ABSTRACT
I The question has often been raised as to why Keynes in The General Theory accepted the Marshallian framework of perfect competition.1 This is surprising because he was actively involved in the discussions following the publication of Sraffa’s ‘The laws of returns under competitive conditions’ (Sraffa, 1926) which had launched a severe attack on the Marshallian theory. Indeed, he was responsible for making Sraffa’s argument known to English-speaking readers. In the Spring of 1926, Keynes invited Sraffa to write for the Economic Journal an English version of an earlier article written in 1925. The Italian article, Keynes wrote, had very favourably impressed his co-editor, F.Y. Edgeworth, who had died in February 1926, before he could make the invitation himself. Sraffa’s critique of the assumption of decreasing and increasing returns, under conditions of perfect competition, led to the conclusion that the ‘old-fashioned’ theory of cost of production, based on constant returns, was a better foundation for a theory of value (Sraffa, 1925). However, in accepting Keynes’s invitation (in a letter dated 6 June 1926), Sraffa outlined the content of the proposed ‘sequel’ on a different basis, i.e. in terms of an assumed ‘generality of some elements of monopoly’. He intended to prove that as soon as imperfection is introduced into a system of competition ‘equilibrium is reached in a way extremely similar to that of monopoly and very far from that of competition’ (Roncaglia, 1978, p. 13). In fact, the article published in 1926 proposed, as a way out of the Marshallian inconsistencies, ‘to abandon the path of free competition and turn in the opposite direction, namely towards monopoly’ (Sraffa, 1926, p. 542).2 It was in developing the implications of Sraffa’s suggestion that the theory of imperfect competition was elaborated in the late 1920s and early 1930s. The attitude of Keynes towards imperfect competition was described by Joan Robinson as a dismissive one: ‘Neither Roy Harrod nor I could get Maynard to take an interest in “marginal revenue” ’ (Robinson, 1979, p. 173). However, we have evidence that he had endorsed Sraffa’s argument: when Harrod presented the concept of marginal revenue in an article submitted to the Economic Journal
in the Summer of 1928,3 Keynes rejected the article with a letter of 1 August 1928, but the same letter reveals a complete endorsement of Sraffa’s point:
Dear Roy, I hope you will save something out of the article. But if you pursue the matter on the basis of assuming full competition will you not be up against all the usual increasing and decreasing returns difficulties-a difficulty, that is to say, as to how firms of different sizes can exist together in conditions of full competition unless there is constant return? I am still an adherent of the theory put forward by Sraffa in his Journal article to the effect that observed results could only be explained by assuming that each producer has within certain limitations his own private and local market.4
