The Sleepwalkers: Setting the Stage for a Changing Vision

The 1960s was an exciting decade to be in Cambridge (Harcourt 2009: 204).

Distribution, growth, employment, money, and pricing, the five defining themes of the economics of GCH, are those that were developed in the changing visions of the basic works of those whose pioneering classics provided the soil for the fertile, sympathetic, generous, and consistently open-ended^[1] theoretical structures he erected—and has left for posterity.

Distribution in its functional aspects, from Ricardo to Sraffa, growth as accumulation due to a judicious choice of techniques, from Smith to Robinson, employment stimulated by effective demand and due to a rejection of Say’s Law, from his great Jesus College predecessor, Malthus, to his Cambridge apostle, Keynes, money (and finance),^[2] from Marshall and the Cambridge tradition of Pigou, Keynes, and Kahn, and pricing in the Kaleckian lineage, were the five building blocks of the economics of GCH. Equilibrium in the short and long periods, expectations (in the sense in which the ‘Swedes’ of the 1920s and 1930s used anticipations—neither straitjacketed by the strictures of any kind of formal probability theory), returns to scale of many varieties, at many levels of aggregation and an indiscriminate switch between period and continuous analysis at the macroeconomic level and that between algebra and geometry, together with copious arbitrariness on transition dynamics, were those accoutrements to the fundamental five concepts. These accoutrements were the main causes of the paradoxes, perplexities, and confusions that plagued a full appreciation of the basics of the five defining themes.

In that ‘exciting decade of the 1960s’, GCH produced six of his undisputed and pioneering contributions to macroeconomic—and microeconomic, especially with regard to pricing and market forms—analysis—in

Harcourt (1963, 1965b, 1966, 1968,^[3] 1969b) and Harcourt and Kenyon (1976).^[4] These six, together with the four classics discussed in the previous section, form the core contributions of GCH to the changing visions of the economics that, eventually, became, in his competent and magnanimous ways, Post Keynesian Economics (PKE).^[5]

GCH himself considers Harcourt (1965b) his ‘favourite theoretical paper’! In an e-mail to me, dated 1 October 2015, GCH wrote as follows (italics added):

I was interested in your ranking of The Accountant in a Golden Age and the Two-Sector Model... I suppose I lean towards the second because the ideas in it came from me, whereas [Harold Lydall] had asked me to look into his puzzle with the accounting rate of profit vs the economic rate of profit. I’ve always felt that I should have asked him and Deane Terrell to be co-authors. I’ve also been grateful to Ken Wright who went after Kay when Kay criticised the OEP paper—I asked Ken to do so as I did not have the required maths to do so myself.

Of these six, the traumas of publishing three of them—namely, those of 1963, 1966, and 1976—are documented by GCH in one or another of his many splendoured writings. I want, however, to highlight just one of them— that associated with the important critique of Kaldor’s distribution mechanism, in a full employment growth context, first published as Kaldor (1955-1956) and Kaldor (1957).^[6]

Much of the unfortunate details of the story of the beautiful and thoroughly competent—from a serious ‘economics of Keynes’ point of view—Harcourt foray into the thorny field of functional distribution of income theory has been told by GCH himself, in his highly readable essay in the Shepherd edited volume (Harcourt 1995b: 73-74). The ‘punch line’ of the story is when GCH, immediately after presenting the paper at the King’s College seminar for research students run by Kaldor, ‘remembering going up to one particularly arse-licking graduate student afterwards’ (ibid.), and picking ‘him up by the lapels and [saying]: “Look, son, if you want to grease Kaldor at my expense, at least do your bloody home work first”’. I have never thought it good manners to ask GCH to reveal the identity of this particular ‘arse-licking graduate student’, nor has he, true to his civil nature, volunteered to identify the person. All I wish to add is that this kind of behaviour is, I think, more prevalent in our sorry profession now, than it was then—and I am a personal witness to this increasing ‘trend’.

GCH ends this particular story by his characteristically magnanimous statement (ibid.: 75): ‘I should add that Nicky and I subsequently became firm friends. You will find an affectionate and appreciative evaluation of him and his work by me in Economica, May 1988’. However, he does ‘add’, in Harcourt (2006a: 148, fn. 1): ‘It took me about 10 years to get our relationship back on an even keel and we ended up friends. Just before he died in 1986 I received an appreciative note from him for my review.. .of his Okun Lectures.. .in the Economic Journal’.

On the technical side, GCH was fundamentally correct in pointing out the inconsistency implied in the pricing process between sectors, within and between periods in Kaldor’s model (particularly in Kaldor 1957). But the period assumption in Kaldor’s growth model is fundamentally unfaithful to the stance Keynes took on this issue. In his letter to Robertson of 20 May 1933, and in the first section of the draft Chapter 7 in The Monetary Theory of Production—appropriately, from the point of view of GCH—titled ‘The Concept of an Accounting Period’ (Keynes 1979: 26, 73-76), there is ample evidence that GCH was comprehensively correct in this aspect of his criticism of Kaldor.

In Kaldor (1957) there are two common infelicities, mainly because Kaldor was not careful (as usual!). They were not formally taken up by GCH, although he does so implicitly, but discussed in detail by Champernowne (1971, especially the appendix) and McCallum (1969), who both analyse local stability^[7] and also for the continuous-time version of a period model, in which the justification in Kaldor was economic—and its inconsistency was, of course, pointed out by GCH. To analyse local stability and claim it has validity for the dynamics of a capitalist system is the customary vice of ‘pretty polite techniques’. To not know that even a ‘little bit of nonlinearity’—of which all of the referred to Kaldorian models are copiously endowed, and neglected—can destroy any kind of formal equivalence between a continuous and discrete time dynamic model, is inexcusable.^[8]

Finally, I should at least mention the highly innovative ‘Keio paper’ (Harcourt 1969b).^[9] It is a teaching model of Keynesian economics in the grand tradition of Hicks (1937) and Lindahl (1953—who uses the geometry of Goodwin’s innovative four-quadrant alternative to the more famous two-quadrant, IS-LM diagram (cf. Harcourt 1985: 415-416).^[10] In the opening paragraph GCH extols the virtues of the ‘algebraic method’ over the ‘geo- metric’-‘quadrant approach’. He adds (ibid.: 87; italics added):

The preference for the use of algebra rather than geometry arises from the view that the ‘quadrant’ approach can mislead students, who may settle for mechanical drill; it may confuse them about the applicability of their results and geometry does not always bring out clearly the limitations of the methodology used. The dangers are more easily avoided, it is believed, when algebra is used.

These strictures against the geometric method, and in favour of an algebraic method, can easily be reversed, even referring to the pernicious practice of the current frontiers in economic theory.^[11]

More importantly, GCH is, as always, absolutely candid on his comparative static analysis—essentially based on a judicious use of Marshallian ceteris paribus assumptions—and there is no pretence whatsoever of any kind of transition dynamics or its stability (ibid.: 87-88; italics in original):

Nothing is said, formally about the process of getting from one equilibrium position to another or whether the economy will actually do so, and any statement about changes as opposed to differences requires an act of faith (which is common to all believers but is not always made explicit). That comparative statics results are so applied to process situations is not stressed enough in the text books.

How very true! I only wish my teachers, and the textbooks they chose to teach from—before Goodwin—had stressed this important fact, or at least advised us to keep a barrel of salt by our sides, when authors pontificated about dynamics and stability. I envy those who were taught Keynesian economics from this particular teaching model.

[1] There are no complete theories, as GCH emphasized the Robinsonian precept, conjoined to Keynes’srule that there is a spectrum of methodologies to invoke, in theorizing for policy as an economist, just asthere is the spectrum of philosophies, to understand ‘man’s changing vision of the universe’.
[2] GCH was ‘there’, long before the fashionable heterodox tradition of relying on Minsky to have originatedthat supreme notion of the finance motive, in Keynes. A theorist who ‘soiled’ his hands in the messy worldof company finance, inflation, and depreciation, always coupled money and finance, money with finance—thus another invisible bridge was forged between Keynes and Kalecki, in the economics of GCH.
[3] In a precise analytical sense, this paper is a link between the ‘Yogi’ and the ‘Commissar’ of the previoussection and the five issues that form the core of this section. In particular, it provides the crucial linkbetween theory and policy.
[4] But this really belongs to the vintage (sic!) GCH productions of the 1960s (Harcourt 2012a: 28): ‘Thearticle had a long gestation period. I wrote the first draft in 1966’.
[5] To the best of this author’s knowledge, the reference to PKE by GCH was first made in footnote 1 ofHarcourt and Kenyon (ibid.).
[6] It was in relation to these that Paul Samuelson (1964) first characterized—perhaps chided is a betterdescription—a ‘fictitious’ being called Jean-Baptiste Kaldor. GCH has referred to this fictitious characterin many of his writings, where he is critical—as in Keynes (1936) [1973]—of any kind of Say’s Law (ofmarkets). In Harcourt (1991: 160) he reminds his readers that Paul Davidson had critically discussedJean-Baptiste Kaldor’s theories of full employment distribution. Davidson (1968: 259) had, in turn,referred to the priority of Samuelson on this point.
[7] It is more than intriguing that the two titles of the papers by Champernowne and McCallum, in leading journals, contain the words stability and instability, referring to the same economic model! Does it notindicate that the naive technical notion of stability is inadequate to encapsulate the rich connotationswhich it implies in an economic context? As Keynes wrote to Gerald Shove, on 21 April 1936(Keynes 1973: 2—replacing ‘the influence of expectations and of transitory experience’ with ‘stability’—or ‘instability’): As soon as one is dealing with [stability], one is, in the nature of things, outside the realmof the formally exact’. I suspect that GCH is in agreement with this view of Keynes—as I am!
[8] As my maestro, and GCH’s good friend, Richard Goodwin, pointed out, 65 years ago (Goodwin 1950:319): ‘Combining the difficulties of difference equations with those of non-linear theory, we get an animal of a ferocious character and it is wise not to place too much confidence in our conclusions as tobehavior’.
[9] I shall refer to the version reprinted in Harcourt (2012b), as Essay 3.
[10] In a personal letter to me, dated 2 January 1985, the late Bent Hansen wrote that this particular paperby Lindahl was prepared for a lecture visit to Australia—and it was in preparing this that Lindahl had firstread Keynes (1936) [1973], fully over a decade and a half after it was first published!
[11] In Harcourt (1985: 418; italics added) he reports on Goodwin’s preference for geometry as follows:‘[Goodwin] wanted to be the Marshall of the twentieth century by teaching linear models entirely ingeometry (he still prefers geometry and geometric proofs to algebra)’. I should point out that Goodwinused geometry for non-linear model analysis and algebra for linear analysis! In an e-mail to me, dated 24September 2012, GCH wrote (italics added): ‘In my lectures on macro to 3rd year Ugads in 1967 atAdelaide I used a linear geometric version of Hugh [Hudson's] model, see Essay 3 of On Skidelsky’s Keynesand Other Essays’.