Hostname: page-component-78c5997874-94fs2 Total loading time: 0 Render date: 2024-10-31T10:56:35.720Z Has data issue: false hasContentIssue false

The Ricardian Theory of Production and Distribution1

Published online by Cambridge University Press:  07 November 2014

Frank H. Knight*
Affiliation:
The University of Chicago
Get access

Extract

The “correct” approach to the theory of distribution in economics is by way of a theory of productive organization; in fact, a sound distribution theory is hardly more than a corollary or footnote to an exposition of the mechanism by which resources are apportioned among different uses, and organized in each use, under the forces of price competition. The “sense” in economics as a subject for study is derived from the idea of increasing the efficiency of action (i.e., of “economy of resources”) through mutual rendering of services by individuals. Historically, the development of the science may be imputed to the realization that purchase and sale represent and mediate such an exchange of services; with a few individual exceptions, men seem never to have had this realization before the eighteenth century, but to have regarded it as axiomatic that what anyone gains in a pecuniary transaction, someone else must lose. Under analysis, a socially general effort to profit by exchange relations takes the form of a tendency toward a maximum (viewed as a state of equilibrium for the society as a whole) through the attraction of every increment of service capacity into its “best” use for the mutual advantage of both parties, i.e., the owner of the increment of capacity and the consumer of its service (who gets the service in exchange for that of some increment of capacity owned by him). If it is seen that this urge on the part of individuals as owners of service capacity is the organizing force in all price relations, and in the economic life which prices mediate and reflect, it becomes truistical to remark that the distribution of the product of all service capacity, considered as acting jointly, is on the basis of increments of yield.

Type
Articles
Copyright
Copyright © Canadian Political Science Association 1935

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

Footnotes

1

This is the second and concluding instalment of a paper, the first instalment (parts I and II) of which appeared in the preceding issue of this Journal (February, 1935). As in the previous instalment, references in the text to authors by pages relate to the following works:

Smith, Wealth of Nations (Cannan ed.), 2 vols. Vol. I unless specified.

Ricardo, Principles of Political Economy and Taxation (Gonner ed.).

Senior, Political Economy (octavo ed.).

Mill, Principles of Political Economy (Ashley ed.).

In a few cases I have used decimals in page references to indicate the approximate position of the statement in question on the page, where this seemed desirable.

Some apology seems in order for the absence of references to previous writers. A critique of previous critiques would be another labour equal to the survey itself, and I should not care to undertake it. Acknowledgements are also beyond my ability to state. I have learned from the better known historians of economic thought, but am not conscious of indebtedness to any one in particular for any of the main interpretive ideas (such as they are) presented here. Except for running through a few chapters of Cannan's history of production and distribution theories, I have made no direct use of published work in preparing the essay, which presents the results of an effort through years of teaching to get at “what it is about”, and makes no pretence to scholarship. I should be more interested in making acknowledgements to students who have written suggestive “papers”, but that also is out of the question. For a general check on the argument and aid in connection with references, I have relied upon an able giaduate assistant, Miss Rose Director, who has lately been working over the material intensively in connection with a more specialized study in the history of the theory of capital.

References

2 Services may be rendered “personally” or by granting to another the use of some external thing which one owns. There is no important analytical difference between the two sorts, either in economic theory, in social psychology, or in ethics.

3 All such statements should be accompanied by others emphasizing the considerations which tend to make “nonsense” (instead of sense) out of the study if they are not carefully taken into account: (1) The individual, including productive capacity in personal and property form on the one hand, and knowledge, taste, and judgment as consumer on the other, is taken as an absolute datum. (The severe limitations of this procedure cannot be developed here.) (2) The individual as a datum includes also his comparisons of present and future, and (3) either includes his attitude toward uncertainty (security vs. adventure), which cannot conceivably be formulated accurately, or else we are forced to make the impossible assumption of universal omniscience of the future. (4) It is assumed that individuals behave in a purely individualistic way, without collusion and without antagonism. This assumption, too, is highly unrealistic; in particular, it is theoretically impossible for two individuals in a system simultaneously to act intelligently in a way which at once takes the system as a set of given conditions of action and modifies it in any way, unless the action is fully preconcerted, which, of course, means collusion. It should be noted, too, that practically any aggregate of individuals within an economic system can profit by collusion; the only question is whether the gain will be greater than the cost of “organization”, which is a conception so subtle that analysis in general terms cannot be carried very far. (5) Finally, and perhaps “worst” of all, it is not true that the things really and finally wanted by human beings are typically, or in a major degree, physically measurable or that value-experience is a mathematical function of physically measurable causes or sources. Economic values are really instrumental to values inherent in process rather than result, and in a social context rather than the physical world.

4 See Wealth of Nations, book I, chapter vii, “Of the Natural and Market Price of Commodities”.

5 The question of the nature of cost of production, especially its relation to pain, or sacrifice, was, of course, one of the main sources of controversy and of confusion in the whole body of classical literature. Two questions were really involved: (1) what it is that the money outlays of the producers really pay for, and (2) the meaning of sacrifice. On both points, the classical writers were in general wrong. What any outlay pays for is a service to the person making the payment, and not a pain; and to the person rendering the service a sacrifice which is voluntary, or connected with a choice, is, in any rational view, simply the giving up of one alternative to get another, which must be equal or greater, or the choice would not be made.

Regarding pain as the basis of production and of value, it is interesting to note that even Ricardo clearly recognized (ch. i, sec. 2) that the wage payments themselves do not correspond at all to the pain of the labourer, and Mill emphasized (p. 388) the tendency to an inverse correspondence where the wages go to different labourers in different wage groups. In the face of the unescapable facts, the related dogmas that “labour” is a homogeneous factor, rather than an indefinite plurality of kinds, and that prices must correspond with a meaningless pain cost, forced the writers into absurdity. The culmination in connection with labour was Cairnes's theory of value in terms of the “non-competing groups”. As already noted, Senior's great “contribution” was the interpretation of capital as pain in the form of abstinence; the dimensional confusions in this notion were also pointed out above (see this Journal, February, 1935, pp. 23 ff.).

6 Chapter iv in Ricardo gives the argument corresponding to book I, chapter vii in Smith, under practically the same title, “Natural and Market Price”. In Senior, we have no separate section on the subject, but see his emphatic statements, especially on pages 97, 101 ff. (He uses here the expression “labour and capital”, but the argument as a whole will show that the meaning is not different from that of Ricardo; the heart of the matter is that a fixed amount of capital always “supports” a fixed amount of labour.) In Mill, the equalization process receives relatively little attention in connection with the discussion of value, but is discussed at length in connection with labour and capital (wages and profits) separately (see below).

Mill's discussion in the first section of his chapter on cost of production (pp. 451-2) is thoroughly muddled as to the meaning of cost. He says (p. 452.6): “As a general rule, then, things tend to exchange for one another at such values as will enable each producer to be repaid the cost of production with the ordinary profit; in other words, such as will give to all producers the same rate of profit on their outlay. But in order that the profit may be equal where the outlay, that is, the cost of production, is equal, things must on the average exchange for one another in the ratio of their cost of production: things of which the cost of production is the same, must be of the same value.” This seems to be a reworking of Ricardo's statement (p. 50) to the effect that if a manufacturer should continue in a trade, “it would be only on condition that he should derive from it the usual and general rate of profits on stock; and that could only happen when his commodity sold for a price proportional to the quantity of labour bestowed on its production”.

Ricardo seems at least to make his meaning intelligible, however untenable his doctrine, which can hardly be said of Mill. Ricardo's position, as we must repeatedly emphasize, is that a fixed amount of capital always employs a fixed amount of labour; capital cost and wages paid are identical, and measure “labour cost”.

7 This omission may be connected with Smith's earlier acceptance of the dogma that labour produces all wealth (related to the doctrine of pain cost) with which the chapter as a whole is in outright contradiction. The inferiority of the treatment in the later writers may also be due to their being still more influenced by this dogma.

8 See above, this Journal (February, 1935, pp. 22 ff.).

9 Except the idea of an increment of product associated with an increment of resource, of which there seems to be no clear trace in his thought.

10 If this doctrine meant what it said, that rent does not “enter into” or “form a part of” (Hume) price, it would imply that the aggregate value of the produce would be equal only to the sum of wages and profit, and the source from which rent would come is a mystery. What was meant was, of course, that the money payment for rent, and land use itself, are not causal in relation to price.

The general problem of interpreting Ricardo is too large and difficult to be attacked here, but one or two observations may be worth noting. Ricardo's primary interest in value theory was to make the theory of value “monistic”, and we might say “absolute”, in contrast with what seemed to him the pluralism and circularity (relativity) of Smith's reasoning. The task he set himself was two-fold: to get rid of all use of prices to explain price, and to find some one thing, not itself a price, to which the price of a product must be equal. For fairly simple reasons, he picked quantity of labour. To explain price in this way, he had to explain away the price-determining roles of capital and land (or their remunerations) and to overcome the further difficulty that for common sense the price of labour as paid by the entrepreneur does not correspond to its quantity. Only one aspect of his procedure calls for notice here. He emphatically asserts (ch. i, sec. 2), that he is not concerned to explain the “absolute” value of commodities, but only the “variations in the relative value”, which he says are not affected by the differences in remuneration of different grades of labour. The misuse of words is characteristic; he means that he is concerned with changes in relative value, not with actual relative value at any given time. In other words, his contention is that he is concerned with historical changes and not with equilibrium relations or tendencies, a position characteristic of many modern writers. The difficulty (as with his later emulators) is that a large part of what he says makes sense only in connection with a theory of equilibrium.

If we could view Ricardo's system in a purely historical light, as an explanation of contemporary events, it would, in a general way, fit the facts as to the outstanding changes in England during the half-century prior to the date of his book. The main fact was an increase of several fold in the prices of agricultural relative to manufactured products. And it would be correct to explain this, in the main, and speaking superficially, by the increase in the efficiency of “labour” in manufactures through inventions, and its decrease in agriculture, in consequence of the pushing outward of the margin of cultivation. (See especially Ricardo's Essay on the Influence of a Low Price of Corn on the Profits of Stock.) The latter change was a consequence of the increase of population and the disruption of foreign trade by the Wars. The first was a phase of the Industrial Revolution, which also underlay the growth of population.

11 See especially Senior, pp. 93, 94; Mill, pp. 417-8; in Ricardo and Smith the assumption is clear beyond doubt. We find abundant recognition of lending at interest, but no separation of entrepreneur and capitalist functions. Karl Marx, who in so many respects is more classical than the classicals themselves, had abundant historical justification for calling—i.e., miscalling—the modern economic order “capitalism”. Ricardo and his followers certainly thought of the system as centring around the employment and control of labour by the capitalist. In theory, this is of course “diametrically” wrong. The entrepreneur employs and directs both labour and capital (the latter including land), and labourer and capitalist play the same passive role, over against the active one of the entrepreneur. It is true that entrepreneurship is not completely separable from the function of the capitalist, but neither is it completely separable from that of labour. The superficial observer is typically confused by the ambiguity of the concept of ownership. The “owner” of an enterprise may not “own” any of the “property” employed in it; and further reflection will show that the same item of property may in different senses be owned entirely, or in widely overlapping degrees, by a considerable number of proprietors.

12 The use of the terms surplus and residuum in this technical sense appears later, though all our writers refer to rent as the surplus product of land. See, e.g., Smith, pp. 146, 147; Ricardo, p. 52; Senior, pp. 90, 91; Mill, pp. 427, 472.

13 Ricardo says expressly (p. 59.6) that using less capital is the same thing as using less labour. As far as the duality of the capital-and-labour factor is concerned, we must recognize that if the proportions are assumed to be fixed (which generally seems to be the assumption, in the discussion of rent), or if the use of more capital with a given quantity of labour through the payment of higher wages occurs but does not increase the total product (which is the assumption at some other points), then the grouping of two factors as one would not be illicit. (See next section.) The fallacy of treating labour as a factor would remain, and the treatment of capital in the same way would have to be justified by a clear explanation of capital quantity, which is certainly not found in the classical writers (and can be justified only with complex and fairly serious limitations).

14 His doctrine as a whole undoubtedly requires that only labour be considered productive, though in the preface and in the second paragraph of the chapter on rent, he refers to “the produce of the earth”. Other classical writers make occasional statements more explicitly to the effect that rent is the value of the land-use, or its “product”, which, of course, is a lapse into common sense inconsistent with the general position. (See Smith, p. 344). Senior (p. 136) in one of his characteristic flashes of insight really commits himself to a productivity theory (cf. also p. 181), though he gives no hint of an imputation process. But on page 140 he says that the amount is subject to no general rule, and has neither a maximum nor a minimum.

15 The notion of “dosing” an entire supply of one factor or group of factors on to the given total supply of other factors is as unnecessary as it is unrealistic. Even the postulate of continuity in the proportions of combination is superfluous. The correct notion is one of the increment of total effect secured by adding any concrete agency, or bargaining unit of a “factor”, to an organization as a whole. (The consequences of the actual finite size of the increment of productive capacity is a question to be examined on its merits in view of the facts in any case.) As for the residual theory in general, it practically reduces to saying that the effect of any change in a causal complex is the difference between what it is and zero, and that this fact is a causal explanation.

It may be worth noting that the analysis of distribution in the form of a double dichotomy with both rent and profit as “residual” does not involve the fallacy of two residuals in the same system, which is found in some later expositions, egregiously in that of F. A. Walker. (My own teaching was wrong on this point for years.) But as we shall see, the fallacy of plural residuals does occur in the classical theory of profits and wages, which reduces to the absurdity that each claimant gets what is left after the other is paid.

16 Cf. Young, Allyn A., in Ely's, Outlines of Economics, ed. 4, p. 410 (end of fine print section).Google Scholar

17 In a graph showing total product from a given amount of some fixed factor or factors, as a function of the amount of a variable factor, the proportional law finds expression in the fact that the curve at a given point has a positive slope less than that of a line connecting the point with the origin of co-ordinates, i.e., a tangent at the point cuts the vertical or variable-factor axis. The incremental law is expressed by the fact that the curve of total product bends toward the base line (the fixed-factor axis).

18 Monopoly is, of course, almost always a matter of degree, a fact which has lately attracted a somewhat strange amount of attention. The views of the classical writers, including Ricardo, on monopoly value or price were absurdly confused (and they have no discussion of monopoly in connection with distribution); and of the authors under consideration here, only Ricardo failed explicitly to treat land as a monopoly.

19 He does refer in the chapter on rent (p. 429) to a “portion” of capital applied to agriculture yielding “only the ordinary profits”; and in his chapter on “Distribution as Affected by Exchange” his language is “the least productive instalment of the capital employed on the better lands”. Just what should be read into this is a typical problem of interpretation. It may be argued that the process of dosing and equalization is taken for granted. But should it, or would it, be in such an exposition? The difficulty, as already pointed out, is that practically all the fundamentals of economic analysis are too obvious to be really overlooked by any observer who thinks about the problems at all.

20 There is an obviously accidental exception on page 181. The first sentence on the page says that “the produce of land increases, caeteris paribus, in a diminishing ratio to the increase in the labour employed.…” This is really the incremental principle. But before finishing the next sentence the author quotes H. C. Cary as holding that “the produce increas[es] in a greater ratio than the labour”. Senior also makes a similar accidentally correct statement of the incremental principle (pp. 105-6).

Both Senior and Mill treat capital-and-labour as a single homogeneous factor, referring to it indifferently as capital, or labour, or joining the two words in either order. (I have not noticed a statement by either paralleling that of Ricardo (p. 59.6) saying expressly that the two always vary together.)

21 See final paragraph of introduction to book I, chapter xi; also the change made in the second edition in chapter vi (p. 51, n. 7). Ashley suggests that this was done at the suggestion of Hume (see his abridged edition, p. 46, note). In the chapter on rent, Smith relapses by implication into an “absolute” labour theory of value, like that which Ricardo assumes in his book subsequent to chapter i, where the explicit argument is subject to a much more reasonable interpretation. Smith nowhere really argues for such a labour theory as cause or determinant of value, not even in chapter v. Moreover, his chapter on rent itself runs largely in other terms, and is much more realistic than those of the other writers.

22 Perhaps an exception should be made for two or three sentences relating to economic progress in colonies in the chapter on profits (book I, ch. ix, p. 94.7).

23 The references are chiefly to vineyard land on the one hand, and to competition between tillage and grazing on the other. Similar passages occur in the discussion of natural and market price (book I, ch. vii, pp. 62-3) and in the chapter on rent (book I, ch. xi, p. 156 on vineyard land; pp. 149-50, 150, 153, 157-9 on competing uses). The passage dealing with payment of rent above a “natural rate” (p. 63) is especially significant. The argument as to which particular use “determines” the value in all other uses is, of course, confused.

It might be noted, too, that in at least one place (in book II, see p. 344) Smith takes a productivity view of rent, saying that it “may be considered as the produce of those powers of nature, the use of which the landlord lends to the farmer”.

24 Cairnes's brief parenthetical reference to rent (Leading Principles, pp. 58-9) might well be placed alongside Mill's oft-quoted reference to value theory as a topic on which nothing remained to be said. But as a matter of fact, the “controversy” in question raged around the composition of the “wages-fund”, which is by no means the central conception in the doctrine of wages and capital; and on the other hand, as shown above, the different writers did not really state the same theory of rent, though they were unconscious of the fact.

It will be recalled that it was really after the “downfall” of the older classical school (under the impact of the theory of subjective value) that Cairnes wrote his main treatise; also, that he wrote it for the purpose of rehabilitating the wages-fund theory (after the recantation by Mill), along with the other “leading fallacy” in the classical position, the labour theory of value.

25 On the treatment of capital, it might perhaps be observed first of all that the writers have nothing to say about the yield or income of concrete “capital goods”. This would, of course, tie up with rent theory; it conforms to the sense in which the term rent is used in everyday life, and recognition of the realities here would have gone far toward straightening out the entire system of concepts. (If “rent” is to be used in economics in a more specialized sense, to represent a theoretically peculiar sort of income, it should be the “price-determined” payment for the services of completely specialized productive agencies—which would naturally include a large fraction of wages.) But while the classical writers have much to say in various connections about various capital instruments—machines, buildings, etc.—they neither have anything to say about the earnings of such concrete things nor give any explanation of how they are quantified as abstract capital. The same statement holds for human beings, and also for land; everything is abstract quantity, yet, of course, there is no indication of definitions in quantitative terms. Even the conception of rent as a surplus falling or accruing to the “original and indestructible powers of the soil” treats these powers as a homogeneous fund, though an attribute possessed in different degrees by different examples of “land”.

26 That is, they seemed to themselves to do so; to be accurate, they would still have required a rate of interest to define a quantity of capital. This is true under any conditions in which capital exists at all, and is to be kept in mind as a special reservation whenever we speak of the incremental productivity of capital.

27 Of course, this notion of what is “necessary” to enable the labourer to produce is indefinite at best. As remarked at the outset, the classical writers did not understand the notion of functionally related variation and cannot be expected to recognize a detail like difference in degree. Logically, the line would be drawn at consumption which increased production by more than the product consumed, and so yielded a profit. Such principles would have significance for a slave economy under perfectly rational exploitation, but it is hard to see their relevance to any free society. Needless to say, the writers, being human beings and endowed with common sense, did not stick to their definitions. Even apart from extensive animadversions on machines and other forms of capital which are not products in the sense here in question, and are not fitted into the theory of a rate of return, they always treated as capital whatever is actually paid (“advanced”) to labourers as wages.

In discussing the notion of unproductive consumption, Senior makes an observation (p. 56) which contains much realistic economic insight, and deserves especially to be brought to the attention of social reformers devoted to the equalitarian ideal: “We do not, of course, mean it to be inferred that all personal expenditure beyond mere necessaries is necessarily unproductive. The duties of those who fill the higher ranks in society can seldom be well performed unless they conciliate the respect of the vulgar by a certain display of opulence.” Senior illustrates the point by the magistrate and ambassador, but the extreme paradoxical example is the minister of such a religion as Christianity.

28 Malthus held a diametrically opposite view, which was “refuted” by Senior (p. 64). Cf. also Smith, p. 58.

Exclusion from capital of the share of produce consumed by the landlord (when rent is paid) would be easier to justify, since the landlord was supposed to play a strictly passive role. (In fact, our authors discuss the division between labour and capital only at the land margin.) From the standpoint of a correct analysis, the landlord, of course, furnishes productive services to industry in a sense theoretically indistinguishable from similar activities on the part of the labourer, and is subject to a “sacrifice”, in the same sense, of giving up one valuable alternative to get another. The capitalist, as shown above (part II), is in a somewhat different position, in so far as he not merely furnishes the use of capital, but “abstains” from consuming the capital itself, and perhaps saves it in the first place. But, of course, where land is private property in the complete sense, the individual land owner is in identically the same position as the owner of any other salable productive agency (ignoring the deeper similarities between land and other agencies). But, from the standpoint of a really correct analysis, all of the contributors to production receive their remuneration instantly out of current production and not out of “advances” or “capital”.

Some writers, but not in general those here being considered, identified the rate of return with the fraction of the produce accruing to the capitalist. The rent share is, of course, ignored, or eliminated by thinking of production at a land margin, but the statement is still inaccurate, unless the capitalist's own share is considered as capital. But the statement is approximately true since the profit share may be assumed to be small in comparison with that of wages.

29 The expositions of Senior (pp. 51-7) and of Mill (book I, ch. iii, sees. 5-6) are some-what easier to follow. Smith has a different arrangement of the material, but except for the “refinement”, his wages and profit docttine is that of the later writers. His main discussion of the phenomena of capital is segregated from that of price and distribution theory in a short second book (book II, “Of the Nature, Accumulation and Employment of Stock”), in chapter iii of which the doctrine in question is developed. His chapter on profits in book I is brief and says little about general theory. (Its main theoretical point will be noticed later.)

30 Fixing them, of course, at the economic value of the service to the entrepreneur, which is their value to the consumer, under the given conditions. But no ethical implication attaches to the notion of the economic value of a service as a remuneration for the service, without a special examination of the whole problem from an ethical standpoint. All competitive values depend, as a matter of course, on the distribution of ownership of productive capacity (human and non-human) and on the knowledge and tastes of its owners as they affect the supply of services for various uses and the demand for various products. Persons who do not own some productive capacity in some form are outside the economic system altogether. See also, above, p. 172, n. 3.

31 With the interesting exception, that since the workers are not actually slaves by inheritance, there is no reason why the individual employer should provide the worker with maintenance for a family. But this might be explained by assuming that the employer cannot prevent the worker from having a family and sharing his wages with them, and so must support the family to get the services of the worker. From a social-ethical standpoint, it is an interesting feature of the system that provision for all contingencies in the lives of the workers and their families becomes a responsibility of the workers individually, and one which they cannot adequately meet. But all “individualistic” societies alleviate some of the worst consequences of this inadequacy through various organizations for relief.

32 Senior and Mill make similar appeals, but Ricardo's preaching generally relates to public policy such as taxation, protective duties, and the poor law, rather than to private behaviour. We should not leave the topic without noting that Ricardo and Mill also expressly state that profit must be at a level which will enable the capitalist to live! (Ricardo, p. 100.9; Mill, p. 450.5). Smith also calls profit the capitalist's fund of subsistence (pp. 57-8).

33 See below, on the classical long-run subsistence theory based on Malthusian population doctrine. This “Malthusian” doctrine is quite adequately stated by Smith (ch. viii, esp. pp. 81-3), along with characteristically sensible observations on the direction of the causal relation between scale of living and wages at a given time (pp. 77-8).

34 The practical conclusion which motivates the wish-thinking back of this theory—namely, that to raise wages it is only necessary to find some way to “build a fire under” the individual who pays them—illustrates a basic human trait. We seem to crave some “enemy” whom we can blame for anything that seems to be wrong, and can attack, as a method of correcting the evil. It is often implied, if not explicitly stated, that it would only be necessary to eliminate (“liquidate”?) the employer and place “the workers themselves” (meaning the propagandists as their “representatives”) in charge of production in order to pay wages equal to the income which (successful!) employers receive under competition.

35 As already noted, they make it clear that competition compels all employers to pay the same wages, and equalizes the rate of profit among them all. (Generally in a separate discussion, not dealing with the general levels of wages and profits, but cf. Senior, pp. 191.6 ff. Ricardo, in chapter iv, talks constantly about the competition of different uses for capital, but only once uses the word, at p. 68.9.) This, of course, clashes directly with the notion of an arbitrary division between the employer and labour—but in no wise prevents their asserting the latter fact.

Smith repeatedly speaks of increased competition lowering the rate of profit. See pp. 89, 129-30, 335.3, 335.6. At the point of the last reference, he also has the competition raise the wages of labour (cf. also Mill, p. 343). The paragraph in Smith containing the last two references occurs in the chapter on interest in book II. It is especially interesting as giving what is perhaps the fullest discussion to be found in his work of a demand function for a productive service, and as completely failing to give any indication of the real mechanism of competition, based on variable proportions and the significance of increments.

36 One of the puzzling features of the classical exposition is the absence of any clear suggestion that an increased amount of capital, even in the primitive use of raising wages, should within some limits increase the output of given labourers. Smith's remarks at page 88.4 do not invalidate this observation, as it is the increased division of labour which increases the product, and the increase in wages referred to simply raises the question, “why”, discussed in the text. At page 259.5f. there is a parallel statement with no reference to an increase in wages. On p. 84, Smith raises the question whether higher real wages lower efficiency, as the mercantilist writers generally held. Senior's discussion in particular (p. 188 ff.) is put in a form to give him every opportunity and suggestion for mentioning such an effect, but it is conspicuously absent. In order to rationalize a raise in wages from the employers' standpoint, it would seem necessary to assume either that the labour supply responds instantly (see below) or that they deliberately maintain the higher rate for at least a generation in order to make possible an increase in the supply of labourers (which is supposed always to be tending to happen). But such conduct is out of the question under an individualistic system.

In the view of the classical conception of capital, an increase in wages is the same thing as an increase in the proportion of capital to labour. Only Senior of the writers under consideration seems to refer expressly to changes in the proportions of capital and labour as a condition affecting the division between them (pp. 140, 188, 189, 190); but even he makes no mention in this connection of a principle of diminishing returns or the significance of increments. Smith, in his discussion of colonial conditions, in his chapter on profits, mentions the proportion of stock to territory and of people to stock (p. 94). (The classical discussions of changes in the proportions of fixed and circulating capital have possible implications for this topic, and for the theory of capital in general; but the problem is too intricate to take up here. But see below.)

37 The writers generally did not know whether they were talking about quantities or rates of change. When Smith says, at the beginning of the second paragraph of the chapter on profits, that “the increase of stock … raises wages”, he means a “sufficient” rate of increase. The third answer above may sometimes be used to interpret the other two, but often the context makes this very difficult, definitely implying that more labour will always be on hand to consume more capital.

For the purpose of this essay, we must cut short the discussion of the wages fund with a note on its relation to the facts. The writers' objective, as already observed, was to show that wages cannot be raised by arbitrary action, or by any method which does not increase the amount of capital in proportion to the labouring population—or by “inventions”, which they treated in a strangely parenthetical tone. Of course, the application of inventions is generally connected with an increased use of capital, and with this qualification, the conclusion of the older economists was sound, though their reasoning was fallacious. Their major premise was that the employer's share in the product is either negligible or fixed. In recent years, statistical investigation has more and more impressed us with the narrow limits of the possibility of raising wages at the expense of interest and profit. With rent (in accord with any probable estimate) excluded, and allowance made for taxes, gifts for charitable and benevolent purposes, and necessary savings, the total share now falling to capital and entrepreneurship would not raise total wages by any large percentage, and it would have been much smaller a century or two ago. Of course, deductions from the capital share would make more difference if used to raise only the lowest wages. The amount of saving which ought to be provided for is indeed a large political question. As to rent, the modern disposition (with the exception of some Roman Catholic writers) would be to throw the landlord to the wolves first of all. But in Europe, under the feudal tradition, the rent of land was the chief means of payment for public services and for the maintenance of all cultural activities.

Even the distinction between types of product consumed by labourers and by capitalists, suggested by Senior, has vastly more significance for the problem of transfer of income than he recognized; a large part of the difference between large and small incomes would disappear under equalization in relative changes in the prices of “choice cuts” and those less choice, where the proportions in production are relatively inflexible.

From the standpoint of economic theory, the conception of capital as a fund for the support of labour over an interval between production and consumption is still profoundly important. It was reworked by Bohm-Bawerk, who combined elements from Ricardo and Senior to make the amount of capital a matter of the length of the production process, and in his form is still generally accepted and taught. For the fundamental fallacies involved, see above, part II and references, this Journal, vol. I, no. 1, p. 25, note.

38 See Ricardo, , “Essay on the Influence of a Low Price of Corn on the Profits of Stock, etc.”, in Works, p. 372 and second footnoteGoogle Scholar; Senior, p. 188.7; Mill treats agriculture and manufactures together, failing to see that his theory approximately fits only the former (book II, ch. xv).

39 It could at most be approximately correct if the demand for capital in the particular part of the market supposed to control were overwhelmingly large in comparison with all other demand. (It seems legitimate to say that the level of the water in the ocean “determines” the level in some particular bay.) Moreover, Ricardo's argument that a change in the level of profit in agriculture will change the level in manufacturing industry without a movement of capital from one field to the other is rank fallacy (see Essay, in Works, pp. 379-80).

40 They have been broached above in part II. The first requisite which comes to mind is that the physical cycles should not overlap, the capital-creation phase of one coinciding with the capital consumption phase of the next. We may revert again to Böhm-Bawerk, to note that his error was the same as that of the classical writers; that is, it lay not in assimilating the other forms of capital to the wages fund, but in failing to give an accurate analysis fitting any case, under anything like real conditions.

41 In a deeper analysis, the error in the whole classical position (including, of course, Böhm-Bawerk) roots in the special character and role assigned to labour. More generally still, it consists in confusing conceptual analysis with ethical evaluation. From the former standpoint, labour and capital instruments, including land, are all alike, simply productive resources. It is in the contrast between capital as “supporting” and “aiding” labour, and labour as “using” and “reproducing” capital, that the immediate fallacy lies. The relation is symmetrically mutual. All the productive agencies use each other (or are used jointly by the enterprise), and all are jointly maintained and reproduced by all, in a continuing process.