such a research program. In a very real sense, the program will be whatever those who choose to work in a Hayekian framework make it.
Above all else, Hayek emphasized a microeconomic approach to economic questions. Conventional macroeconomic models with constant functional relationships that can be mechanically manipulated are virtually ruled out of such a program.*35 It is doubtful that there can ever be a Hayekian alternative to the Hicksian cross.*36 This is true, not because Hayek's theory is needlessly complex, but because the world is too complex to picture it accurately in simplistic formulae. So long as economists are wedded to the contrary idea, so long will progress in the theory of economic fluctuations be stalled. In this, I should say, there is affinity between Hayek's views and those of Professor G. L. S. Shackle and the other Keynesians who agree with Shackle's radical interpretation of Keynes.*37
Conversely, one area that seemingly cries out for attention is the sequence of effects in an inflationary expansion. Here one must distinguish carefully between Hayek's theory of economic fluctuations and his empirical hypotheses concerning the effects of an inflation of the money stock. His theory is in the Cantillon tradition, which, broadly speaking, emphasizes distribution effects. Hayek's hypothesis concerns where and how injections of money and credit enter the economy. He looked to private investment as the key variable. It would not be surprising if since 1931 there had been important changes in the paths taken in the inflation process. As an example, federal spending is a much more important part of the economy than it was in 1931. A substantial part of increases in the money stock, broadly or narrowly defined, is typically due to monetization of the debt. What is the net effect of government spending on the "consumption-investment" ratio? Does it stimulate investment or consumption relatively more? And particular kinds of private investment more than others? Do particular sectors, industries, or even firms usually gain first in such an inflation? These are the kinds of questions that will undoubtedly be addressed in a Hayekian research program in monetary theory.*38
Finally, on the most general level a Hayekian research program must surely be concerned with the coordination of economic activities and the emergence of an undesigned ("spontaneous") order.*39
Hayek recently commented: "But though there is no longer a distinct Austrian School, I believe there is still a distinct Austrian tradition from which we may hope for many further contributions to the future development of economic theory."*40
There certainly are areas in which Hayek and his fellow economists have had an identifiable impact on current economic thought. One could scarcely call orthodox cost theory "Austrian," but some versions contain strong Austrian elements. As James Buchanan has argued, there is really a separate and distinct approach to cost theory, in which the subjective element is strongly emphasized.*41 Though never incorporated entirely into orthodox economic theory as such, this approach, heavily influenced by the twentieth-century Austrians, represents a lively alternative.
Elements of Austrian capital theory have been incorporated into many theoretical discussions about capital—especially in the period of investment approach. One could argue, of course, that other essential features of Austrian capital theory (for example, its emphasis on time preference and the subjective factors) are missing in these contemporary treatments. Yet Austrian capital theory, too, has been kept alive as an independent approach, though little has been done with it since World War II.*42
A sort of modern version of the Austrian school is apparently emerging. In periods of intellectual convergence, such as we have had since the Keynesian research program was developed, there is comparatively little scope for distinct schools in a science. In periods of intellectual divergence, in which we now appear to be, there is comparatively more scope for fundamental disagreement and for distinct schools of thought.*43 A number of new articles and books have been written in this emergent school.*44 There is every indication that once again the intellectual climate exists in which the theoretical conundrums can be solved; and economists who consider themselves as part of the Austrian tradition will be addressing themselves to the pressing theoretical and practical problems of the day. Economics can only gain from this development, for in intellectual endeavors, too, increased competition is beneficial.
Notes for this chapter
The theoretical connection between Ricardian classical political economy and Walrasian neoclassical economics has been recently noted by Robert Eagly: "Janus-like, the Walrasian system is situated between two great systems of economic theory. It forms the capstone to classical theory on the one side, and on the other the cornerstone to the modern post-classical theory. It provided answers to questions posed by the normal progression of theoretical inquiry within the classical framework. But at the same time it posed new questions that were to occupy the attention and time of economists during the following century" (Eagly, Structure of Classical Economic Theory [New York: Oxford University Press, 1974], p. 134).
New York: Cambridge University Press, 1960.
Nuti has recently characterized the approach of Sraffa and others as "a general equilibrium approach with the preference side chopped off." And he argued that "the approach...has no overwhelming advantages over the general equilibrium approach" (Dominco Mario Nuti, "On the Rates of Return on Investment," Kyklos 27 : 357). Nuti likewise identifies the Sraffa approach as "'classical'" (Nuti, 357-58).
The term "Neo-Ricardian" is borrowed from L. M. Lachmann, Macro-economic Thinking and the Market Economy (London: The Institute of Economic Affairs, 1973).
Philip H. Wicksteed, "The Scope and Method of Political Economy," in Readings in Price Theory, George J. Stigler and Kenneth E. Boulding, eds., (Homewood, Ill.: Richard D. Irwin, 1952), p. 19n.
Recent works articulating this view are James M. Buchanan, Cost and Choice (Chicago: Markham Publishing Co., 1969); and James M. Buchanan and G. F. Thirlby, eds., L.S.E. Essays on Cost (London: Weidenfield Nicolsen, 1973).
Leijonhufvud has recently made a persuasive case that the use of "neoclassical" is more confusing than illuminating. As much separated the "'neoclassical' grants" as bound them together. He argued that we dispense entirely with the term. See Axel Leijonhufvud, "The Varieties of Price Theory: What Microfoundations for Macrotheory?" U.C.L.A. Discussion Paper Number 44 (Los Angeles: mimeographed, 1974). William Jaffé has taken up the same theme recently in "Menger, Jevons, and Walras De-Homogenized," Economic Inquiry 14 (December 1976):511-24. In my defense, I would note that I have tried to limit the use of the term to those cases where the similarities of the various neoclassical schools are greatest. I recognize, however, that these similarities have been greatly exaggerated in recent years.
Mises, Theory of Money and Credit, new ed., trans. H. E. Batson (Irvington-on-Hudson, N. Y.: The Foundation for Economic Education, 1971).
See Don Patinkin, Money, Interest, and Prices, 2d ed. (New York: Harper & Row, 1965), pp. 79, 574-75.
See Joan Robinson, "What Has Become of the Keynesian Revolution?" in Essays on John Maynard Keynes, ed., Milo Keynes (New York: Cambridge University Press, 1975), p. 125.
See Axel Leijonhufvud, On Keynesian Economics and the Economics of Keynes (New York: Oxford University Press, 1968), p. 24. On the other hand, Patinkin apparently sees little role for changes in relative prices in Keynes's theoretical vision. (Don Patinkin, "Keynes' Monetary Thought," History of Political Economy 8 [Spring 1976], 45).
For the methodological subjectivist, it is an essential feature of human affairs that this be so. And the social scientist must take account of this, as Hayek long ago noted: "In the social sciences the things are what people think they are. Money is money, a word is a word, a cosmetic is a cosmetic, if and because somebody thinks they are" ("The Facts of the Social Sciences," in Individualism and Economics Order [Chicago: University of Chicago Press, 1948], p. 60).
Some papers in this development appear in Edwin G. Dolan, ed., The Foundations of Modern Austrian Economics (Kansas City: Sheed & Ward, 1976). More will appear in a forthcoming proceedings of a conference on Austrian Economic Theory held at Windsor Castle in 1976.
See Hayek, "Personal Recollections of Keynes," in A Tiger by the Tail, ed. Sudha R. Shenoy (London: Institute of Economic Affairs, 1972), pp. 101-2.
On this, see Roger Garrison's, "Austrian Macroeconomies," a paper prepared for the 1976 Symposium on Austrian Economics at Windsor Castle (Menlo Park, Calif., 1976). Forthcoming.
For a synopsis of Shackle's views, see his "Keynes and Today's Establishment in Economic Theory: A View," Journal of Economic Literature 11 (June 1973): 516-19.
The American Geographical Society is currently engaged in a study of the spatial diffusion of inflation in the United States. This is very much a topic of the kind that I am discussing.
See Gerald P. O'Driscoll, Jr., "Spontaneous Order and the Coordination of Economic Activities," a paper prepared for the 1976 Symposium on Austrian Economics at Windsor Castle (Menlo Park, Calif., 1976). Forthcoming.
Hayek, "The Place of Menger's Grundsätze in the History of Economic Thought," in Carl Menger and the Austrian School of Economics, J. R. Hicks and W. Weber, eds. (Oxford: Oxford University Press, The Clarendon Press, 1973), p. 13.
See the references cited in note 6.
Two works written in this tradition are Ludwig M. Lachmann, Capital and its Structure (London: London School of Economics, 1956); and Israel M. Kirzner, An Essay on Capital (New York: Augustus M. Kelley, 1966).
Professor Lachmann first offered this analysis in a talk at the University of Delaware in June 1976.
Some of these can be found in the seected bibliography of the Dolan book cited in note 13.
End of Notes
Return to top