Keynes and Keynesianism, May 1946

On April 21, 1946, John Maynard Keynes died, almost exactly ten years after the publication of the General Theory, and Samuelson was invited, on Marschak’s recommendation, to write a commemorative article for the July issue of Econometrica.^26,1 Frisch hoped that Samuelson’s article would run alongside an article by William Beveridge, offering personal reflections on Keynes, but Beveridge’s article failed to materialize and Samuelson’s article was published on its own.²⁷ Though Samuelson had engaged with Keynesian theory, a semi-historical article like this represented a new departure.

Samuelson's article began with his remark, quoted in chapter 12 earlier, about economists under thirty-five having had no resistance to the General Theory. However, rather than praise Keynes, he went on to find fault with the book in language that was as critical as that used by his Harvard teachers, Leontief and Schumpeter, a decade earlier.

It is a badly written book, poorly organized; any layman who, beguiled by the author's previous reputation, bought the book was cheated of his 5 shillings. It is not well suited for classroom use. It is arrogant, bad-tempered, polemical, and not overly-generous in its acknowledg­ments. It abounds in mares nests and confusions: involuntary unem­ployment, wage units, the equality of savings and investment, the timing of the multiplier, interactions of marginal efficiency upon the rate of interest, forced savings, own rates of interest, and many oth­ers....

The result was that Keynes had set up “an indoor guessing game”: “Wherein lies the essential contribution of the General Theory and its distinguish­ing characteristic from the classical writings?” The General Theory was “an obscure book,” the obscurity and polemical character of which would, Samuelson conjectured, serve to maximize its long influence. And yet it was, Samuelson claimed, a work of genius, its analysis both obvious and new.

But what was the book's novelty? He dismissed Keynes's theory of liquid­ity preference on grounds that the rate of interest was of minor importance, and he argued that though Keynes had brilliantly called attention to the importance of expectations, he had hardly provided any theory. The novelty of the General Theory lay not in its theory of liquidity preference or its concep­tion of expectations but in its analysis of effective demand.

I myself believe the broad significance of the General Theory to be in the fact that it provides a relatively realistic, complete system for ana­lyzing the level of effective demand and its fluctuations. More nar­rowly, I conceive the heart of its contribution to be in that subset of its equations which relate to the propensity to consume and to saving in relation to offsets-to-saving.

There was no mechanism to ensure that investment was equal to saving at full employment. The General Theory might have been an obscure book, needing a companion volume that could provide a guide to its contents, but it provided a new system, and that was what was needed to defeat the clas­sical theory.^m

Samuelson wrote favorably of Keynes's wartime analysis of inflation using the concept of the “inflationary gap” as the modern theory of inflation. This theory might have been the reason why the inflation rate was much bet­ter during the Second World War than in previous major wars.

The book had been influential because economists in Britain rapidly real­ized that effective demand was not to prove a passing fad, but was part of the “wave of the future,” and American economists soon followed.

Obviously, exactly the same words cannot be used to describe the analysis of income determination of, say, Lange, Hart, Harris, Ellis, Hansen, Bissell, Haberler, Slichter, J. M. Clark, or myself. And yet the Keynesian taint is unmistakably there upon every one of us. (I hasten to add—as who does not?—that I am not myself a Keynesian, although some of my best friends are.)³⁰

Having established that Keynes provided a system, and that his econom­ics did amount to a general theory, applicable in times of prosperity as well as depression, Samuelson sought to establish that his philosophy was “pro­foundly capitalistic” in nature, aimed at saving the existing system. Keynes

m. A few years later, Hansen (1953) was to write such a guide. himself was an “urbane and cosmopolitan provincial English liberal” who found nothing in “the turbid rubbish of the Red bookshops.”^31,n

Samuelson then turned to the development of Keynes's ideas, arguing that the General Theory could not have been anticipated, given his previous work. Though he criticized orthodoxy, such as Irving Fisher's quantity theory of money, Keynes was not an original economic theorist, his work being notable only for its “political novelty and persuasiveness.” He had even made a num­ber of serious mistakes, such his views on population and his arguments with Bertil Ohlin on reparations payments.

So why did Keynes have so little interest in economic theory, and why was he so bad at formal theorizing? Samuelson's answer was,

Perhaps because he was exposed to economics too young, or perhaps because he arrived at maturity in the stultifying backwash of Marshall's influence upon economic theory—for whatever reason, Keynes seems never to have had any genuine interest in pure economic theory. It is remarkable that so active a brain would have failed to make any con­tribution to value theory; and yet except for his discussion of index numbers in Volume I of the Treatise and for a few remarks concerning “user cost,” which are novel at best only in terminology and emphasis, he seems to have left no mark on pure value theory.³⁵

Reading between the lines of this article is a complex task. Given that Samuelson was well aware of his own precocity, it is remarkable that he criti­cized Keynes for coming to economics too early. His likening of Keynesianism to a disease suggests that either one caught it or one did not; yet in distanc­ing himself from Keynes he implied that, while he might not have been

n. Other than that he wanted to recognize Keynes's qualities without conceding too much, it is not clear precisely what he meant by describing him as cosmopolitan and provincial.

o. Samuelson was later to hold a more positive view of Ramsey's theory of saving, which in the 1980s came to be seen as one of the main foundations of macroeconomic theory. immune, he never succumbed completely. The vehemence of his attack on Marshall, which goes beyond anything in Foundations, makes it impossible to avoid conjecturing that this was an article on which he sought Schumpeter's advice. In describing Keynes as someone who “stumbled upon and formu­lated a new system of analysis,” with no interest in economic theory and as reasoning incorrectly, Samuelson presented him as having brilliant intuitions but leaving a gap that Samuelson was to fill with the “Keynesian savings-

36

investment-income cross.

A large part of the explanation of Samuelson's attitude toward Keynes is that he came to “Keynesian” problems through Alvin Hansen. Hansen had come to terms with Keynes's ideas by incorporating them into his own theory. This explains the paradox that, as Samuelson put it, “The Keynesian notions are old and new at the same time.”³⁷ Samuelson sometimes disagreed with his mentor, as when he had tried to explain that consumption spend­ing could have the same impact on demand as investment, or when he had to argue dogmatically that it made no difference whether one argued in terms of saving and investment or in terms of income (saving plus consump­tion) and expenditure (investment plus consumption), a point Hansen could not understand as late as 1947. However, Samuelson and Hansen remained extremely close, personally and intellectually.³⁸ This explains why, in 1946, though admiring Keynes, Samuelson could still distance himself from him, putting forward an interpretation of the General Theory that stressed precisely those elements that fit into Hansen's way of thinking. The myth of Hansen's conversion to Keynes that Samuelson was to propagate later was not yet part of his thinking.³⁹

There had, however, been a change in Samuelson's position in relation to Keynes.

It has seemed to me that what I call the “income flow analysis,” of which yours is the most noted presentation, has done something which has not been done in comparable degree since Ricardo and Marx; namely constructed a coherent logical theoretical system or formula having the quality of a mechanism, growing directly out of current conditions

and problems which are of paramount importance and furnishing a key for working out definite answers in terms of policy.⁴⁰

Though he would not have seen this letter, this was Samuelson’s position: that the Keynesian theory of income determination, if made properly dynamic, could formalize the income flow analysis of Clark, Hansen, and others. Though he was no mathematician, Clark conceived of economics as dynamic and concerned with the relation of fluctuating demand to a relatively inelas­tic supply. In Studies in the Economics of Overhead Costs, he had written that his book was “a study of the discrepancies between an ever fluctuating demand and a relatively inelastic fund of productive capacity, resulting in wastes of partial idleness and many other economic disturbances.”⁴¹ He might talk of “unused capacity” rather than unemployment, and he might stress the role of imperfect competition, but the subject was the same. In 1946, Samuelson had occasion to reread Studies in the Economics of Overhead Costs, and he wrote Clark, praising it.^p

I cannot refrain from writing to tell you what a seminal work it is.

A dozen or more years ago, as an undergraduate at the University of Chicago, I first read it. I have been astonished to discover how many ideas in my mind trace back to that first reading—and incidentally how much of the current discussion of unemployment reflects its influ­ence. I was also amused to note that as an undergraduate, steeped in the orthodox Chicago tradition, I jotted down critical marginal notes which today I would no longer hold against you.⁴²

By the time he wrote these words, Samuelson’s views had changed dra­matically in a way that enabled him to appreciate Clark’s work; but though Keynes’s ideas had been crucial to that transformation, he had not simply converted to Keynes, any more than his mentor had. When Lewis Haney, at New York University, who was to become the harshest academic critic of his textbook, classified him as a Keynesian in his textbook on the his­tory of economic thought, Samuelson replied that after being hostile to the General Theory, he “finally came round to the view that his work is character­ized by numerous logical flaws and omissions but that his tools represent an important addition to our economic knowledge” and that he held “no

p. Overhead costs are costs that have to be incurred whatever the level of production. As output rises, such costs will be spread over a larger quantity of output, with the result that cost per unit of output will fall. Facing such cost conditions, companies could not be in competitive equilibrium and there must be an element of monopoly. particular brief for the particular policies advocated by Keynes or for his general weltanschaung.”⁴³

Though they might prove useful for other reasons, given a postwar politi­cal climate that was increasingly hostile to Keynesianism, there is no reason to think that such statements about where he stood in relation to Keynes were anything other than an accurate reflection of a position he had reached over several years during which he had been engaged in the urgent practical problems related to the war and its aftermath.^q This version of Keynesian economics, rooted as much in the ideas of Clark and Hansen as in the General Theory, was shortly to be popularized in what became a bestselling textbook.

q. He could also joke about his views. When Paul Douglas, as president of the American Economic Association, asked Samuelson to participate in a session on Keynesian economics in 1947, for which he wanted to include a true defender of Keynes, a strong opponent, and two others with intermediate positions, Samuelson replied that he was not sure “as to which of the backfield posts—‘true defender,' ‘strong opponent,' or ‘intermediate position' ”—he was slated for. P. H. Douglas, April 28, 1947, Letter to Paul A. Samuelson, PASP 24 (D {1942-64]); P. A. Samuelson, May 20, 1947, Letter to Paul H. Douglas, PASP 24 (D {1942-64]).