While Keynes Lived (Pre-1946)

The early history of the Cambridge Post Keynesians is—mercifully—rather less complicated. It begins in 1930, with the publication of Keynes’s A Treatise on Money and the emergence of an informal discussion group consisting of some of his younger colleagues that came to be known (after the complex intersection in the Charing Cross Road in Central London) as the Cambridge Circus. Its members included Kahn, Austin and Joan Robinson, Sraffa and Meade, but not the rather younger Kaldor, who was still a student at the London School of Economics (LSE). They attempted, eventually successfully, to persuade Keynes to abandon the Wicksellian theoretical foundations of the Treatise, in which macroeconomic adjustment takes place through movements in the rate of interest and the price level but not in the volume of output and the level of employment. In the circumstances of 1930-1931, with the Great Depression intensifying and unemployment rising sharply in every capitalist country in the world, this was easier than it would have been a few years earlier.

The work of the Circus is recorded in some detail in the two principal biographies of Keynes (Moggridge 1992: 531-536; Skidelsky 1992: 531— 536), in earlier work by Harcourt (1985, 1994) and in the reminiscences of two of the participants (Kahn 1984; Robinson 1978). It ceased to operate in mid-1931, having accomplished its mission. Between then and the publication of The General Theory in February 1936, the young economists extended their work on the fundamentals of macroeconomic theory and continued to exert some influence over Keynes. The contributions of Harrod and Meade in the mid-1930s do not belong here, since both were now at Oxford rather than Cambridge (as Harrod had always been), and the peculiar nature of Sraffa’s work will be discussed below. Most important were Richard Kahn and Joan Robinson. Kahn’s great contribution to the writing of The General Theory is recorded in the chapter devoted to him in this volume. Robinson’s Essays in the Theory of Employment were substantially complete by the end of 1935 (Robinson 1937). Somewhat uneven in quality, they dealt with the labour market, the macroeconomics of the long period, the open economy and issues of economic policy and methodology. At their best, these essays are ‘full of insight and imagination, shedding light on obscure corners of the General Theory, probing some of its crucial weaknesses, and anticipating many of the awkward questions which Sidney Weintraub, Paul Davidson and their North American colleagues were later to ask of the neoclassical “Keynesian” synthesis’ (King 1996: 165).

As for Sraffa, he mounted a brief but incisive attack on the Wicksellian economics of Friedrich von Hayek, denying the existence of a single ‘natural’ rate of interest and questioning both the operation of any mechanism that established macroeconomic equilibrium at full employment and the validity of marginalist theory more generally (Hayek 1931, 1932; Sraffa 1932a, b; Roncaglia 2009: 30-34). Thereafter Sraffa retreated into himself, concentrating on the editorial work on David Ricardo for which he became famous but publishing nothing on this or any other topic until volume I of Ricardo’s Collected Works finally appeared in 1951.

Sraffas views on The General Theory remain a matter of some controversy. Based on a close examination of his unpublished papers, Heinz Kurz concludes that Sraffa was severely critical of Keynes on several theoretical questions. His objections related, first, to the argument in A Treatise on Money that the price level of consumption goods was determined independently of the price level of investment goods; second, to the notion of the marginal efficiency of capital set out in The General Theory; third, to the view that the money supply was controlled by the banking system, and therefore determined exogenously; and, fourth, to the liquidity preference theory of interest. ‘In important respects’, Kurz concludes, ‘he felt that Keynes had granted too much to received economic theory. Keynes’s new theory exhibited several loose ends and contradictions and retained in new garb marginalist concepts that Sraffa deemed untenable’ (Kurz 2010: 201).

I have two reservations about this conclusion. First, while these issues are significant they are not really central to the main concerns of The General Theory. They do not go to the principle of effective demand, the explanation of involuntary unemployment or the critique of Say’s Law (all different ways of summarising Keynes’s fundamental message). Each one of Sraffas criticisms could be accepted without seriously damaging the core of Keynes’s argument, and indeed his third—endogenous money—has been developed by Post Keynesian theorists like Kaldor and Basil Moore to strengthen the revolutionary core of Keynesian macroeconomics. Whether Sraffa himself realised this, and what he thought of that analytical core, cannot be easily inferred from Kurzs account of the archival evidence.

Second, Sraffa’s criticisms have little or no bearing on the main policy implications of The General Theory, variously characterised as cheap money (Tily 2007), counter-cyclical fiscal policy based on ‘functional finance’ rather than ‘sound finance’ (Lerner 1943) and a substantial degree of direct regulation of private sector investment by the state, the ‘somewhat comprehensive socialisation of investment’ alluded to by Keynes himself in The General Theory (Keynes 1936 [1973]: 378). Again, Sraffa seems to have remained silent on all these questions, in print and out of it. He was certainly a man of the Left, and Pasinetti assumes that he was also a Keynesian on matters of macroeconomic policy, but there is again no direct evidence, either way.

Meanwhile, an outsider had briefly joined the ranks of the Post Keynesian pioneers. Once described by Geoff Harcourt as probably the best all-round economist of the twentieth century (ahead of Keynes), the Polish emigre Michai Kalecki turned up in Cambridge in November 1937, looking for work. The brief success and eventually conclusive failure of his quest are described in some detail in the recent biography of Kalecki by Jan Toporowski (2013: Chapters 11-14). Kalecki had already achieved a great deal before his arrival in Cambridge, not least the publication of a remarkable mathematical model of the trade cycle in the newly established journal Econometrica (Kalecki 1935), and Keynes described him as ‘something of a genius’ (Keynes quoted in Toporowski ibid.: 109). But he was also a foreigner, had no teaching experience and no university degree (merely a diploma in engineering). Kalecki spoke dreadful English and had few, if any, social graces; being Jewish would probably also not have helped him (though it seems to have done Kahn and Sraffa no great harm).

At all events, Kalecki was not a strong candidate for a college fellowship. Instead he needed a research job, and at the beginning of 1938 Cambridge had none to offer. Keynes, Kahn and Robinson did their best to help him find employment elsewhere, and Sraffa tided him over with a loan. Finally, at the end of the year, the Cambridge Research Scheme began operation, with Kalecki as the only full-time employee. He already had some enemies, not least the powerful chair of the Faculty Board, Dennis Robertson, who regarded trade cycle theory as his personal property and did not welcome interlopers. The microeconomic focus of the research that he was called on to perform did not suit Kalecki, and by mid-1939 criticism of his work from Arthur Bowley, Richard Stone and even from Keynes himself produced a response from Kalecki of ‘barely suppressed outrage’ (ibid.: 133).

He soon left Cambridge for the more congenial surroundings of the Oxford Institute of Statistics, where he worked most productively during the war years before leaving Britain in 1945 to spend the remainder of his career briefly in Canada and then in the USA and Poland. Oxford’s gain was very clearly Cambridge’s loss, and in Section 6 I shall speculate on the benefits that the Cambridge Post Keynesians might have obtained if Kalecki (who died in 1970) had remained there for the final three decades of his life.

The General Theory was a poorly structured and in some ways also a poorly written book, which left many questions unanswered and many theoretical issues unresolved. Keynes was aware of the problems that this posed, writing a summary and clarification of the arguments for his American readers (Keynes 1937) and promising Ralph Hawtrey that he would produce a set of ‘footnotes’ to the book dealing with the more important criticisms that had been made. It is tempting to believe that these ‘footnotes’ might have been expanded into a substantially revised second edition of The General Theory. Keynes’s ill health—he suffered a severe heart attack early in 1937—and his work as a more than full-time public servant during the Second World War prevented him from even beginning work on any such project before his death in 1946. We can only guess at what he might have achieved, and what subsequent difficulties might have been prevented. The contributors to the two-volume ‘second edition’ of The General Theory, edited by Harcourt and Riach (1997), offer some valuable suggestions.

The war did have one beneficial effect for the Cambridge Post Keynesians. This was the arrival of Nicholas Kaldor, with the relocation of the LSE to Cambridge. Kaldor had already abandoned his support for Hayekian macroeconomics, and between 1940 and 1944 he made important contributions in a Post Keynesian vein to both theory and policy, publishing an ambitious model of the trade cycle that attempted to formalise the informal insights that he had found in Chapter 22 of The General Theory and producing a series of detailed empirical studies that estimated the necessary fiscal policy conditions for maintaining full employment after the war (see King 2009: Chapter 3). A war Circus began to operate, involving Kaldor, Robinson, Sraffa and (when he could escape from official duties in London) Kahn. Kaldor and Robinson became especially close. They cooperated on plans for the socialist reconstruction of Britain in peacetime, adding a microeconomic dimension to the existing programme for a macroeconomics of full employment without inflation (Harcourt and Kerr 2009: Chapter 5; King 2004).

Sraffa’s involvement in the war Circus was more social than intellectual. We have seen that there are grounds for believing that he was secretly sceptical about some aspects of The General Theory. There is even less evidence concerning

Keynes’s opinion of the Sraffian system. Here the responsibility is entirely Sraffas. Deeply, obsessively secretive, Sraffa after 1928 seems to have disclosed nothing to anyone (except the mathematician Abram Besicovitch) about the formal rehabilitation of classical political economy that he would eventually publish 32 years later as Production of Commodities by Means of Commodities (Sraffa 1960). We do not know what Keynes made of the early disclosures (nor indeed how detailed they were), and we can only guess what he would have made of Production ofCommodities had an early draft been offered to him before his death.

We can, however, make an informed guess. Sraffa believed (correctly) that he had fatally undermined the Marshallian theory of value and distribution. This belief could be inferred, more or less, from his well-known 1926 Economic Journal article, which had been published by Keynes himself as editor (Sraffa 1926). Yet a strong case can be made that Keynes remained a Marshallian right down to The General Theory and beyond, as Paul Davidson and other ‘fundamentalist Keynesians’ have always maintained (see Hayes 2006 for a very cogent restatement of this view), along with authors as diverse as Leijonhufvud (2006) and Kurz (2010: 185). Certainly, Keynes had very little time for Ricardo. Sraffa must have known all this, and he must have disapproved of it. As for Keynes, it is hard to see him responding favourably to Sraffas system and its ‘prelude to a critique’—of Marshall, above all others.

From his arrival in Cambridge in 1927, Sraffa was one of Keynes’s closest friends. Just how much time they spent in each other’s company in the late 1920s and 1930s will never be known, and the folklore on this question is not entirely convincing: as one who has spent more than his fair share of hours trawling through the second-hand bookshops of Melbourne (population four million), I find it hard to believe that Keynes and Sraffa could have occupied very much time in this way in interwar Cambridge (population 50,000), as is sometimes claimed. If they really never did speak about economic theory—or policy—on their excursions, it can only have been as the result of an (unspoken?) agreement not to jeopardise a valued friendship by potentially acrimonious debate. As Bob Cord reminds me (personal communication, 4 June 2015), this was something of a Cambridge tradition: A.C. Pigou refused to talk about economics unless it was absolutely necessary.

Pasinetti’s verdict is that it ‘seems reasonable to conclude that it is difficult to find clear evidence supporting the view of a relevant influence, on scientific grounds, of Keynes on Sraffa; and at the same time that it is equally difficult to find clear evidence of a substantial influence of Sraffa on Keynes’ (Pasinetti 2007: 194). But this misses the crucial point, which is not about influence.

Did Keynes and Sraffa actually agree on fundamentals? That is the central issue, which Pasinetti simply avoids. I shall return to it in Section 6.