Franco Donzelli Topics in the History of Equilibrium

Franco Donzelli Topics in the History of Equilibrium Analysis Lesson 5 Hicks and Hayek on the Epistemological Status of GET from the Late 1920 s to the Mid-1940 s Ph. D Program in Economics University of York February–March 2008 Franco Donzelli Lesson 5 - Hicks and Hayek on GET

Introduction n What is the epistemological status of General Equilibrium Theory? ¨ Is it a positive or a normative theory? ¨ What is its relation to the real world? ¨ What is its empirical content? ¨ What its explanatory or predictive power? n A historical perspective: the status of GET in the third phase of its development (1920 -1950) n Hicks on the epistemological status of GET in the 1930 s and 1940 s n Hayek on the epistemological status of GET from the late 1920 s to the mid-1940 s n A fundamental ambiguity in the epistemological foundations of GET Franco Donzelli Lesson 5 - Hicks and Hayek on GET 2

Hicks: tâtonnement and equilibrium 1 n Referring to the tâtonnement process in Walras’s exchange model (freely interpreted) Hicks (1934, 342 -343) writes: “What, however, Walras does not make really clear is whether any exchanges do or do not actually take place at the prices originally proposed, when those prices are not equilibrium prices. If there is no actual exchange until the equilibrium prices are reached by bidding, then Walras’ argument is beyond reproach on the score of logical consistency, though it may be called unrealistic. (The market then proceeds under Edgeworth’s principle of “recontract”, or provisional contract. ) But if such exchanges do take place, then, in general, the final equilibrium prices will be affected by them. […] Neither Walras nor Pareto faced up to this difficulty; when we do so, it is impossible to avoid the conclusion that the “Lausanne equations” are of rather less significance than they imagined. The equations of Walras are not by any means a complete solution of the problem of exchange; but they remain a very significant step towards such a solution. Franco Donzelli Lesson 5 - Hicks and Hayek on GET 3

Hicks: tâtonnement and equilibrium 2 [continued] For Walras’ system of prices will be reached, either if contracts are made provisionally or (a more important case) if people come on to the market on successive “days” with the same dispositions to trade, and there is no carry-over of stocks (or a constant carry-over) from one day to the next. When it is understood in the last sense, theory of static equilibrium of exchange takes its place as a step towards the development of a complete theory with which future exposition is unlikely to dispense. ” n Hicks contrasts the two alternative interpretations of the tâtonnement process: the virtual one in “logical” time (Edgeworth’s “provisional contracts” or Walras’ “no-trade-out-of-equilibrium” assumption), leading to an instantaneous equilibrium concept, and the observable one in “real” time, leading to a stationary equilibrium concept (though necessarily confined to a pure-flow economy) n In 1934 Hicks leans towards the second interpretation and the associated stationary equilibrium concept: “a more important case” Franco Donzelli Lesson 5 - Hicks and Hayek on GET 4

Hicks: tâtonnement and equilibrium 3 n Moreover, in 1934 Hicks (346, fn. 19) qualifies as “rather pathetic” the passage where Walras, by means of the “hypothèse des bons”, generalizes to the entire economy (namely, to the comprehensive model of exchange, production, capital formation, circulation and money) both the virtual interpretation of the tâtonnement process in “logical” time, and the associated notion of instantaneous equilibrium n Yet, just five years later, in Value and Capital (1939, 122 ff. ), Hicks will dramatically change his mind as to which of the two possible interpretations of the tâtonnement process and the associated equilibrium concept ought to be regarded as the “more important” one. For: Hicks’s exclusion of all sort of “trading at false prices” generalizes Walras’ notrade-out-of-equilibrium assumption; ¨ Hicks’s celebrated “Monday-week” distinction is simply a pseudo-realistic, Marshallian way of reintroducing that distinction between the “phase des tâtonnements préliminaires”, the “phase statique”, and the “phase dynamique” that Walras had accomplished by means of his “hypothèse des bons”; ¨ ¨ Hicks’s allegedly “new” notion of “temporary equilibrium” is nothing but a particular version of Walras’ pre-existing notion of “instantaneous” equilibrium. Franco Donzelli Lesson 5 - Hicks and Hayek on GET 5

Hicks: the epistemological status of GET n This is not the end of the story, however. For, a quarter of a century later, in Capital and Growth (1965, 73 -76), Hicks will once again radically change his mind about the explanatory power and empirical content of the “temporary equilibrium” version of GET: “Both in Lindahl version and in the Value and Capital version [of the Temporary Equilibrium model] it is necessary to assume that prices remain unchanged throughout the single period; and that these prices are equilibrium prices which, within the single period, equate supplies and demands. In order to visualize this, some such construction as my “week” and my “Monday” appears to be necessary; but the artificiality of such construction is only too obvious. They do deliberate violence to the order in which in the real world (in any real world) events occur. […] The fundamental weakness of the Temporary Equilibrium Method is the assumption, which it is obliged to make, that the market is in equilibrium in the very short period, which is what its single period must be taken to be. […] Even in a very competitive economy, such very short-run equilibration is hard to swallow; in relation to modern manufacturing industry, it is very hard to swallow indeed. ” Franco Donzelli Lesson 5 - Hicks and Hayek on GET 6

Hayek: the empirical content of GET in the early 1930 s 1 n In 1933 Hayek entertains a highly “realistic” vision of GET, especially of its trend of development (1939, 135, 138 -9): “The most characteristic feature of the work of our generation of economists is probably the general endeavor to apply the methods and results of the pure theory of equilibrium to the elucidation of more complicated “dynamic” phenomena. […] A great part of this work will certainly consist in the elaboration of particular chapters of general theory, especially of theory of capital and theory of money, in the direction of a more careful analysis of the processes resulting from any change in the data. It is, however, the common peculiarity of all such attempts to make theory more realistic that they soon bring us back to the fundamental problem of all economic theory, that is to the question of the significance of the concept of equilibrium and its relevance to the explanation of a process which takes place in time. Franco Donzelli Lesson 5 - Hicks and Hayek on GET 7

Hayek: the empirical content of GET in the early 1930 s 2 [continued] There can be no doubt here that some of the formulations of theory of equilibrium prove to be of little use and that not only their particular content but also the idea of equilibrium as such which they use will require a certain amount of revision. […] The main difficulty of the traditional approach is its complete abstraction from time. A concept of equilibrium which essentially was applicable only to an economic system conceived as timeless could not be of great value. ” n Equilibrium theory ought to be interpreted as a realistic theory, endowed with an empirical content and applicable to all sorts of real-world, dynamic phenomena (business cycles, capital, investment, money, price movements, etc. ) n But it can be so interpreted only if an equilibrium state is viewed as a result of “a process which takes place in time”, what has not always been done before. n What sort of process should be supposed to support an equilibrium state? The answer is given in “Economics and Knowledge” (1937, 43 -44) Franco Donzelli Lesson 5 - Hicks and Hayek on GET 8

Hayek: the empirical content of GET in 1937 1 “We shall not get much further here unless we ask for the reasons for our concern for the admittedly fictitious state of equilibrium. Whatever may occasionally have been said by over-pure economists, there seems to be no possible doubt that the only justification for this is the supposed existence of a tendency towards equilibrium. It is only with this assertion that economics ceases to be an exercise in pure logic and becomes an empirical science. […] In the light of our analysis of the meaning of a state of equilibrium it should be easy to say what is the real content of the assertion that a tendency towards equilibrium exists. It can hardly mean anything but that under certain conditions the knowledge and intentions of the different members of society are supposed to come more and more into agreement […]. The only trouble is that we are still pretty much in the dark about (a) the conditions under which this tendency is supposed to exist, and (b) the nature of the process by which individual knowledge is changed. ” Franco Donzelli Lesson 5 - Hicks and Hayek on GET 9

Hayek: the empirical content of GET in 1937 2 n An economic equilibrium is a state in which the plans of the individual members of the economy are mutually compatible, hence executable. n The individuals choose their plans on the basis of their respective information, knowledge, and preferences, which are subjective, fragmentary, dispersed, and often inconsistent. n The empirical content of GET rests on the existence of a tendency towards equilibrium, which in turn depends on the existence of reliable mechanisms of diffusion of knowledge, through which individual plans can be reconciled. n But the conditions under which the social processes of diffusion of knowledge are effective are difficult or even impossible to state. n The idea that the knowledge dispersed in society can be centralized in a single mind and used to plan the whole economy, as claimed or implied by some economists siding with the advocates of socialist planning during the Socialist Calculation Debate of the 1930 s, is completely wrong. Franco Donzelli Lesson 5 - Hicks and Hayek on GET 10

Hayek: the empirical content of GET in the early 1940 s n Hence, only four years after “Economics and Knowledge”, in writing The Pure Theory of Capital, Hayek (1941, 28) proves to be much more pessimistic about the empirical content of GET: “The explanation of why things ever should, and under what conditions and to what extent they ever can, be expected to approximate to [a state of equilibrium], requires a different technique, that of the causal explanation of events proceeding in time. But the fact that it is probably impossible to formulate any conditions under which such a state would ever be fully realised does not destroy its value as an intellectual tool. On the contrary it seems to be a weakness of the traditional use of the concept of equilibrium that it has been confined to cases where some specious “reality” could be claimed for it. In order to derive full advantage from this technique we must abandon every pretence that it possesses reality, in the sense that we can state the conditions under which a particular state of equilibrium would come about. Its function is simply to serve as a guide to the analysis of concrete situations, showing what their relations would be under “ideal” conditions […]. ” Franco Donzelli Lesson 5 - Hicks and Hayek on GET 11

Hayek: “explanations merely of the principle” vs. “explanations of the precise result” 1 n But then, if GET is not an empirical theory in the usual sense, what sort of explanations does it provide? What is its epistemological status? Hayek (1942, 290 -291) writes: [T]he number of separate variables which in any particular social phenomenon will determine the result of a given change will as a rule be far too large for any human mind to master and manipulate them effectively. In consequence our knowledge of the principle by which these phenomena are produced will rarely if ever enable us to predict the precise result of any concrete situation. […] The distinction between an explanation merely of the principle on which a phenomenon is produced an explanation which enables us to predict the precise result is of great importance for the understanding of theoretical methods of the social sciences. […] Franco Donzelli Lesson 5 - Hicks and Hayek on GET 12

Hayek: “explanations merely of the principle” vs. “explanations of the precise result” 2 n n n [continued] The best illustration [of the first type of explanation] in the field of the social sciences is probably the general theory of prices as represented, e. g. , by the Walrasian or Paretian system of equations. These systems show merely the principle of coherence between the prices of the various commodities of which the system is composed; but without knowledge of the numerical values of all the constants which occur in it and which we never do know, this does not enable us to predict the precise results which any particular change will have. ” Hayek’s distinction between types of explanation is reminiscent of Pareto’s distinction between degrees or stages of explanation. But Hayek’s “explanations merely of the principle” are more similar to Pareto’s stage α than to Pareto’s stage β explanations. Moreover, unlike Pareto, Hayek does not even trust statistical explanations or predictions, because an individualistic theory can only start from individual “data”. Franco Donzelli Lesson 5 - Hicks and Hayek on GET 13

Hayek: the epistemological status of GET n In the early 1940 s Hayek arrives at the conclusion that GET is not an empirical theory in the usual sense of the word, since: ¨ it cannot provide “explanations of the precise result”, that is, numerically specified, deterministic explanations or predictions; ¨ it cannot even provide statistical explanations or predictions. n But, if GET can only offer “explanations of the principle”, why is it structured as a potentially deterministic theory, apparently capable to provide very detailed “explanations of the precise result”? n The potentially deterministic structure of GET is dangerously misleading, for it wrongly suggests that theoretical apparatus can be employed to directly calculate (or indirectly elicit) the equilibrium values of the endogenous variables, as indeed occurred during the Socialist Calculation Debate of the 1930 s. n This is why Hayek eventually drops the equilibrium notion, and corresponding theory, in favor of the milder notion, and theory, of “spontaneous order”. Franco Donzelli Lesson 5 - Hicks and Hayek on GET 14

Conclusions 1 n The epistemological status of GET is highly ambiguous from beginning to end. n For Walras GET is an empirical theory, but its empirical content is made to depend on a theoretical construct, the tâtonnement, whose empirical content is itself controversial: starting from a highly realistic interpretation of the tâtonnement, Walras progressively realizes theoretical difficulties connected with such interpretation, slowly turning the tâtonnement into a less and less realistic construct. n Pareto drops theory of the tâtonnement altogether, focusing on the static notion of equilibrium exclusively: for Pareto it is “absurd” to deem that the system of equations defining a static equilibrium state can be numerically solved, but such system still retains a sort of intermediate explanatory power, as a first approximation to a highly complex world; if sufficient statistical data are collected, the use of GET to produce quantitative predictions of a statistical kind cannot be ruled out in the future. Franco Donzelli Lesson 5 - Hicks and Hayek on GET 15

Conclusions 2 n Hicks, as Walras, is apparently convinced that the explanatory power and empirical content of GET depend on showing that an equilibrium state is the final outcome of an adjustment process in real time. Yet, after realizing the difficulties and limitations connected with such a tâtonnement process in real time, he switches, once again following Walras, to the so-called Temporary Equilibrium Method, thereafter continuously oscillating between accepting and rejecting such Method as a tool for investigating the functioning of a real economy. n Hayek initially endorses a highly realistic interpretation of GET, explicitly stating that its empirical content rests on “the supposed existence of a tendency towards equilibrium”. But then, after realizing that “the conditions under which such tendency is supposed to exist” cannot be easily specified, he first reinterprets GET as providing “explanations merely of the principle” and then, after reflection on the misuse of GET during the Socialist Calculation Debate, he abandons the equilibrium notion altogether. Franco Donzelli Lesson 5 - Hicks and Hayek on GET 16

References 1 Donzelli, Franco. 1993. "The Influence of the Socialist Calculation Debate on Hayek's View of General Equilibrium Theory", Revue européenne des sciences sociales, Tome XXXI, n. 96, 47 -83. — 1997. "Pareto's Mechanical Dream", History of Economic Ideas, Vol. V, n. 3, 127 -178. — 2005. “Equilibrium and tâtonnement in Walras’ Eléments”, mimeo. Hayek, F. A. (ed). 1935 a. Collectivist Economic Planning: Critical Studies on the Possibilities of Socialism, George Routledge & Sons, London. — 1935 b. “The Nature and History of the Problem”, in Hayek F. A. 1935 a. — 1935 c. “The Present State of the Debate”, in Hayek F. A. 1935 a. — 1937. Economics and Knowledge, Economica, n. s. , vol. 4, February, 33 -54. — 1940. Socialist Calculation: The Competitive ‘Solution’, Economica, n. s. , vol. 7, 125149. — 1941. The Pure Theory of Capital, Routledge and Kegan Paul, London and Henley. — 1942. Scientism and the Study of Society, Part I, Economica, n. s. , vol. 9, August, 267291. Hicks, John R. 1934. "Léon Walras", Econometrica, 2, October, 338 -48. — 1939. Value and Capital. An Inquiry into Some Fundamental Principles of Economic Theory, Oxford University Press, Oxford. — 1965. Capital and Growth, Oxford University Press, Oxford. Jaffé, William. 1980. “Walras’s Economics as Others See It”, Journal of Economic Literature, 18, June, 528 -549. Franco Donzelli Lesson 5 - Hicks and Hayek on GET 17

References 2 Pareto, Vilfredo. 1896 -1897. Cours d'économie politique. Nouvelle édition par G. -H. Bousquet et G. Busino, Librairie Droz, Genève, 1964. — 1906. Manuale di economia politica, Edizioni Bizzarri, Roma, 1965. Walras, Léon. 1874. "Principe d'une théorie mathématique de l'échange", Séances et travaux de l'Académie des sciences morales et politiques (Institut de France), Collection, 33 e année, nouvelle série, 101, tome I, Paris, Alphonse Picard, janvier, 97 -116; as reprinted in Walras, Léon. 1993. — 1876 a. "Equations de l'éxchange", Bulletin de la Société vaudoise des sciences naturelles, 2 e série, 14, no. 76, octobre, 365 -94; as reprinted in Walras, Léon. 1993. — 1876 b. "Equations de la production", Bulletin de la Société vaudoise des sciences naturelles, 2 e série, 14, no. 76, octobre, 395 -430; as reprinted in Walras, Léon. 1993. — 1877. "Equations de la capitalisation", Bulletin de la Société vaudoise des sciences naturelles, 2 e série, 15, no. 77, mars; as reprinted in Walras, Léon. 1993. — 1954. Elements of Pure Economics, translated by William Jaffé from the 5 th ed. , Irwin, Homewood, Ill. — 1988. Eléments d'économie politique pure ou Théorie de la richesse sociale, édition comparée, in Auguste et Léon Walras, Œuvres économiques complètes, 8, Economica, Paris. — 1993. Théorie mathématique de la richesse sociale et autres écrits d'économie pure, in Auguste et Léon Walras, Œuvres économiques complètes, 11, Economica, Paris. Franco Donzelli Lesson 5 - Hicks and Hayek on GET 18