Professor Samuelson on Sraffa and the classical economists

In this article, part of an ongoing discussion, Samuelson (2000) is taken as the occasion for a critical examination of Samuelson's work on the classical economists and Sraffa, a subject of continuing interest for that author, especially after Sraffa (1960). The article argues for the existence in Smith and Ricardo of an alternative approach to distribution and prices, and it aims at a critique of Samuelson's contention that 'Smith, Ricardo and J.S. Mill used essentially the same logical paradigm as did Walras and Arrow Debreu' (2000: 140). In the first two sections, the attempt by Arrow (1991) to detect in Ricardo a theory of prices independent of demand - and founded instead on a real wage determined separately from, though not necessarily independently of, prices and the non-wage distributive variables - is considered with its implication of the wage entering the determination of the latter as an 'intermediate datum' of the theory. This then makes it possible to outline the characteristic analysis we find in Smith and Ricardo, where the wage as 'intermediate datum' entails a similar treatment of the output levels. The resulting theoretical structure is then used in order to answer, in sections III and IV, the two basic criticism, that Samuelson has advanced against Sraffa (1960). While the claimed dependence of the (1960) prices on an assumption of constant returns is voided by the mentioned treatment of outputs as intermediate data, the relevance of the Standard commodity, as well as that of Ricardo's 'invariable measure of value' is explained by the needs of determining non-wage incomes as a difference or 'residual', the essence of the theoretical structure under consideration. Section V then deals more directly with Samuelson's denial of the existence of a classical paradigm of economic theory. His arguments and interpretations are found to be in contrast with central features of Smith and Ricardo's work and, in particular, with their theory of wages. Thus, the admission of labour unemployment in 'normal' competitive positions compels Samuelson to a highly questionable interpretation of the chapter 'On Machinery' in Ricardo's Principles. In section VI, finally, the attribution to 'Sraffian literature' of a central concern for what Samuelson sees as 'steady states', but are in fact the traditional 'normal positions' of the economy leads the article to the deficiencies of neoclassical theory - an issue inevitably underlying the debate on the Classical paradigm. The dependence of the traditional versions of the theory, based on normal positions, on the notion of capital as a single magnitude - which forced the generalized abandonment of those versions in pure theory after the early stages of the capital controversies - is argued to emerge as equally present in the contemporary reformulations of the theory, thus affecting them, it is argued, no less than it did the abandoned earlier versions.