of wages: laborers produce the whole produce; the wage problem is to show why they do not get the whole produce but have to submit to certain ‘deductions’; hence the wage problem is automatically solved as soon as these deductions have found their explanations. But even for A.Smith himself, and also for James Mill, Sismondi, and Marx, who went further in the direction of Smith’s lead than did any other leading economist, analysis of the upper and lower limits of what ‘can’ or ‘must’ go to labor is so much more important for their treatment of wage problems than is their general philosophy that it is more instructive to deal with them without further reference to the latter. This agrees with what I believe to be the common opinion of a majority of historians. But I cannot agree with the classification of wage theories, which many of them have adopted, into minimum-of-existence theories, supply and demand theories, and productivity theories. For these are not distinct, let alone incompatible, explanations of wage incomes. The first is no theory of wages at all but simply a theorem on the long-run equilibrium level of wages. 37 The supply and demand apparatus is necessary for any wage theory whatsoever and does not identify any particular one. 38 Ricardians (including Marx) failed indeed to recognize this, in the case of wages as in all other cases, for the determination of long-run normals, but even they allowed wages like other prices to be determined by supply and demand in a short run. But with wages, this meant something very different from what it meant with other prices. For if the long-run normal be made dependent upon adjustments of the population, the short run extends over at least fifteen years. 39 For short runs of this and even greater length—indeed for spells of ‘indefinite’ length—the Ricardians relied on the particular form that the supply and demand apparatus takes in the wage-fund doctrine. But in a different (the normal) form, the supply and demand apparatus was also used, for long-run and short-run problems, by all the other leaders, Say and Malthus in particular. Here the demand for labor may be represented by a 37 On the concept of average rate of wages and the objection that has been raised against it, see below, this subsection. As regards the difference between a theory of wages in the sense of fundamental analysis of the phenomenon and an equilibrium theorem about it, note the analogies with the so-called quantity theory of money and with Ricardo’s law of value. 38 There was indeed some opposition to applying the supply and demand apparatus to labor on the ground that it involved treating human beings like commodities—on the Continent, especially, the English ‘classics’ were sometimes indicted for this outrage to human dignity. Nothing of the sort is, of course, involved in this application of the concepts of supply and demand. It should be observed, however, that there was occasionally more in this than a rather cheap emotionalism: the ‘commodity labor’ does present peculiarities that are relevant even for the most matter-of-fact analysis. 39 This was pointed out by Barton (see below, subsec. h). Since an increase in real (in our sense) per capita wage incomes does not increase the birth rate immediately; since it takes a prolonged spell of high per capita wage income to produce a quantitatively significant effect; and, finally, since during so long a time new standards of life develop, the case for the ‘classic’ long-run theory of wages is really much worse than indicated in the text. History of economic analysis 632 schedule that simply describes the quantity of labor taken by employers at varying wage rates. Say’s notion of demand for and supply of services of labor implies this. But a demand schedule of this kind was made explicit and was actually drawn by Fleeming Jenkin. 40 And such a demand schedule, in turn, implies an embryonic marginal productivity theory. The latter, though worked out by Longfield and von Thünen within the period, remained in abeyance, however, so far as the profession at large is concerned. Hence no more need be said about these beginnings of later wage analysis except that the element of productivity, also, must enter into any complete wage theory (in some shape or other) and therefore should not, per se, be identified with any particular one. We have, then, this situation before us: in some form, practically all the economists of the period attacked the wage problem by means of a more or less well-understood supply and demand analysis. 41 The element of productivity, not adequately worked out by those who got a hearing, is visible in the picture but only dimly. The foreground is dominated by two particular results that may be derived from a supply and demand analysis on the insertion of certain additional factual hypotheses (‘restrictions’): the minimum-of- existence theorem as regards longest-run normals and the wage-fund doctrine as regards shorter-run deviations. The minimum-of-existence theorem, as we know, was an essential part of the teaching of Quesnay and of Turgot. It was, as we also know, handled with care by A.Smith—with so much care in fact that there was not a great deal left of it. Malthus’ Essay in its original form put, however, a different face upon the matter, though in the later editions of the Essay and in the Principles we find qualifications that should have produced, but did not produce, recantation. But for Ricardo a strict formulation of the tendency of wages toward ‘that price which is necessary to enable the labourers…to subsist and to perpetuate their race, without either increase or diminution’ (Principles, ch. 5), enforced by an equally strict acceptance of Malthus’ law of population, is really required or else the long-run level of wages becomes indeterminate. The quotation shows that, at least by 1817, 42 Ricardo was 40 Fleeming Jenkin’s papers on ‘Trade Unions’ and ‘The Graphic Representation of the Laws of Supply and Demand and their Application to Labour’ appeared in 1868 and 1870, London School Reprint, 1931. His demand function (x being price, D the quantity bought at that price, and A a constant, takes the form: 41 This also holds for Marx, for the proposition that wages tend to equal the value of labor power, which in turn is identical with the labor embodied in it, implies the play of supply and demand. Of course, Marx’s theory of wages does not consist in this proposition only. On the contrary, it forms an extremely complex whole which covers practically all the aspects of the wage phenomenon and includes careful investigations into the deviation of wages from the level determined by the ‘value’ of labor, especially into the cyclical deviations. This whole must be pieced together from many parts of his writings and cannot be reconstructed here. 42 There is nothing to correspond to the sentence quoted in the Essay on the Influence of a Low Price of Corn on the Profits of Stock (1815). General economics 633 aware of this, but the subsequent argument of his chapter on wages also shows that he knew the necessary theorem to be untenable. Following Torrens, 43 he substituted for the ‘physical minimum’ what it became usual, later on, to call the ‘social minimum of existence,’ which in Torrens’ words means ‘such a quantity of the necessaries and comforts of life, as, from the nature of the climate and the habits of the country, are necessary to support the labourer…’ A little reflection will show that this amounts to accepting customary wages as an institutional datum. This is always possible: anything can be labeled as a datum, which simply means that we give up the hunt for a purely economic explanation of whatever it is we so describe. 44 And it seems more realistic to look upon the ‘classic’ long-run wage ‘theory’ in this light rather than in the light of the physical-minimum theorem, which they disavowed themselves and which is of little importance anyhow, owing to the fact that in the matter of wages a huge ‘short run’ practically replaced the long run. As stated, the supply and demand apparatus actually used by the English ‘classics’ for the treatment of wage problems was of a peculiar kind, traditionally described as the wage-fund doctrine. 45 For the sake of simplicity, we shall neglect the supply of and the demand for labor that play in the income sphere—the supply of and demand for directly consumed services of servants, teachers, and so on—and confine ourselves to the supply of and demand for industrial labor (this, of course, in the widest sense: all labor from the employment of which ‘profit is expected’ as A.Smith put it) as if there were no other employments. Further, we follow ‘classic’ practice in assuming that, at any given point of time, there is always the given supply of labor represented by a given number of laborers: there is no shifting between laborers and self-employed, no variation in the age at which laborers enter or leave the labor market, no change in hours per day or week, and, except for a qualification to be inserted later, no reserve price below which laborers refuse to accept employment. Without doubt, these simplifications, even though not always strictly adhered to, helped to discredit the wage-fund doctrine. But all that matters to us is that they are not more than simplifications that could be dropped without great difficulty. We then have not a supply schedule of labor, but only a given quantity supplied and, so we have assumed, supplied unconditionally for a ‘short run’ of at least fifteen years. Demand is represented in the wage-fund theory in a somewhat unusual manner, namely, by in- 43 Essay on the External Corn Trade…(1815; pp. 58–63). 44 There are limitations to this, of course: if we have an economically determined system and then decide to make a datum of some of its variables, we must drop an equal number of equilibrium conditions or else the system becomes overdetermined. 45 Note, however, (1) that, as we have seen in our discussion of Say’s law, the supply and demand apparatus does not admit of unqualified application to a commodity as important as labor, whose variations in price influence all social aggregates; and (2) that the wage-fund doctrine may be considered as a clumsy attempt to take account of this. History of economic analysis 634 dicating a ‘sum in real terms’ 46 —wage goods, means of subsistence, variable capital 47 — that capitalists have decided to spend on labor. This ‘demand,’ also, is no schedule, at any given moment, but a given quantity. And again—as with workers on the supply side who have no reserve price below which they refuse to go—there is no price of labor beyond which ‘capitalists’ will refuse to go: having decided what to reserve for their own consumption, they cannot, given this decision, spend more than that sum (the wage fund); and, never allowing capital to be idle, they (normally) will not spend less. 48 Since the quantity of labor supplied is given at each moment, since the ‘sum’ to be spent on it is also given—from considerations that are, as it were, behind the scene—at each moment, and since in equilibrium the quantity of labor demanded must be equal to quantity of labor supplied, we have an equation that will uniquely determine a magnitude that is called the average wage rate. 49 If actual wages are fixed above this rate, there is unemployment; if below it, unsatisfied demand for labor. We shall refer to this as the short-run wage-fund theory. But nobody held, of course, that the supply of labor and the wage fund were actually given constants. On the contrary, propositions about their variation over time were not only a part, but the most important part, of the doctrine. The factor that governed the labor supply was either the Malthusian law or else simply the ‘habits’ of the working class. The factor that governed the variation of the wage fund, hence of demand, was 46 This may be understood either in our own sense, namely in the sense of a monetary magnitude corrected by a cost of living index, or in the Ricardian sense of labor embodied in the wage goods. The ‘classics’ meant sometimes the one and sometimes the other. This has given rise to misunderstandings. Stuart Wood (‘A Critique of Wages Theories,’ Annals of the American Academy of Political and Social Science, 1890) accused the ‘classics’ of having held that ‘no assiduity on the part of the laborers, no improvement in production could raise wages,’ i.e. increase the wage fund. Sometimes they did hold this (except that improvements in the production of wage goods will raise profits, hence increase savings and in consequence also the wage fund) but only in the Ricardian sense and not in any sense in which it would be wrong to do so. 47 Always keep in mind that Marx’s variable capital is exactly the same as the ‘bourgeois’ wage fund. 48 Observe: this is an equilibrium proposition, for ‘capitalists’ could spend more or less; only if they did they would not be satisfied with the result, hence not be in equilibrium. But it is easy to understand that, with the ‘classics’’ sloppiness of both thought and exposition, this did not stand out as it should have, even to themselves, let alone to opponents. Note the analogy of this situation with the situation of the quantity theory of money, the more imperfect formulations of which also read as if it were assumed that people must spend, on consumers’ goods or investment, every penny they get hold of. 49 The ‘classics’ were not blind to the problems involved in speaking of an average wage rate; witness their concern with the differences of wage rates in different employments. Nevertheless, in fundamental wage theory they used the concept of average wage rate quite uncritically. In order not to increase our difficulties, we shall do the same, assuming that there is one kind and quality of labor only, equally remunerated in all occupations. It is important to note in defense of this ‘classic’ practice that it does not involve anything that could be called an error. General economics 635 saving. Therefore, given the productive efficiency of the economic process, the course over time of real wage rates (in our sense) and of the per capita real income of the working class depends upon the latter’s rate of propagation and upon the community’s rate of saving. 50 We shall refer to this as the long-run wage-fund theory. We may now combine the argument above with what we have found out in previous sections about the notion of the wage fund. In doing so we shall add or recall the necessary minimum of historical references. The basis of the wage-fund doctrine is the proposition that (industrial) wages are ‘advanced’ from capital. This fundamental proposition goes far back, at least to Cantillon and Quesnay. He who accepts it cannot oppose the wage-fund doctrine, root and branch, however much fault he may find with details, simplifications, or applications. For the long-run wage-fund theory it is equally important that these advances should depend upon saving as their source: this point was driven home by Turgot and A.Smith. 51 We still naturally credit Malthus with contributing to the long-run wage-fund theory ‘his’ law of population, but it is only haziness about what this theory really says that has induced some historians to list him as a wage-fund theorist in any other respect. 52 50 The short-run wage-fund theory, so we have seen, is really no supply and demand theory at all in the usual sense in which this means operating with supply and demand schedules. But the long-run analysis above might be couched in terms of such schedules. I shall only indicate how this could be done: labor supply, by virtue of the Malthusian law, could be represented as a function of real (in our sense) wage rates; the problem is to represent the quantities of labor that ‘capitalists’ demand, also as a function of real wage rates. Since, at any moment, these wage rates depend on the size of the wage fund, since this wage fund’s variations are governed by the rate of saving, since, given everybody’s propensity to save (Mill’s ‘effectual desire of accumulation’), savings depend (mainly) on ‘capitalists’’ incomes, hence on ‘profits’; and since according to Ricardo, profits depend upon wages…and so on. Not that I think much of this construction. But it has two virtues. First, it brings out an aspect of the wage-fund theory—the dependence of future wages on present profits—which is as important as it is apt to be unpopular and which was no doubt very much in the minds of the wage-fund theorists. Second, it clears up a matter that might bother the careful student. The wage- fund theory has sometimes been rendered by saying that, according to it, the elasticity of expenditure on labor with respect to wage rates is zero (elasticity of demand for labor equal to one). This statement is not felicitous. In the long run it is not true, and in the short run it is misleading. 51 A.Smith was the first, I think, to speak of ‘funds destined for the maintenance of labour.’ This phrase was copied by many sponsors of the wage-fund doctrine and gave much offense to its opponents, since it seems to beg the question. J.S.Mill’s recantation (described below) issued in the statement that there is no such fund that is ‘destined’ once for all for the maintenance of labor. But if the objection was that the ‘classics’ simply postulated the existence of this fund without investigating how it was determined, then there was no point to the objection. For funds are, in the ‘classic’ theory, ‘destined’ for the maintenance of productive labor by the decision of the saver and thus are determined if annual savings are. 52 Thus, Malthus’ statement in the Essay on Population, that a poor man, if he receive an additional sum of money, while total output of the country remains the History of economic analysis 636 Ricardo, in his chapter on wages, emphasized strongly enough that it is increase in capital that carries the market rate of wages above the natural rate ‘for an indefinite period.’ In as much as in this chapter he defined capital so as to include ‘food, clothing, raw materials, etc.’ he may be credited with having introduced another element that is, according to our interpretation, characteristic of both the short-run and the long-run wage-fund doctrine, namely, the assumption that we can treat the ratio between wage and non-wage capital as constant on the understanding that its variation is to be treated apart. 53 Thus, though adding edge here and there, he did not in this respect really go beyond A.Smith. But he did something else. He infected his followers with the Ricardian Vice, that is, with the habit of establishing simple relations between aggregates that then acquire a spurious halo of causal importance, whereas all the really important (and, unfortunately, complicated) things are being bundled away in or behind these aggregates. Thus, James Mill declared, as Mrs. Marcet had done before both him and Ricardo: ‘Universally, then, we may affirm, other things remaining the same, that if the ratio which capital and population bear to one another remains the same, wages will remain the same’ (Elements, ch. 2, 2). 54 Had somebody objected that the quantity of labor demanded, hence the wage, may obviously vary even if the sum available for paying wages remain constant—or something of the kind—he would have replied: ‘Oh yes, but we have settled all that behind the scenes just as we have determined that sum previously. As we have shaped our model, there are no other Proximate Causes of the wage rate but that ratio. Everything else acts through them only. For instance, fertility of the soil has nothing to do with the real wages at which labor can be employed. Of course, it affords the means of rapidly accumulating capital and this will same, cannot acquire a larger share in this output without diminishing the share of others, has been held to imply the wage-fund doctrine! That, from first to last, he used the Smithian phrase—funds specifically destined for the maintenance of labor—evidently proves nothing. 53 Of course, we may also interpret him to the effect that he resolved all advances into advances to labor or, which is the same thing, that he resolved the whole of capital into wage capital. J.S.Mill (Principles, Book II, ch. 15, 6) elaborated this explicitly. According to this interpretation, Ricardo would be still more definitely a forerunner of Jevons, Böhm-Bawerk, Taussig, and Wicksell than he is according to our own. But I feel unable to reconcile this with the text of his chapter on wages. In any case, we should have to say—and this is in fact an acceptable compromise—that the larger conception involved in a resolution of non-wage capital into wage capital, while present at the back of his mind, did not influence his analysis of wages but that, even so, Böhm-Bawerk’s aversion to having the ‘classic’ wage-fund theory confused with his own was justified. That Ricardo never co- ordinated the capital theory of his first chapter, Section 4, with the wage-fund theory of his chapter on wages is abundantly clear from the fact that neither he nor his followers ever referred their wage fund to a variable period of time. 54 And this though he had defined capital so as to exclude the ‘subsistence or consumption of the labourer.’ General economics 637 of course raise wages in the future. But, formally, this does not constitute any objection to my theory—going a step further and keeping population constant, we even can say that wages depend upon capital.’ McCulloch then established himself as the leading exponent of the wage-fund doctrine. 55 But he added nothing. Torrens 56 did add something, though this should have been obvious from the first, namely, that the wage-fund theories offer no reason for denying that a combination of the whole working force can raise wages so as to swallow up not only profits but also depreciation allowances. 57 J.S.Mill’s case is quite different. By emphasizing the sequence-analysis aspect of the wage-fund doctrine, he in fact made a point for it and, considering the general level of his technique, there is little to object to in his use of this aggregate that stands, as a sort of intermediate datum, for processes that his technique did not enable him to analyze more satisfactorily. For we must never forget: not only was the wage-fund doctrine, properly stated, not ‘wrong’ logically; not only did it emphasize, though too narrowly, certain important aspects of the wage problem; but, in addition to this, it was an analytic tool that, within the analytic structure of its time, was distinctly useful, and there is no sense in criticizing it in abstracto, that is, without reference to the general-value theory of its time. Nor was there any need for fighting it except in one way: by proffering better tools and letting this one peacefully rust away. All the more surprising must seem J.S.Mill’s ‘recantation.’ He paid no attention to the attacks, if he knew of them, by Jones and Longe. 58 But in response to the elaborate restatement of the latter’s arguments by William Thornton, he wrote a review article that did not indeed spell complete surrender 59 and, in particular, did not induce him to change any of those pas- 55 His first statement of it, never substantially altered, occurs in his Encyclopaedia Britannica article ‘Political Economy’ (1823), his second in the Principles (1825). His Essay on…Wages was published in 1826, and an enlarged edition under the title of Treatise on Wages in 1854. 56 On Wages and Combinations (1834). I do not stay to notice Senior’s insignificant contribution. 57 Observe, however, that James Mill could have replied: ‘Oh no. It is not the combination that effects this but the temporary increase it enforces in the funds destined for the maintenance of labor: it is only by influencing these that the combination has any effect at all.’ It is perhaps unnecessary to point to modern argument of the same nature. 58 Richard Jones (Literary Remains, publ. 1859) accepted the wage-fund doctrine fully for wages that are being paid out by modern business, but he denied, even then, that this was the only important case. This ‘historical’ objection received little credit at the time but all the more later on when opposition to the English ‘classics’ came to be in itself a title to praise. F.D.Longe, A Refutation of the Wage-Fund Theory… (1866; reprint in Professor Hollander’s series, 1904). Other contemporaneous attacks do not present any additional points of interest. 59 William T.Thornton, On Labour…(1869). The word Restatement is to denote a fact but is not to insinuate a charge of plagiarism, though Longe did complain that neither Thornton nor Mill mentioned him—inferring (the optimist!) from his History of economic analysis 638 sages—for example, the Fourth Proposition on Capital—that ought to go out if the wage fund does. But it surrendered a phrase, which was all the public took in. The Longe- Thornton argument, as accepted by J.S.Mill, did not amount to more 60 than denying that there exists any definite quantity of wage goods 61 that ‘must’ under any circumstances go to labor. If, after all that has been explained above, we are not to dismiss this argument as a childish misunderstanding—which is how it appeared to Cairnes 62 —we must interpret it to mean that there is little point in inserting aggregate wages as a ‘proximate cause’ that, as such, plays a role of its own. But if that was all, why all the fuss about this point of theoretical detail and all the excitement about Mill’s alleged recantation? Well, it was not all so far as the public—even the professional public—was concerned. A thing had happened that happens so frequently in our field. The public had caught hold of the surface meaning of a word and this was all it was interested in. Fund—how definite that sounds! Labor must get it and cannot ever get more! Popular writers, if of a certain color, made this mean that raising wages is ‘scientifically impossible.’ Popular writers, if of another color, foamed with indignation at so vile an attempt to thwart labor’s hopes. How absurd all this was should be obvious. 63 No less obvious should it be having sent his pamphlet to Mill that the latter had read it. Moreover, though Longe anticipated the substance of Thornton’s criticism of the wage fund theory, the latter’s book contained several points that were new. Outstanding among them is his emphasis on expected consumers’ demand as the true guide of the producers. In view of the importance the element of expectations has gained of late, Thornton’s book must be allocated a place in the history of analysis that is quite independent of the particular wage-fund issue. J.S.Mill’s review article appeared in the Fortnightly Review, May 1869, and was, besides being a remarkable display of good feeling in the face of what other men would have taken as a provocation, a gentle correction of superficial misunderstandings rather than a retraction. 60 Longe and Thornton brought forth also other criticisms than the one we are going to notice. They criticized, e.g., the concept of an average rate of wages and the way in which the wage-fund theory handled the supply and demand apparatus. There is something in these other criticisms but they could all be met without giving up the theory itself and do not go to the bottom of the issue. 61 Both Longe and Thornton blur the issue—and incidentally betray inadequate grasp of the ‘classic’ analysis—by speaking of ‘money’ without making sure that this money stands for physical goods. This is even more true of H.D.Macleod’s argument in Elements of Political Economy (1858; 3rd ed., Elements of Economics, 2 vols., 1881–6). 62 Cairnes, Leading Principles, Part II, ch. 1, especially pp. 214 et seq. But (p. 186) he interpreted the wage fund in a way that left little to defend. This does not apply, however, to his views on Thornton’s attack upon Supply and Demand. 63 But let me repeat: first, that the wage-fund theory implies nothing of the sort; second, that, if it did imply it, it would still be irrelevant to 99 per cent of all wage struggles, in which claims to higher wages are rationalized by arguments that have nothing to do with any equilibrium rate of wages, but contend that, for reasons such as friction or weak bargaining power, workers do not get these equilibrium wages. General economics 639 that the ‘practical’ diagnosis behind most of the wage-fund theorizing, even if coarsened for the benefit of the public, is not more than common sense: it makes (real) wage rates and (real) wage income dependent upon the efficiency of the productive process, ‘habits’ (high or low customary standard of living and, in connection with this, rate of propagation), free trade in food and other necessaries, and the rate of saving—all of which was no doubt tailored to the prevailing English situation but on the whole quite reasonable. 64 If the importance of changes in money wage rates was discounted—as it is being discounted in Keynesian economics—this was but an additional merit. And so were such warnings as may be found in it against irresponsible ‘wage policies.’ J.S.Mill did not renounce any warnings he had uttered himself. Still there it was—the first English economist of the age had disavowed the hateful scarecrow. But that emotionalism and the not less absurd belief that ‘theories’ guide policy 65 lent zest and glamour to what on its merits should have been a dry-as-dust discussion on a technical point. This had its repercussions on scientific literature. In England and the United States, killing the wage-fund ‘theory’ became a favorite sport: the names of F.A.Walker and H.Sidgwick suffice to illustrate this. On the Continent, especially in Germany, Hermann’s opinion—quite all right in itself, but a mistake if framed as an objection—prevailed on the whole: though Rau’s textbook (in its 8th ed., 1868) upheld the wage-fund doctrine as it did other heirlooms, Roscher (1854) followed Hermann as did Roesler in his fairly influential history of wage theories and L. Brentano. 66 (g) Rent. Whereas the so-called wage theories that were then current did not invoke different explanatory principles but were nothing but more or less valuable parts of a more comprehensive theory of ‘wages and capital’ that failed to mature, the period’s explanations of the rent of land (generalized into the rent of natural agents) really were different theories based upon different principles. We shall refer to them as the monopoly theory, the productivity theory, and the diminishing-return theory. This is not to deny the presence of a unifying principle. Ricardo himself began his discussion of the subject by defining rent as ‘that portion of the produce of the earth which is paid to the landlord for the use of the original and indestructible powers of the soil’ 67 (Principles, ch. 2), and J.S.Mill started his by recalling the triad 64 The only element in this that is really inacceptable to the modern radical is the relation between wages and saving. 65 In the pages of the American Economic Review there has been an interesting discussion concerning the actual influence of the wage-fund theory on popular thought and political action. 66 C.F.H.Roesler, Zur Kritik der Lehre vom Arbeitslohn (1861). Lujo Brentano, ‘Die Lehre von den Lohnsteigerungen,’ Jahrbücher für Nationalökonomie (1871). 67 Let us briefly note that the writers of the period still struggled with the problem of identifying the phenomenon to be explained. A.Smith, as Ricardo noticed, had been vague on the subject and had not always clearly distinguished between pure rent and the total income from the ownership of land, which also includes the return from History of economic analysis 640 of the requisites of production, which comes to the same thing (Principles, Book II, ch. 16). This points toward supply and demand—the principle that not only unifies those three theories but also assimilates rent with all other kinds of incomes generated by a stationary business process. But the great majority of economists did not take this route so that it is historically more realistic to speak of three distinct theories after all. The monopoly theory, espoused by A.Smith, 68 counted adherents, then as always, among politicians and pamphleteers. But its role in the scientific literature was not nearly so important as it seems to be, at first sight, owing to the frequent occurrences of the term monopoly in this connection. The examples of Senior and J.S.Mill will suffice to show this: when we analyze their use of the term, we discover immediately that they did not mean to assert that landowners formed cartels and that the services of land were priced— as a general rule; there is of course the unique mine or vineyard—according to the rules of monopoly theory. All they meant was that rent constitutes a case of pricing ‘costless’ things that exist in definitely limited quantities, which their defective theory of price led them to identify with the genuine monopoly case. J.S.Mill even wrote of a ‘monopolized’ thing among the holders of which there is ‘competition’ (Book II, ch. 16, 2); and both Mill and Senior really adopted, though quite illogically, the diminishing-return theory to be discussed presently. The reader may well ask whether anyone actually held a monopoly theory of rent that was more than an agitatorial phrase, in view of the fact that monopoly that may be present in any case of pricing is constitutionally incapable of explaining the nature of a return. The writer who, so far as I have been able to make out, came nearest to doing so was T.P. Thompson. 69 improvements, such as drainage, fencing, and the like—what Marshall was to call quasi-rent. The distinction, clearly indicated by A.Smith, established itself quickly however. (Thünen called the total income from land Gutsrente and the pure rent Grundrente.) Another question arose concerning exhaustible natural agents such as mines, return from which Ricardo’s definition excludes. But the similarity of the two cases being easily recognized, no trouble arose about that (Principles, ch. 3). The short-run similarity between rent in this sense and returns from any appliance, the quantity of which cannot be changed within the time span that constitutes the short run was, however, not clearly seen before Marshall, and this did entail some consequences of importance: whoever sees this similarity and, hence, that in the short run there is no difference between rent and quasi-rent, is bound, sooner or later, to ask himself whether the yield of physical capital goods is really the same thing as interest. 68 But after having interpreted the rent of land as a monopoly gain, A.Smith declared that rent ‘enters into the composition of the price of commodities in a different way from wages and profits. High or low wages and profit are the cause of high or low price; high or low rent is the effect of it’ (Wealth, Book I, ch. 11). He does not seem to have observed that this contradicts his monopoly theory of rent for, if rent were a monopoly gain, it would enter into price. However, this blundering sentence may have given a clue to Ricardo, with whose analysis it agrees much better than with Smith’s. 69 Thomas Perronet Thompson, The True Theory of Rent (1826), an anti-corn-law pamphlet. I wish I had space to say something on this vital and most interesting man— General economics 641 . guide of the producers. In view of the importance the element of expectations has gained of late, Thornton’s book must be allocated a place in the history of analysis that is quite independent of. rent and the total income from the ownership of land, which also includes the return from History of economic analysis 640 of the requisites of production, which comes to the same thing. many parts of his writings and cannot be reconstructed here. 42 There is nothing to correspond to the sentence quoted in the Essay on the Influence of a Low Price of Corn on the Profits of Stock