Machlup on Equilibrium

10 November 2006 at 3:07 pm Nicolai Foss Leave a comment

| Nicolai Foss |

The notion of equilibrium in economics has been discussed on an earlier occasion here at O&M (here). Perhaps as a reaction to this, former guest blogger, Joe Mahoney, has mailed a long but important quotation from Fritz Machlup’s (1967) Essays in Economic Semantics (pp. 44-5, 54, 56-7) which is reproduced below. (Joe notes that “as a humorous aside the original 1963 version of the book was titled: Essays ON Economic Semantics, but since only one chapter of the book was actually ON the topic of economic semantics, the 1967 paperback version title was changed to Essays IN Economic Semantics” — very interesting!) Here is the quotation (and a few comment below):

The most prevalent use of the equilibrium concept in economics is probably as a methodological device in abstract theory. Here, “equilibrium” is employed in connection with “models” containing several interrelated. . . . It is a different use of the equilibrium idea when it is employed to refer to concrete economic situations: here it is supposed to characterize a historical situation as one that has lasted or will last for a relatively long time without significant change. . . . The jump from equilibrium as a methodological device (useful fiction, purely mental construct) to equilibrium as a characterization of a concrete historical situation (operational concept) is a big one; that many take it without noticing any strain and without noticing the difference, is attributable . . . to their failure to recognize the function of the analytical concept, and is conductive to considerable confusion. . . . In light of the preceding discussion, we may define equilibrium in economic analysis, as a constellation of selected interrelated variables so adjusted to one another that no inherent tendency to change prevails in the model that they constitute. The model as well as its equilibria, are, of course, mental constructions (based on abstraction and invention). . . . Equilibrium as a tool for theoretical analysis is not an operational concept; and attempts to develop operational counterparts to the construct have been unsuccessful. . . . The “real” world surely has infinitely more variables than any abstract model, and their actual interrelations are neither known nor, I fear, knowable (partly because they probably change unpredictably over time). It follows that the equilibrium between selected variables could not be observed even if each of the variables had an observable counterpart in the real world. . . . Only after the variables are selected, and their interrelations assumed, can we speak of equilibrium and disequilibrium in the sense in which these terms are used in economic analysis. Despite the operation alsub-set, the model as a whole and its equilibrium are not observable, not operational; they remain mental constructions. . . . To characterize a concrete situation “observed” in reality as one of “equilibrium” is to commit the fallacy of misplaced concreteness. At best, the observer may mean to assert that in his opinion the observed and duly identified situation corresponds to a model in his mind in which a set of selected variables determine a certain outcome, and that he finds no inherent cause of change — that is, that he believes only an outside disturbance, not in evidence at the moment, could produce a change in these variables. This, of course, is a personal judgment, meaningful only if the variables are fully enumerated and the assumptions about their interrelations are clearly stated. As matters stand, any concrete economic situation may correspond at the same time to an equilibrium of one model and a disequilibrium of another.

Several things are noticeable about Machlup’s position on equilibrium; here are some of these:

It seems very different from that of fellow Austrian, Friedrich von Hayek. In the the 1937 “Economics and Knowledge” paper Hayek insists that it equilibrium is only a useful analytical tool if we are indeed convinced (and can show) that there is in fact a tendency towards equilibrium in the real economy. In other words, Hayek does not seem to make the strong distinction between equilibrium “as a tool for analysis” and as an “operational concept” that Machlup introduces.
Still, Machlup applies an essentially Hayekian argument in arguing that there is this strong distinction: “The ‘real’ world surely has infinitely more variables than any abstract model, and their actual interrelations are neither known nor, I fear, knowable (partly because they probably change unpredictably over time).”
One wonders what Machlup would have thought of computable general equilibrium models.