On Walras's Concept of Equilibrium

Abstract

The view that Walras's equilibrium concept refers to a temporary general equilibrium with stationary expectations has come to be the conventional opinion within the history of economics. This interpretation overlooks salient aspects of Walras's original equilibrium concept: (i) that it refers to a centre of gravitation, thus equilibrium must be assessed together with the adjustment mechanisms that are supposed to bring the economy towards its position of rest; (ii) that it attempts to represent a persistent position of the economic variables; and (iii) that it must not be regarded as a mere artificial model disconnected from reality, but as a device to understand how actual economies work. We therefore conclude that the scope of Walras's work can be better grasped by interpreting his equilibrium system as aiming to describe a long-period position of the economy.

Acknowledgment

Special thanks are due to Fabio Petri for his invaluable help and careful reading of earlier versions of this article. We also thank Enrico Bellino, Amit Bhaduri, Roberto Ciccone, Franco Donzelli, Saverio Fratini and Geoffrey Harcourt for useful discussions on previous drafts. We are also grateful to an anonymous referee for valuable comments. We extend our gratitude to the late Pierangelo Garegnani, to whose memory we dedicate this paper.

Notes

¹If the endowments of capital goods are part of the data, substitution among different processes of production will generally be ruled out. It might well happen, as a result, that the factor demand curves for capital goods would be highly rigid, yielding a zero rental price in equilibrium. The elasticity of labour demand might also be very low ‘entailing an analogously implausible equilibrium real wage’ (Petri, 2004, p. 44).

²Modern general equilibrium theory, also known as neo-Walrasian, covers both the TGE approach and Intertemporal General Equilibrium theory.

³For example, Bliss (1975, p. 28, emphasis added) has acknowledged that ‘even if the equilibrium were to be stable there might not be enough time within the space of a “week” for prices to adjust to an equilibrium […] In the face of the foregoing problems we could regard the object of our investigations not as “the economy” but as an “economic equilibrium”.’

⁴The reason for the choice of authors will become evident in what follows. We shall focus mainly on writers who have engaged in the highly significant, though not very well known, Italian branch of the debate on Walras's theory triggered by Pierangelo Garegnani's (1960) book on Il capitale nelle teorie della distribuzione. The recently published provisional draft of a 1962 paper by Garegnani on Walras's capital theory is indicative of the continuing relevance of this part of the debate (see Garegnani, 2008). The main reason for our concentration on this set of texts is that they address the issue of how the interpretation of Walras as a TGE theorist took root.

⁵We are grateful to an anonymous referee who suggested this second classification for the evidence.

⁶Zaghini (1968, p. 371) argues that ‘by increasing the available amounts of capital goods [yielding the highest rate of return] with respect to the quantities of the other ones, the prices of their services will diminish, and it will also decrease their corresponding rate of return.’

⁷Walras ([1954] 1977, p. 112) allowed for the possibility that an unstable position might satisfy the condition that demand matches supply. However, in order for any position to be a solution to the problem of equilibrium, market-clearing is only a necessary, but not a sufficient condition: Walras insisted that a genuine equilibrium must be stable. His contemporary, Knut Wicksell ([1893] 1970, p. 85), held the same view: ‘the equality of supply and demand is indeed necessary, but at least from the theoretical point of view, not a sufficient condition for equilibrium on the market, supposing the latter to be stable—if, that is to say, the proportion of exchange would automatically return to (approximately) the same position after an accidental shifting’.

⁸This statement applies straightforwardly to economies in which final goods are produced by different types of land and labour. In such economies, out-of-equilibrium trading does not significantly affect the ownership pattern of resource endowments because sales and purchases of factor stocks are neglected. However, it could be extended, as we argue below, to progressive economies producing capital goods, as long as the quantity of money capital in the economy and its distribution among the population are included as part of the data of the equilibrium.

⁹For an assessment of the Arrow problem, see Petri (2004, pp. 187–189).

¹⁰This is so, because it might well happen that if the capital good yielding the highest return is produced with a relative high proportion of itself, the increase in its demand will raise its rental price relatively to its cost, so the rate of return will increase even more (see De Vivo, 1976; Garegnani, 1960, pp. 103–114; 1990, pp. 16–19). This is why there is no basis to support Zaghini's statement that ‘within the single equilibrium period [there will be] a tendency towards uniformity of the rates of returns of all capital goods’ (Zaghini, 1968, p. 377, emphasis in original).

¹¹As aptly pointed out by an anonymous referee, in any case, a single admission on Walras's part that the uniform rate of return condition cannot be satisfied does not in the least mean that he was prepared to drop the condition itself.