Keynes and Macroeconomics
A Treatise on Money had a long gestation period: its first draft table of contents was dated 14 July 1924. By 2 August 1929, Keynes had received proofs of 320 pages of a single volume version of the Treatise. Working to a publication date of 1 October, Keynes began proof corrections and final rewriting. But he became increasingly dissatisfied with what he had done and began to attempt more substantial revisions. By the end of the month, he was writing his American publisher Alfred Harcourt that the revisions were so drastic that he was going to go over to a two-volume version and could not foresee publication before January 1930 (XIII: 117).[1]
His public life added complications. On 4 November 1929, Philip Snowden, the Chancellor of the Exchequer, announced the appointment of a Committee on Finance and Industry under Lord Macmillan with Keynes as a member. Four days later, he had lunch with the Prime Minister, Ramsay MacDonald. Two further meetings followed in early December. The result was the appointment of an EAC with Keynes again a member. The Macmillan Committee and the EAC made large demands on Keynes’s time: between November 1929 and the dispatch of the final text of the Treatise to the printers, his diary records 65 meetings of both bodies. Revised proofs started coming at the end of 1929, but it was April 1930 before a final set was available. Keynes was so busy that he did not have time to read some of the comments his readers made on the proofs: in the case of Ralph Hawtrey, he did not reply to some comments until two months after the book was published!
Keynes sent his book for publication on 14 September 1930, when he wrote to his mother:
This evening, at last, I have finished my book. It has occupied me 7 years off and on,—and one parts from it with mixed feelings ... Artistically it is a failure—I have changed my mind too much during the course of it for it to be a proper unity. But I think it contains an abundance of ideas and material (XIII: 176).
Keynes set out much of his Treatise analysis in terms of his ‘Fundamental Equations’ with P being the price level of consumption goods, R their real output, E the community’s nominal income, O the output of consumption goods and investment goods, T he income earned in the investment goods industries, I the value of new investment, S nominal earnings, and П the price level of output as a whole:
These ‘purely formal’ equations were ‘identities, truisms which tell us nothing in themselves. In this respect they resemble all other versions of the quantity theory of money’ (V: 125). Because saving was defined as ‘the sum of the difference between the money income of individuals and their money expenditure on current consumption’ (ibid.: 113) and investment as ‘the value of the increment of capital during the period’ (ibid.: 114), it followed that ‘the value of current investment...will be equal to the aggregate value of savings and profits’ (ibid.). Keynes’s definition of income was chosen to permit his measures of savings and investment to differ from one another. This was desirable, given that ‘decisions which determine savings and investment respectively are taken by two different sets of people influenced by different sets of motives, each not paying very much attention to the other’ (ibid.: 250).
Hawtrey, for example, pointed out that since the Fundamental Equations were tautologies, Keynes could not use them to suggest that an excess of investment over saving caused a rise in prices relative to costs, since he had defined saving and investment in such a way that the existence of such an excess meant that prices had risen (XIII: 132-133). That assumption was also challenged by his younger Cambridge colleagues as the widow’s cruse fallacy: ‘Thus profits, as a source of capital investment for entrepreneurs, are a widow’s cruse which remains undepleted however much of them may be devoted to riotous living’ (V: 125). The point was picked up by Robertson, who acknowledged the contribution of Meade in ‘putting me on the right track here’ (Robertson 1931: 406, fn. 7), and Hayek (1932: 3).
Keynes’s view of the relationship among saving, investment, and interest rates was similar to those of Swedish contemporaries. However, Keynes, unlike the Swedes or Hayek, was aware of the underlying capital theoretic difficulties.
Keynes’s Treatise adoption of a saving-investment framework increased the importance of the long-term rate of interest. This was coupled with a diminished emphasis on the effects of changes in short-term rates, which was the result of his increased knowledge of commodity and capital markets where the volume of investment in working capital was relatively insensitive to changes in the short-term rate of interest.
Given his own previous and his contemporaries’ emphasis on changes in Bank rate, and the fact it remained the most visible and readily available instrument of monetary policy, Keynes was faced with the problem of explaining how changes in short-term rates could affect the long-term rate. He developed a theory of what is known as the term structure of interest rates, that is, the relationship between interest rates on securities with different terms to maturity, a problem not previously discussed in the literature. This showed much concern for ‘conventional’ expectational factors which impeded the flexibility of long-term rates.
In the Treatise, Keynes shifted his ground from the Tract and accepted the desirability of an international standard of value which would be an improvement on the gold standard. The principal disadvantage of any international standard with perpetually fixed exchange rates was the dilemma it could pose for a country whose domestic circumstances called for a monetary policy different from that which adherence to the standard entailed. The dilemma could become acute for a member of such a system with international capital mobility whose efficiency wages (money wages adjusted for productivity) were too high to allow it to achieve full employment at existing international interest rates, although these rates might be appropriate for the system as a whole. The authorities in the problem country could not use monetary policy to increase employment because lower interest rates would induce capital outflows and a loss of reserves which would force the authorities to raise interest rates. In these circumstances, there was a reserve weapon available—loan-financed public works or capital development schemes with an accommodating credit policy to meet the needs for working capital, so that investment and employment could expand at existing interest rates. This was the ‘special case’ for public works (VI: 337). In the general case, monetary policy concentrating on the long-term rate of interest and the adjustment of supplies of credit to the varying needs of the economy was the appropriate technique of aggregate economic management.
Late 1930 to early 1932 was the period during which professional economists absorbed and reacted to the Treatise. By late 1931, Keynes made a series of statements to others, including Japanese readers of the Treatise (V: xxvii), of his intention that he was ‘endeavouring to express the whole thing over again more clearly and from a different angle; and in 2 years’ time I may feel able to publish a revised and completer version’ (XIII: 336).
The sources of criticism varied. One had been available before publication— Hawtrey’s comments on Keynes’s proofs were eventually published as Chapter VI of The Art of Central Banking in 1932 and did not get Keynes’s attention until after publication. Other academic reviews and comments included Hayek’s
‘Reflections on the Pure Theory of Money of Mr. J.M. Keynes’, Robertson’s ‘Mr. Keynes’s Theory of Money’ and further discussions with Robertson and Pigou. Then there were discussions with younger Cambridge colleagues, academic visitors, and students.
These discussions reflected the arrival of a younger generation in Cambridge. Richard Kahn and Piero Sraffa had already had some impact on the Treatise. The others included Austin Robinson, a recently appointed University Lecturer and Fellow of Sidney Sussex College, his wife, Joan Robinson, about to commence a career of prolific publication, and James Meade, recently appointed a Fellow of Hertford College, Oxford, who was spending the year at Trinity College studying economics. This group began to meet in the Michaelmas Term of 1930 in Kahn’s rooms in King’s to discuss the Treatise. Their discussion continued into the Lent Term before they became a more formal ‘seminar’ held in the Old Combination Room of Trinity. To this ‘seminar’, participation was by invitation, with undergraduates wishing to receive invitations having to satisfy an interviewing board of Kahn, Austin Robinson, and Sraffa. The two sets of discussion took on the name of the Circus, which has an alternative definition of a scene of lively action and a group of people engaged in common activity—in this case trying to understand the Treatise. Keynes did not attend. Kahn played, in Margaret Meade’s words on a later occasion, the role of the messenger angel, ‘who brought messages and problems from Keynes to the “Circus” and who went back to heaven with the results of our deliberations’ (XIII: 338-339; Lambert 1969: 250).
As noted, the criticisms that Keynes received from different sources varied. In many cases, they reinforced each other. If they came from outside and they were taken up or echoed by members of the Circus or merely Kahn, who from this time was Keynes’s intellectual confessor on economic-theoretical issues, they were likely to strike home. Kahn’s role in the process was critical for, as a Fellow of King’s, he could nag Keynes over an issue until Keynes saw where the difficulty lay. From the Circus and other sources of dissatisfaction to the beginnings of creation, where Kahn also played an important role, was but a short step.
In addition, Keynes was receiving suggestions of an alternative focus that would successfully take account of movements in output and employment. Just before Keynes finished the Treatise, Kahn, puzzled by Chapter 6 of Can Lloyd George Do It? (IX: 102-110) which dealt with the amount of employment that would be provided by the Liberal Party’s 1929 election proposals for public works, developed a framework for estimating the effects of loan-financed government expenditure on a public works programme
(now known as the employment multiplier). He presented his early results to the Committee of Economists of the EAC, of which he was secretary and Keynes the chair, in September 1930. In the months that followed, in conjunction with Colin Clark of the EAC and Meade in Cambridge, Kahn worked out the implications more fully before publishing them in the Economic Journal for June 1931 (Kahn 1931). Kahn’s article had two implications for Keynes’s work. Focusing on the supply curve for output as a whole, Kahn realised that the Treatise’s Fundamental Equations were a limiting case. Keynes had assumed that output as a whole was inelastic to changes in demand. Relaxing that assumption, as Hawtrey had also realised in his criticisms of the Treatise, markedly affected the analysis and brought it ‘very much closer to the actual conditions that prevail to-day [1931]’, although neither Hawtrey in 1930 nor Kahn in 1931 perceived how much that difference would be (Kahn 1972: 9-10). Kahn also provided in ‘Mr. Meade’s Relation’ an outline statement of the resources available for this investment when the supply of output was less than completely elastic to demand changes. The argument and its corollaries were not yet clear to Kahn—much less to Keynes.
Thus in the summer of 1931 Keynes had started to work it out all over again. In the course of a visit to Tilton which ended on the weekend of Britain’s departure from the gold standard on 21 September, he and Kahn had been discussing some of the issues (XIII: 373-376). On 22 November, Keynes reported to his wife Lydia that after the usual distractions of the King’s audit were over, ‘I have begun writing in my chair about monetary theory’ (Keynes quoted in Moggridge 1992: 536). He sitting in that chair in the famous cartoon by David Low (XIII: xvi). By late February 1933, with The Means to Prosperity (IX: 335-366) ready for publication in The Times, he had moved a considerable way from his position of 1930-1931.
To sort out how far Keynes was towards what became The General Theory at any time up to his sending it to his publisher for the last time at the end of 1935, we have to have some notion of the distinctively ‘new’ elements in the final product. I should note, however, that despite the initial discussion in the late 1930s and immediately after the Second World War, most of this interpretative literature dates from 1961. It followed a period of over 20 years of professional agreement as to what The General Theory was about. Many of the seemingly endless subsequent discussions reflect attempts to look at Keynes’s work in the light of recent theoretical preoccupations or to claim Keynes’s paternity for one’s own heterodox ideas. But it also reflects the richness of a book that was breaking new ground—witness Keynes’s own ‘long struggle of escape’ (VII: xxiii) from earlier modes of analysis.
After the publication of the book, in virtually identical letters to Roy Harrod and Abba Lerner, both of whom had published evaluations, Keynes set out the stages of his quest:
You don’t mention effective demand or, more precisely the demand schedule for output as a whole, except in so far as it is implicit in the multiplier. To me, regarded historically, the most extraordinary thing is the complete disappearance of the theory of the demand and supply of output as a whole, i.e. the theory of employment, after it had been the most discussed thing in economics. One of the most important transitions for me, after my Treatise on Money had been published, was suddenly recognising this. It only came after I had enunciated to myself the psychological law that, when income increases, the gap between income and consumption will increase,—a conclusion of vast importance to my own thinking but not apparently, expressed just like that, to anyone else’s. Then appreciably later came the notion of interest being the measure of liquidity preference, which became quite clear in my mind the moment I thought of it. And last of all, after an immense lot of muddling and many drafts, the precise definition of the marginal efficiency of capital linked one thing with another (Keynes to Harrod, XIV: 85; Keynes to Lerner, XXIX: 215; underlining in original).
This account of the stepping stones to The General Theory has four steps: (1) the consumption function, (2) the theory of effective demand, (3) ‘the notion of interest being the measure of liquidity preference’, and (4) the marginal efficiency of capital.
Scholarly accounts of the making of The General Theory have normally started from this list and sometimes added or subtracted items. Peter Clarke more or less stuck to the list, although his discussion adds another element, the notion that the whole might be more than the sum of its parts, which had been in Keynes’s thought since his undergraduate contact with G.E. Moore and can hardly be ‘new’ to The General Theory. However, Keynes may have recognised the greater significance of organic unities in economics (Clarke 1988: 259-59, 269-272; O’Donnell 1989: 177-178). Don Patinkin, arguing that the notions of liquidity preference and the marginal efficiency of capital were both previously available - the former in the Treatise and the latter in the work of Irving Fisher - concentrated on the development of the theory of effective demand, most notably the equilibrating role of changes in income implicit in the multiplier (Patinkin 1976). But it is clear from the available texts that Keynes did not come to the marginal efficiency through Fisher but rather worked it out for himself and only had Fisher’s work pointed out after the fact (Kent 2014). Robert Dimand (1988), in contrast, adds to Keynes’s list the notion that the labour market may fail to clear because workers cannot adjust their real wage by bargaining for nominal wages.
The dating of the development of Keynes’s thought may also be subject to scholarly disagreement for another reason—the nature of the evidence required to indicate that he had taken a particular idea into his mode of thought. At one extreme, one could take Patinkins approach in Anticipations of the General Theory (1982) towards others’ claims to have anticipated The General Theory— the exposition of the doctrine in a form intended for a professional reader- ship—a criterion suitable for the question of crediting a person with a scientific discovery. In his Keynes’s Monetary Thought (1976), when coming to settling an upper limit to Keynes’s own discovery of the central message of The General Theory, Patinkin used the evidence of lecture notes and other unpublished material (Patinkin 1982: especially 11, 16, 85; 1976). At the other extreme is Clarke’s (1988: 258) criterion of ‘indications of developments in [Keynes’s] thinking which represented his initial insight, even if they were only disjointed flashes of illumination’. Such a criterion will produce an earlier set of dates than either of Patinkin’s criteria, but one needs additional evidence to ensure that the disjointed flashes were more than just the ‘mirages’ that Keynes remembered:
To the extent to which one sees one’s destination before one discovers the route is the most obscure problem in the psychology of original work. In a sense, it is the destination that one sees first. But then a good many of the destinations so seen turn out to be mirages. Only a small proportion of one’s original intuitions survive the struggle of trying to find the route to them (Keynes to Falk,
19 February 1936, quoted in Moggridge 1992: 551).
The evidence from the documentary remains of the preparation for The General Theory is far larger than for the Treatise. There are draft tables of contents, fewer in number than for the Treatise, but a larger number of the related draft chapters surviving even from the earliest stages. Fragments, often running over several chapters, exist for over half a dozen drafts of the book, and there are three almost complete sets of proofs over which Keynes corresponded with other economists. Several sets of student lecture notes survive from his annual series of eight lectures. These were collected, transcribed, and collated by T.K. Rymes in his Keynes’s Lectures 1932-35: Notes of Students (Rymes 1989a) and distilled into Keynes’s Lectures 1932-35: Notes of a Representative Student (Rymes 1989b). In Keynes’s case, the eight sets of more or less independently taken notes from good students, most of whom became professional economists, show high levels of agreement over what seems to have been said. On occasion there are fragments of the draft from which Keynes lectured. Much more correspondence also survives from this period than from the years of the Treatise. Finally, there are Keynes’s own publications, the publications of others involved in the process of creation, and subsequent memoirs and reconstruction of what went on. It is probably the case that with The General Theory, scholars have the most voluminous record surrounding the creation of any classic work in economics.
What does this evidence show? If one takes the view that the concept of the marginal efficiency of capital originated with Fisher, the emphasis of the story would be on output equilibration and liquidity preference: there are traces of the former in the autumn 1932 lectures. If one takes Keynes’s letter to Harrod literally, that liquidity preference came after output equilibration, he must have had the former in place. This is Clarke’s argument (1988: especially 164). At this point, any discussion of the chronology of Keynes’s ideas runs into problems, for we have few dates for any fragments of manuscript until Keynes gives his next set of University lectures in the autumn of 1933, again under the title ‘The Monetary Theory of Production’. By then it is clear that he had seen the equilibrating role of fluctuations in output associated with the theory of effective demand. If we want to pick an earlier date in 1933, we have to rely on circumstantial evidence such as a chapter, ‘The Parameters of a Monetary Economy’, which could have been written soon after the Michaelmas 1932 lectures, or the revisions to his essay on Malthus for Essays in Biography taking advantage of Sraffa’s contemporary work on Ricardo which involved his correspondence with Malthus (X: 87-91, 94-100). Was it a coincidence that these passages on effective demand and Malthus’s relations with Ricardo were added in October and November 1932 or that the concluding passages concerning savings and investment and the rate of interest were added in later December 1932 and early January 1933? Then there is The Means to Prosperity, a series of articles which appeared in The Times, completed on 6 March 1933, and published between 13 and 16 March. After publication, Keynes decided to think of expanding the argument in several directions, the first being ‘the train of reasoning by which I arrive at the multiplier relating secondary employment to primary employment’ (XXII: 63). An article on ‘The Multiplier’ appeared in the New Statesman for 1 April 1933 and was incorporated into the American version of The Means to Prosperity. Both Patinkin (1976: 79, 1982: 33-34) and Dimand (1988: 158-159) dismissed The Means as evidence because it does not contain the necessary formal theoretical justification—which was hardly appropriate for the place where the article appeared. That came later in Keynes’s 1933 lectures and, for the marginal efficiency of capital, the summer of 1934. Even though he had started sending chapters to the printer on 13 September 1934, it was another 15 months before he returned the corrected proofs of the Preface dated 13 December 1935, and the index. The General Theory came out on 4 February 1936.
It was priced at five shillings to make it readily available to students and, as it turned out, ensured that Keynes made almost no money from it. The book was reviewed everywhere. Roger Backhouse (1999) collected a selection of reviews in English during the first year after publication which turned up examples in such unusual places as the Kingston Ontario’s Whig Standard. Of the reviews Keynes noted that he had ‘no companions...in my own generation, either of earliest teachers or of earliest pupils’ (XIV: 85): Dennis Robertson and Hubert Henderson were unimpressed; A.C. Pigou was positively irritated. The same could be said of senior American economists Frank Knight (born 1885) and Joseph Schumpeter (born 1883). He received a more sympathetic reception from Jacob Viner (born 1892). Finally, Alvin Hansen (born 1887) reviewed the book twice. His tone was sympathetic but suggested it would be no better than the Treatise, given its defective theory of interest and ‘its special notion of equilibrium at less than full employment’ (Hansen 1936: 83). Ironically, Hansen, who moved from the University of Minnesota to Harvard in 1937, would become Keynes’s major senior interpreter in the USA.
On both sides of the Atlantic, it was younger economists who took to the book, analysed it, and tried to develop it. Some of the enthusiasts had helped in the development of the theory—Harrod, Kahn, Meade, Austin, and Joan Robinson. Others had been Keynes’s supervision pupils—David Champernowne and Brian Reddaway—or had heard the theory developed in Keynes’s lectures—R.B. Bryce, Alec Cairncross, Walter Salant, and Lorie Tarshis—or had heard it at the London and Cambridge research student seminars where Joan Robinson and Kahn tried to spread the word—Abba Lerner. Others had taken it from the book itself—Kaldor and Hicks (newly appointed to a Cambridge Lectureship from the London School of Economics who reviewed it for the Economic Journal). In 1936-1937, often in fruitful collaboration and interaction, as at the September 1936 meeting of the Econometric Society at Oxford, many of the younger economists discussed the book (Brown 1988; Young 1987). Keynes read most of the papers and ‘approved’ of all: we know of no adverse comments. But then as he had told Hicks in 1935: ‘I deliberately refrain in my book from pushing anything very far, my main object being to press home as forcibly as possible certain fundamental opinions—and no more’ (Keynes quoted in Hicks 1973: 8).
Many of the expositions attempted to formalise the model and did so along remarkably similar lines, as Oscar Lange, an early user of such models, pointed out in 1938 (Lange 1938). All worked in terms of a simple model that was to become the mainstay of undergraduate expositions of the theory for the rest of the twentieth century.
Keynes’s reaction to the reviews and expositions was interesting. He replied only to Viner’s views at any length. In private, he expressed his views on some of the others, finding Pigou’s review ‘profoundly frivolous in substance’, and bemused by what he took to be Knight’s two main conclusions—‘namely that my book caused him intense irritation, and that he had great difficulty in understanding it’ (XIV: 87, XXIX: 257). Of the younger people’s reviews and expositions he was tolerant, perhaps because he believed that ‘if the simple basic ideas can become familiar and acceptable, time and experience and the collaboration of many minds will discover the best way of expressing them’ (XIV: 111).
- [1] All citations to The Collected Writings of John Maynard Keynes take the form of volume number followedby page number.
