A Rehabilitation of Say’s Law

X “EFFECTIVE DEMAND” IN RELATION TO SAVING PREFERENCE AND LIQUIDITY PREFERENCE

It is puzzling to me that such acute critics of Keynes as Leijonhufvud, Clower and Yeager still seem not to have shaken off the notion (expressed in my terminology) that some supplies of non-money do not become “effective” demands because, in the aggregate, sellers demand (temporarily) additional money instead of other non-money; or, as it is sometimes expressed, that conditions can arise in which an “excess demand” for money can develop—an “excess” which (in Keynes’ words) can be “choked off” only by depression and unemployment. In such circumstances, it is held, “aggregate demand” becomes “deficient” (some portion of that demand being “ineffective”).

For example, using Keynes’ phrases from his Treatise on Money (1930) and not from the General Theory, following him in linking saving preference with liquidity preference, and stating that savers want “wealth as such,” described as “the potentiality of consuming an unspecified article at an unspecified time,”1 Leijonhufvud appears to think that Say’s law is upset, and enduring unemployment disequilibrium is made possible because, on occasion, savers (see my definition, p. 16) invest in additional money rather than in non-money.2

But non-savers or even dissavers may also want to acquire “unspecified things at unspecified times” to a greater extent than previously and so increase their holdings of money or the proportion of money in their total assets. In general demands for the services of money are independent of the saving process.

The general notion that I am questioning seems to be that some demands are “ineffective” because they are not expressed through the offer of money. We have seen, however, that a self-employed farmer may, without purchasing, demand and consume the inputs of his own labor and those of his equipment.3 But the real point at issue is surely that when a person buys, he normally demands with money’s worth, not with money. He demands with money only when he happens to be reducing his investment in it (i.e., not concurrently replenishing his money holdings), for he can always obtain money costlessly by realizing his inputs or outputs (services or assets) at their money’s worth. (An awareness of my definitions of “money’s worth,” “market-clearing price,” pp. 19-20, and “withheld capacity”, 20-21, as well as my further explanation of these concepts in Chapter V, is, I think, essential for a full understanding of this and the following paragraph). I say that he can; but it may not be profitable for him to do so; for the market-clearing price of the full output of a thing exceeds money’s worth when investment in inventories of it promises yields in excess of interest. The reader is reminded that, when a person is asking more than the money’s worth of the whole output of what he and his assets produce, he may be regarded as investing in the output; while if he asks more than the market-clearing value, then he is withholding supply and contributing to the cumulative contraction of activity that we have been considering.

The holding of money is not costless. Just as with all other assets held, interest is a cost. And money is productive. The incentives to invest in it (as I stressed above) are no different from those for investment in non-money assets, with the ever-present possibility of speculative gain or loss to supplement or detract from the non-speculative yield. It is the acquisition and spending of money which is costless. It follows that money is as incidental (and as important) as cash registers and cashiers in the demanding and supplying process. I hope that, in stressing this truth (which the pre-Keynesian economists understood), I shall not be accused of attempting to explain “monetary phenomena by real causes.” (See pp. 110-111).

Leijonhufvud’s repeated references to “purchasing power” stress the implication I am challenging, namely, that the use of money to pay for labor’s inputs somehow either prevents those inputs from being supplied, or if supplied, from being an effective contribution to demands for non-competing things. I have been unable to find any satisfactory explanation of why the need for a temporary depletion of money stocks in the course of paying for labor, or the need for the replenishment of money stocks (as with other inventories) in order to be in a position to pay for labor at the next payday, can either frustrate the will to supply labor, or reduce (or destroy) the profitableness of demanding labor. In short, I have discerned no reasons why the employment of the monetary mechanism should prevent potential demands for non-competing outputs from becoming actual demands.

The fact that people in general (savers, non-savers and dissavers) will at times judge money to become more desirable than previously relatively to non-money does not seem to me to be relevant. They will do so, of course, when they predict that present yields from investment in additional non-money assets are less favorable than future yields from non-money investments are destined to be; and this usually implies that they are expecting (a) downward cost adjustments, or (b) downward price adjustments, or (c) a rise in the market rate of interest at some time in the future. In the case of (a) and (b) we have the phenomenon of what I have called “unstable price or cost rigidities.” And far from the market pressures exerted when speculative investment in liquidity is so induced being of an equilibrium-disturbing type, those pressures are tending to break through inertias, or other unstable barriers, and facilitate equilibrium.

Moreover, under any monetary standard of the commodity type (e.g., the gold standard), speculative demand for liquidity may supply an additional, deflationary compulsion for the price and wage-rate adjustments towards the market-clearing levels which alone can restore the wages and income flow. This does not mean that, at times, speculative investment in money may not over-correct, or that attempts to correct through a policy of “monetary flexibility” may never over-correct. But under effective monetary flexibility, although deflationary pressures to price adjustment will be avoided,4 other pressures tending to break down barriers to recovery—unstable cost and price rigidities—will still build up (through the absence of hopes of any inflationary validation of unstably rigid costs). Such pressures will increase in force because it will become increasingly obvious that it is not in the interests of would-be wage earners or the owners of assets that their valuable (i.e., potentially demanded) productive capacity shall remain idle.

We can now return to Leijonhufvud’s conviction (following Keynes) that recession is aggravated inter alia by provision for the future (saving) rising in people’s scales of preference when activity declines and by the superiority of money as a store of value when this occurs. He assumes that the prospective productiveness of money increases, not for speculative reasons owing to downward cost rigidities (and price rigidities) being judged to be unstable, but for other reasons.

There are two separate questions raised here. Firstly, because the condition of recession implies a decline in real income or a decline in the rate of growth of real income, why should consumption in the distant future come to rise in scales of preference in relation to consumption in the immediate future? One would expect both the absolute aggregate value of the replacement of consumption and the absolute value of additions to the stock of assets to decline but with the latter declining relatively; for it is surely easier to postpone provision for the future than to postpone consumption. On the other hand, apprehensions of straitened circumstances might conceivably induce people in general to decide to save a larger proportion of a falling real income. For the sake of argument, let us assume that they do.

We are then brought back to the second question which we encountered earlier (pp. 67-69) Why should savers in particular (as distinct from other owners of assets) regard money as a superior store of value to non-money assets? Leijonhufvud’s idea that savers “want wealth as such,” in the sense of wanting “the potentiality of consuming an unspecified article at an unspecified time,” does not explain why, for the time being, savers judge the productivity of money to rise relatively to the productivity of non-money: it is simply a way of asserting it. Leijonhufvud does not refer in this context to such factors as I have just mentioned.

It is true, of course that savers want “wealth as such.” But all assets are wealth. All assets and not merely money are a “store of value.” And while at times money may seem to be an exceptionally efficient store of value, at other times it is a prospectively poor store of value. As I have already insisted, all assets and services constitute power to demand through buying at their money’s worth. (See definition, p. 19) But I can conceive of no reason why, (i) during periods of increasing thrift, or (ii) (a wholly separate issue) during recession, or (iii) during (i) and (ii) together, the non-speculative prospective yield to money held should rise relatively to yields to non-money.

The non-speculative prospective productivity of money held as a producer good will tend to fall, not rise, as profitable outputs decline. Inventories of money will not be fully replaced for the same reasons that inventories of materials will not be fully replaced in recession. Post-Keynesian economists seem to forget that the rate of interest is a cost of holding all assets. In my own confessedly heretical judgment, when the market rate of interest is abnormally low, the only reason why the relative productivity of money should increase is that the probability of a rise in the rate of interest is then likely to be higher and hence likely to induce a speculative increase in demand for money.

In the case of recession, only (a) downward cost and price rigidities judged to be unstable, or (b) deflation [which may temporarily aggravate the factors under (a)], or (c) a forecast rise in the rate of interest, can explain an increase in speculative demand for money which more than offsets the decline in non-speculative demand for it (due to the contraction of “supplies” in general). And even so, the bidding up of the aggregate real value of money for such reasons (or for any other imaginary reason, such as mere inertia in the acquisition of non-money while savings are increasing) need not be deflationary. That is, one is not compelled to assume any particular monetary policy.

The misconceptions which, I think, abound on this subject may be largely due to misleading terminology. If any portion of “aggregate demand” is described as “ineffective” when it is directed to money, in spite of the fact that the demander of money obtains “wealth as such”—assets equal in value to what he sells in order to invest in money—it is a most eccentric use of the term “ineffective,” to say the least. The demander obtains what he demands! And when that portion of “aggregate demand” which is directed to money is simultaneously called “excess,” it implies not only that that portion of demand is effective but that it is too effective! Then if so, what portion? The portion which would cause deflation under monetary rigidity? If that is the implication, why is it not expressed that way? An increase in the relative demand for money simply means that the aggregate real value of money rises relatively to the aggregate real value of non-money. But this is truistical. We do not say that some portion of the demand for rye is “ineffective” because some former purchasers of it demand wheat instead.

In Keynes’ equations, M (the money supply) is constant. But this explicit assumption of monetary rigidity for expositional purposes tends not to be abandoned when policy conclusions are inferred; and in post-Keynesian economics, it often intrudes, I think, as a tacit, implied assumption. Yet it is fantastically unrealistic. For instance, what eventually became Keynes’ bête noire, the pre-1914 gold standard, provided a very high degree of monetary flexibility.5

The real issues are, I suggest, most successfully clarified under the opposite and less unrealistic assumption, namely, of perfect monetary flexibility. Under “monetary flexibility,” the money valuation of real income varies in proportion to real income. Hence no “gap” can arise in “the stream of spending” except through the withholding of “supplies” and the consequential contraction of income. Yet under this imaginary condition recession, or even a serious depression can still develop. Typical prices of non-money (an appropriate price index) will not fall, but an economy can run down just as it can under anticipated inflation. Hence the notion of what I have called “the supply,” as distinct from the “potential supply,” of one thing becoming “ineffective” as a contribution to the source of demands for non-competing things (other than money) has no meaning. “Monetary flexibility” (in contrast to “price flexibility”) alone is incapable of correcting the “automatic” process which throws men and assets into idleness.

The reality is, then, that depression can emerge only through an interruption of the stream of “supplies” any interruption of the “stream of spending” being an effect—not a cause.6 But if the “money supply”—the value of the money stock in terms of actual money units—is caused to lag in relation to an increase in the aggregate real value of that stock (the case of monetary rigidity) the scale of prices must fall and, under the additional assumption of widespread wage-rate and price rigidity, depression will be inevitable. In those circumstances, deflationary monetary policy, not the use of money, explains everything. Of course, periodic rectifying deflations (offsetting an unintended inflationary situation), i.e., what Leijonhufvud presumably means by “recurrent attacks of central bank perversity,”7 must cause some avoidable dislocations and possible strains on the adjustment mechanism. But any depressive consequences will still be due to the withholding of supplies. The notion of demands being “ineffective” explains nothing.

When there is an “excess supply” of labor, says Leijonhufvud, “the offer of labor services (by the unemployed) does not constitute exercise of purchasing power over commodities, i.e., it does not constitute effective demand.”8, 9 “Workers find that their labor is not a source of direct purchasing power over output.”10 Well, admittedly, a worker’s skills and muscle power do not become “purchasing power” (or barter power!) if they are priced (or valued) above the market-clearing level; and in this case that means inconsistently with prospective yields from investment in them; or, expressed somewhat differently, if they are not priced in conformity with the community’s predicted ability to acquire the output, which includes the case in which they are priced inconsistently with price expectations. Labor is “supplied” only when the workers who offer it accept a wage-rate which entrepreneurs believe will return (through subsequent sale of the resulting output) not less than it has cost (the “means of payment” for it); then, through the labor inputs supplied and the outputs into which they are embodied having been priced at their money’s worth, they become “money-spending power”; and if there has been (a) no monopolistic withholding of labor supplied, and (b) neither monopsonistic constraint in the purchase of labor’s inputs nor monopolistic constraint in the pricing of the outputs, the money’s worth of both inputs and outputs will have corresponded to their market-clearing prices.

Now the “constraints” to which I have just referred appear to be seen by Leijonhufvud as arising out of the process of depleting and replenishing money. Referring to “producers” (meaning the residual claimants on the value of outputs, and not the workers), he says they “find that their output is not a means of payment for the purchase of labor’s inputs”.11 But while neither inputs nor outputs are money, they do have money’s worth, which, ceteris paribus, will be higher the larger the potential supplies that are “withheld.” And as I have already stressed, at their money’s worth the exchange into money is costless, as is the subsequent exchange of money for a different form of non-money. There is, as I have insisted above, a cost of holding money, not a cost of using it to sell or buy.

Because all firms must sell their outputs in order to pay wages and settle debts, they would not have risked providing the fixed assets needed for the employment of workers unless they had been able to plan with confidence the replenishment of their money stocks as they met wages claims, just as they had planned also for replenishment of inventories of materials. That is, in a money economy, entrepreneurs try to proceed so as always to achieve a sufficient stock or flow of money’s worth to settle all debts incurred, including those incurred with contractual claimants on the value of outputs, as well as to earn a residue and avoid losses.

It is equally obvious that some of the goods piling up in warehouses or shops will not be making their full potential contribution to the source of demands if they are being priced higher than is destined to induce their sale at the optimal rate, i.e., higher than will maximize the yield to the capital invested in them. What may be called “normal” inventories, at all stages of production, consist of goods of which the full production process is incomplete at the stage in question, because all the services demanded by potential customers (including the production of “availability utilities,” “assembly utilities” and “time utilities”) have not yet been incorporated into them. Customers demand “reserves” of non-money and money—that is, they demand (and remunerate) the services of the capital invested in inventories as well as the fixed capital and the risk-bearing function. If allowance is made for this complication, it seems clear that services and goods are not “supplied,” i.e., they do not enter into the source of demands, until (a) their owners invest in them for their own purposes12 (in which case the owners both supply and demand them), or (b) ask a value in exchange or price which induces exchange or purchase (for investment or consumption). What needs to be explained is why some services and some goods should not be so valued or priced.

Leijonhufvud’s continuation of the argument does not appear to me to clear up the difficulties. He says that because “workers ask for money wages, . . . from the standpoint of prospective employers . . . , the offer of labor services is not directly connected with a demand for additional output.”13 I am not sure of the exact meaning of the words, “not directly connected”, or of the phrase, quoted before (p. 74), “labor is not a source of direct purchasing power over output.” A firm taking on formerly unemployed workers will usually plan to invest simultaneously (a) in the services of this additional labor, (b) in additional inventories of materials and work in progress, (c) in additional inventories of money, and (d) (possibly) in additional fixed assets. And as the firm will thereby increase its output, it will be adding to the source of demands for additional outputs of non-competing producers as its products are priced for clearance. Moreover, it will know (or ought to know) that if firms in non-competing sectors of the economy are doing the same thing, there will be for that reason an additional justification for an increase in its output. General recovery in activity gives rise to general and justified entrepreneurial optimism. Such optimism will raise the wage-rates at which employment for additional workers will be prospectively profitable. But less optimistic expectations will not mean that all workers wanting to be employed cannot supply their labor. At appropriately lower initial wage-costs, it will always be profitable to employ them unless their services are offered in a field in which they are completely valueless. I shall return to this point.

The surviving idea in post-Keynesian writings that the exercise of positive time preference is depressive does not always rest on the belief that the use of money creates a hiatus, but that the desire to provide for the future means a cessation of “demand”. Because savers may not now know exactly how they will, at “some unspecified time”, finally satisfy their time preference (i.e., choose the presently “unspecified” services or commodities they are destined ultimately to consume) does not mean that they cease to demand current inputs and outputs. A rise in saving-preference does not mean that savers’ demands are postponed. It means that such demands are directed to different services and assets. Savers demand either (a) productive services of men and assets for embodiment into assets of different prospective lifespans, or (b) when, through this process, the costs of productive services have been bid up to a certain value, they demand existing assets (because investment in them offers, in prospect, a larger income stream).

 

 

14The reference to consumption here is not easily interpreted. Of course the ultimate purpose of all economic activity is consumption. But one is more likely to invest in money temporarily in order later to acquire assets which are expected to yield an income-stream (i.e. long-life assets—producer goods) than for the purpose of acquiring short-life assets (consumer goods) or services.

15Walras suggested that such a person should be regarded as two persons, one who invests in the equipment and one who consumes its services.

16A policy of “monetary flexibility” will permit the development of latent inflation as long as the demand for money is abnormally high, but the monetary authority will be committed to avoiding open inflation when the demand for money becomes normal, i.e., committed to allowing the “money supply” to run down as individual money holdings are reduced (M to decline as V increases) unless, of course, expansion of real income should happen to offset the inflationary tendency.

17Although, in a better system of government than the pseudo-democracy of the 20th century, the “orthodox” economists would have welcomed a system with an even higher flexibility. (See pp. 118-120, 128-132).

18In Leland Yeager’s contribution to Clower’s Monetary Theory, he accepted my reasoning on this point. In his paper, “The Keynesian Diversion”,19 however, he says he now thinks he “conceded too much to the non-monetary view of depression.” But the latter view is truly the explanation of all depression. When deflation is the initiating factor (under downward cost or price rigidity), the economy still runs down through the cumulative consequences of the withdrawal of supplies of non-money.

20Leijonhufvud’s diagnosis of “ineffective demand” for, or “excess supply” of, labor as the chief condition that constrains demands for goods is influenced by and closely resembles Clower’s diagnosis. But the former’s treatment discloses more clearly what I feel is the core of the controversy and I have found it most fruitful therefore to direct my criticism mainly at Leijonhufvud’s text.

21“Their own purposes” covers the intention to consume the goods. (See definition of “supply”, pp. 14-15, 34.

 

 

22  Leijonhufvud, Keynes and the Classics, p. 37.

23  Yeager, “The Keynesian Diversion”, Western Economic Journal, 1973, p. 8.

24  Leijonhufvud, On Keynesian Economics and the Economics of Keynes, New York, O.U.P., 1968, p.399.

25  Leijonhufvud, Keynes and the Classics, p. 35.

26  Ibid., p. 35.

27  Ibid., p. 35.

28  Ibid., p. 35.

• 1*T. Sowell, Say’s Law, Oxford University Press, 1973.
• 2 W. H. Hutt, Economists and the Public (London, Jonathan Cape, 1936), p. 34.
• 3*R. Clower and A. Leijonhufvud, Say’s Principle, What It Means and What It Does Not Mean (Intermountain Economic Review, Vol. IV, No. 2, Fall, 1973).
• 4*Even so, Say’s early phrasings of the “law of markets,” some of which are quoted in Chapter III, are laudably clear and unequivocal: and his detractors during the earlier period can be not unfairly regarded as “amateurs” in economics, in comparison with those who, like Thornton, the Mills, Senior, Ricardo and McCulloch, were learned scholars in the subject. (Compare Sowell, Say’s Law, pp. 220,232).
• 5*James Mill referred to Say as “a late French writer” and quoted an anecdote of Say’s to illustrate the absurdity of regarding consumption as the source of demand.
• 6*This statement holds without qualification in the absence of saving or dissaving. When saving is occurring, assets are being demanded through the offer of services (capital by the offer of income). When dissaving is occurring, services for consumption are being demanded through the offer of assets, or assets are themselves being consumed.
• 7 J. M. Keynes, The General Theory (See note above), p. 19.
• 8 T. Sowell, Say’s Law (Princeton University Press, 1973), pp. 3, 4 (twice), 12, 19, 201.
• 9*I am here thinking of such things as the confusion of saving-preference with liquidity-preference, the confusion of spending with demanding, the confusion of spending with consumption, the fear of general overproduction, the notion that consumption is the source of demands.
• 10 J. A. Schumpeter, History of Economic Analysis (Oxford University Press, 1954), p. 625.
• 11 Ibid., p. 625.
• 12*I use the word “may” because I think that, apart from Turgot, Say’s main inspiration came from Hume, Adam Smith and (possibly) James Mill, in whose works the confusion to which I have referred is absent. Of these, only James Mill stated the principle in general terms (in Commerce Defended, 1808), and his statement was better than Say’s had been in the first edition of the latter’s Political Economy.
• 13 James Mill, Commerce Defended (London, C. and R. Baldwin, 1808), p. 76.
• 14Since this essay was submitted to the publisher, a major work on an important aspect of the same subject has appeared: Say’s Law by Tom Sowell.* This is a work of distinction. Its scholarly treatment of the controversies which Say’s ideas provoked during the first half of the nineteenth century is, I judge, an important contribution to the history of economic thought. But Sowell’s aim and mine are not quite the same. I shall be trying to rehabilitate the “law of markets” by showing “what Say really meant” or “what Say ought to have meant.”
• 15At the galley proof stage, I became aware of yet another relevant contribution, which may turn out to be of path-breaking importance. It is a recent article entitled “Say’s Principle”, by R. Clower and A. Leijonhufvud*. It is essential to draw attention to it. Previous works by these authors suggested that they were hesitant about the importance of Say’s law. I discuss their earlier views at some length. But their treatment of “Say’s principle” (which, as now explained, refers to a vital implication of Say’s law rather than to the law itself) seems to me to record a significant development. The position of these economists is now much nearer to mine than appeared from their previous writings. The student is urged to read their latest treatment as corrective or supplementary to the argument I submit. Holding that “there is hardly a single problem in macro theory (or for that matter, micro theory) that can be consistently analyzed without . . .” a recognition of “Say’s principle,” they maintain that “the extensive literature since Keynes . . . has got bogged down in a mire of conceptual and semantic confusion.” Their words here explain my incentive and why I devote the whole of Chapter II to definitions. I am trying to clear up the confusion.
• 16Controversy about the validity of this law (which occurred mainly in two periods—from 1803-1850 and from 1936 until today) was described by the great Joseph Schumpeter as “a discussion which reflects little credit on all parties concerned.” It has “dragged on to this day,” he said, “with people armed with superior technique still chewing on the same old cud. . . .” Well, more than two decades after the posthumous appearance of Schumpeter’s monumental History of Economic Analysis, from which this scathing comment is quoted, I have been moved to keep the cud-chewing going. This is in spite of my agreement with Schumpeter (a) that Say did not fully understand the “law” to which his name has become attached; and (b) that in his defence of the “law” against acute but economically illiterate critics, Say was sometimes very “woolly,” slipping, at times, into serious fallacy.*
• 17An occasional eccentricity of the erudite Schumpeter was reckless exaggeration. It was unjust of him to have written that Say “hardly understood his discovery himself.” Certainly Say did not fully understand “the law of markets” which, I think, he regarded as accepted doctrine and not as his discovery. That is, I suggest, why James Mill did not refer to Say’s chapter on markets in his Commerce Defended (a most impressive exposition of the “law”), while Say, in his second edition, did not mention James Mill’s contribution which had appeared in between.* Both were explaining what they believed to be unchallengeable. Both attempted to convey to the business world, to governments and to students what the economists had learned from the Physiocrats, Hume and Adam Smith.
• 18“The source of demand” means “the power to demand.” In other words, the power to demand any part of the flow of productive services springs from the flow of non-competing productive services.* Moreover, all producers of goods or services are always in a position to value or price their contribution so as to induce a demand from others for that portion of their inputs or outputs as they (the producers) do not demand (i.e., retain) for themselves, as investors or consumers (unless what they intended to be production turns out to be valueless and wasted—inadvertent consumption).
• 19 M. Blaug, Economic Theory in Retrospect (Homewood, Ill., Irwin, 1968), p. 154 n.
• 20The basic notion of Say’s law had been expressed several decades before Say’s effective enunciation of it. For instance, the Physiocrats generally, as J. J. Spengler has pointed out, had recognized that “commerce consists, not in buying and selling, but in the exchange of goods and services for goods and services,” and that is a very broad yet apt statement of “the law of markets.” Josiah Tucker came near to stating the principle when he wrote, “the work of one man gives employment to another.” And Mercier de la Rivière, although accepting other ideas inconsistent with it, came even closer when he wrote, “no one can buy except by reason of what he sells.” But prior to Say, in spite of this recognition that one purchases things with money’s worth (and not with money, unless one’s money holdings are being depleted), certain mercantilist, Keynes-like fallacies, largely contradictory with this insight, seem to have prevented any clear perception of the full implications.*
• 21Turgot, however, discerned the confusion due to supposing that “saving and hoarding are synonymous,” and insisted that although “savings” are made “in the form of money,” the savers have no use for money “than to convert it immediately into different kinds of effects. . . . ” Turgot’s statement is, of course, an exaggeration because it tacitly assumes that the prospective productivity of money is not changing in relation to that of non-money. But his tacit assumption is reasonable enough for most circumstances and his criticism of his contemporaries on the point had, I suggest, validity. It was because of the Physiocrats’ misconceptions on the very issue Turgot had in mind that, in developing his “theory of markets,” Say strongly criticized the school of economists from whom his inspiration may have originated.*
• 22While Keynes was writing his famous work I was engaged on a book that appeared a month or two later than his. In my contribution I maintained that “every independent and serious economist . . . must, if his beliefs lie in the path of ‘orthodox’ or Classical tradition, be aware of a periodic recurrence of a sense of utter helplessness. . . . He recognizes that in spheres in which policy and action can be influenced, he is doomed to virtual dumbness today.”* Thus, where Johnson blames the prevailing orthodoxy of the 1930s, I blamed (in 1936) the ignoring of its basic teachings for the emergence and perpetuation of depression.
• 23I can think of no economist who has shown greater intellectual courage than has Harry Johnson in abandoning previously held positions. Where he now stands in respect of the “law of markets,” I do not know. But I believe that “the classical economists” who “from Smith to Pigou” (in Keynes’ view and Mark Blaug’s words) “. . . fell victims to Say’s law” were better economists because they were so victimized. I believe also that with all their shortcomings the non-Keynesians of the inter-war period had a deeper insight into the origins of chronic unemployment and depression than the overwhelming majority of today’s professional economists. I shall try to substantiate this thesis after a restatement of Say’s “law of markets” and an examination of typical criticisms of that law. At the end of my essay, in Chapters XIV to XVII, I shall briefly discuss the British situation in 1936 as well as “classical” teachings and government policies in Britain between the first world war and 1936. The environment of that period was one which somehow inspired the most influential book in the history of economic thought, the book which engendered the so-called “new economics” or the “Keynesian revolution” and, at a crucial period, destroyed the authority of the economists whom Johnson still disparages but towards whose position he seems (to me at any rate) to have been veering. I shall finally submit my own brief diagnosis of the grave sickness which was plaguing the British economy during the 1930s; and I shall suggest, (a) that exactly the same fundamental problem is baffling British politicians and opinion makers today that was baffling them in the 1920s and 1930s; and (b) that the solution is to be found through a perception of the light which Say’s law throws upon the nature of co-operation in a free economy.
• 24In defending the pre-Keynesian “classical economists,” and “Say’s law,” I shall insist that this law dominated the prevailing orthodoxy of the 1930s so naturally that it was seldom if ever stated explicitly and (prior to the “Keynesian revolution”) virtually never challenged. But it was ideas we associate with J. B. Say which constitute the vital insight of the “orthodoxy” that Keynes attacked. In one passage Keynes explicitly recognized that the teachings he was trying to refute involved reliance on Say’s law, although in that passage he referred not to Say himself but to J. S. Mill, from whom he quoted an excerpt out of a not very satisfactory exposition of the principle. (See p. 24).
• 25*I believe that Keynes (General Theory, p. 18) was the first to use this phrase. It has been repeated in these very words by virtually all other economists who have since referred to Say’s law. Even erudite and careful historians of economic thought repeat Keynes’ words, e.g., Schumpeter, Mark Blaug and Tom Sowell, the latter opening his essay with Keynes’ summary of the law and subsequently using the phrase several times. But I cannot trace its ever having been used before Keynes. Sowell refers to it as a “familiar assertion,” which he admits is not a “direct quotation” from Say or James Mill. Leijonhufvud uses it in quotation marks without attributing it to Keynes. (On Keynesian Economics and the Economics of Keynes, p. 280).
• 26Controversy about the validity of this law (which occurred mainly in two periods—from 1803-1850 and from 1936 until today) was described by the great Joseph Schumpeter as “a discussion which reflects little credit on all parties concerned.” It has “dragged on to this day,” he said, “with people armed with superior technique still chewing on the same old cud. . . .” Well, more than two decades after the posthumous appearance of Schumpeter’s monumental History of Economic Analysis, from which this scathing comment is quoted, I have been moved to keep the cud-chewing going. This is in spite of my agreement with Schumpeter (a) that Say did not fully understand the “law” to which his name has become attached; and (b) that in his defence of the “law” against acute but economically illiterate critics, Say was sometimes very “woolly,” slipping, at times, into serious fallacy.*
• 27Controversy about the validity of this law (which occurred mainly in two periods—from 1803-1850 and from 1936 until today) was described by the great Joseph Schumpeter as “a discussion which reflects little credit on all parties concerned.” It has “dragged on to this day,” he said, “with people armed with superior technique still chewing on the same old cud. . . .” Well, more than two decades after the posthumous appearance of Schumpeter’s monumental History of Economic Analysis, from which this scathing comment is quoted, I have been moved to keep the cud-chewing going. This is in spite of my agreement with Schumpeter (a) that Say did not fully understand the “law” to which his name has become attached; and (b) that in his defence of the “law” against acute but economically illiterate critics, Say was sometimes very “woolly,” slipping, at times, into serious fallacy.*
• 28*James Mill referred to Say as “a late French writer” and quoted an anecdote of Say’s to illustrate the absurdity of regarding consumption as the source of demand.