economics of the epoch. Its authors stood to the professed economists on a footing of cool and reciprocated indifference. Yet one should think that each group might have derived help from the other. 13 But Tooke and Overstone did influence opinion within the fold—and were in turn influenced by it—and their work succeeded in setting on foot what may be described as a new analysis of ‘the’ business cycle. 14 Also, they influenced one another more than they realized or, at any rate, were prepared to admit, and the affinity between their methods and results is more important than are the differences. The prevailing impression to the contrary is due, in the first place, to their antagonism in matters of central-bank policy, especially to their controversy over Peel’s Act. In the second place, they were very different types and would express the same fact or result so differently that it would look like two different facts or results. In the third place, they did differ in several points of theory and factual diagnosis, which both of them stressed unduly but which, so far as business-cycle analysis is concerned, amount to less than they seem to. In the state of research in the 1830’s, the mere fact that they saw and 13 I mention, first, John Wade, who was a complete outsider and whose politeness toward ‘political economy’ barely veils feelings akin to contempt. In his History of the Middle and Working Classes…(1833), he developed a pretty comprehensive theory of ‘the commercial cycle of depression and prosperity,’ to which he attributed an average length of between five and seven years, in terms, chiefly, of prices and employment. Faulty and inconclusive though his reasoning is, it is of some interest as a primitive instance of an endogenous dynamic model that reproduces alternation of depression and prosperity by virtue of a lagged relation between prices and consumption. The second work to be mentioned, Hyde Clarke’s ‘Physical Economy’…(Railway Register, 1847) I know only from Jevons’ report (Investigations in Currency and Finance, pp. 222– 3). He had a ten-year cycle (1796, 1806, 1817, 1827, 1837, and 1847 being the crisis dates, which suggest a little manhandling) and in addition a longer period of about 54 years, a striking anticipation of the major cycles or spans of later days, especially of Kondratieff’s long waves (see below, Part IV, ch. 8). But his attempts at explanation in terms of meteorological facts came to nothing. Next, I want to advert to interesting work published in the Transactions of the Manchester Statistical Society, notably the papers by W.Langton (1857–8) and John Mills (1867–8). Both present some evidence of a decennial cycle which both associate rather vaguely with psychological (‘moral’) factors. The former, in addition, anticipated Jevons’ analysis of the ‘autumnal drain’ and noticed the fact that the third quarter of the year is particularly favorable to the outbreak of crises; the latter specifically labeled his cycles as credit cycles. In the United States, periodicity in the sense of recurrence was recognized quite early. For the rest, the question whether or not bank credit was the cause of cycles was zealously debated (see, e.g., the discussion in C.Raguet’s Treatise on Currency and Banking, 1839; this writer produced also an interesting but inadequately motivated overconsumption theory of cycles). R.Hare (‘Do Banks Increase Loanable Capital?’, Hunt’s Merchants’ Magazine, 1852) was one of the earliest of the few writers who attributed to cycles the function of speeding up economic advance. 14 Some contemporaneous writers, Hyde Clarke and Langton especially, recognized a multiplicity of cycles that run their courses simultaneously. Tooke and Overstone, however, knew just one type of cyclical fluctuations. History of economic analysis 712 understood—intuitively at least—the phenomenon of cyclical variations in business situations constitutes in itself fundamental affinity. But the manner in which they gave expression to their vision illustrates very well the difference in their mental set-up that induced so many historians to overlook all they had in common. With Tooke’s method of arriving at results from discussion of individual situations, the perception of the phenomenon was merged so completely into the ocean of his details that it nowhere stands out clearly, and the very fact that he saw it needs to be established against high authority. 15 Lord Overstone, who theorized—though no doubt also from facts, especially the facts of his experience as a banker—boldly and purposively set forth that the ‘state of trade’ (his quotes) ‘revolves apparently in an established cycle’ that he divides into states of quiescence, improvement, growing confidence, prosperity, excitement, overtrading, convulsion, pressure, stagnation, and distress, ‘ending again in quiescence.’ 16 No importance attaches to these ten phases any more than to Tooke’s two or three. But the sequence makes sense all the same. Neither author made any conscious attempt to associate with his phases general characteristics that would have produced a standard picture of the cycle. But it could be shown that they saw all those that experienced practitioners of business would see and practically all that our wealth of statistics has taught us to see. Prices, interest, credit, gold movements, speculation, and investment, in their relation to business activity and overtrading, naturally were foremost in their minds. There is this difference though: preoccupied as he was with the historical facts of successive situations, Tooke presented a rich assortment of relevant elements that is entirely absent from Overstone’s publications and presumably was not fully present in his thought. In that assortment, two things merit particular attention. First, Tooke emphasized throughout the importance of the ‘corn trade’ and, in connection with this, of harvests. We cannot credit him with a harvest-theory of the cycle—any such one-factor theory was quite alien to his way of thinking. But we should, I think, credit him with having kept this element before the eyes of students 15 That authority is Sir T.E.Gregory. But Tooke’s description (as has been pointed out by Kepper) in chs. 9 and 10 of the History (2nd vol., particularly the last paragraph of sec. 2, ch. 9) of the developments from 1828 to 1837 shows conclusively his awareness of a definite cyclical mechanism which, in this place, he describes in terms of a lag of supply behind consumption during ‘a state of rising markets’ and of the reverse in the subsequent phase of ‘stagnation.’ ‘Phases, within which the changes and alternations between periods of confidence and discredit, of the spirit of enterprise and despondency’ revolve, are noticed on p. 175 of vol. I. There is even a suggestion of a ten-year cycle. 16 This famous passage, that has been often quoted, occurs in his ‘Reflections suggested by a Perusal of Mr. J.Horsley Palmer’s Pamphlet on the Causes and Consequences of the Pressure on the Money Market,’ 1837 (republ. Tracts, p. 31). But it is only from unsystematic comments dispersed over the whole of his collected tracts, letters, evidences (see above, sec. 4b) that a conspectus of his opinions can be gained. There remain many loose ends and also some contradictions that cannot be resolved—it is fair to assume that he never completely thought out his ideas. Money, credit, and cycles 713 and with having given an impulse to this theory which commanded some support even before Jevons wrote. 17 Second, he emphasized that periods of prosperity are associated with investment in fixed capital—this particularly in connection with the boom in railroad construction during the 1840’s—and technological change. Emphasis on these two elements constitutes, of course, an important step in causal analysis. The cycle theory of both Tooke and Overstone is primarily an ‘endogenous’ theory, that is to say, both authors tried to show how each phase of the cyclical process is induced by the conditions prevailing in the preceding one. But neither was content with this. Though Tooke’s method produced a much larger array of explanatory, auxiliary, and random factors, Lord Overstone also recognized the more important categories, especially technological improvement, which he came near to considering as the most important of the causes of upswings. It is, therefore, quite wrong to attribute to him personally that purely monetary theory of the cycle which saw nothing in the latter but the vagaries of an ill-regulated currency and credit system and which no doubt counted adherents among his English companions-in-arms and still more in the United States. 18 Overstone himself explicitly stated that it is not the policy of banks which produces upswings. 19 The sense in which this explanation is, nevertheless, possible for modern exponents of monetary theories of the cycle—Hawtrey and von Mises in particular—may be defined in terms of two propositions. First, no matter what he thought about the problem of ultimate causation, Lord Overstone certainly believed that expansion of bank loans, by means of bank notes and ‘created’ deposits, beyond the boundary of ‘real’ capital 20 17 Von Bergmann (op. cit. p. 239) mentioned a French author, Briaune (Des Crises commerciales…, 1840; Du Prix des grains, du libre échange et des réserves, 1857), who presented a clearcut harvest theory in the sense that the cycle is fundamentally nothing but the effect of spells of good and bad harvests upon society’s total income. 18 Even G.W.Norman (Remarks upon some Prevalent Errors with respect to Currency and Banking), who, of writers of some reputation, came nearest to holding a purely monetary theory of cycles in the sense above, qualifies it drastically by admitting many other causal factors. For American examples, see H.E.Miller, op. cit. p. 193 et seq. 19 See, especially, his A Letter to J.B.Smith (1840). No doubt he sometimes expressed himself rather carelessly on the subject, but I do not think it correct to say that the statements alluded to are contradicted by others, or that they are no more than the sort of illogical concessions which people engaged in political controversy are frequently compelled to make. 20 This is the phrase he used in his Evidence before the House of Commons Committee on Bank Acts, 1857, and the appendix thereto (republ. 1858). I think it safe to identify this real capital with the stock of purchasing power that the banks absorb from the savings of the public or gain through imports of gold, so that his argument aimed at Ricardo’s ‘fictitious’ capital. Possibly, this distinction linked up, in his banker’s mind, with the different but related distinction between funds that are available for long-period investment and funds which, though available for short periods only, are nevertheless used for the financing of long-period investment. History of economic analysis 714 is responsible for a course of events that differs qualitatively from what would happen if lending always remained within these boundaries. Miscarriages, so he argued, would also happen in the latter case; but, however frequent, they would be each of them individual occurrences, not necessarily connected with one another, hence capable of being currently absorbed. But if credit has been substantially expanded beyond that boundary, then the whole structure of the economic process is distorted. Investment by firms is generally increased to an extent that the underlying conditions of the economy do not warrant and which, therefore, justifies itself only so long as this inflation goes on. This is more than is implied in the statement, which Tooke never denied, that excessive easy money will facilitate ‘overtrading’ and accentuate its consequences. Second, Lord Overstone explained the turn of affairs from ‘overtrading’ to ‘convulsion, pressure, stagnation’ by a purely or predominantly monetary mechanism: recession was the reaction to the preceding boom, but it was primarily a reaction to the credit expansion of the boom. This credit expansion raised prices, thereby causing drains of cash (into circulation as well as to foreign countries) and threatening convertibility of bank notes. This was bound to raise interest rates and this again to shake confidence and to contract bank deposits and the amount of commercial bills outstanding (Tracts, 1857, p. 264 et seq.). Nothing of all this was worked out with the care and thoroughness to which later analysis, aided by hostile criticism, has accustomed us. But the general import is clear enough: it is money and credit which, themselves unstable, unstabilize economic progress, and it was bank reform which was needed in order to stabilize it, not indeed completely—Overstone repeatedly disclaimed this—but so far as it is capable of being stabilized at all. Tooke criticized all this adversely: he did not believe in the existence or, at all events, in the importance of ‘fictitious’ capital; he minimized the role of interest in the cycle; he did not think that the contraction of credit was the most important factor in causing the downturn. This was indeed enough for him to arrive at different conclusions as to policy. But when we take account, on the one hand, of the qualifications that Overstone applied to his argument, and, on the other hand, of all the qualifications that Tooke applied to his denials, we find the range of their disagreements considerably narrowed. Thus, a rich crop of ideas and of analytic performances may be garnered and put to the credit of that period’s account. We have noticed the overproduction theories and at the same time the elimination of their most naïve types; we have noticed several underconsumption theories and, in their case also, critical work that exposed their errors; we have noticed the random-disturbance theories in the most varied editions; we have noticed the discovery of the business cycle and the emergence of both monetary and investment theories of it; even an overconsumption and the harvest theory were not absent; above all, we have noticed the beginnings of statistical work on the problem. But, strange to say, nobody seems to have known all these bricks or to have understood that they were bricks awaiting the hand that would combine them in a comprehensive structure—comprehensive though provisional— before the period was out. J.S.Mill failed at the task though he offered more of a synthesis than appears at first sight. 21 He described the cyclical mechanism in terms of expectations of profit—induced by favorable or unfavorable occurrences—that act upon dealers’ stocks, hence upon prices which eventually go on rising for no better reason than that they have risen and, when it is realized that the rise has gone beyond the extent warranted by the initiating occurrence, Money, credit, and cycles 715 begin to fall until they go on falling because they have fallen. Carefully pointing out that this could happen even ‘in a community to which credit was unknown,’ he then emphasized the fact that readily extensible credit will greatly increase the violence of such fluctuations. But commercial crises—defined as situations in which ‘a great number of merchants and traders at once either have, or apprehend that they shall have, a difficulty in meeting their engagements’—may also arise without ‘particular extension of credit’ when a large proportion of the capital which usually supplies the loan market is absorbed by unusual demands for foreign payments, fixed investments, and the like. He argued against naïve overproduction and underconsumption theories, especially the oversaving theory, yet found, within the events of depression, place for both excess supply and underspending. Interest also gets its modest place, and so does the purely monetary mechanism of internal and external drains. Nor is periodicity (in the wide sense of the word) absent. I think this is enough to impart to readers the sense of flatness I experienced myself in trying to reconstruct Mill’s analysis of the cycle. But though commonplace, all this is also common sense and not a bad foundation for further work. In perusing what A.Marshall said on the subject, 22 we find indeed much more material and do not have the same sense of flatness; but substantially his treatment does not amount to more than an elaboration of J.S. Mill’s suggestions. Many other students were influenced or even initiated by Mill. Even Marx may have learned something from him. Marx’s analysis of business cycles is an ‘unwritten chapter’ and no coherent picture of it has emerged, or is likely to emerge, that would command the approval of all, or even of all orthodox, Marxologists. 23 Several methodolog- 21 This is in part because Mill nowhere concentrated all he had to say about crises or cycles. He dealt perfunctorily with the subject in Book III, ch. 12, 3 of the Principles. But pertinent material is found in many other places, especially in chs. 14 and 23 of Book III and in ch. 4 of Book IV. 22 The sedes materiae is Book IV on ‘Fluctuations of Industry, Trade, and Credit’ in Money, Credit, and Commerce, published in 1923 but consisting chiefly of results of much earlier work (some of which dated from the 1880’s). In addition, there are several relevant passages in Marshall’s Principles. The phrase above that ‘prices rise or fall because they have risen or fallen’ is Marshall’s. 23 On the Marxist literature on the subject, see P.M.Sweezy Theory of Capitalist Development, Part III. More nearly correct than Dr. Sweezy’s own interpretation seems to me to be the one by H.Smith, ‘Marx and the Trade Cycle,’ Review of Economic Studies, June 1937, to which readers are particularly referred—a reference that is to excuse in part the brevity of the comments which follow. The only other excuse I have is the impossibility of presenting, from the vast mass of relevant material, a tolerable account within the available space. Most of this material is to be found (the History of economic analysis 716 ical features call first for our attention. As always, Marx was conscious of reasoning— sometimes on the same page—on widely different levels of abstraction. In the matter of cycles, this is particularly important to note because, each cycle being a historical individual and in part conditioned by circumstances for which there is no exact analogue in other cycles, we have always to deal with—and even to construct ad hoc theories for— facts the relevance of which varies according to the level of abstraction on which we wish to move: a cycle theory may still aim at being general or fairly general, and yet contain elements that are nonessential from the standpoint of a pure model. This greatly increases the difficulties of interpretation. Moreover, Marx attended carefully to the vital distinction between general institutional conditions that permit cyclical movements and ‘causes’ or factors that actually produce them. For instance, the famous ‘anarchy’ of capitalist society, the intervention of money between ‘real’ transactions, and the vagaries of bank credit were for him facts to be taken account of, but as permissive—though necessary—conditions only, and not as ‘causes’: he perfectly realized the emptiness of any ‘theory’ that contents itself with pointing to these and similar facts. Finally, he distinguished, from both conditions and causes, another set of facts, the symptoms. 24 It stands to reason that neglect of these distinctions must be a fertile source of errors in analysis and of futile controversy and that this methodological contribution is in itself sufficient to give to Marx high rank among the workers in this field. Next, we must try to appraise the apparent relation, which a passage in the Communist Manifesto seems to suggest, between cycles and the ultimate breakdown of capitalist society. Marx used the notion of a (perhaps decennial) cycle as a matter of course. Crises were never more for him than a phase in the cyclical process. Yet if he did believe, as he seems to have done, that crises tend to become more destructive as the capitalist epoch wears on, it is natural to assume that he associated this supposed fact 25 with the ultimate famous but inadequate passages in the Communist Manifesto and in volume 1 of Das Kapital are really of minor importance) in volumes II and III of Das Kapital and especially in the Theorien über den Mehrwert. Several letters are also essential, e.g. Marx’s correspondence with Engels on the renewal period of durable capital in the English textile industry. 24 On p. 695 of Das Kapital (vol. I, English trans. publ. by Kerr, 1906) occurs the sentence: ‘The superficiality of Political Economy shows itself in the fact that it looks upon the expansion and contraction of credit, which is a mere symptom of the periodic changes of the industrial cycle, as their cause.’ Of course, Political Economy as a whole does not do this. Yet there is a lot of truth in what Marx meant to express. 25 There is a lot to be said both about this ‘fact’ itself and about Marx’s belief in it. That he held such belief can be ‘proved’ in the same way that we can ‘prove’ that Marx believed in a violent and spectacular breakdown of the capitalist order of things. It may be that he held, and toward the end of his life abandoned, both these beliefs. For us it is more important to note (a) that the thesis that crises increase in intensity is not logically inherent in his general theory, and (b) that some later Marxists, and especially Hilferding, repudiated it until the events of 1929–32 provided the semblance of a verification. Money, credit, and cycles 717 breakdown or even that he expected that capitalism would break down in a final crisis that would be so disastrous as to set fire to the framework of capitalist society. It is more fair to Marx’s fundamental conception, however, to neglect such evidence as there is for his having taken this view and to emphasize that, in his analysis, the cyclical process per se and the trend that points toward breakdown—especially if the breakdown amounts to not more than stagnation—are as a matter of fact two distinct phenomena, each of which might exist without the other. There was nothing in this to have prevented him from looking upon recurrent crises as ‘contributory causes’ to an ultimately untenable social situation. Finally, we must try to collect Marx’s contributions to a fundamental or ‘causal’ explanation of the cycle, trying to find out, as so many others have tried before us, whether any definite theory of it can be attributed to Marx even though he never penned one explicitly. The first step is easy. Marx clearly visualized that the ‘decennial cycle’ that (‘interrupted by smaller oscillations’) runs along in a sequence of phases (or ‘periods,’ as he said) of average activity, prosperity, overproduction, crisis, and stagnation 26 is ‘characteristic of modern industry’ and not merely the result of a series of incidents or accidents. And he definitely located its source in the process of accumulation. But beyond this, one thing only is certain, namely, that he treated this process, including the increase in productive capacity it brings about and the ‘industrial reserve army it creates,’ as a movement away from equilibrium, and crises as the catastrophes which periodically re-establish equilibrium and, by means of radical destruction of capital values, recreate the conditions for profitability of business. This is a promising approach that avoids many possible errors and irrelevancies and purposively leads up to the question that remains: why should the process of accumulation be essentially 27 disequilibrating? Since Marx considered the cycle as an essential form of capitalist life, we cannot accept a random-disturbance theory as an answer. Since he passed a contemptuous judgment on credit-theories of the cycle, we can exclude these, however much he made of speculation and other excesses that are facilitated by an expansible credit system. He certainly was not an adherent of any naïve overproduction theory of crises in the sense of Fourier’s crises pléthoriques. 28 Nor should he be saddled, as he frequently has been, both by 26 Das Kapital, vol. I, ch. 25, sec. 3 (p. 694 of English trans., 1906). 27 There is no difficulty in understanding why, in real life, it is exposed to disequilibrating factors, such as speculative manias, errors, miscarriages of all sorts. But these factors do not solve the ultimate theoretical problem why cyclical fluctuations should be inherent in the logic of capitalism—as Marx well knew. 28 The reader must not allow himself to be misled by the frequency with which the phrase overproduction occurs in Marx’s writings—as we have seen, it occurs even in his sequence of phases. It has with him no meaning other than a descriptive one. In his phenomenology of cycles all-round unsalability of goods, of course, does play a role But he was above allocating any causal importance to it. Some followers (e.g. K.Kautsky in Das Erfurter Programm, 1891) were not. History of economic analysis 718 friends and foes, with that underconsumption theory which associates crises with inadequacy of labor’s purchasing power and which, to the layman, seems so closely connected with exploitation. 29 But this theory belongs to Rodbertus and not to Marx, who, like the good economist he was, was quite aware of its weakness and repudiated it in so many words. 30 Thus, finally, we seem to be left with the falling rate of profit—the consequence, with Marx, not of accumulation per se but of the relative increase in constant as against variable capital—and several possibilities do in fact come into view of harnessing this ‘law’ to serve the purpose in hand. To begin with, this ‘law’ can live on the highest level of abstraction. Further, there is no question but that prosperity periods are periods of supernormal investment and that the resulting increase in productive capacity has an effect upon prices and profits that need not be of causative, but must always be of considerable, importance. 31 Finally, Marxist accumulation leads to unemployment and tends to undermine the industrial structure that exists at any time (destruction of smaller and less efficient firms and so on). Marx seems to have realized, however, that none of these elements will readily explain the cyclical form of the process of accumulation and still less the occurrence of crises. In any case, perhaps wisely, he did not commit himself to an explanatory hypothesis clearly based upon any or all of them. 32 29 See above, this sec., note 4. Since nobody has ever attributed to Marx an underconsumption theory of what in the passage referred to we have called the nonspending (Keynesian) type, it should be superfluous to insist on the fact that Marxist capitalists are always in a hurry to invest and that hence this element of the case has no place in his system. Since these capitalists invest because they have got to—owing to the pressure of competition—the same reasoning applies to Malthusian underconsumption. 30 See on this H.Smith, op. cit. pp. 193–5. Underconsumption of workers does come in but only in an indirect and secondary manner, not as the fundamental cause: if wages were higher, i.e. the degree of exploitation smaller, the rate of accumulation would also be smaller; since accumulation is responsible for the cycles, we might therefore expect that these would be less pronounced in that case. 31 This fact can, of course, be expressed by the phrase Overproduction of Capital. But this does not make Marx a sponsor of either an overproduction or a disproportionality theory. 32 We have had to omit many features of Marx’s speculations on cycles, e.g. his brief and superficial references to the existence of a self-generating mechanism that works by virtue of its momentum; we have, however, mentioned his interest in the replacement cycle of durable capital. This search for additional facts seems to support the guess that the ultimate problem remained unsolved in his mind. Money, credit, and cycles 719 PART IV FROM 1870 TO 1914 (AND LATER) CHAPTER 1 Introduction and Plan 1. COVERAGE THIS PART IS to cover the history of analytic work from about 1870 to 1914. For justification of the first date I invoke a fact that few economists will deny, namely, that it was around 1870 that a new interest in social reform, a new spirit of ‘historicism,’ and a new activity in the field of economic ‘theory’ began to assert themselves; or, that there occurred breaks with tradition as distinct as we can ever expect to observe in what must always be fundamentally a continuous process. The justification for the second date is the thesis that the First World War was an ‘external factor’ powerful enough for its outbreak to be made a terminal point, though the influences that were to put an end to that epoch of economic analysis and to usher in another were all clearly visible before and though they did not conquer until another decade or so had elapsed. All this must be taken with the same qualifications that apply to any attempt to periodize anything, and in particular with qualifications similar to those with which we found it necessary to safeguard our conception of the preceding period. A number of men and of works ride astride both periods and cannot be assigned to either without much arbitrariness; and there were many overlaps in views, attitudes, and methods. Partly because of this, some men and works that belong chronologically to either the preceding or the following periods have been allocated to this Part. There is, however, also another reason for referring, sometimes rather fully, to developments of our own period and for carrying our story, in some matters, down to date (1949): modern developments will be but cursorily treated in Part V, and it seems desirable to use opportunities as they arise for indicating, at least in a number of important points, how modern work harks back to the work of 1870–1914—how far it is on the foundation laid by the latter that we ourselves are building. But all the qualifications that are necessary in order to prevent periodization from becoming misleading—or downright nonsense—should not blind us to the fact that the period we are about to discuss actually forms a real unit, which would have to be recognized quite irrespective of the claims of expository convenience. The breaks with tradition around 1870 were meant to be breaks by the men whose names are associated with them: they may have looked to those men more abrupt and more important than they do to the historian, but this does not mean that they were wholly imaginary. Upon these ‘revolutions’ followed two decades of struggle and more or less heated discussions. And from these again emerged, in the nineties, a typical classical situation in our sense, the leading works of which exhibited a large expanse of common ground and suggest a feeling of repose, both of which created, in the superficial observer, an impression of finality—the finality of a Greek temple that spreads its perfect lines against a cloudless . Overstone, however, knew just one type of cyclical fluctuations. History of economic analysis 712 understood—intuitively at least—the phenomenon of cyclical variations in business situations. Kepper) in chs. 9 and 10 of the History (2nd vol., particularly the last paragraph of sec. 2, ch. 9) of the developments from 1828 to 1837 shows conclusively his awareness of a definite cyclical. only, are nevertheless used for the financing of long-period investment. History of economic analysis 714 is responsible for a course of events that differs qualitatively from what would