The credo that Keynes hit the hardest in The General Theory was Say's Law. The idea that supply creates its own demand was overthrown so effortlessly that it now seems ridiculous to economists and students of economics. And yet the primacy of supply was simply replaced by faith in the primacy of demand. The tyranny of one six-lettered word gave way to that of another. The writings of most respectable economists now insist of the primacy of demand in the short run though none is quite so unsophisticated as to chant that demand creates its own supply. Ironically, the new emphasis is structurally a replica of the old and represents a similar distortion.
Keynes' strong emphasis on demand was, we suggest, strongly influenced by his political preconceptions. As most mainstream economists in the west share these preconceptions they have not observed certain unusual lapses and assertions in Keynes' argument. Even attempts to escape from Keynes' influence or to rewrite his analysis still appear to be affected by his framework and phraseology. In an era when economists are increasingly fascinated by algebra, it is perhaps not surprising that they should be easily surprised and seduced by words; and Keynes was a master of words.
Keynes' genius with words is exemplified in his use of the phrase "effective demand". He argued that effective demand formed "the substance of the General Theory of Employment".(1) Effective demand raised or lowered the level of employment; it determined the welfare of industrial nations. So important was the concept that it appeared by name in at least seventy pages of The General Theory of Employment, Interest Rates and Money.
As The General Theory nowadays is not read widely, and as the phrase "effective demand" seems simple enough, it is important to stress that the phrase did not convey to Keynes the meaning which it now usually conveys in the economic literature. To Keynes it embraced not simply demand; it also held a componant of supply. The supply componant was, theoretically, as important as the demand componant, for Keynes defined effective demand as the point where the aggregate demand function was intersected by the aggregate supply function. At the same time Keynes, while paying his theoretical respects to supply, placed the greater emphasis on demand. In selecting "effective demand" as the key phrase in his brilliant book, he reinforced the idea that demand was dominant in determining the level of employment.
Names are capsules that are easily swallowed; they can also numb the senses. The name "effective demand" has perhaps numbed our senses for a generation. After all, Keynes would have been equally justified, when describing that point where the aggregate demand function is intersected by the aggregate supply function, in selecting the phrase "effective supply" and making it the vital phrase in his book. In the light of that alternative the preference for the phrase "effective demand" must be one of the most cavalier decisions in the history of economic thought.(2) Where two concepts represent opposite poles it is unusual to give to the point where they intersect the name of one of the poles.
The emphasis in demand in The General Theory came also from the relative space which Keynes allocated to demand. While in several passages he argued that the aggregate supply function is as important as the aggregate demand function in determining the volume of employment, he explained that the supply side required little further comment. It involved, he wrote, "few considerations that are not already familiar" (3). When at last on page 280 we reach the first of the dozen pages devoted to aggregate supply we are informed that six of them can be skipped by those who - "rightly" he said - '"dislike algerbra". That, incidentally, appears to be the book's only invitation to skip. The remaining five and a quarter pages on aggregate supply are not concerned with the central determinants of supply. Instead they examine the proportion of total demand which is devoted towards each industry. In the following chapter, Keynes discusses how aggregate supply gradually becomes more inelastic as successive industries reach the bottleneck phase. Thus the primacy of aggreagate demand is modified by the constraint of the existing factors of production, and this modification is echoed by the best of the modern macroeconometric models. But neither the model builders, nor the master from whom directly or indirectly they draw their inspiration, consider the possibility of autonomous change from the supply side. Their aggregate supply curves are purely ex post constructions (4).
From Keynes' book it is obviously easy to glean, in many obvious or subtle ways, the message that demand is the dominant factor in determining income. Effective demand was his key phrase: aggregate demand occupied the large part of the book. It may be that some kind of over-statement is to be expected when an author of originality, indeed genius, is fighting strong preconceptions in his readers' minds. Likewise over-statements or blurred statements may also be the outcome of the acute difficulty of devising a general theory and of experiencing the painful reshaping of opinions which that kind of intellectual journey entails. Nevertheless, even when we allow for his scattered assertions that aggregate supply is important, we receive the impression that demand is more important. Often Keynes regarded aggregate demand and effective demand as identical. He regarded the propensity to consume and the inducement to invest as the two constituants of aggregate demand, and in places he treated them as the sole constituants of effective demand:
* "The propensity to consume and the rate of new investment determine between then the volume of employment ... If the propensity to consume and the rate of new investment result in a deficient effective demand, the actual level of employment will fall short of the supply of labour potentially available at the existing real wage ... "(5).
* "Employment is a function of the expected consumption and the expected investment".(6)
* "The theories which we have examined above are directed, in substance, to the constituants of effective demand which depends on the sufficiency of the inducement to invest. It is no new thing, however, to ascribe the evils of unemployment to the insufficiency of the other constituant, namely, the insufficiency of the propensity to consume." (7)
Today the phrase "effective demand" has lost even the meaning that Keynes begged in his formal deficition in The General Theory. The deep dye of supply has been leached from the phrase. The cruder, more extreme disciples of Keynes concentrate solely on aggregate demand and eliminate aggregate supply from their economic models. Even the moderate Keynesians have come to equate the phrase "effective demand" simply with aggregate demand. Likewise an unhesitating emphasis on demand is reflected in the two best-known biographies of Keynes. Sir Roy Harrod interprets Keynes' views thus:
"Employment is determined by total "effective demand", depending on the propensity to consume and the propensity to invest". (8)
Robert Lekachman begins a paragraph on Keynes' views by explaining that the level of employment rests at "the point of intersection between the aggregate demand function and the aggregate supply function" but ends the paragraph - in unconscious contradiction - by explaining that the level of employment is determined by aggregate demand. (9)
In textbooks we see demand ascendant, indeed rampant. "The equilibrium level of the national income" explains the latest edition of Samuelson's Economics, "is at the intersection of the C + I schedules of desired total spending with the 45 degree line depicting the value of total output." In a footnote, he observes that some economists call the C + I curve an aggregate demand curve and the 45 degree line an aggregate supply curve. "The equilibrium level of income", he adds, "is where the aggregate demand and supply curves intersect". (10) Finally, the comment of Robert Mundell is worth quoting not only because he sensibly regrets the failure of many economists to assess the relevance of the aggregate supply schedule but because he himself equates "effective demand" with aggregate demand, thereby showing unconsciously how pervasive is the confusion to which he points:
"Post-World War II literature has, however, frequently assumed that an increase in effective demand leads to an expansion of employment and of output up to the point of full employment; further increases will cause inflation. The underlying implicit assumption is that the aggregate supply curve, reflecting the marginal costs of firms (on the assumption of competition), is perfectly elastic up to the point of full employment, and completely inelastic after that point, at any given monthly wage.
Is it logical to place such emphasis on aggregate demand as the determinant of employment? Is it realistic either to ignore aggregate supply completely or to assign it merely a subsidiary role as a mechanistic constraint? We suspect that the framework of analysis of economists is shaky.
It seems dubious even to accept the limited relevance which some economists see in Keynes' "general" theory of employment. His theory, it is sometimes argued, was most applicable to the unusual economic situation in Britain and the advanced industrial nations in the early 1930's. As machinary stood idle in factories, as the stocks or raw materials and semi-processed goods lay unwanted in warehouses, and as the operative skills could be recruited at street corners, all that was needed to galvanise these idle ingredients of supply was increased demand. As Hicks observed, when originally reviewing The General Theory, Keynes relied heavily on "the assumption of a high elasticity of consumption goods". (12) Similarly, Joan Robinson observed that the "unemployment that concerned Keynes was accompanied by under-utilisation of capacity already in existance." (13) In such a situation the capacity could "obviously" be utilised as demand was increased. Since supply appears to be present, demand can be singled out as the missing ingredient. But when that diagnosis is translated into theory, and when the shorthand of everyday speech and conversational causation is put aside, the emphasis on demand ceases to be logical. Demand was exceptionally stimulating only because supply was exceptionally ready to respond to stimulus. Even in the unusual situation of the early 1930's a theory that enthroned demand was dethroning reality. It is not unlike a small child who saw for the first time a hammer crush a walnut. He might initially marvel at the power of the hammer, but when he saw that the same hammer was unable to crush a ball-bearing he might suspect that the material to be hammered was causally as important as the hammer itself. Do economists learn the same question more slowly?
If theory, whether in economics or hydraulics, should explain the principles and assumptions which underlie an area of knowledge, then the emphasis on the power of demand during the world depression is a dubious theory. It is dubious because it calls itself general theory but concentrates on a particular case; and it is dubious because it is silent about one of the assumptions which it should explain - the responsiveness of supply.
Supply is not in fact as malleable as Keynes and the Keynesians have assumed. In analysing unemployment they make much of the vital distinction that purchasing power and purchasing are not the same. In The General Theory Keynes remarks wistfully that Alfred Marshall wrote the sentance - "though men have the power to purchase, they may not choose to use it" (14) - but did not realise its full implications for Say's Law. In essence, supply could hardly be said to create its own demand if men preferred not to use every unit of the demand conferred on their wages, dividends and profits. As that statement helps to explain - in Keynesian terms - why supply cannot create its own demand, it is worth devising a parallel statement that can help to explain why demand cannot create its own supply. The parallel statement can be fitted neatly into the shell of Marshall's sentance: Though men have the power to produce, they may not choose to use it. In other words, maunfacturers during a depression could be favoured by all the physical conditions of supply, and could be confronted with increasing consumers' demand, but they could still refuse to manufacture more goods until they have sold virtually of their existing stockpile. In such a situation their gloomy expectations of the short-term future could dissuade them from expanding output and employing more men just as gloomy expectations dissuaded many customers from using some of their idle purchasing power. Certainly the response of manufacturers to signs of increasing consumer demand was far from automatic in the era of laissez faire capitalism which Keynes diagnosed. (15)
It is even harder to believe in the primacy of demand when Keynes' definition of aggregate demand is examined closely. In his eyes aggregate demand was "a function of the expected consumption and the expected investment". (16) But the distinction between consumption and investment - or between consumer goods and capital goods - is a relevant to aggregate supply as to aggregate demand. Just as aggregate demand can be divided along the Keynesian lines into consumption and investment, so aggregate supply can be divided into the production of consumer goods and producer goods. It is not inconsistent with Keynes' own framework to treat investment and consumption as componants of aggregate supply, because Keynes explains that entrepreneurs' decisions to set production in motion are based on expectations of what consumption will be. In that sense the expected production in the short term and the expected consumption which motivated that production are so similar that they can be placed on the side either of aggregate demand or aggregate supply. This is not to deny Keynes' distinction between consumption and investment: it is rather to deny the validity of his emphasis on aggregate demand and his relative neglect of aggregate supply. Indeed, a careful reading of The General Theory suggests that in Keynes' mind the distinction between supply and demand is blurred or arbitrary. (17) At one point of his argument he writes positively that employment is "determined by the producer's short-term expectations"; (18) and one does not need to be sharp-eyed to notice how close that sentance comes to suggesting that aggregate supply determines employment.
In microeconomics the concepts of supply and demand are clear-cut and can easily be personified. To carry these concepts into macro-economics is inevitably to blur them in the eternal circles of supplying and demanding. Keynes, it would almost seem, was content to blur them for most of his argument. Indeed the central figure in Keynes' protrait of the economy is the organiser of manufacturing, the entrepreneur, and he implicitly straddles both supply and demand. He demands the factors of production - raw materials, fuel, labor - and he supplies goods to the consumers. Above all the expectations of the entrepreneurs are based on their assesments of supply and demand and influenced by others' assesments of supply and demand. Since effective demand is aggregate expectation - the point where "the entrepreneurs' expectation of profit will be maximised" (19) - it must embrace supply and demand.
The method of classification and naming by Keynes has had astonishing success. In his use of words, however, there is sometimes the touch of the mountebank. His phrase "effective demand" can equally mean "effective supply" though not once did he hint as much. All those economists and financial journalists who call for "increased demand" or "diminished demand" to adjust the level of income are in one sense the victims of a verbal conjurer. Has it again come to pass - as Keynes wrote in his final paragraph - that those in authority, and their academic advisors, "are distilling their frenzy from some academic scribbler of a few years back?" (20)
Why then did Keynes decide to give such emphasis to demand? While his Treatise on Money of 1930 and the writings published the following year in Essays in Persuasion point towards the conclusions he was to put forwards in The General Theory, they offer no clue, so far as we can see, no inkling that those conclusions were to be fitted into a demand-centred framework. His emphasis on demand perhaps has at least two academic sources. One may have been a reaction against the emphasis which classical economists tended to give to supply. Another may have been a desire to draw attention to his general theory, to emphasise its novelty, by placing it in an unfamiliar framework. We suspect, however, that these were not the main reasons for Keynes' emphasis on demand. Probably the main reasons were political and social more than economic. The subtle implications of demand were more attractive to Keynes' own system of values. To stress demand as the elixer was in normal parlance to enthrone the consumer: to stress supply would have been to point to socialism rather than the restrained state intervention which Keynes desired. Production had long dominated economic thinking and for Keynes therefore to state too positively that there were deficiencies in production and supply rather than in demand would have been to smite capitalism harder than he had intended. And to smite capitalism too hard would have been to elevate socialism or the new "authoritarian state systems" which Keynes felt were inferior to western democracy and capitalism. They were inferior, he argued, because they neglected the importance of personal choice, "the loss of which is the greatest of all the losses of the homogeneous or totalitarian state." (21) Is it then to be wondered that Keynes preferred the word "demand" with its strong connotation of customer, consumer, liberty, and the act of personal choice? (22)
In the last chapter of the book Keynes revealed the kind of personal reasons which would make him, consciously or unconsciously, stress demand more than supply. The last chapter, however, tended to be neglected by economists because at first sight it is concerned less with economics than with political economy. The chapter has the curious title of "Concluding Notes on the Social Philosophy towards which the General Theory might Lead"; and such a title conveys the impression that the conclusions in The General Theory have arisen from a neutral, academic, weighing of hypotheses and evidence which had been carefully sterilised before reaching human hand and mind. Ironically it is in economics, of all the social sciences, that such a chapter title was most likely to be accepted without comment, even though its author was well known as a man of affairs with clear preferences and priorities in national and international politics. Hugh Stretton, in his perceptive book The Political Sciences (23) (of which economics is one), wondered how much the social philosophy of Keynes had led him towards the framework and conclusions of The General Theory. The role which Keynes gave to demand suggests that the influence of his social philosophy was considerable.
If some Kondratieff, returning from some Siberia, were to survey the history of economic thought, he might notice "long swings" in the mind of successive generations of economists. The economic system itself has hardly changed sufficiently to justify these radical reappraisals. The reappraisals themselves seem to be extreme reactions to extreme positions. The extremists, in their zeal for part of the truth, blot out another part; at last a new generation sees the significance of the hidden part and in their reaction blot out the part that previously was venerated.
We suggest that it is incorrect for economists to focus their analysis on either aggregate demand or aggregate supply. Existing "short-run" models of the determination of national income assert the primacy of demand, and in this respect they follow Keynes closely. "Best practice" econometric models determine output by the level of demand, subject to the technological constraint imposed by the production function and the factors of production, and in this respect echo the discussion in Chapter 21 of The General Theory. In neoclassic models (24), this short run primacy of demand is modified by the tendency of changing prices to alter demand until in the long run it equates to the desired supply. Amidst all this concentration on either the short-run or the long-run we wonder what happens in the medium run. Is the medium run a treacherous swamp where the economists rarely venture for fear of meeting the truth that supply and demand are equally important?
Since the level of output is influenced by the decisions of producers of consumption and investment goods, as well as the decisions of those who demand consumption and investment goods, these contemporary models appear to be seriously in error. If the models are erroneous, the policies they propose are also likely to be in error. The separate effects of demand and supply may be difficult to sort out, but continued reliance on oversimplified models is sure to be misleading. If economists persist in using only one hand to grapple with vital problems, they would be wise to remember that the hand has only five fingers.
Geoffrey Blainey, University of Melbourne
Peter Jonson, Sydney
Footnotes and References
1. JM Keynes, The General Theory of Employment, Interest and Money, London, 1964 printing, P 25.
2. In prominent places of his book Keynes implied that he and Thomas Malthus shared the same notion of effective demand. It is clear, however, to those who read Keynes’ biographical essay on Malthus (Essays in Biography, London, 1933, esp. Pp 141-6), that the phrase “effective demand” did not mean the same to Keynes as it had meant to Malthus; Malthus’ definition seems to have contained no component of supply. In fact, Keynes may have quoted Malthus in an attempt to establish a respectable ancestry for his main theoretical innovation.
3. Op. Cit. P 89.
4. In the extensive literature on the “Keynesian supply curve” are attempts to detect in The General Theory an ex ante aggregate supply curve. The latest attempt is by JR Millar, “Aggregate Supply and Demand functions”, Economic Journal, June 1972, P 607.
5. Op. Cit. P 30.
6. Op. Cit. P 98.
7. Op. Cit. P 358.
8. RF Harrod, The Life of John Maynard Keynes, London, 1951.
9. Robert Lekachman, The Age of Keynes: A Biographical Study, London, 1969, P 78.
10. PA Samuelson, Economics, New York, 8th Edition, 1970.
11. RA Mundell, International Economics, New York, 1968, P 204.
12. JR Hicks, “Mr Keynes Theory of Employment”, Economic Journal, June 1936, P 245.
13. Joan Robinson, Economic Philosophy, London, Penguin, 1964, P 112.
14. Op. Cit., P 19n.
15. Nor is the response, apparently, automatic today. Note F Cairncross’ comment on Britain’s unemployment of 4.3 % in The Observer, 23 January 1972, P 11; “Consumer spending, in particular, has been rising at a spanking pace ... Normally a substantial increase in consumer spending would persuade manufacturers to take on more men and step up production. But so far production remains stagnant.”
16. Op. Cit. P 98.
17. This may explain Millar’s difficulty with the concepts. Millar, Op. Cit. Especially P 607.
18. Op. Cit., P 51.
19. Op. Cit., P 25. See also P 55.
20. Op. Cit. P 383.
21. Op. Cit. P 380.
22. The above interpretation possibly illuminates one of the strangest sentences in The General Theory (p 379) where Keynes argues that capitalism had broken down “in determining the volume, not the direction, of actual employment”. That sentence is consistent with the political theorising in his book, but it is inconsistent with the economic theorising in his book. For if capitalism had directed more men into investment industries and fewer into consumption industries in the late 1920s there could not have been – according to Keynes’ main premise – an inadequate volume of employment.
23. H Stretton, The Political Sciences: General Principles of Selection in Social Sciences and History, London 1969, Pp 343-6.
24. As expounded, for example, in D. Patinkin, Money, Interest and Prices, New York, Second Edition, 1965.
25. It is interesting to note that the political implications of Patinkin’s model are more conservative than Keynes’; the system tends to an equilibrium in which everyone is voluntarily employed. In practice, however, Patinkin acknowledges the probable need for Keynesian policies to speed up the equilibrating process by operating on aggregate demand.