Commodity-Trade Theory of international finance (or of international payments or of international gold movements). The commodity-trade theory of international finance is thus open to the criticism—as is the theory of international values—that its conception of the phenomena with which it undertakes to deal is much too narrow. 2 Also, it must be pointed out that its particular assumptions disqualify it for direct practical application. But there is another criticism that strikes at it within its own precincts and should be mentioned at once because it has received undue prominence during the last twenty years or so. A theory of international finance that pivots on commodity trade will naturally emphasize the equilibrating role of variations in relative prices. It has been pointed out, first by Wicksell, that adjustments to disturbances of trade relations may and often do take place without actual changes in prices and also without actual gold movements. This is true, of course, and no classic writer, least of all Ricardo, would have denied it. But if the ‘classic’ theory be nevertheless criticized on the ground that it put an altogether unjustifiable burden upon the price mechanism and in doing so failed to notice other equilibrating factors, then the critic is wrong because price variations of the kind the ‘classic’ theory visualizes imply shifts of demand curves which in turn imply variations in income, as we shall see presently. Moreover, in the pattern which the ‘classics’ chose for analysis, price variations in fact do hold the key position. All the critic can rightfully say is that this is no longer so in patterns in which prices are rigid and capital movements dominant. Finally, several writers of the period under survey explicitly introduced the factors that critics miss in the classical picture. 3 In the third place, the ‘classic’ theory of international finance was not fundamentally new. Thornton, who blocked it out, referred—approvingly and critically—to Locke, Hume, and A.Smith, and Hume’s analysis undoubtedly was the starting point of the period’s work. But Hume himself did no more than formulate effectively the result of a long development in the course of which ‘mercantilist’ work had been slowly moving toward the doctrine of the ‘classics.’ Thornton’s teaching prevailed, more or less, with most of the leading writers of that and the subsequent period, from Malthus 4 through Tooke to J.S.Mill and Cairnes and eventually to Taussig. But Wheatley dissented from it and was followed by Ricardo. 5 We proceed to consider the point at issue. 2 On this, see J.H.Williams, ‘The Theory of International Trade Reconsidered,’ Economic Journal, 1929, reprinted in Postwar Monetary Plans and Other Essays (1944), Part IV. 3 This has been shown by Viner (op. cit. pp. 293 et seq.), who in particular mentions Longfield, Torrens, and Joplin. 4 Edinburgh Review, 1811, the review articles referred to already on another occasion (above, sec. 3). 5 To some extent, as McCulloch has pointed out, they were anticipated by Barbon. See above, Part II, chs. 6 and 7. History of economic analysis 702 For this purpose, we start from a state of monetary equilibrium between two countries. 6 Being in possession of the theory of international values and of the equilibrium condition given by the equation of reciprocal demand (which, as we know, comprises the principle of comparative cost), we readily see that the condition of monetary equilibrium which we must now add is simply that, claims from commodity transactions canceling one another, there should be, under our assumptions, no movement of gold from one country to the other. We proceed to investigate the properties of this equilibrium by assuming it to be disturbed and by analyzing the adjustments that will follow. First, we assume a disturbance occurring in the sphere of money: we assume, like Hume, that in one of the two countries everybody’s holding of monetary gold is suddenly doubled. Without committing ourselves to any strict quantity theory, we may aver that in this country incomes and business funds in terms of gold, hence expenditure, will increase; that demand schedules for all commodities will shift upwards; that gold prices will rise in consequence; that exports will decrease; and that gold will flow out until equilibrium is reestablished. Nobody ever challenged this, though the highly artificial process assumed might have given plenty of scope for quibbling. 7 Second, instead of assuming that gold has increased in one of the two countries, assume that commodities have decreased, for example, because of a bad harvest. The reader will be tempted to argue that, the ‘need’ for food imports having increased, an unfavorable balance of trade will ensue and produce an export of gold through which adjustment will, for the time being, 8 be achieved. But on reflection, he will realize that this is not strictly correct and does not constitute a generalization of Hume’s argument but a deviation from it. For the bad harvest per se does not produce an unfavorable balance. Needs are not incompressible. So far as the need for food imports is imperative, other imports can be curtailed: in other words, all that has happened is that the people in the country that experienced the bad harvest are, for the time being, poorer than they were before and have to readjust their consumption and investment to a lower level of real income; but on this lower level the balance of trade can be, and in the absence of money and credit would have to be, just as much in equilibrium as before. Actually, however, we do get 6 We make the following assumption partly in order to reproduce the ‘classic’ pattern and partly in order to simplify exposition: no international economic relations except commodity trade with perfectly flexible (and competitive) prices, hence incomes; no credit whatsoever; perfectly free international gold standard; two countries only, differing not too much in size, for neither of which foreign trade is of negligible importance; no gold mining; gold, though considered as a commodity, all absorbed in the monetary function; no cost, risk, or loss of time involved in its transportation or in the transportation of commodities. Evidently a fairly complete theory can be derived by dropping these assumptions one by one. 7 I have reduced the scope for quibbling by reasoning in terms of income and expenditure rather than of mere quantity of money. But there still remains some. 8 The next step would be an increase of incomes and prices in the other country, which will counteract the process even before normal harvests, re-establishing previous conditions, reverse it entirely. Money, credit, and cycles 703 the unfavorable balance also if we reason correctly per analogiam of Hume’s argument, only we get it as the effect rather than as the cause of the export of gold. Since gold has not been decreased by the failure of the harvest and since we may assume that money incomes and money expenditures have not decreased either, but since there is now less of commodities to buy, prices will rise or gold in terms of commodities will get cheaper or, as we may also say, from the standpoint of the previous price level, gold has become redundant. This curtails exports and fosters imports of commodities other than gold exactly as if gold and incomes and expenditures had increased, output having remained unchanged. 9 We have thus reduced the disturbance that arose in the commodity sphere to a disturbance in the monetary sphere. Thornton, who began his investigation into the nature of monetary equilibrium in international trade by presenting the example of the crop failure (Paper Credit, p. 143), seemed to argue in the way that has just been shown to be open to objection. It is true that in other places (e.g., ibid. pp. 244, 247), his argument indicates that he understood the point I have been trying to make. But he was so hazy and so hesitant about it that Wheatley and later on Ricardo were right in asserting that the factors that operate on the value or purchasing power of gold are one thing and that the operation of the value or purchasing power of gold is another thing. But they so mismanaged their case as to leave contemporary as well as later writers to wonder whether they had any case at all. 10 9 It should be superfluous to point out that the sequence of events described is only to express the logic of the process and need not always be actually observable. But this does not constitute an objection any more than it constitutes an objection to the usual theory of the effects of a specific tax imposed upon a commodity—that its price need not, as the explanatory schema seems to postulate, first rise by the whole amount of the tax, then fall again owing to the consequent reduction of quantity demanded, and so on until it settles at the new equilibrium level, and that in practice some steps may be omitted. Similarly the gold may start to flow at once in payment of the additional grain import, and the influence of the crop failure on the price level in the affected country may never show fully: this need not effect the role of price variations in the explanatory schema. 10 On this controversy, see Viner, Canada’s Balance of International Indebtedness, 1900–1913 (1924), ch. 9. Wheatley and Ricardo introduced not only irrelevant or non-essential but also erroneous arguments. For instance, both Wheatley and Ricardo denied that a crop failure will create ‘redundancy’ of currency, though Ricardo admitted this in a letter to Malthus (Letters, p. 13). But Wheatley, perhaps because he had a clearer conception of the price level, came much closer than did Ricardo to grasping the principle involved. Thus, he said boldly that, in spite of all the subsidies and other sums sent abroad during the Napoleonic Wars, it would have been possible to enforce ‘influx of money to any extent’ (Essay on the Theory of Money…, 1st vol., p. 194). Barring the obvious exaggeration, this clearly implies, though it does not state, the principle that it is a monetary mechanism which is immediately behind exchanges and gold flows and that these are never determined uniquely by the factors—such as political payments or conditions that determine the demands for individual commodities—that operate at one remove. History of economic analysis 704 But had they? Is it really more than hairsplitting when we insist that gold flows out because it is the ‘cheapest exportable commodity’ and not because of a bad harvest when the bad harvest makes gold the ‘cheapest exportable commodity’? Instead of any other answer, I shall merely point to a fact of considerable importance both for the history of economic analysis and the history of economic thought. It is natural for bankers or writers who place themselves in the position of the individual banker to say that banks cannot expand credit beyond limits that are given to them irrespective of their own behavior. It is not less natural for bankers and students of the problems of the individual bank to start from the obvious fact that outflows or inflows of gold result from unfavorable or favorable exchanges which in turn result from the demand for and supply of claims on foreign places. Excepting the quality of the individual paper, the factors behind demand and supply seem to be all that the banker has to analyze for purposes of diagnosis and forecasting: political factors, business situations, state of the crops, and so on; and this is in fact the standpoint on which Goschen placed himself in writing his famous Theory of Foreign Exchanges (1861). 11 Since demand for and supply of foreign paper reflect a country’s current (and prospective) balance of payments, we may call this the Balance-of-Payments Theory of foreign exchanges. Decades of quiet conditions may pass without anyone’s becoming aware of the fact that there is anything missing in this theory. ‘But if people keep on applying it in conditions of violent disturbance, the presence of another factor becomes obvious that cannot be resolved into those which we may unearth by analyzing the individual items of the balance of payments, namely, the value (purchasing power) of the monetary units in which balances of payments are expressed. We may label as Relative Inflation the variations in the value of a country’s monetary unit, in relation to the value of other countries’ monetary units, and speak accordingly of an Inflation Theory of Foreign Exchanges. We shall return to this subject in Part IV, Chapter 8. Now I wish merely to point out that the first rumble of the prolonged battle between these two theories—though it should be clear that they do not amount to alternative explanations—is audible in the controversy between Thornton and Wheatley-Ricardo: when, in the latter’s phraseology, gold becomes ‘redundant’ in a country or ‘the cheapest exportable commodity,’ then this country experiences ‘relative gold inflation.’ So Wheatley and Ricardo did have a case and one that was more than chopping-logic, though, so far as 11 George J. (afterwards Lord) Goschen (1831–1907), Chancellor of the Exchequer in the second Salisbury administration (1886–92) and of historical importance as the last of the ministers of finance in the pure tradition of classic liberalism (it is highly significant though that this liberalism had taken shelter in a conservative cabinet), was a banker of German extraction. His book describes extremely well what a highly educated and intelligent dealer in foreign exchange would know about foreign exchanges. As a piece of analysis, the performance, which nowhere goes below a well-observed surface, does not rank highly. But it explained things about which politicians and academic economists are likely to know but little and hence was a boon for both. The success of the book was sweeping and it is still worth reading. Money, credit, and cycles 705 their attack upon Thornton is concerned, they may have been unjust because Thornton’s fine mind paid but little tribute to the fallacy in the balance-of-payments argument. When countries are in monetary equilibrium with reference to one another, then, so it has been stated above, gold is distributed between them in such a way that there is no profit in transferring any part of a country’s holding to any other country. We may express this by saying that the purchasing power of gold is internationally at par and also, from the standpoint of the inflation theory of foreign exchange, that this parity and its variations are the (immediately) determining factors in the foreign-exchange market. This Purchasing-Power Parity theory, or some rudimentary form of it, goes far back and can, as we have seen above, certainly be attributed to Malynes. During the First World War, a particular coinage of it became associated with the name of Cassel. But the principle involved must also be attributed to Wheatley and Ricardo 12 in whose work it appears, as it was to appear in Cassel, in characteristic association with a strict (and crude) quantity theory. 13 The ‘classic’ reasoning about gold movements and exchange rates can be generalized without much difficulty to irredeemable paper. 14 Its application to the cases of loans, subsidies, and absenteeism 15 —a standard topic of the economics of the period—presents more difficulties. Of course, the divergence of views just described carries over into the discussion of these cases. But this is not the only problem. All of these cases, but especially international loans, 12 This is not the opinion of Professor Viner (op. cit. pp. 126 and 382 et seq.). But this is only because he reserves the term for the Casselian form of the principle. This, of course, cannot be attributed to Ricardo, who always fought shy of the price level concept that is essential to the Casselian form of the principle though not to the principle itself. The struggles of economists with the emerging price-level concept have been discussed above. 13 On a certain level of monetary theory, quantity theory and purchasing-power parity theory are simply complements or even two different aspects of the same thing. It is, however, possible to show that, on other levels, they may be so formulated as to constitute logically independent, though still related, propositions. 14 Wheatley (Essay on the Theory of Money, 1807) saw more clearly than did others that even in this case it was ‘redundancy’—i.e. pressure upon the price level—and not any occurrence in the world of commodities per se that would cause an unfavorable turn of exchange rates. There is an echo of this in J.S.Mill’s rather inadequate treatment of the subject (Principles, Book III, ch. 22, 3). 15 Irish landlords, living in England on the rents of their Irish estates, naturally were being increasingly discussed, the main purely economic question being whether or not their living and spending in England and not Ireland made any difference to the Irish people. McCulloch returned a negative answer on the ground that it does not matter where a man consumes what he consumes (‘Essay Showing the Erroneousness of the Prevailing Opinions in regard to Absenteeism,’ Edinburgh Review, 1825, reprinted in Treatises and Essays on Money, Exchange, Interest, 1859); Senior a weakly affirmative one (Edinburgh Review, 1825; see also Senior’s Outline, p. 156). M.Longfield’s performance (Three Lectures on Commerce and One on Absenteeism, 1835, London-School Reprint, 1937) is of some analytic interest. History of economic analysis 706 raise questions that cannot be satisfactorily treated by a schema derived from the miraculous increase in the gold stock of a country or from the failure of a harvest: among other things, income effects begin to play a qualitatively different role and interest a decisive one. Results were correspondingly unsatisfactory. Nevertheless modern criticism—qua criticism—often errs by not paying sufficient attention to the particular conditions of the transactions that individual authors envisaged and the sequence of events that were imposed by these conditions. Take J.S.Mill’s famous treatment of unilateral political payments—say, an annual tribute—that served as a starting point and evoked much criticism in the discussion on German reparations after 1920 (Principles, Book III, ch. 21, 4). The treatment is short and oversimplified, but substantially correct so far as the one case considered is concerned, namely the case where the recipient insists on receiving annual sums of money, the first of which the debtor country has no choice but to collect from the pockets of its citizens. Here the gold movement is not a question of automatic mechanisms at all but is simply imposed by the initial conditions of the problem. Under these conditions, a fall in prices in the paying country can hardly fail to come about. This will increase exports and decrease imports and would bring back the gold but, as Mill constructed the case, the paying country’s claims to this gold will be absorbed by the receiving country’s claim to the next instalment of the tribute, so that gold stock, incomes, expenditures, and prices in the paying country are kept down and its excess exports are kept up. Another case might no doubt be constructed that would produce a different sequence of events and contain no gold movements and no price changes at all but only income changes and commodity movements. But either case serves to illustrate what from the ‘classic’ standpoint is the essential thing, namely, that the true equilibrating factor is the commodity transfer. And neither is very realistic. 6. ‘THE’ BUSINESS CYCLE One of the most important achievements of the period under survey, and one of the few that were truly original, was the discovery and preliminary analysis of business cycles. It is true that the crises of 1815, 1825, 1836–9, 1847–8, 1857, and 1866 pressed the phenomenon upon the attention of even the most academic of economists. But similar breakdowns had occurred, with similar regularity in the eighteenth century, and nevertheless nobody had gone deeply into the matter: nobody had distinguished them clearly from the effects of war and other external disturbances or seen in them anything but chance misfortunes or the results of manias or errors or misconduct. The first suggestion that there might be deeper causes to these breakdowns, causes which are inherent in the economic process, are indeed to be found in the ‘mercantilist’ literature, mainly in connection with the ideas that were later on worked up into the various underconsumption theories. But these ideas were not made explicit before the controversy on gluts, during and after the Napoleonic Wars, which we know already and which ended, for the time be-ing, in their defeat. After some additional comments on this Money, credit, and cycles 707 controversy, we shall consider the analyses of business cycles that are primarily due to Tooke and Lord Overstone and then end up with the contribution of Marx. For an extensive treatment of the subject, the reader is referred to the work by von Bergmann. 1 The facts about crises that the press and the public primarily notice and to which they naturally attribute such effects as bankruptcies and unemployment are the collapse of credit and the unsalability of commodities: press and public are inveterate adherents of monetary and overproduction ‘theories.’ 2 It was against the popular ideas of the latter kind that J.B.Say argued in his chapter on the ‘Law of Markets.’ As has been stated already, so far as the subject of crises is concerned, the main merit of that law was a negative one. Say showed successfully that, however large the phenomenon of overproduction may loom in the historical picture of individual crises, no causal explanation can be derived from it: there is no sense in saying that there is a crisis because ‘too much’ has been produced all round. Though negative, this contribution was very important. It may be said to stand at the fountainhead of the scientific analysis of cycles and to mark the point at which the latter broke away from pre-analytic thought. But the positive application that Say attempted to make of his law was much less valuable. He inferred from it erroneously, though with apparent logic, that, if general overproduction was no explanation, then partial overproduction must be at the root of the trouble—that some commodities are unsalable because their complements are lacking, or that the apparent overproduction of some was really underproduction of others. This is the Disproportionality Theory of Crises, 3 as it was called 1 Eugen von Bergmann, Die Wirtschaftskrisen: Geschichte der nationalökonomischen Krisentheorieen (1895). Most authors of systematic works on business cycles present some information on the history of business-cycle analysis, and there are also a few other histories. Analytically, Professor Friedrich Lutz’s Das Konjunkturproblem in der Nationalökonomie (1932) moves on a much higher level than does von Bergmann’s work. Nevertheless the latter is the only one I know who presents the results of extensive research in the literature of the period—a fact which is surprising in the case of a subject that looms so large in modern work. Scholarly effort has been directed toward the work of individual authors or groups of authors or toward individual issues or theories rather than toward a comprehensive survey. On Tooke and Overstone, e.g., there is the excellent book by Georg Kepper, Die Konjunkturlehren der Banking—und der Currencyschule (1933). On American work, see H.E.Miller, Banking Theories in the United States before 1860 (1927, ch. 16). 2 Today, the ‘theory’ of business cycles means very much more than explanatory hypotheses: it means a whole apparatus of the theoretical and statistical tools of analysis. For the nineteenth century, however, it is approximately correct to say that hypotheses as regards the ‘causes’ of crises or cycles were the main, if not the only, contents of what was meant by theories of crises. 3 The disproportionality that Say envisaged was primarily a disequilibrium within the same stage of the productive process: overproduction of shoes with reference to the production of coats. We had better confine the term to this meaning—there is, of course, no objection to forming larger categories than shoes and coats—and distinguish History of economic analysis 708 later on, which died from lack of vitality in the course of the nineteenth century, though individual adherents might be named throughout. One of them was Ricardo. Slightly improving the idea in Chapter 19 of his Principles, he made a reasonable though, of course, inadequate case for Sudden Changes in the Channels of Trade as the most important single cause of disturbance. As we know, Sismondi and Malthus (followed by Chalmers) were the leaders in the campaign against Say’s law—some of their arguments having been anticipated by earlier writers, especially Lauderdale. It is very difficult to label their theories, which neither of them systematized completely and which with both of them, but especially with Malthus, were theories of stagnation and prolonged unemployment rather than theories of ‘crises.’ Malthus, however, came much nearer to definiteness and may I think be credited—or debited—with an underconsumption theory of the oversaving type: 4 stagnation ensues when people save and invest to such an extent as ‘to leave no motive to a further increase of production’ owing to the incident fall in prices and profits. 5 It cannot be emphasized too often that this argument, whatever its incidental merits—and one of them is that it locates the source of stagnation in the saving-investment process—is, as it stands, definitely erroneous if intended to explain ‘crises,’ though not if merely intended to show the possibility of production’s becoming stationary. But Sismondi offers such a multi- it from theories which attribute cycles or crises to disproportionality between stages taken as a whole, e.g. between investment goods production and consumers’ goods production. For disproportionalities of the latter kind are always linked to other factors, e.g., monetary ones or oversaving, and therefore are symptoms or consequences rather than ‘causes.’ 4 In a sense, of course, underconsumption can always be described as overproduction. Accordingly, von Bergmann labeled Malthus’ theory a ‘motivated overproduction theory.’ It seems more conducive to clear distinctions to avoid the latter phrase whenever an author locates the seat of the trouble with the behavior of consumers, even if the result is also some sort of overproduction—just as, for the same reason, we have adopted a strict definition of the phrase Disproportionality. We shall distinguish three types of underconsumption theories, all of which put in an appearance during that period. There is, first, the oversaving type just mentioned, of which Malthus was the chief exponent. There is, second, the nonspending type that emphasizes disturbances which arise from saving decisions that are not offset by decisions to invest. Malthus, as we have seen, glanced at this idea, which is an old one—to be attributed, e.g., to Quesnay and several of his French predecessors—but which did not play any great role in modern economics until our own time. And there is, third, the mass-poverty type that attributes gluts to the inability of labor, owing to low wages, to ‘buy its own product.’ The most important sponsors of this theory were Sismondi and, much more definitely, Rodbertus. This theory, as Marx well knew, is beneath discussion since it involves neglect of the elementary fact that inadequacy or even increasing inadequacy of the wage income to buy the whole product at cost-covering prices would not prevent hitchless production in response to the demand of non-wage earners either for ‘luxury’ goods or for investment. 5 See Malthus’ letter in J.M.Keynes, Essays in Biography, p. 143; also the contemporary German discussion of the controversy between Malthus and Say by K.H. Rau, Malthus und Say über die Ursachen der jetzigen Handelsstockung (1821). Money, credit, and cycles 709 plicity of responsible factors that he cannot be classified satisfactorily. The oversaving argument is no doubt present and forms the core of his analysis of disequilibria of production and consumption. 6 But underconsumption owing to low wages is still more prominent, both because of the ‘vicious’ distribution of incomes per se and because of the unemployment created by labor-saving machinery. Then there is, incident to his sequence analysis, the idea that increasing outputs meet totals of purchasing power that have been earned some time before by participation in the production of a smaller output. Further, Sismondi made much, and rightly, of all the random vicissitudes through which the road leads to the theorists’ smooth ultimate long-run normals. Thus he became the patron saint of all those ‘explanations’ that are content to talk about the anarchy of capitalist production, the lack of knowledge of what the other fellow does and of what buyers want, and so on, though all the crudities that are to be found in the literature of this kind must not be attributed to him. The phenomena of the post-Napoleonic depressions suggested to him a rich array of sources of trouble of all sorts that was easier to shape into an indictment than into an analytic organon. Thus, he stands also in the current of ideas that produced a ‘theory’ which was to command much support by very able economists from, roughly, 1850 to the end of the nineteenth century, and will have to be mentioned again. In a nutshell, it may be expressed by saying that crises will occur when anything of sufficient importance goes wrong. One of the chief representatives of this view was Roscher. 7 But in addition to this common-sense, if somewhat commonplace, theory Roscher presented what can only be described as a fricassee of most of the ideas that were current at the time he wrote. Emasculating all of them, he accepted Say’s law but reduced it to an identity; 8 he accepted and amplified Ricardo’s sudden changes in the channels of trade; he cautiously accepted Malthus’ oversaving factor though he said that Malthus 6 See, especially, besides Sismondi’s article in Brewster’s Edinburgh Encyclopaedia and the Nouveaux Principes, his article on ‘Balance des consommations avec les productions’ in the Revue encyclopédique, May 1824. In this same periodical (June and July 1827) he also crossed swords on the subject with Dunoyer. 7 Principles (Grundlagen, 1st ed. 1854; English trans., 1878), Book IV, 216–17 on ‘Commercial Crises,’ and 220, entitled ‘When Saving is Injurious.’ The theory that every circumstance that suddenly and largely increases production or decreases consumption or ‘disturbs the ordinary course of industry, must bring with it a commercial crisis’ is still more fully explained in Ansichten der Volkswirtschaft (1861). This view of the matter, sometimes glorified by a shrewd and instructive analysis of individual situations, was very common in France. Courcelle-Seneuil, Chevalier, and many other authors who differ only in the relative emphasis they put on the circumstances that are particularly apt to play a role—credit expansion, e.g.—could be quoted to this effect. A very typical example will suffice, however: Joseph Garnier’s Éléments (1845; later Traité) and especially his article on ‘Crises commerciales’ in the Dictionnaire universel théorique et pratique du commerce et de la navigation (1859). 8 In doing so, however, he hit upon a formulation that is not inelegant and will ring familiarly to modern ears though, as we know, it misses Say’s meaning completely: he said that Say’s law is true for all commodities including money. History of economic analysis 710 had overstressed his point; he admitted that saving is ‘injurious’ if savings are not invested (Principles, 220); he accepted several points that had been made by Sismondi; finally, perhaps under the influence of J.S.Mill, he recognized the role of absorption of funds in fixed investments 9 —all this without any effort at rigorous formulation or co- ordination. The situation that produced performances like this invited factual investigation and there were several good monographs on individual crises, but I shall only mention the comprehensive and very successful history of crises by Wirth. 10 Of much greater interest than the work we have been surveying so far was the cycle analysis of Tooke and Lord Overstone. Though ‘crises’ commanded the scene throughout the century, it occurred to many observers from the 1820’s on—among whom, not much to their credit, the scientific leaders of the profession were not conspicuous—that crises are but phases in a more fundamental wavelike movement and cannot be really understood except within this broader setting. From the first, writers used the term ‘cycle’ or ‘commercial cycle’ in order to denote the units of this movement 11 and spoke of a ‘periodicity’ of these cycles, by which most of them meant not more, however, than a definite sequence of phases irrespective of duration. 12 Some, however, did suggest approximate, if not exact, equality of duration and among these the ‘ten-year cycle’ eventually gained a certain popularity—even Marx experimented with it in a noncommittal manner. This pioneer work produced within the period the seminal performances of Jevons and Juglar, which will however be more conveniently considered in Part IV. In the footnote below, I mention a few others that have been almost forgotten. Observe that there is no relation between this work and the earlier discussions on gluts. It grew up independently and owes little if anything to the general 9 This theory was elaborated at that time by several authors, among others by V. Bonnet, Questions économiques et financières à propos des crises (1859). 10 Max Wirth, Geschichte der Handelskrisen (1858). His contributions to analysis are insignificant. But he was one of the first to attempt descriptive classification of crises (credit crises, capital crises, crises of speculation, etc.), an approach to the problem that appealed to many students in Germany. Also he emphasized the international aspects of crises. 11 The idea was new but the word was not. Sir William Petty used it with reference to the sequence of good and bad harvests (‘dearths and plenties’) in his Treatise of Taxes and Contributions (1662), in the course of an attempt to evaluate the normal rent of land. There is no evidence that he had any notion of a general economic cycle or that he wished to explain it by the variations in crops. 12 Some confusion about this has arisen from the fact that some modern writers who use the phrase ‘periodicity’ in the strict sense—recurrence in constant periods—attribute the same sense to all writers who use the word and then speak of assertion or denial of periodicity when they should speak of assertion or denial of periods of constant duration. This must be borne in mind throughout. Lord Overstone spoke of ‘conditions which are periodically returning’ but did not assert that they were recurring in equal periods. Juglar (see below, Part IV, ch. 8, sec. 9a) spoke of the retour périodique of crises, but his dating displays very unequal time distances between them. Moreover, he expressly denied that the material suggests the presence of anv definite period. Money, credit, and cycles 711 . exportable commodity’? Instead of any other answer, I shall merely point to a fact of considerable importance both for the history of economic analysis and the history of economic thought. It is natural. Part II, chs. 6 and 7. History of economic analysis 702 For this purpose, we start from a state of monetary equilibrium between two countries. 6 Being in possession of the theory of. History of economic analysis 708 later on, which died from lack of vitality in the course of the nineteenth century, though individual adherents might be named throughout. One of them