how very interesting this is and how revelatory of ‘the ways of the human mind.’ It implies, of course, that Ricardo was completely blind to the nature, and the logical place in economic theory, of the supply-and-demand apparatus and that he took it to represent a theory of value distinct from and opposed to his own. This reflects little credit on him as a theorist. 26 For it should be clear that his own theorem on equilibrium values is only tenable, so far as it is tenable at all, by virtue of the interplay of supply and demand. Ricardo could not have failed to discover this, had he tried to deduce that theorem rationally instead of merely positing it intuitively. That is to say, had he but stopped to ask why exchange values of commodities should be proportional to the quantities of standard labor embodied in them, he would, in answering this question, have found himself using the supply-and-demand apparatus by which alone (under appropriate assumptions) that ‘law’ of value can be established. Then he could never have denied the validity of the ‘law’ of demand and supply for the long-run normal prices of the goods the quantities of which can be indefinitely increased by human industry, while admitting its validity for short-run market prices and for the prices of monopolized or ‘scarce’ goods. For, as Malthus pointed out painstakingly (Principles, 1st ed., ch. 2, 2 and 3), supply and demand come in quite generally 27 to determine prices in both the long-run and the short-run cases, and the difference between them consists only in the level at which supply and demand fix them, which has certain properties in the one case that are absent in the other. In other words, the concepts of supply and demand apply to a mechanism that is compatible with any theory of value and indeed is required by all. But so great was Ricardo’s personal authority with some later writers that traces of this mistake of his may be found not only in J.S.Mill’s but even in A. Marshall’s Principles. However illogically, the supply-and-demand mechanism actually drifted into the place of a theory of value, 28 the exponents of which may even be said to have held the fort against the labor-quantity theory throughout the period. This was not only due to Ricardo’s carelessness but also to their own. We have seen that they failed in their analysis of the element of utility, though they brushed against it time and again. No more than Ricardo did they trouble to work out a theory of exchange which, in their case as in his, among other things, accounts for faulty handling of the concept of scarcity—the basic importance of which for the whole field of value theory was, however, asserted by 26 The same applies to Marx, who took the same view without observing that his exploitation theory presupposes that supply and demand do their work. 27 This is not strictly true except in pure competition in Professor Chamberlin’s sense. In the case of monopoly there is no supply function; in the case of monopolistic competition, also in Professor Chamberlin’s sense, there is neither a demand nor a supply function of the kind that exists in the case of pure competition. To be true, the statement above must be confined to the latter case. 28 Malthus (Principles, 1st ed., p. 495) went so far as to call ‘the principle of supply and demand’ the ‘first, greatest, and most universal principle’ of Political Economy. [All future references are to the first edition of the Principles.] History of economic analysis 572 Lauderdale, Malthus, and Senior—and for failure to understand monopolistic pricing. 29 But the sponsors of supply and demand, again with the unnoticed exception of Cournot (and very few others, such as C.Ellet and D.Lardner), even experienced difficulty in setting on its feet the very supply-and-demand apparatus, the claims of which to a place in economic theory they tried to assert. They talked of desires or desires backed by purchasing power, of ‘extent’ of demand and ‘intensity’ of demand, of quantities and prices, and did not quite know how to relate these things to one another. The concepts, so familiar to every beginner of our own days, of demand schedules or curves of willingness to buy (under certain general conditions) specified quantities of a commodity at specified prices, and of supply schedules or curves of willingness to sell (under certain general conditions) specified quantities of a commodity at specified prices, proved unbelievably hard to discover and to distinguish from the concepts—quantity demanded and quantity supplied. Malthus made indeed some progress toward clarification. But the reader needs only to look up Senior (Outline, pp. 14 et seq.) to satisfy himself of the blundering way in which he tried to explain these simple matters. Or were they so simple after all? Is it not a fact, which stares at us from the histories of all sciences, that it is much more difficult for the human mind to forge the most elementary conceptual schemes than it is to elaborate the most complicated superstructures when those elements are well in hand? Lauderdale, Say, Malthus, and others all asked themselves the question how cost of production fits with supply and demand. Say’s contribution is enshrined in the proposition that cost of production is nothing else than the value of the productive services that are consumed in production; and that the value of the productive services is nothing else than the value of the commodity which is the result—another of those sayings of his that indicate possible insights without making them explicit enough for contemporaries and later critics to understand. Malthus, however, though he probed less deeply, explained things much better so far as he saw them. In particular, he nicely indicates the locus of cost of production, which ‘only determines the prices of commodities, as the payment of it is the necessary condition of their supply’ (Principles, ch. 2, 3)—a turn of phrase that points far ahead toward Jevonian teaching. Another lesson to conclude with. Many circumstances combined to keep the theory of those writers in a state that cannot be described as anything but primitive; but one of them was obviously the lack of the appropriate technique: essentially quantitative relations cannot be stated satisfactorily without mathematics. It is the same defect that also marred J.S. Mill’s attempt at summing up. 29 This is the more remarkable because Cournot produced within the period (1838) his classic theory of monopoly, which, however, passed unnoticed. One of the consequences of this state of things was the prevalence of very loose ideas on what monopoly really is. Even Senior spoke of a ‘monopoly in land.’ But in his case nothing more was involved than misleading terminology: he did not mean more than scarcity of land and did not actually try to explain rent by a nonexistent monopoly in it. Others did, however, and it is not always easy to tell whether a given author only used current phraseology in order to denote operation of scarcity in the case of a ‘costless’ factor of production or actually meant to assert what would be true only if landowners acted like a single seller. General economics 573 (c) J.S.Mill’s Half-Way House. ‘Happily, there is nothing in the laws of Value which remains for the present or any future writer to clear up; the theory of the subject is complete.’ So J.S.Mill wrote in 1848 (Principles, Book III, ch. 1, 1)—evidently well pleased with the analytic structure he was about to erect from existing material. Actually, the structure is not an attractive residence. Its main merit consists in the fact that it showed up its defects so clearly as to make even casual visitors desirous of remodeling it. On the one hand, there is no doubt that Mill himself sincerely wished to restate Ricardian doctrine in an improved form. And so his work in this field has been and is being interpreted to this day. Leaning heavily on De Quincey’s exposition of that doctrine, Mill accepted Utility and Difficulty of Attainment as conditions of exchange value. But the energy with which he insisted on the relative character of the latter completely annihilated Ricardo’s Real Value and reduced other Ricardianisms to insipid innocuousness. Also, abstinence takes its place along with quantity of labor as an element in ‘cost.’ In other points, shifts of emphasis do the rest to destroy what it was Mill’s intention to rebuild. But on the other hand, Mill’s own main contribution was to develop the supply-and- demand analysis so fully that, as Marshall himself was to indicate, there remained not so very much to do beyond removing loose ends and adding rigor in order to arrive at something not far distant from Marshallian analysis. He did not achieve perfect clarity 30 or in fact a complete and correct statement of the theory of supply and demand. But he went much further than the majority of economists before him—always excepting Cournot—and may be said to have been the first to teach its essentials. In particular, he wrote out, in words, the Equation of Demand and Supply, and he made full use of it in his chapter on international values, which is discussed below. It is quite true that he paid token tribute to Ricardo’s shadow by introducing supply and demand in the modest role of determinants of value in the case of commodities that are ‘absolutely limited in quantity’ (Book III, ch. 2)—with which he classed, erroneously of course, monopolized commodities, whereas he let the commodities ‘which are susceptible of indefinite multiplication without increase of cost’ be determined by this cost (ibid. ch. 3) and commodities ‘which are susceptible of indefinite multiplication, but not without increase of cost’ by ‘cost of production in the most unfavourable existing circumstances’ (ibid. ch. 5). But he was concerned not so much with supply and demand per se as with the level at which supply and demand 31 will fix equilibrium price in each of those cases. And he was truer to his thought when he formulated the ‘law of supply 30 See, e.g., his comments upon Senior’s statement that limitation of supply is essential to the value of labor itself (‘Notes on N.W.Senior’s Political Economy; by John Stuart Mill,’ publ. by Professor F.A.von Hayek in Economica, August 1945). Mill replied that ‘labour being painful, would not be incurred without some sort of equivalent pleasure and advantage even if labourers could be multiplied indefinitely by a volition or if every man could work a hundred thousand hours in the four and twenty.’ But disutility of labor is only relevant because it operates to limit the supply of labor. 31 [Throughout this book J.A.S. refers to ‘supply and demand’ whereas Mill and Marshall usually wrote of ‘demand and supply.’] History of economic analysis 574 and demand’ quite generally as he did in his ‘Notes on Senior,’ defining supply and demand as quantity supplied and quantity demanded: ‘the value of a commodity in any market will always 32 be such that the demand shall be exactly equal to the supply.’ And I maintain that this, in fact if not in intention, replaces Ricardo’s law of equilibrium values and, incidentally, completes the scrapping of Ricardo’s central concept of real value. This interpretation is reinforced by a passage in the chapter on international values: whenever the ‘law of cost of production is not applicable,’ we must ‘fall back upon an antecedent law, that of supply and demand’ (Book III, ch. 18, 1). Should this not mean that Mill embraced—without being fully aware of it—the very same analysis that was anathema to Ricardo, then the sense of this passage escapes me. Nor is there anything to oppose this interpretation in the heap of blundering propositions that Mill called ‘Summary of the Theory of Value’ (Book III, ch. 6). Quite unimportant concessions are made to the labor-quantity theory 33 (see especially propositions XIII and XV). On the other hand, definitely anti-Ricardian doctrine is repeatedly asserted (see especially propositions I, V, VIII). And the Ricardian theorem that rent is not an element in the cost of production is upheld with qualifications which, if correctly stated and developed (which Mill did not do), amount to renouncing it (see proposition IX) and point toward the opportunity cost theory. 34 A muddle, all this, no doubt. But it was not a hopeless muddle. Let us rather call it a fertile one—for this muddle contained all the elements that were necessary to straighten it out. 35 Cairnes was the first to try to do this, though without any great success. Marshall did succeed in doing this, though not without invoking ideas from outside Mill’s range of vision (see Part IV, chs. 5 and 6). 32 The proposition is true only for competitive equilibrium, which is all he meant by ‘natural’ or ‘necessary’ price. Yet Mill, while perfectly aware of this, used the word ‘always.’ I mention this because we shall meet, in sec. 4 on Say’s Law, a similar difficulty of interpretation. Let me therefore point out at once that ‘always’ or ‘necessarily,’ when they occur in those old writers who were so delightfully lacking in precision, do not necessarily mean assertions of identities. Mill obviously meant an equation and not an identity. He meant ‘always in equilibrium.’ And so may have Say. 33 But a disutility-plus-abstinence theory fits better into his general system of thought. It would be almost though not quite correct to say that Mill (and Cairnes) transformed the Ricardian labor- quantity theory into the Marshallian ‘real-cost’ theory. 34 See also Book III, ch. 16, ‘Of Some Peculiar Cases of Value,’ which, another proof of how hastily the work was put together, strays far from the matters to which it is germane and in which we read ( 1): ‘Since cost of production here fails us, we must revert to a law of value anterior to cost of production, and more fundamental, the law of demand and supply’ [my italics, J.A.S.], and this after a statement, made only three paragraphs before, which seems to exclude cases of free competition from the operation of the law of supply and demand, precisely the only ones to which it applies strictly. Indeed, for Mill Malthus had written in vain. 35 As regards methods that had been new in his youth, he was surprisingly backward: for instance, ch. 15, ‘Of a Measure of Value’ does not contain a single reference to price index numbers—an indication of narrowness of outlook that might be illustrated also by other examples. General economics 575 3. THE THEORY OF INTERNATIONAL VALUES Some of the aspects of that period’s policy of international trade have already been dealt with (chs. 2 and 5). Its monetary aspects will be surveyed in the next chapter. Here we shall consider, with the utmost brevity, 1 the purely theoretical core of the ‘classic’ teaching in international trade for which J.S.Mill introduced the phrase Theory of International Values. We are primarily interested in two things: in the contributions this theory made during that period to the analysis of international trade; and in the relations of these contributions to the theory of ‘domestic’ value sketched out above. The ‘classic’ writers, being most of them ardent free traders, were no doubt much concerned with pointing out the advantages or ‘gains’ that accrue to a country from international trade. Therefore, much of what they had to say on the subject pertains to the field of welfare economics and constitutes in fact their most important exploit in this field. But this is of secondary importance from the standpoint of this section. As regards contributions to the analysis of international economic relations— remember, we now neglect the monetary angle 2 —we have three novelties to record: (1) a distinct theory of international values; (2) the theorem of Comparative Cost; and (3) the theory of Reciprocal Demand. The first was that a distinct theory of international values emerged at all. This was, in a sense, in conformity with old tradition, since already the mercantilist writers had looked upon foreign trade as something that differed essentially, in nature and effects, from domestic trade. But for the ‘classics,’ who did not accept the rationale of the mercantilist distinction, it was by no means evident that there was any theoretically—or even practically—relevant difference or, if there was, what it consisted in. In fact, economists have never quite agreed on this. 3 The group in which Ricardo was the brightest light selected immobility of factors of production as a criterion. That is to say, they defined domestic trade as the trade relations of industries or firms between which capital and labor move without hindrance, thus assuring, in equilibrium, equal rates of return to investment and work of the same difficulty, riskiness, and so on—which was quite essential to their ‘domestic’ theory; and they defined foreign trade as the trade 1 The resulting inadequacy is more tolerable than it would be otherwise because readers may be referred confidently to an excellent study of the subject in Professor J.Viner’s Studies in the Theory of International Trade, chs. VIII and IX. 2 The fact that it is possible to neglect the monetary angle and to consider the barter angle separately is of course due to a property of the ‘classic’ pattern of economic analysis that has been discussed above. It would not be possible within every system of economic theory. We may add that in the conditions of that epoch it was somewhat less unrealistic than it would be now to identify international economic relations with trade in commodities and services and to consider this trade as a barter of commodities against commodities—though, on principle, it was just as inadmissible as it would be now. 3 Perhaps the most obvious difference between foreign and domestic trade follows from the fact that most people take different attitudes toward the advantage of their own and the advantage of a foreign country. The common habit of expressing oneself as if it were nations as such that trade (and not individuals) is in part due to this difference in attitude. But some authors have stressed the importance of national monetary and credit systems. Others have looked upon problems of location as the core of the theory of international economic relations. History of economic analysis 576 relations of industries or firms between which—for reasons such as distance, 4 difference in language, difference in legal institutions,unfamiliarity with conditions of life and habits of business—capital and labor do not move freely. This has often been misunderstood. The ‘classics’ were, of course, not unaware of the facts of international migration of both labor and capital, just as they were not unaware of the fact that neither is completely ‘mobile’ within a country. All they did was to set up, for purposes of analytic convenience, the two limiting cases as ‘ideal types’ that, though neither actually occurs in real life, represent important constituents of what does occur in real life. It is another question how the lack of realism involved affects the practical applicability of this schema. It could be shown, however, that, so long as there is any difference at all between domestic and international mobility, a theory based upon this schema will retain relevance. It can also be shown, moreover, that what the ‘classic’ theory of international value thereby loses in applicability in the field of international relations it gains in the field of domestic relations, where imperfect mobility prevails. Cairnes (Leading Principles, Part I, ch. 3) conceptualized this by introducing the terms Industrial and Commercial Competition. The former term denotes trade relations with mobility and the latter, trade relations without mobility. He also introduced the concept Non-competing Groups to denote groups of workmen (local and occupational) or of firms, the members of each of which will not or can not normally move into any of the others. Using this terminology, we may say that the ‘classics’ really developed, in addition to what purported to be a general theory of value, a theory of value for the case of non-competing groups or of commercial competition. They did this, no doubt, because they thought primarily of the application to the analysis of international trade; but the theoretical characteristic of their new doctrine is, all the same, not confined to this practical purpose. The second contribution, as everybody knows, was the theorem of Comparative Costs. As Professor Viner (op. cit. p. 440) has pointed out, A.Smith never went beyond stating that under free trade everything would be produced in the place where costs (taking account of transportation costs) were lowest. He also has pointed out that some earlier writers had formulated the more general proposition that, under free trade, commodities would be imported whenever they can be obtained most cheaply in this way. This includes the case of exports that cost less than it would cost to produce the corresponding imports at home, and thus implies the theorem of Comparative Costs. 5 I also follow Viner, however, in believing that there was distinctive merit in stating explicitly that imports can be profitable, even though the commodities imported can be produced at less cost at home than abroad. This merit belongs to Torrens (The Economists Refuted, 1808) and to Ricardo. The former baptized the theorem, the latter elaborated it and fought for it 4 The element of distance thus did enter the picture. But it entered the picture only in this and in no other way. The ‘classic’ writers did not make distance per se, i.e. cost of transportation, the center of their picture. This was done by some of their later followers or critics, Sidgwick in particular, but must not be confused with the quite different and much more modest role that was assigned to distance by Ricardo and Mill. 5 To the instance mentioned by Viner may perhaps be added M.Delfico’s argument in his memorandum Sulla liberta del commercio (1797). General economics 577 victoriously. 6 The simplest way of conveying it is to let Ricardo’s famous example do duty once more. Take two countries, England and Portugal, and two commodities, wine and cloth. Portugal, being more efficient than England in both lines of production, can produce a certain quantity of wine by the labor of 80 men and a certain quantity of cloth by the labor of 90 men, whereas in England the production of the same quantities of wine and cloth takes, respectively, the labor of 120 and of 100 men. Under these circumstances, Portugal will advantageously ‘specialize’ in wine and import cloth, while England will ‘specialize’ in cloth and import wine provided, of course, that wine and cloth exchange on any terms between the limits of one unit of English cloth for of a unit of Portuguese wine and one unit of English cloth for units of Portuguese wine. In the former case, all the advantage goes to England, and Portugal is no better off than she would be without trade; in the latter case all advantage goes to Portugal, and England is no better off than she would be without trade. So far as this goes, any intermediate exchange ratio is possible with advantage to both countries, and if traders in both countries acted as monopolists, the exchange ratio would be indeterminate between those limits. Ricardo and his immediate followers did not worry about this but glibly assumed that the advantage would be halved—which may have spelled error but also may have been merely carelessness. Other writers, among them Torrens, realized however that the indeterminateness of the terms of trade or exchange ratios would be in general removed, at least under conditions of perfect competition (or of one-sided monopoly), by the mechanism of what Torrens was, I think, the first to call Reciprocal Demand (in print). J.S.Mill, surpassing himself in generosity, not only defended Ricardo against any charge of having committed a mistake but also disclaimed credit for the original conception of this idea, although he had developed it in all essentials in an essay written as early as 1829–30 but not published before 1844 (in his Some Unsettled Questions). From it, he took the substance of Sections 1–5 of the famous Chapter 18 of his Principles (Book III) 7 that to all intents and purposes set the theory of reciprocal demand on its feet—the third novelty that was contributed during that period to the general analysis of international economic relations. 6 In spite of weak resistance that was in part supported by incompetent argument, the theorem may be said to have conquered in England. In the United States it did not catch on so well and still less did it do so on the continent of Europe, where it was widely misunderstood even among free traders. But Cherbuliez gave a good account of it. And von Mangoldt improved it, or carried it further, in one very important point (see Viner, op. cit. pp. 458 et seq.; but if the reader refers to the original text, he should turn to the first edition of Mangoldt’s Grundriss, which appeared in 1863 and contains the relevant appendix that the editor of the posthumous second edition of 1871 thought fit to omit; see above, ch. 4, sec. 5). 7 It is only these five sections of that chapter that have attained such fame. The rest of the chapter, added in the third edition, in deference to the ‘intelligent criticisms’ of friends, has had no share in the applause and has been voted ‘laborious and confusing’ even by good Millians such as Bastable and Edgeworth. I cannot quite share either opinion. There are valuable contributions in the rest. For instance, Mill never came so close to a grasp of the nature and use of the concept of elasticity of demand (which he called ‘extensibility’) as he did in 8 of that chapter. Some of the criticisms that were leveled at it rest on nothing but the clumsiness and ambiguity of expression that are unavoidable in a verbal presentation of this topic reinforced by nothing except numerical examples. History of economic analysis 578 The problem being complex and quite beyond his command of technique, J.S.Mill dealt with it by means of a number of simplifying assumptions some of which he tried to remove in Sections 6–9 of the chapter. In particular, he confined his argument at first to the case of only two commodities and of two countries—the latter, it should be added, of similar size and productive capacity—and it is in fact in this case that the principle involved can be best displayed. In order to determine the point at which, within the limits set by comparative costs, the exchange ratio or the terms of trade between the two countries and commodities will tend to be fixed, Mill fell back once more upon the ‘antecedent’ (logically fundamental) law of supply and demand. He perceived that (under fairly comprehensive assumptions) the equilibrium exchange ratio would be determined by the condition that the quantity of each of the two products that the importing country is willing to take at this ratio be equal to the quantity that the exporting country is willing to give at this ratio (Equation of International Demand). 8 It is assumed that, if the one country is willing to take more or less at this ratio than the other is willing to give, competition of ‘buyers’ or ‘sellers’ will adjust the exchange ratio until it fulfils this condition. 9 It should be recorded to Mill’s credit that he saw that this will not exclude multiple equilibria, 10 and there are more delicate questions that cannot be touched upon here. Also it should be recorded that he put the apparatus he created to good use. His treatment in Section 5 of the effects of technological improvement in an export industry that are not necessarily favorable to the exporting country deserves to be mentioned in particular. For further light on the matter the reader is primarily referred to Professor von Haberler’s well-known treatise. 11 Let us note at once that in this field Marshall did not do more than to polish and develop Mill’s meaning. He cast it into an elegant geometrical model (The Pure Theory of Foreign Trade, 1879) that greatly clarified the theory. 12 But he was well aware (see 8 The equivalent formula that the ratio is such as to equalize the values of exports and imports is simpler but brings out less well than does ours that the proposition is an equilibrium condition and not an identity. 9 This implicit assumption of Mill’s really constitutes an additional condition, the so-called secondary or stability condition. 10 The matter is a little complicated. On the one hand, as Professor Viner has pointed out (op. cit. p. 537), Mill had the correct idea of the nature of his equation of supply and demand, i.e., he saw and asserted against objectors that it was an equilibrium condition and not an ‘identical proposition,’ which would, of course, be incapable of determining an equilibrium point. This passage from a letter of his to Cairnes (Letters, ed. by Hugh S.R.Elliot, 1910) should be borne in mind precisely because it proves that he understood this difference perfectly. But, on the other hand, he stated ( 6 of the chapter on international values) that ‘it is conceivable that the conditions [of the equation of international demand] might be equally satisfied by every numerical rate which could be supposed,’ and this would make an identity of that ‘equation.’ If, however, we read this passage in its context, we readily realize that it does not really mean more than the perception of the possibility of the existence of more equilibrium positions than one, and then a distinct merit emerges to claim recognition instead of the demerit to which critics—including Edgeworth—have called attention. 11 G.von Haberler, The Theory of International Trade (1936, chs. 9–12). The reference also covers the subject of comparative costs and is intended to help all those readers who will find my brief account unsatisfactory or even ununderstandable. 12 On this model, see Haberler, op. cit. pp. 153 et seq.; also Appendix J of Marshall’s Money, Credit and Commerce. There is a London School Reprint (1930) of the papers of 1879 under the title, Pure Theory (Foreign Trade—Domestic Values). General economics 579 Memorials of Alfred Marshall, ed. by A.C.Pigou, 1925, p. 451) that his curves ‘were set to a definite tune, that called by Mill.’ This applies even to the geometrical apparatus: Mill’s reads almost like a somewhat clumsy instruction for choosing these curves rather than any others. Edgeworth’s famous restatement (‘The Pure Theory of International Values,’ Economic Journal, 1894, reprinted in Papers Relating to Political Economy, vol. II) added many interesting details but also did not go beyond Mill in fundamentals. Serious attacks do not antedate the 1920’s, and even then leading masters in the field substantially adhered to his teaching. Since advocacy of free-trade policy was the main practical purpose the ‘classical’ writers had in mind when they developed their theory of international values, they were naturally much interested in displaying the ‘gains’ that accrue to a nation from foreign trade. We have noticed elsewhere the bias which this imparted to their argument and their tendency to underestimate the possibilities of unilateral gain from protection. Here we are more interested in finding out how they defined these gains and how they tried to quantify them. Of course, in the early stages of the discussion it was quite sufficient to say that foreign trade will supply a nation with commodities which it could not produce at all or could produce only at higher cost. The latter element having been reinforced by the introduction of the comparative-cost principle, it was not less natural for Ricardo to stress the resulting saving in cost per unit of product. There are two aspects to this. On the one hand, this comes to the same thing as stressing the gain in quantity of product per unit of costs. 13 Ricardo recognized, of course, that foreign trade cannot increase the sum total of real value (in his sense) in a country, but ‘it will very powerfully contribute to increase the mass of commodities, and therefore the sum of enjoyments’ (Principles, ch. 7). There he stops because he strongly believed that utility (value in use) cannot be measured. 14 But still we might express Ricardo’s meaning by saying that foreign trade increases enjoyment per unit of his real value. In any case, this is as far as he went into the welfare economics of foreign trade—further, however, than is commonly believed. On the other hand, foreign trade does bear upon the structure of Ricardian real value in this way: if, as was the case with England, imports consist to a considerable extent of foodstuffs and other necessities—such as cotton—that enter largely into the consumption of the working class, then the share of the latter in total value will fall and the real value of profits and the rate of profit will rise. Needless to say, this is an essential part of Ricardo’s free-trade argument: foreign trade increases indeed the ‘happiness of mankind’ by improving the allocation of resources and by giving ‘incentives to saving and to the accumulation of 13 Referring to the example by which Ricardo explained the operation of the comparative-cost principle, we observe that if England and Portugal each produced one unit quantity of cloth and one unit quantity of wine without trade, this would take 390 labor units in all whereas, after specialization through free trade, these same four unit quantities would take only 360 labor units. 14 But since he said that foreign trade will increase the sum of enjoyments (the word ‘sum’ is out of place, of course, if utility is not measurable), he should not have said that utilities cannot be compared. The latter is in fact implied by the comparative-cost principle, which is relevant only because comparison of utilities is possible. At a stretch, we might credit Ricardo with the modern idea that there is such a thing as ‘ordinal utility,’ although there is no such thing as ‘cardinal utility’—that a utility is capable of being greater or smaller than another, although it is not capable of being a multiple of another. History of economic analysis 580 capital’—‘by the abundance and cheapness of commodities’ which it brings about—but does not, except temporarily, raise profits unless it is instrumental in reducing the Ricardian real value of wage goods just as would a technological improvement in their production. So far as there is anything at all in Malthus’ argument on the subject, it does not contradict Ricardo’s. Of course, as Professor Viner has pointed out (op. cit. p. 531), he might have said that the ‘sum of enjoyments’ is a treacherous concept to use because foreign trade will influence the distribution of incomes, conceivably in a direction that may be unfavorable to small incomes. But he did not say this. Nobody did at the time excepting some politicians who argued on this line, in the English corn law controversy, on behalf of farmers. I do not maintain, of course, that either Ricardo or Mill handled the welfare aspects of foreign trade satisfactorily. Objectively, Mill’s theory of reciprocal demand was a step in advance because it pointed more directly toward welfare (utility) aspects. But Mill himself did not exploit the possibilities, such as they were, that his approach suggests. This was reserved for Marshall and Edgeworth, who developed methods which, though they have become obsolete by now, gave satisfaction to many in the 1890’s (see below, Part IV, ch. 7, Appendix). They, Edgeworth especially, criticized Mill for estimating benefit from foreign trade exclusively by the criterion of exchange value (terms of trade). 15 In view of Ricardo’s emphasis on increase of means of enjoyment, this criticism hardly applies to him. In the case of Mill there is more to it, but not much more. Both saw the nature of the ‘social gains of trade’ correctly. It is truer to say that they did not attempt to measure them at all—and there is something to be said for stopping at what Cairnes regretfully called an ‘indefinite and vague result’ (Leading Principles, p. 506 of English ed.; the pagination differs in the American ed.)—than it is to say that they attempted to estimate them from the terms of trade. Let us now ask the question how were the theories of comparative cost and of reciprocal demand related to Ricardo’s and Mill’s general theories of value or, to put it in the usual way, what was the relation between their theories of foreign and domestic value? First of all, what was the relation of the theories of comparative cost and of reciprocal demand to one another? Mill’s generosity has obscured the obvious answer. As we have seen, in his Essays on Some Unsettled Questions, 1844 (Essay I, ‘Of the Laws of Interchange between Nations’), he presented his equation of reciprocal demand as a modest supplement to Ricardo’s comparative-cost principle that the great pioneer had had no time to add himself. Most historians and critics have taken the same view. But it should be clear that this view is entirely wrong. The demand-supply schedules, whose intersection gives the geometrical picture of the equation of reciprocal demand, represent an approach that Ricardo always rejected except for temporary fluctuations and for monopolized commodities. They introduce a new and more general principle just as, in 15 That is to say, Edgeworth and others accused Mill of believing that a country’s total gain from trade always increased or decreased as its gain per unit of exports (the quantity of German linen that England received for a unit of her cloth) increased and decreased, just as writers on wages sometimes assume that the national pay roll always increases or decreases when wage rates do. General economics 581 . importance of national monetary and credit systems. Others have looked upon problems of location as the core of the theory of international economic relations. History of economic analysis 576. consumption of the working class, then the share of the latter in total value will fall and the real value of profits and the rate of profit will rise. Needless to say, this is an essential part of Ricardo’s. capable of being greater or smaller than another, although it is not capable of being a multiple of another. History of economic analysis 580 capital’—‘by the abundance and cheapness of commodities’