competition—which according to him is productive of disorganized markets in which every participant is proletarized. But better things were to follow in the eighteenth century. We shall confine ourselves to the peak achievements of Beccaria, Turgot, and Isnard, and then consider the manner in which the Wealth of Nations codified the whole of the value and price theory of the epoch. Beccaria dealt with value and price in Part IV, Chapter 1 (‘Del commercio’) of his Elementi (publ. posthumously, 1804): the subject stands there pretty much where it was to stand in J.S.Mill’s Principles. He explains the phenomenon of value, as mentioned already, by utility and scarcity, and then proceeds to investigate the modus operandi of a hypothetical market in which wine is bartered for wheat (cf. Marshall’s apples and nuts). 13 He recognized clearly that the exchange ratio is indeterminate in the case of isolated exchange (between two individuals) and that determinateness is brought about by competition through the ‘higgling of the market’: fluctuations must eventually lead to the price at which quantity demanded equals quantity supplied. His careful treatment of the exchange of three commodities against one another, in which he insists on the phenomenon (and necessity) of indirect exchange, is particularly satisfactory. This is about as much as the average economist had to say a century later. Beccaria’s performance has been chosen for comment because of its comparative fullness, but it had been strikingly anticipated by Turgot’s Réflexions, XXXIII–XXXV (written in 1766, published in 1769–70). After having deduced trading (commerce) from besoins réciproques, Turgot, too, touches upon the case of isolated exchange and then introduces the determining force, competition. His description of the market mechanism is very similar to that of Böhm-Bawerk (see below, Part IV, ch. 5, sec. 4). The resulting market price (prix courant) is then made to vary under the impact of forces acting through demand or supply. The crowning achievement of the epoch in this line of analysis is Isnard’s. 14 In his not otherwise remarkable book there is an elementary system of equations that—barring the difference in technique—describes the interdependence within the universe of prices in a way suggestive of Walras. 13 I hope that we need not take too seriously his proposition that the exchange value of one commodity relatively to the other (the exchange ratio) will be in ragione reciproca delle loro quantità. Possibly there is a connection between this and his friend Verri’s hyperbolical demand law, which however is not open to the same objection but, on the contrary, must be recorded as the first attempt to give a precise form to the demand curve: if p be price (in money), q quantity, and c a constant, then, according to Verri’s law, pq=c. 14 Achille Nicolas Isnard’s Traité des richesses appeared in 1781, and thus forms no part of the material ‘codified’ by A.Smith. There is, however, a question of the latter’s influence upon the former. Isnard’s treatise, also the contribution mentioned in the text, could have arisen from a perusal of the Wealth of Nations. Isnard does not mention A.Smith, however, unless, indeed, I have overlooked the reference. The title of the book is included in Jevons’ list of mathematical writings, which is how I came to know of it. I have found no traces of its influence. History of economic analysis 292 [(d) Codification of Value and Price Theory in the Wealth of Nations.] A.Smith’s ‘chief work was to combine and develop the speculations of his French and English contemporaries and predecessors as to value.’ 15 Also, it is quite true that he made ‘a careful and scientific inquiry into the manner in which value measures human motive,’ that is to say, I take it, that he made exchange value (price or, at all events, relative price) the centerpiece of a primitive system of equilibrium. But he was not, as Marshall held, the first to do so; moreover, in codifying, he dropped or sterilized many of the most promising suggestions contained in the work of his immediate predecessors. Of course, he may not have known Turgot’s Réflexions, and he cannot have known Beccaria’s Elementi: Pufendorf and then Cantillon, Harris, Locke, Barbon, Petty—these last five are mentioned by Marshall—and Quesnay were presumably his principal guides so that his ‘subjective’ performance was greater than was his ‘objective’ achievement. But he ‘developed’ this material less successfully than had Turgot and Beccaria. The blame is at his door for much that is unsatisfactory in the economic theory of the subsequent hundred years, and for many controversies that would have been unnecessary had he summed up in a different manner. The reader should refresh his recollection of the Reader’s Guide presented above. 16 A.Smith’s exposition in the first Book surges purposefully up to the phenomenon of price and down again into the component parts of commodity prices, which components are the cost and income categories, wages, profit, and rent. This is, to repeat, a primitive way of describing the universal interdependence of the magnitudes that constitute the economic cosmos; but it is an effective way. Critics who did not understand that the theory of price is but another name for theory of economic logic—including, among other things, all the principles of allocation of resources and of formation of incomes—blamed him for having adopted the narrow point of view of the businessman. Other critics who did not understand the nature of a system of interdependent magnitudes accused him of circular 15 A.Marshall, Principles, 4th ed., p. 58. The reader will realize that the opinion of a workman such as Marshall is worth a ton of philosophizing by less workmanlike people, and we cannot do better than use that dictum of a master of economic theory as a motto. Also, the reader will realize that even Marshall, whose admiration for Smith was unbounded, does not go beyond what our own term ‘codification’ implies. Though he was far from attributing to Smith any original ideas, he nevertheless arrived at an estimate of the performance that seems much higher than ours. One reason for this may be that he was speaking of a brother—for as has been and will be emphasized, there are many similarities in the performances and in the historical positions of the two. Another may be that he was speaking of a countryman—for Marshall was very insular. A third one may be that he was speaking of a fellow liberal—for Marshall, too, was a strong free trader. But whatever the reason, the reader should observe that, so far as Marshall’s very brief comments enable us to judge, there is no difference as to the facts of the case except this: Smith may certainly be said, in a sense, to have ‘developed’ existing doctrines of value and price; but whereas Marshall approved unconditionally of the manner of this ‘development,’ I have some fault to find with it. 16 [After writing this, J.A.S. apparently removed several pages on A.Smith, including the Reader’s Guide, from the manuscript. This material has been restored by the editor and is to be found in ch. 3, sec. 4e.] Value and money 293 reasoning. His shade easily wins out against these and other criticisms. It is this part of his performance that constitutes his chief merit in this field. There are others. As primitive but as distinctly visible as is his concept of universal interdependence is his concept of equilibrium or ‘natural’ price. This equilibrium price is simply the price at which it is possible to supply, in the long run, each commodity in a quantity that will equal ‘effective demand’ at that price. This again is the price that will, in the long run, just cover costs. And these, in turn, are equal to the sum total of the wages, profits, and rents that have to be paid or imputed at their ‘ordinary or average rates.’ Thus, we also get a glimpse of Marshall’s distinction between short-run and long-run phenomena, A.Smith’s market price being essentially a short-run phenomenon, his ‘natural’ price a long-run phenomenon—Marshall’s long-run normal. ‘It is all in A.Smith’ was a favorite saying of Marshall’s. But we may also say: ‘It is all in the scholastics.’ There is no theory of monopoly. The proposition (Book I, ch. 7) that ‘the price of monopoly is upon every occasion the highest which can be got’ might be the product of a not very intelligent layman—taken literally, it is not even true. But neither is the mechanism of competition made the subject of more searching analysis. In consequence, A.Smith fails to prove satisfactorily his proposition that the competitive price is ‘the lowest which the sellers can commonly afford to take’—to the modern reader it is a source of wonder what kind of argument he took for proof. Still less did he attempt to prove that competition tends to minimize costs, though it is evident that he must have believed it. But what was A.Smith’s theory of value in the narrow sense of the phrase, meaning his views on the problem of causal explanation of the phenomenon of value? Since during the subsequent century economists were much interested in that problem, they eagerly discussed Smith’s views about it and for this very reason we cannot pass it by. In itself, the answer is plain enough. First of all, if the reader will look up the last paragraphs of Book I, Chapter 4, he will be able to satisfy himself of two things. On the one hand, A. Smith declares there that he is going to inquire into the rules which ‘men naturally observe in exchanging’ goods ‘either for money or for one another.’ This means that he was not primarily interested in the problem of value in the sense just defined. What he wanted was a price theory by which to establish certain propositions that do not require going into the background of the value phenomenon at all. Evidently this was also Marshall’s opinion. On the other hand, having distinguished value in use and value in exchange, he dismisses the former by pointing to what has been called above the ‘paradox of value’—which he evidently did believe to be a bar to progress on this line—thereby barring, for the next two or three generations, the door so auspiciously opened by his French and Italian predecessors. No talk about his ‘recognizing the role of demand’ can alter this fact. Second, in Book I, Chapter 6, A.Smith expressly states: ‘Wages, profit, and rent, are the three original [my italics] sources of all revenue as well as of all exchangeable value.’ If words mean anything, this is conclusive. His theory of value was what later on came to be called a cost-of-production theory. This is indeed the opinion of many students. But, third, the matter is complicated by the fact that a very large number of passages in the Wealth of Nations seem to point to a labor theory of value or rather to several. 17 17 It is still not sufficiently recognized that the term labor theory of value covers several distinct meanings. The subject has, however, been exhaustively dealt with by H.J.Davenport, Value and Distribution, 1908. History of economic analysis 294 In Book I, Chapter 5, of the Wealth of Nations occurs the proposition: ‘The real price of everything, what everything really costs to the man who wants to acquire it, is the toil and trouble of acquiring it’—one of those treacherous platitudes that may mean anything and nothing. On the face of it, however, it indicates a tendency to base the value phenomenon upon the irksomeness or disutility of work, or to adopt a labor-disutility theory of value. This theory, however, may be discarded, because A.Smith makes no use whatever of it. Again, at the beginning of Book I, Chapter 6, Smith produces the famous example about the beaver: ‘if…it usually costs twice the labour to kill a beaver which it does to kill a deer,’ one beaver would naturally sell for as much as two deer. There it is quantity of labor that ‘regulates’ value and not toil and trouble, which is, of course, not the same thing. No doubt is possible but that this passage is the root of Ricardo’s and Marx’s labor-quantity theories of value. But A.Smith limits this theory to ‘that early and rude state of society which precedes both the accumulation of stock and the appropriation of land,’ which, interpreted charitably, means that competitive prices of commodities will, in equilibrium, be proportional to the labor entering into their production if the labor is all of the same ‘natural’ quality and if there are no other scarce means of production. This is true but does not in itself constitute a labor-quantity theory, or any labor theory, of value, because, for this special case, all theories of value would arrive at the same result. Finally, as we have already had occasion to notice, A.Smith (Book I, ch. 5) considers the quantity of labor a commodity can command in the market the most useful substitute for its price in money, that is to say, he chooses labor for numéraire. On principle, there can be no objection to this decision, which in itself no more commits him to a labor theory of value than the choice of oxen for numéraire would commit us to an ox theory of value. But he tries to motivate his decision by so many arguments that seem to claim deeper meaning for it—such as that ‘labour alone…never varying in its own value, is alone [sic] the ultimate and real standard’ of the values of all commodities or that ‘it is their real price’ or ‘the only universal as well as the only accurate measure of value,’ which are all wrong—and seems himself so little clear about what is and what is not implied in choosing something for numéraire that it is almost excusable if many later economists misunderstood what he actually did mean and that they, among them Ricardo, 18 accused him of having confused the quantity of labor that enters a commodity with the quantity of labor it will buy. This indictment fails, however, and it is important that it does, for it amounts to accusing Smith of an absurdity: taking what a commodity exchanges for, no matter what it is, as the explanation of its value would be one of the worst slips in the history of theory. It should be added that, if choosing an hour’s or day’s labor as the unit in which to express prices does not imply accepting a labor theory of value, no more does emphasis upon labor’s role in production or upon labor’s claims or wrongs. As has been mentioned already, there is plenty of this in the Wealth of Nations, much of it, perhaps, inspired by Locke. ‘The produce of labour constitutes the natural recompence or wages of labour’ (Book I, ch. 8). It is the laborer who raises the crop and the landowner, having appropriated the land, demands a share of it. Profit makes a second deduction from the 18 Ricardo did not misunderstand him always, however. Ricardo also argued that the exchange value of labor is no more exempt from fluctuations than is the exchange value of anything else. But this is relevant only with regard to making labor a numéraire that is to function over time. Value and money 295 ‘produce of labour.’ To this day, it has remained difficult to make the philosophy- minded see that all this is completely irrelevant for a theory of value—considered not as a profession of faith or as an argument in social ethics, but as a tool of analysis of economic reality. 4. THE QUANTITY THEORY It will not surprise the reader to learn that the effects of the violent price revolutions of the fifteenth, sixteenth, and seventeenth centuries should have been zealously discussed. But it might surprise him to learn that there was any question about their causes. For debasement of the currency—devaluations by governments as well as fraudulent clipping of coins by individuals—and the torrent of American gold and especially silver were before everyone’s eyes; and not even the most sophisticated theorist of today could find fault with the obvious diagnosis, seeing that the monetary units newly created by either the debasement of the coinage or the influx of the American silver were very promptly spent, while the very wars on which they were mainly spent greatly interfered with production. Nevertheless, though it is probably possible to find early arguments that more or less distinctly imply this obvious diagnosis, 1 it seems to be the fact that no explicit, full, and—so far as it went—theoretically satisfactory presentation of it appeared before 1568, when Bodin published his Response to the Paradoxes sur le faict des Monnoyes (1566) of M.de Malestroict. (There is a translation of Bodin’s Response in A.E.Monroe’s Early Economic Thought.) On the strength of this, he is universally voted the ‘discoverer’ of the Quantity Theory of Money. Since the matter has received attention quite out of proportion to its importance, we shall go into it briefly ourselves. [(a) Bodin’s Explanation of the Price Revolution.] Jehan Cherruyt de Malestroict had argued that the universal rise in prices was due to debasement and that, expressed in full-weight coin, prices had not risen. Bodin replied— 1 I say ‘probably’ because I do not myself know of any clear instances. The cases mentioned by Professor Seligman (article, ‘Bullionists,’ Encyclopaedia of the Social Sciences) are not convincing. Some historians who trace this kind of ‘quantity theory’ to the Middle Ages and even to the Roman jurist Paulus have misunderstood the word quantitas, which must not be translated by ‘quantity’ (see above, ch. 1). The best example I can mention—but it followed by a year Bodin’s work, which I am about to mention in the text—is in the Summa de tratos y contratos of Tomás de Mercado (1st ed. 1569; ed. used 1571). By the end of the century, recognition of the effects on prices of the American silver was widespread, if not universal, as well it might have been. Luíz Valle de la Cerda (Fundación, 1593; Desempeño, 1600) called the price rise the ‘efecto muy natural de la rapida multiplicacion de los signos y moneda.’ The words ‘very natural’ in this passage have nothing to do with natural law but simply stand for ‘indubitable’ or ‘obvious.’ The contrary impression, voiced by Professor Hamilton, viz. that the influence of the inflow of silver upon prices was being persistently ignored or denied, seems due to the facts that many later authors had reasons for emphasizing the other factors in the price rise which were operative, especially in the case of Spain, and also that the common run of writers was then, as it is now, quite impervious to even the simplest economic truths. History of economic analysis 296 and then repeated in Les six livres de la République, 1576—that this argument overlooked the influence of American silver. The price revolution, according to him, was due to (1) the increase in the supply of gold and silver; (2) the prevalence of monopolies; (3) depredations that reduced the stream of available commodities; (4) the expenditure of kings and princes on the objects of their desires; and (5) the debasements that were the only factor considered by his opponent. He added, moreover, that the first cause was the most important one. The reader will observe that this analysis needs but little readjustment or generosity of interpretation to be a correct diagnosis of the historical case as it presented itself in 1568. Even as regards general theoretical content, it is superior to much later work. In fact, Bodin’s analysis escapes several of the typical objections that were to be raised against the quantity theory in the nineteenth century. But does it state or imply this theory? The question may seem surprising, but it is well worth asking. Let us embrace, for a moment, uncompromising metallism and consider the case of perfect gold monometallism—gold metallism such that gold can move freely in and out of the monetary system—from this standpoint. Gold being a commodity like any other, the value in terms of commodities of the golden monetary unit will fall, other things being equal, if gold production increases, just as the price of eggs will fall, other things being equal, if egg production increases. Any rise in prices in terms of gold that may occur is here explained as a consequence of increased supply. Let us note that the extent of this fall (in the value of gold) will simply depend upon the shape of the demand schedule for gold as a commodity in terms of some other standard, and that the operative ‘quantity’ in question is the total amount of the increase. In consequence, there is no reason to suppose that, however equal the other things, the fall will be proportional to the increase. It will be seen that no special hypothesis enters into the argument, which flows smoothly from the metallist basis and would have been accepted by the scholastics as a matter of course. But recognition of the relevance of ‘quantity’ to the value of money in this sense and for this reason has no more to do with the Quantity Theory of Money than that the word quantity occurs in both arguments. And no more than this is required for Bodin’s argument or, let us add at once, for that of A.Smith. [(b) Implications of the Quantity Theorem.] In order to make this clear, let us look at the same case from the standpoint of the quantity theory. To facilitate exposition we shall assume that there is an absolutely fixed collection of goods that must be sold for whatever money buyers have, and that these buyers feel compelled to spend promptly all of whatever money they have upon that collection of goods. Also we shall henceforth speak, instead of the quantity theory, of the quantity theorem, because it is not a complete theory of money but merely a proposition about the exchange value of money. Keeping, then, everything else severely as it is, we let gold production increase. As in the simple metallist argument, we infer that this will make the unit of gold less valuable, that is, raise all prices in terms of it. The reason for this is the same as before so far as that part of the increase is concerned that goes into industrial uses. But that part of the increase which spills over into circulation now operates in a different way and produces a fall in the exchange value of the monetary gold—a rise in commodity prices—for a different reason: the fall, under our highly Value and money 297 artificial assumptions, is exactly proportional to the increase in the quantity of the monetary gold stock; and the immediate reason for this is not the fall in the commodity value of the gold—which is relevant indeed but only at one remove, that is, by virtue of the fact that it will determine the extent to which the quantity of monetary gold will be increased—but the increase of the quantity of coins per se. It is the increase in this quantity which, the purchasing power of the total monetary stock remaining constant, is the immediate cause of the resulting fall in the exchange value of the monetary unit. And this fall will be the same as if this stock, without being increased, had been split into units of smaller gold content, because in either case there is now less of every commodity per coin. Operation of the new gold in the commodity use may be likened to the effects of adding workers of the same skill to a given plant and equipment. Operation of the new gold in the monetary use may be likened to the effects of replacing the working force operating a given plant and equipment by more workers of proportionately less skill. Thus the quantity theorem does three things: first, it recognizes the fact that the monetary function will affect the value of the commodity chosen for money and is a logically distinct—though not independent—source of the exchange value of gold (this, of course, we can recognize without committing ourselves to the next steps); second, it recognizes that the mechanism that determines the value of gold in circulation is different from the mechanism that determines the value of the industrial gold or of any other commodity; third, it offers a specific schema—very primitive but also very simple—of that mechanism. The apparent difficulty of this really simple matter is due to the facts that in the case of perfect gold monometallism the two different mechanisms must, of course, produce the same values of gold in the monetary and in the industrial sphere; and that the influences of an increase in gold production upon the commodity value and upon the monetary value of gold so intertwine that we do not see either quite clearly. But it is one of the strong points of the quantity theory that it can be applied to the case of paper money without any auxiliary construction. And in this case—when there is no commodity value of the material to cause ambiguity about what quantity we mean and what modus operandi we attribute to it—all becomes perfectly clear. This logical affinity of the quantity theorem with theoretical cartalism should be borne in mind: the theorem essentially amounts to treating money not as a commodity but as a voucher for buying goods, though not everyone who does consider money in this light need accept the specific schema offered by the quantity theorem. It is the more important to remember this point because later developments tended to obliterate it. There is no trace of considerations of this type in Bodin. But there is in Davanzati (1588, see above, sec. 2), who confronted the mass of commodities with the mass of money—stock with stock—and would have to be credited with superior formulation of the quantity theorem in its most primitive form even if we interpreted Bodin’s argument in the same sense. Subsequent advance on this line was slow. Mere recognition of the effect upon prices of American gold and silver imports or of any increase in a country’s stock of gold and silver, of course, soon became commonplace. It is not always easy to tell from the uncouth writings of the less literate ‘mercantilists’ what it was they had in mind, but some of them, especially Malynes and Mun (see below, ch. 7), tried, I think, to convey the genuine quantity-theory idea—though in a quite rudimentary form—while History of economic analysis 298 others, perhaps the majority, were content with ‘simple metallism.’ 2 However, Davanzati at long last found a successor in Montanari (1680), and in England instances become frequent in the second half of the seventeenth century. Among these Briscoe (1694) deserves special notice, 3 because he was the first, so far as I know, to write an equation of exchange in the unsatisfactory form: stock of money equals prices times real income. 4 In the course of the eighteenth century it was the genuine quan-tity theorem that, sometimes in the crudest possible form, became a commonplace for many of the leaders. It is taken for granted by Genovesi, Galiani, Beccaria, and Justi, and Hume reasserted it with an emphasis that was hardly necessary (1752). All the more significant is it that A.Smith did not definitely commit himself to more than simple metallism. 2 Malynes says categorically: ‘Plentie of money maketh generally things deare’ (Tudor Economic Documents III, 387), which at least admits the interpretation above. Mun’s case is similar. An illustration of a recognition of the effect of increased ‘treasure’ that does not imply the genuine quantity-theory idea is in Sir Robert Cotton’s speech on the ‘Alteration of coine,’ 1626: ‘Gold and silver…are Commodities valuing each other according to the plenty or scarcity; and so all other Commodities by them’ (Reprinted in McCulloch’s Scarce and Valuable Tracts on Money). It should be observed, however, that what we have called the theory of ‘simple metallism,’ i.e. theoretical metallism without the specifically quantity-theorem element in it, is sufficient to protect ‘mercantilists’ from the indictments that in general they failed to perceive the effect of their beloved gold and silver imports upon prices and that, whenever they did, they really refuted their argument for export surpluses. The specific or genuine quantity theorem is not necessary for perceiving that effect. And the proposition that export surpluses are desirable because they will bring ‘treasure’ into the country, whatever its demerits may be, is not refuted by the argument that this inflow of gold and silver will raise prices and thus stop the exports. For, first, much treasure can be collected on the way to this consummation. Second, the treasure imported would have that effect only if it entered circulation, which was not always the idea. There are also other possible lines of defense that will appear as we go along. 3 Montanari and Briscoe are discussed in sec. 2, above. 4 Since this is the first time that we meet this analytic tool, it is convenient to make at once a few comments that will be of help to the beginner. Other comments will be added later (Part IV, ch. 8). The equation of exchange (also called the Fisher equation after the most eminent of those modern economists who used it as a starting point of the theory of money) is now usually written: MV=PT, where M means quantity of money, V velocity, P price level, and T physical volume of transactions. Briscoe’s equation becomes identical with this by putting V=1. The first thing to note is that M, V, P, T can each of them be given any of a number of different meanings, which must, of course, be made to correspond; for instance, M may mean only full-weight coin, or only legal tender (including government paper), or only legal tender plus bank notes minus the reserves held against them, or legal tender plus bank notes plus demand deposits minus the reserves of all banks. Similarly, T may mean all transactions, or only the transactions incident to production and distribution, or only the transactions consisting of income payments and income expenditure on consumers’ goods—the last being the definition adopted by Briscoe. Second, in the case of V, there is a distinction of a different kind to be made that is of the utmost importance. On the one hand, we may put V=PT/M by definition. If we do this, then the equation of exchange must evidently hold always and under all circumstances. It is, as we say, a mere tautology or identity and should really be written MV≡PT. But we need not do this. We can also define V independently of the three other magnitudes, for instance, by the number of times a dollar can on the average be paid to an income receiver in the institutional arrangements of a given society. Value and money 299 But Briscoe’s equation of exchange was already obsolete when he published it: 5 a major step forward had been taken before. The most primitive way of looking at the relation between quantity of money and prices, but to the primitive mind the most natural way, is to compare a stock or fund of money with a stock or fund of goods that are supposed to be exchanged against one another. The next idea to occur to one’s mind, when one comes to think of it more carefully, is that this stock of goods is a rather doubtful entity: the total of the coins may indeed be thought of as a definite stock of pieces that, unless demonetized or exported, are also permanent; but the commodities that are being currently exchanged for these coins are not each time the same individual pieces—the individual units of bread and wine and cloth and so on disappear from the market for good and are currently replaced by other units to meet, on the next market day, the same coins again. Therefore, comparison is between a stock and a flow. The obvious way to reduce them to comparability is to choose a unit period and to multiply the stock by a coefficient that tells us how often in this period the stock meets the flow, that is, how often per period the money does what the goods can do only once. The problem is greatly simplified, though its solution loses much in value, if we assume that all the coins are spent—none held back—and spent only once each market day—equal quantities of all goods being offered on each day—and that there are no other transactions: then the ‘velocity’ or ‘rapidity of circulation’ of money will equal the number of market days per unit period. If this number be 12 per year, the stock of money will support the same price level that a stock 12 times as great would support with but a single market day per year. Taken in this sense, ‘velocity’ is peculiar to money and neither has nor can have any analogue in the world of commodities. 6 Perception of this fact and its insertion into the analytic engine was mainly the achievement of three men—Petty, Locke, and Cantillon. Its importance warrants an inquiry into the manner in which the ‘discovery’ was made. Neither Petty nor Locke proceeded in the logical way, that is to say, by deducing the phenomenon of velocity from the nature of money—the way adumbrated above. They ran up against it in the course of their attempts to answer a practical question which they thought important. This question was: what is the quantity of money that a given country 5 Still more obsolete was, of course, a similar equation published in 1771 by H. Lloyd in his Essay on the Theory of Money. Of this essay I only know the Italian abstract appended to Verri’s Meditazioni in Custodi’s Scrittori Classici. 6 I have taken space I can ill afford to spare to explain this matter in its most elementary aspect because it is essential for the reader to realize how it presented itself in the beginnings of analysis: every further step leads into mist, ambiguity, difficulty, but that first one is perfectly clear and simple. I take the opportunity to add two points to our knowledge about the equation of exchange. Look again at our example: a definite number of coins settling by payment in specie the commodity transactions of the twelve market days. Now, first, these twelve market days represent a social custom, an institutional arrangement that individuals are powerless to alter. No coin can, under our assumption, have a greater velocity. But what if holders of some of the coins, on any given market day, refuse to spend all the coins they hold? We may then say, of course, that the coins that are not spent on any given market day or are not spent on any of them have a smaller velocity or even the velocity zero. But we may also say…[note unfinished]. History of economic analysis 300 needs? Hume (‘Of Money’ in Political Discourses, 1752) seems to have been the first to show clearly and explicitly that on the level of pure logic this question has no meaning—on the one hand, any quantity of money, however small, will do in an isolated country; on the other hand, with perfect gold currency all round, every country will always tend to have the amount appropriate to its relative position in international trade. But in the sixteenth century people thought differently, and practical sense may in fact be imparted to that question by adding: at the prevailing level of prices. Thus amended, the problem was to determine the requirements of internal circulation under given conditions of time and place with a view either to support up to a point or to combat beyond that point the ‘mercantilist’ policy of enforcing gold and silver imports. The task was primarily of a statistical nature. Petty tackled it from the angle of income payments, that is to say…[unfinished; the two following paragraphs are taken from the brief early treatment on money described in the appendix and hence do not connect exactly with what precedes.] There is one more point. From the standpoint of the theorist it is always a ‘major event’ when an important concept is made explicit and workable, although it was—this is the usual case—implicitly present in previous arguments. The shadow of Velocity of Money may be detected in Davanzati. But it did not acquire substance until the last decades of the seventeenth century. This was a purely English achievement. We know already of Sir William Petty’s exploit in the field. The other sponsor was Locke (in Some Considerations, 1692). He approaches the phenomenon by way of the cash balances which various classes of people are under the practical necessity of holding. The effects of variations in the velocity on prices are not pointed out directly, though they may be said to come in indirectly through the action of the rate of inter-est on idle balances. 7 Cantillon, who, so far as I know, was the first to speak of vitesse de la circulation, was also the first to state in so many words that increase in the velocity of money was equivalent to increase in its quantity. He also drew the conclusion that measures calculated to decrease velocity will counteract the effects of inflation. Neither Hume nor Smith added anything of importance. It will be seen that the concept evolved from the first on both the lines that were followed in its later development. Petty and Locke used the cash balance approach, Cantillon the turnover approach. Locke and Cantillon clearly envisage not only velocity in the strict sense but also the rate of spending. Owing to the prominence that the related concept of Propensity to Consume has gained in connection with the multiplier analysis, it may be interesting to add two examples to show that this concept, too, was perfectly familiar to the economists of that epoch. As we already know, Boisguillebert (Dissertation sur la nature des richesses) pointed out that a coin in the hand of a very small trader is spent much more promptly than a coin in the hand of a rich man who is more likely to shut it up in his coffres—the hoarding rich are evidently no discovery or invention of the last ten years. And Galiani (in the second Dialogue sur le commerce des blés) drew a distinction between the propensity to consume of the farmer, who saves and hoards, and of the artisan who promptly spends (dissipe). 7 On this point as well as on this subject as a whole, see M.W.Holtrop, ‘Theories of the Velocity of Circulation of Money in Earlier Economic Literature,’ Economic History, Supplement to the Economic Journal, January 1929, and Professor Marget’s Theory of Prices, vol. I, passim. Value and money 301 . which is how I came to know of it. I have found no traces of its influence. History of economic analysis 292 [(d) Codification of Value and Price Theory in the Wealth of Nations.] A.Smith’s. understand that the theory of price is but another name for theory of economic logic—including, among other things, all the principles of allocation of resources and of formation of incomes—blamed him. H.J.Davenport, Value and Distribution, 1908. History of economic analysis 294 In Book I, Chapter 5, of the Wealth of Nations occurs the proposition: ‘The real price of everything, what everything really