The Critics of Keynesian Economics

XIX

W. H. HUTT was born in London in 1899. He attended the London School of Economics and after graduation took employment in business. In 1928 he was appointed professor of commerce at the University of Capetown, South Africa, where he is now dean of the Faculty of Commerce. His works include: The Theory of Collective Bargaining, 1930; Economists and the Public, 1936; The Theory of Idle Resources, 1939; and Plan for Reconstruction, 1943.

Professor Hutt’s Theory of Collective Bargaining was a penetrating “history, analysis and criticism of the principal theories” by which economists from Adam Smith down have attempted to prove that unions and “collective bargaining” had raised or could raise the average real level of market wages for the whole body of the workers without causing unemployment. His book, Keynesianism—Retrospect and Prospect, published in 1963, is a thorough, comprehensive, and brilliant refutation of the whole literature of Keynesianism. The following article appeared in The South African Journal of Economics for March, 1954, pages 40-51.

THE SIGNIFICANCE OF PRICE FLEXIBILITY

W. H. HUTT

The period 1932-1953 has witnessed a revolution and counterrevolution in thought on the function and consequences of price flexibility.

In considering this remarkable phase in the history of theory, it is useful to begin by referring to a related field of hardly disturbed agreement. There has been no controversy during the period of our survey among serious economists about the desirability of a system which tends to ensure that different kinds of prices shall stand in a certain optimum relation to one another, or about the desirability, in a changing world, of continuous relative price adjustment in order to bring about some conformance to the ideal relation. From the so-called “socialist economists” of the Lange-Lerner type to the so-called “individualist economists” of the Mises-Röpke type, there has been agreement that the price system has important equilibrating and co-ordinative functions. Moreover, until the appearance of Keynes’s General Theory, in 1936, the measure of agreement about the aims of institutional reform for the better working of the price system seemed to be slowly but definitely growing.

There was not the same marked tendency towards agreement about methods. Some thought that improved pricing could be achieved through a greater centralisation or sectionalisation of economic power, with the final voice to decide both preferences (choice of ends) and productive policy (choice of means) entrusted to elected representatives or syndicates. Others thought that the required reforms involved exactly the reverse—the breaking up and diffusion of economic authority so that the final voice about ends rested with the people as consumers, whilst the final voice about the choice of means rested with those who stood to gain or lose according to the success with which they allocated scarce resources in accordance with consumer-determined ends. But in spite of this apparently basic clash, as soon as explicit plans for the devising of a workable economic system were attempted, even the divergence of opinion about methods appeared to be narrowing. The so-called “socialist economists” were clearly attempting to restore the market and the power of substitution. So much was this so, that I believed the result of their labours would ultimately be the re-building of laissez-faire institutions, in elaborate disguises of name and superficial form, the result being regarded as the perfect socialist pricing system.1

This interesting trend towards unanimity of opinion in several fields was overlapped by and rudely disturbed by Keynes’s General Theory. Since 1936, the economists have become sharply divided about the nature of the price changes which ought, in the interests of “full employment,” to take place in any given situation.2 Consider trade union or State enforced wage-rates. At one extreme, we have the Keynesians who argue that, in maintaining wage-rates, we are maintaining consumer demand, creating a justification for new investment, and so preventing the emergence of depression. At the other extreme, we have those who argue that each successive increase of wage-rates so brought about renders essential a further element of inflation in order to maintain “full employment”—a development which tends permanently to dilute the money unit.

The Keynesian theory on this point proved enormously attractive. The idea as such was not novel; but before The General Theory it had enjoyed a negligible following in respectable economic circles. After 1936, it gave many economists what they seemed to have been waiting for, a non-casuistic argument for the tolerance of the collective enforcement or State fixation of minimum wage-rates.

Curiously enough, Keynes’s challenge was based on a sort of admission of the evils of current collective bargaining and a further admission (by no means explicit, but an inevitable inference 3), that labour in general was unable to benefit in real terms at the expense of other parties to production, by forcing a rise in the price of labour. Gains achieved by individual groups of organised workers were paid out of the pockets of other workers. At the same time, Keynes’s new teachings seemed to support strongly those who cried, “Hands off the unions!” Although his thesis was accompanied by the charge-not wholly without foundation—that orthodox economists had closed their eyes to the consequences of the wage rigidity caused by trade union action, he always seemed to range himself on the side of the unions in their resistance to wage-rate adjustments. The reasons for his views on this question were two-fold.

Firstly, he argued that the price of labour had to be regarded as inevitably rigid. This empirical judgment about economic reality is, of course, not confined to the Keynesians. Where Keynes was original was in the subsidiary and supporting assumption that what other economists have called “the money illusion” was a basic cause of the rigidity.

Secondly, he argued that, in any case, wage-rate flexibility downwards, even if other prices were flexible, would aggravate and not alleviate depression. For even under perfect wage-rate flexibility and perfect flexibility generally, an equilibrium with unemployment could exist.4 As I have previously argued,5 Keynes would have preferred to rely wholly upon the second argument. But he kept the first, as Schumpeter has put it, “on reserve.” In this survey I shall be dealing only with this second argument.

The contention is that wage-rate cuts must in any case be ineffective, as a means of restoring employment in labour, because it is possible to cut money rates only and not real wage-rates. Reduced money rates, Keynes explained, would mean reduced wages in the aggregate and lead to reduced demand. Hence the wage-rate rigidity, which former economists had been inclined to criticise ought, in his opinion, to be regarded as a virtue in times of depression.

At two points, Keynes appeared to have some misgivings about this thesis. He admitted firstly that if the price of labour could be flexible, things would be different, i.e., “if it were always open to labour to reduce its real wage by accepting a reduction in its money wage . . . . ” This condition assumed, he said, “ . . . . free competition among employers and no restrictive combinations among workers.” 6 And he explicitly admitted later that, if there were competition between unemployed workers, “there might be no position of stable equilibrium except in conditions consistent with full employment. . . .” 7 But he did not attempt to reconcile these passages with apparently contradictory passages.

We are left, then, with the principal contention, namely, that changes in wage-rates are “double-edged,” affecting both individual outputs and general demand. As this infectious doctrine has been developed by Keynes’s disciples, costs as a whole are no longer regarded as merely limiting output, but as calling forth output through demand.

The objection to regarding costs as a source of demand can be simply stated. The only cost adjustments which defenders of price flexibility advocate are those which must always increase real income, and hence always increase money income under any system in which the value of the money unit remains constant. If we concentrate attention upon wages, it can be said that, on the reasonable assumption that the growth of real income will not mean a re-distribution against the absolute advantage of the wage-earners, the effect of the wage-rate reductions which are advocated must always mean an increase and not a decrease in aggregate wages received, and hence an increased demand for wage-goods. (The possibility of hoarding being induced is discussed later.)

In part, the Keynesian attempt to handle the problem in terms of the crude concept of “the price of labour” has confused the issue. We are concerned with the prices of different kinds of labour, whilst the index number concept of “the wage level” screens off from scrutiny all the issues which seem to me to be important.8 Throughout Chapter 19 of The General Theory Keynes talked simply of “reduction of money wages.” And he discussed the orthodox view of the desirability of price adjustments as though it was based on a “demand schedule for labour in industry as a whole relating the quantity of employment to different levels of wages.” 9

Through thus thinking rather uncritically about aggregates, the Keynesians appear to have assumed that wage-rate reductions imply reduction of aggregate earnings,10 irrespective of whether the labour price which is cut is that of workers in an exclusive, well-paid trade, or that of workers doing poorly paid work because they are excluded from well-paid opportunities. When the Keynesians do think of adjustments in individual wage-rates, they think of blanket changes. At one point Keynes objected to price flexibility as a remedy for idleness in labour on the grounds that “there is, as a rule, no means of securing a simultaneous and equal reduction of money wages in all industries.”11 But it is not uniform reductions which are wanted, it is selective reductions, the appropriate selection of which can be entrusted to markets when non-market minima have been adjusted.12

But even if equi-proportional wage-cuts were enacted, in a régime in which there was much unemployment, aggregate and average earnings might still tend to increase,13 owing to the redistribution of workers over the different wage-rate groups. It would become profitable to employ more in the higher-paid types of work, whilst in the lower-paid types there would have to be rationing.14 Keynes’s static, short-term methods exclude consideration of these reactions.15 Clarity will not be gained whilst we try to think in terms of “wage levels.” We have to think in terms of changing frequency distributions. This is important enough for the consideration of employment in individual industries, but still more important in relation to employment as a whole.

The Keynesian argument is that it is no use cutting the wage-rates of say, carpenters, if there is unemployment among them because, even if their employment fully recovers, their incomes and expenditure will fall and so cause the demand for the labour of other workers to fall.16 But the case for price flexibility by no means assumes that a moderate fall in carpenters’ wage-rates, together with a corresponding fall in the price of the product will, in itself, greatly increase the employment of carpenters. Such a reaction, although possible, is most unlikely.17

The correct proposition can be put this way. Increased employment among carpenters can be most easily induced as the result of wage-rate and price reductions on the part of those persons who ultimately buy the carpenters’ services. The assumption is that the reductions result in the release of withheld capacity in the industries which do not compete with carpenters, whilst the increasing flow of products becomes demand through being priced to permit its full sale. This is the argument which the Keynesians should answer.

In his Prosperity and Depression, Haberler expressed doubts about this type of argument. He stated the case for it briefly, in a footnote,18 but added that it assumed MV to be constant. I shall try to show that whatever MV may be, the value adjustments needed to secure the consumption or use of all goods and services may still be brought about. Haberler argues also that we cannot infer the truth of the proposition from facts which appear to support it. During the depression, outputs and employment were maintained in the agricultural field, in which the fall of prices could not be effectively resisted, but shrank in industry, in which prices could be effectively maintained. It would seem, then, that full employment and outputs could have been maintained. It would seem, then, that full employment and outputs could have been maintained in industry also, had price competition been effective. That, says Haberler, “has not yet been rigorously proven.”19 But is it not self-evident that, given any monetary policy, selective reduction of the prices of industrial goods would, in general, have made smaller reductions of agricultural prices necessary (in order to secure full employment in that field), whilst the maintenance of outputs as a whole would have eased the task of financing full production without diluting the money unit?20 And is it not equally obvious that, had the price of agricultural products been maintained, so that these products absorbed a greater proportion of the total power to purchase, industrial unemployment would have been still more serious?

The relation between wage-rates and the aggregate pay-roll cannot, I suggest, be effectively considered, except in relation to the price system as a whole. But the Keynesians appear to take the co-ordinative effects of the value mechanism for granted and concentrate upon what they regard as the motive power behind it, namely, money income. They do not continuously envisage and consider the synchronising function of prices, the fact that the prices attaching to individual commodities or services determine the rate of flow at which these commodities or services move into consumption or into the next stage of production. The co-ordination of the rates of flow of materials, services, etc., is brought about through the raising or lowering of prices. Ceteris paribus, a rise in price causes a falling off in the rate of flow, and a fall in price causes a rise in the rate of flow of anything through the stage of production at which it is priced. If certain prices cannot change, other prices (i.e., other rates of flow) must adjust themselves accordingly if the economy is to be synchronised in any sense.21

“Full employment” is secured when all services and products are so priced that they are (i) brought within the reach of people’s pockets (i.e., so that they are purchasable by existing money incomes) or (ii) brought into such a relation to predicted prices, that no postponement of expenditure on them is induced. For instance, the products and services used in the manufacture of investment goods, must be so priced that anticipated future money incomes will be able to buy the services and depreciation of new equipment or replacements.

Admittedly, the view that co-ordinative reductions or increases of wage-rates must always tend to increase real income (and probably real wages in the aggregate also) does not imply that money income (and money wages) will also increase, except on certain assumptions about the nature of the monetary system which exists. Perhaps the pre-Keynesian economists could be criticised for having made tacit instead of explicit assumptions on this point. But orthodox economics (as I understand it) did not overlook what is now called “the income effect.” The tacit assumption22 was that the monetary system was of such a nature that the increased real income due to the release of productive power in individual trades (through the acceptance of lower wage rates) would not result in a reduction of money income. No-one suggested that the monetary system had necessarily to be like that; but from the actual working of the credit system, it seemed to be unnecessary to consider the case in which an expansion of production would not be accompanied by an increase in money income induced by this expansion. The assumption on which Keynes built, namely, that the number of money units is fixed, would have seemed absurd to most pre-Keynesian economists, unless they were considering the economics of a community so primitive that a fixed number of tokens (shells, for instance) served as the sole medium of exchange, whilst no lending or credit of any kind existed.

In a credit economy, there could never be any difficulty, due to the mere fact that outputs had increased, about purchasing the full flow of production at ruling prices. That is, expanding real income could not have, in itself, any price depressing tendencies. Only monetary policy was believed to be able to explain that. But given any monetary policy, they believed that unemployment of any type of labour was due to wage-rates being wrongly related to the “amount of money” existing at any time.23 It followed that downward adjustments of minimum wage-rates and prices could never aggravate—on the contrary would always mitigate—the consequences of any deflationary tendency caused by monetary policy.

Ought we not now to recognise that it is unnecessary to modify this pre-Keynesian view? Under any monetary system, the price situation which permits ideal co-ordination, in the sense which I have explained, must maximise the source of real demand—real income. Whilst this may be clear enough in the case in which monetary policy precipitates primary deflation, it may be less obvious when secondary deflation is induced. But postponements of demand, with their self-perpetuating consequences, arise when current costs or prices are higher than anticipated costs or prices.24

In more general terms, expected changes in costs or prices, unaccompanied by immediate cost and price co-ordination to meet expectations, lead to “secondary” reactions. A cut in costs does not induce demand postponement; nor, indeed, do falling costs have this effect. Postponements arise because it is judged that a cut in costs (or other prices) is less than will eventually have to take place, or because the rate of fall of costs (or other prices) is insufficiently rapid. It follows that “secondary” deflations are attributable to the unstable rigidities which prevent continuous co-ordination of prices. Confusion arises because secondary deflation can be brought to an end, not by true coordination, but at the expense of a prospective permanent sacrifice of real income, i.e., through the imposition of cost and price rigidities (in the form of minima) which are expected to continue indefinitely.25

Now if, for any reason, a change in the value of the money unit becomes the declared object of policy, or the expected consequence of policy, the whole price system is immediately thrown out of co-ordination. Thus, if the value of the money unit is expected to rise, then until the necessary adjustments have all taken place, “willingness to buy” must necessarily fall off—most seriously where values of services and materials in the investment goods industries do not at once respond.26

We turn finally to explicit criticisms of the reasoning on which Keynes based his suggestion of unemployment equilibrium under wage-rate flexibility or, as his disciples were later forced to argue, under price flexibility.

Through the attempts of disciples27 like Lange, Smithies, Tobin, Samuelson, Modigliani and Patinkin to defend or strengthen the new creed, successive refinements have gradually paved the way for the ultimate abandonment, by would-be Keynesians, of the view that wage-rate and price adjustments are powerless to secure full employment. The contributions of these very friendly critics, said Schumpeter, “might have been turned into very serious criticisms” if they had been “less in sympathy with the spirit of Keynesian economics.”28 He added that this is particularly true of Modigliani’s contribution. He could have made the same remark about that of Patinkin, which appeared two years later. But the criticisms of these writers were very serious in any case. Their apparent reluctance to abandon standpoints which their own logic was urging them to reject, clouded their exposition; but it did not weaken the implications of their reasoning.

Modigliani (whose 1944 article29 quietly caused more harm to the Keynesian thesis than any other single contribution) seems, almost unintentionally, to reduce to the absurd the notion of the co-existence of idle resources and price flexibility. He does this by showing that its validity is limited to the position which exists when there is an infinitely elastic demand for money units (“the Keynesian case”). Modigliani does not regard this extreme case as absurd and, indeed, declares that interest in such a possibility is “not purely theoretical.”30 Yet Keynes himself, in dealing explicitly with this case, described it as a “possibility” of which he knew of no example, but which “might become practically important in future,”31 although there are many passages in The General Theory which (as Haberler has pointed out32 ) rely upon the assumption of an infinitely elastic demand. “The New Keynesians” appear to be trying to substitute this “special theory” (Hicks’s description) for the “general theory” which they admit must be abandoned.

It is my present view that any attempt to envisage the “special theory” operating in the concrete realities of the world we know—even under depression conditions—must bring out its inherent absurdity.33 But let us keep the discussion to the theoretical plane. If one can seriously imagine a situation in which heavy net saving persists in spite of its being judged unprofitable to acquire non-money assets, with the aggregate real value of money assets being inflated, and prices being driven down catastrophically, then one may equally legitimately (and equally extravagantly) imagine continuous price co-ordination accompanying the emergence of such a position. We can conceive, that is, of prices falling rapidly, keeping pace with expectations of price changes, but never reaching zero, with full utilisation of resources persisting all the way.34 We do not really need the answer which first Haberler, and then Pigou, gave on this point, namely, that the increase in the real value of cash balances is inversely related to the extent to which the individual (or for that matter the business firm) prefers to save, whilst the rate of saving is a diminishing function of the accumulation of assets which the individual holds.35

I have argued above that the weakness of Keynes’s case rests on his static assumptions; and that once we bring dynamic repercussions into the reckoning (via the co-ordination or discoordination of the economic system) his arguments for unemployment equilibrium under price flexibility fall away. Strangely enough the new Keynesians have themselves transferred the fight to the dynamic field. The position they now seem to assume is that, whilst Keynes’s own analysis (essentially static) cannot be defended, his propositions survive if they are explained through dynamic analysis. But in their attempt to retain Keynes’s conclusions, they have abandoned the very roots of his own reasoning.

Thus, Patinkin36 is equally specific in rejecting the original Keynesian arguments concerning unemployment equilibrium. He says, “it should now be definitely recognised that this is an indefensible position.”37 Even so, Keynes’s errors on this point, and the similar errors of his manifold enthusiastic supporters over the period 1936-1946, are represented by Patinkin as quite unimportant. The truth which the early critics of The General Theory fought so hard to establish (against stubborn opposition at almost every point38), namely, that price flexibility is inconsistent with unemployment, he describes as “uninteresting, unimportant and uninformative about the real problems of economic policy.”39 In spite of the mistakes which led Keynes to his conclusions, he did stumble upon the truth.

Let us consider, then, the conclusions concerning price flexibility of what Patinkin continues to describe as “Keynesian economics” (meaning by that an economics which rejects the logic but retains the conclusions of The General Theory). This version of “the New Keynesianism” contends—again in Patinkin’s words—“that the economic system may be in a position of under-employment disquilibrium (in the sense that wages, prices, and the amount of unemployment are continuously changing over time) for long or even indefinite, periods of time”40 (Patinkin’s italics). “In a dynamic world of uncertainty and adverse anticipations, even if we were to allow an infinite adjustment period, there is no certainty that full employment will be generated. I.e., we may remain indefinitely in a position of under-employment dis-equilibrium.”41

This sounds like pure orthodoxy. Indeed, the use of the word “disequilibrium” implies that some Keynesians have now completely retreated. And the reference to “uncertainty and adverse anticipations” seems to refer to hypothetical situations which, using my own terminology, can be described as follows:

Given price rigidities regarded as unstable, deflation will cause the emergence of withheld capacity. Three cases arise: (a) general expectations (i.e., typical or average expectations) envisage a fall of prices towards a definite ultimate scale which is regarded as most probable; or (b) general expectations are constantly changing so that the generally expected ultimate scale of prices becomes continuously lower; or (c) general expectations envisage a certain rate of decline of the scale of prices in perpetuity.

In case (a), the withholding of capacity will last over a period which will be longer the more slowly the predicted price adjustments come about. In cases (b) and (c), the withholding of capacity will last over an indefinite period, unless downward price adjustments take place as rapidly as or more rapidly than (i) the changes in expectations, or (ii) the generally expected rate of decline, in which case full employment will persist throughout. In short, when the scale of prices is moving or is expected to move in any direction, the notion of perfect price flexibility must envisage current prices being adjusted sufficiently rapidly in the same direction, if the full utilisation of all productive capacity is sought.

In admitting that Keynes cannot be said “to have demonstrated the co-existence of unemployment equilibrium and flexible prices,” Patinkin explains that this is because “flexibility means that the money wage falls with excess supply, and rises with excess demand; and equilibrium means that the system can continue through time without change. Hence, by definition, a system with price flexibility cannot be in equilibrium if there is unemployment.”42 Now if by “excess supply” is meant more than can be sold at current prices, and by “excess demand” more than can be bought at current prices, it remains true, equally “by definition,” that price flexibility so conceived is inconsistent with wasteful idleness, even when we take into account the full dynamic reactions which are theoretically conceivable under a condition of falling or rising prices. For price flexibility then requires that all prices shall be continuously adjusted so as to bring the spot and future values of the money unit into consistency; in other words, to establish harmony between current and expected prices. Under such adjustments, even unemployment dis-equilibrium is ruled out.

Do not the words “adjustment period” in the passage quoted above show that Patinkin, in using the term “disequilibrium,” is in fact still envisaging some price rigidity? What other adjustments, apart from changes in prices and effective exchange values can he be envisaging? How else can the terms “uncertainty” and “adverse expectations” be explained, unless in relation to unstable price rigidities? And the same tacit assumption of rigidity is present in his statement of what he terms, “the Keynesian position, closest to the ‘classics.’” In this position, he says, although price flexibility would eventually “generate” full employment, “the length of time that might be necessary for the adjustment makes the policy impractical.”43 He tells us that this statement (like that in the previous quotation) is not “dependent upon the assumption of wage rigidities.”44 But what “adjustments” other than tardy cuts in rigid wage-rates has he in mind? He must be thinking of unstable price rigidities somewhere in the system.

A critic writes that this argument seems to overlook inevitable rigidities. In practice, contracts cannot be varied constantly, so that costs tend to follow prices with some interval. Thus, copper miners’ wages can hardly change every time the price of copper changes. But for Patinkin’s argument to hold, it would be essential for the wage-rates of the miners to be maintained when actual or expected copper prices had fallen to such an extent that formerly marginal seams became unworkable at current costs. The most complete measure of price flexibility practically attainable involves discontinuities at both the cost and the final product ends.45 But periodic adjustments through recontract (as idleness threatens) can meet that situation.46

In short, the kind of price flexibility for which we can reasonably hope is one in which the price inconsistencies which must exist at any point of time are never in process of material or cumulative worsening. That need not mean unemployment. Contract covers the short run. And inconsistencies need not accumulate: they can be in process of rectification at about the same rate as that at which they arise.

Hence, “the dynamic approach” does not, as Patinkin maintains, obviate the necessity for the assumption of rigidities and revalidate the Keynesian fallacies. On the contrary, it was largely Keynes’s neglect of the dynamic co-ordinative consequences of price adjustment which led him into the error that wage-rate and price adjustments are no remedy for unemployment.47

What are the implications? In my judgment, the abandonment of the theory of unemployment equilibrium under price flexibility means that the Say Law stands once again inviolate as the basic economic reality in the light of which all economic thinking is illuminated. But I do not think that all the critics of Keynes on the point at issue will immediately accept this inference. Indeed, Habeler adheres to a rejection of the Law at the very stage at which his own reasoning seems to be prompting him to recognise it.48

Yet even so extreme a Keynesian as Sweezy has been rash enough (and right enough) to admit, in his obituary article on Keynes, that the arguments of The General Theory “all fall to the ground if the validity of the Say Law is assumed.”49 If my own view is right, then the apparent revolution wrought by Keynes after 1936 has been reversed by a bloodless counterrevolution conducted unwittingly by higher critics who tried very hard to be faithful. Whether some permanent benefit to our science will have made up for the destruction which the revolution left in its train, is a question which economic historians of the future will have to answer.

We are now forced back to the stark truth that the elimination of wasteful idleness in productive capacity is attainable only through the continuous adjustment of prices or the continuous dilution of the money unit. But the latter is a tragically evil method of attempting to rectify disco-ordination due to inertias or sectionalism. For the harmful repercussions of inflation become the more serious (and force an accelerated inflation) the more successfully entrepreneurs and consumers, in the free sectors of the economy, correctly forecast monetary policy. But the new Keynesians, like the old, appear to believe that monetary or fiscal policy, through the control of spending, can act as a universal solvent of all price disharmonies and, like an invisible hand, make unnecessary, or less necessary, the difficult task of overhauling the institutions which make up the price system.

We must remember that the attack on wage-rate adjustment as a policy of securing full employment in labour is an attack on a policy which has never been experimentally tested. For whilst there is a great deal of evidence of wage-rate adjustments forced by depression being followed by recovery, no deliberate attempt to increase income (including the flow of wages) by reducing all prices which appear to be above the natural scarcity level (including wage-rates) so that all prices and wage-rates below the natural scarcity level may rise, has ever been purposely pursued. Actual policies have, for decades, been based precisely upon the politically attractive rule, justified by Keynesian teaching, that disharmony in the wage-rate structure must not be tackled but offset; whilst the current tendency is to assume dogmatically with no examination of the institutional and sociological factors involved, that to advocate wage and price adjustments is to recommend the conquest of the moon.

6 General Theory, page 11.

7 TIbid., p. 253.

8 Compare criticisms of “the wage level” concept by R. A. Gordon (A.E.R. Proceedings, May, 1948, page 354) who refers to “ . . . the concentration of attention upon aggregates and upon distressingly broad and vaguely denned index number concepts—with insufficient attention being paid to those inter-relationships among components which may throw light upon the behaviour of those aggregates . . .”

9 The General Theory, p. 259.

10 It is an interesting commentary on the uncritical nature of current assumptions that Professor Viner has felt it necessary to remind economists that it does not necessarily follow, “and I think that many economists have taken that step without further argument,” that an increase of wage-rates at a time of unemployment will increase the pay-off. “An increase of wage-rates may quite conceivably reduce the pay-roll.” (Viner, op. cit., page 32).

11 The General Theory, page 264. It was partly this which led him to argue that wage-rate adjustment would be possible only in a Communist or Fascist State. (Ibid, page 269).

12 Actually, Professor Pigou has shown that equi-proportional wage-cuts, even under Keynes’s other assumptions, must mean increased employment of labour if the reaction is a reduction of the rate of interest. Professor Pigou suggests that this reaction is “fairly likely.” “Money Wages and Unemployment,” Economic Journal, March, 1938, p. 137.

13 As measured by money units of unchanging value.

14 For simplicity, I am assuming that maxima are enacted.

15 The possibilities of transfers of workers from low-paid to high-paid work are magnified in the long run, because it will be possible to train for the well-paid employment opportunities which are brought within reach of income.

16 Professor K. Boulding has used this actual example and argument in his Economics of Peace, pp. 141-2.

17 Moreover, whilst wage-rate and price adjustments are required to dissolve withheld capacity among carpenters, to adopt that remedy in individual trades and on a small scale would bring severe distributive injustices in its train. Indeed, the aggregate wage receipts of the larger number employed in any trade might be smaller than before the increased employment.

18 Haberler, Prosperity and Depression, p. 493.

19 Ibid., p. 243.

20 I feel that Haberler would now admit this argument, in view of his unequivocal rejection, in 1951, of Keynesian teaching about unemployment equilibrium under price flexibility. “Welfare and Freer Trade,” Economic Journal, Dec, 1951, pp. 779-80. See also his article in The New Economics, pp. 166 et seq.

21 What is commonly expressed as changes in cost-price ratios, i.e., in the price of output in relation to the price of labour, I think of in terms of divergencies from, or conformance with, synchronizing prices at various stages of production. (The last stage is, of course, sale for consumption.)

22 Some economists in the pre-Keynesian era, in attempting to deal with the relations of employment and wage-rates, made explicit, highly simplified assumptions consistent with the assumption as I have worded it, for purposes of aDstract analysis. But I do not know of any economist who has stated the fundamental assumption as I have done. Quite possibly the point was made.

23 Compare F. Modigliani, “Liquidity Preference and the Theory of Interest and Money,” Econometrica, January, 1944, and this symposium, pp. 132-184.

24 My article in the issue of this Journal for December, 1953, is an attempt to deal rigorously with this situation.

25 Imposed cost and price rigidities in the form of maxima (i.e., ceilings) may similarly prevent secondary inflation, but in this case, the effect is the opposite. In so far as the maxima force down monopoly prices nearer to marginal cost, there is a mitigating co-ordinative and deflationary action which creates an incentive to increased outputs (i.e., increased real income).

26 it should be stressed, however, that this is no conclusive argument against policies seeking to increase the value of the money unit, as tardy rectifications of the distributive injustices of inflations. Nor is it a good argument against rectifying price disharmonies which have been allowed to develop and strain the ability to honour a convertibility obligation.

27 I do not include Haberler, whose criticisms have been damaging, as a Keynesian. It is difficult to pick out the other non-Keynesian economists who have been most influential on the point at issue; but Marget, Knight, Viner and Simons must take much of the credit.

28 Schumpeter, in The New Economics, p. 92.

29 “Liquidity Preference and the Theory of Interest and Money,” Econometrica, January, 1944. Reprinted in this symposium, pp. 132-184.

30 Pigou regards the contemplation of this possibility as “an academic exercise.” He describes the situation envisaged (although he is not criticising Modigliani) as extremely improbable, and he adds, “Thus the puzzles we have been considering . . . are academic exercises, of some slight use perhaps for clarifying thought, but with very little chance of ever being posed on the chequer board of actual life.” “Economic Progress in a Stable Environment,” Economica, 1947, pp. 187-8.

31 General Theory, p. 207.

32 Op. cit., p. 221.

33 No condition which even distantly resembles infinite elasticity of demand for money assets has even been recognized, I believe, because general expectations have always envisaged either (a) the attainment in the not too distant future of some definite scale of prices, or (b) so gradual a decline of prices that no cumulative postponement of expenditure has seemed profitable. General expectations appear to have rejected the possibility of a scale of prices which sags without limit, because of such things as convertibility obligations, or the necessity to maintain exchanges- or the political inexpediency of permitting prices to continue to fall.

34 See below, and compare Pigou, op. cit., pp. 183-184; Haberler, Prosperity and Depression, pp. 499-500.

35 in any case, this argument is no answer to the case in which the nature of saving is speculative hoarding. For this reason Haberler claims only that there is “a strong probability” and no “absolute certainty” of there being a lower limit to MV so caused. {Op. cit., p. 390.)

36 Patinkin, “Price Flexibility and Full Employment” (A.E.R., 1948). Quotations are from the revised version in the A.F.A. Readings in Monetary Theory.

37 ibid., p. 279.

38 For an example of the stubbornness, see Keynes’s reply to criticisms in his “Relative Movements of Real Wages and Output,” Economic Journal, March, 1939.

39 Patinkin, op. cit., p. 279.

40 Ibid., p. 280.

41 Ibid., p. 281.

42 ibid., p. 279.

43 Op. cit., p. 282.

44 op. cit., p. 282.

45 That is not, in itself, likely to mean discontinuity in movements of the scale of prices (i.e., in a price index).

46 Sliding scales can render the need for recontract less frequent.

47 The confusion in this field ultimately stems, I feel, from a failure to achieve conceptual clarity, and particularly owing to the absence of a sufficiently rigorous definition of price flexibility.

48 Haberler, in Harris, op. cit., pp. 173-176. The acceptance of the Say Law does not imply, as Haberler suggests, the absurd assumption that the phenomena of hoarding or dishoarding cannot exist. It merely accords to money assets and the services which they provide the same economic status and significance as all other assets and the services which they provide. Nor does the existence of depression or idle resources (under unstable price rigidity) prove that this law does not hold, any more than balloons and aeroplanes invalidate the law of gravity.

49 In Science and Society, 1946, p. 400.