I.Marx was one of the first economists to study a process of capitalist growth within a multi-sectoral framework; he considered a two-sector economy in which one sector produced the means of production and the other one produced consumption goods. He dealt with this problem by using numerical examples in the second volume of Capital [13]. In recent times the Marxian schemes of expanded reproduction have received the attention of many economists who have formalised and generalised them.' Some economists (e.g., Lange [7]) have pointed out the similarities between Marx's schemes and Leontief s input-output technique. Others have gone further and have regarded the schemes as an attempt by Marx to construct a multi-sectoral model of balanced growth. Recent examples of this approach to the Marxian analysis of expanded reproduction are the models by Brody [2], Morishima and Catephores [ 151, [ 161, Roemer [2 13.On one hand, Morishima and Catephores hold that the Marxian theory of expanded reproduction can be properly expressed and generalised in terms of a von Neumann model and they have constructed a model which is called "the Marx-von Neumann approach"; on the other hand, both Brody and Roemer have generalised the Marxian schemes in a way which is very similar to a dynamic Leontief model.The present paper is an attempt to show that, despite some superficial similarities, especially due to the particular numerical example chosen by Marx, the Marxian schemes cannot be interpreted as attempts to construct a model of balanced growth. It will be argued that, in fact, Marx was not interested in the analysis of balanced growth and he constructed his schemes in order to answer questions which are different from those considered by modern models of balanced growth. Marx was not at all interested in studying balanced growth because he regarded this kind of growth as impossible or, at least, highly improbable for the capitalist mode of production. He used the schemes of expanded reproduction in order to illustrate the process of reproduction and circulation at a macroeconomic level in a clear and synthetic way, *I wish to thank S. Bowles, G. Chiodi, C. Gregory, L. Me!dolesi, A. Roncaglia, and an anonymous referee for their helpful comments on an earlier draft of this paper. Any mistake, of course, is my sole responsibility. 'Among the most recent formal versions of Marx's schemes, cf. [7], [2], [4] , [15], [6], and [21].
383
AUSTRALIAN ECONOMIC PAPERS DECEMBERand in order to show that overproduction crises are not only possible but, indeed, highly probable within a capitalist economy.