² The classical cost theory is a strange and interesting mixture of theories based on two different assumptions as to the character of economic life. The pain-cost theory would fit a society in which productive resources are completely immobile or specialized as between different money income yielding uses, but in which each resource is employed in its income yielding use at the “cost” of sacrificing some “non-pecuniary” employment. The most familiar case, and the only one recognized at all clearly by the classical economists, is that of labour and “irksomeness”. Irksomeness reflects the possibility of using one's personal powers for some purpose other than that of creating exchange value. The same principles apply to all, or virtually all, resources. (Regarding the classical doctrine of abstinence as the pain cost involved in the use of capital, see below, p. 23.) On the other hand, there is a large bulk of discussion of valuation more or less explicitly in terms of the mobility of resources between industries producing different commodities, and value determination through the equalization of yield of resources. Both views assume, though in different contexts, and unconsciously, that value is determined by utility, which varies inversely with supply. But, in general, the argument runs in terms of equalization of reward for equal sacrifice rather than of yield on resources. Since in either case the “sacrifices” are generally made by different individuals, this doctrine is subject to sweeping limitations, and Mill even pointed out (p. 388) that, in fact, the values produced by labourers at different wage levels about as typically bear an inverse ratio as a direct one to the presumptive sacrifice (see next footnote).

³ In practice the individual makes a two-fold apportionment of his productive resources: first, between money-earning uses and direct uses, and second, as regards the part used for money-making, among different industries or occupations. In consequence of specialization, especially the division of labour, an individual does not do much apportioning under the second head, and, in fact, the first is rather rigidly determined by circumstances. These and related facts—especially the fact that the principle and motive of economy applies in various degrees to different behaviour, with no definite boundaries—set limits to the clarity and accuracy with which it is possible to formulate cost theory. This argument is briefly developed in the writer's paper, “The Common-Sense of Political Economy: Wicksteed Reprinted” (Journal of Political Economy, October, 1934), and more fully in two articles in the Zeitschrift für Nationalökonomie, the first in the issue for January, 1935.

⁴ It is perhaps more than a fancy to find a source as well as a picture of these two basic misconceptions in the classical system in the “curse of Adam”: “In the sweat of thy face shalt thou eat bread” (Genesis 3:19). This suggests that cost is pain rather than the sacrifice of an alternative, and again, bread is a false and misleading illustration of the nature of economic value or production. Because it has practically no durability, it invites the confusion of consumption of services and consumption of things rendering services.

⁵ As to the meaning of wealth and its increase, and the other problems in connection with the second aberration and the third, see below, part II.

⁶ Political Economy (octavo ed.), p. 12.

⁷ Op. cit., p. 24.

⁸ There is one remarkable partial exception to this general statement in the Ricardian theory of rent which involves a “rudimentary” recognition of the notion of increments, and of imputation. See below, part III.

⁹ In this connection, see the remarkable statement by Ricardo in a letter to McCulloch: “After all, the great questions of Rent, Wages and Profits must be explained by the proportions in which the whole produce is divided between landlords, capitalists and labourers, and which are not essentially connected with the doctrine of value” (Hollander ed., letter XV, p. 72).

¹⁰ Let it be observed here, once for all, that general negative statements are to be taken subject to interpretation. They mean that the “system” expounded by the writers does not take account of the facts in question. To say that they absolutely failed to see anything essential to a sound theory is generally too strong. One of the mysteries of economic theory is the fact that most of the essentials are things of which no fairly intelligsnt adult living in a world where exchange relations are at all developed could possibly be ignorant. The problem is to build these unescapable facts into a coherent system of relations. Such general negative statements are in no wise disproved by citing particular passages in which the fact or principle in question is, or seems to be, recognized. But it is sometimes a matter of interpretation whether they are taken into account where they need to be. In general, our discussion deals with the chapters treating systematically the problems under discussion, and passages from other chapters are doubtfully, if at all, in point.

References to authors by pages are to the following works, unless otherwise stated:

Smith, Wealth of Nations (Cannan ed.), 2 vols.

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

Senior, Political Economy (octavo ed.).

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

¹¹ The classical writers had no theory of consumption, and Mill even excluded it from the province of the economist. See Essays on Some Unsettled Questions: Essay V, “On the Definition of Political Economy, etc.”, p. 132, footnote. They took it for granted that it is wealth which is consumed as well as produced and that wages are paid out of past, not current, product.

In modern discussion we have, of course, the “theory of diminishing utility”, which was really implicit in the classical discussion, and was also given clear statement by Senior, although it was not explicitly built into his version of the classical system.

¹² The necessity of providing for maintenance of capital before counting any result of its use as product was seen, if not understood, by the classical writers (cf. above, p. 7, note 10). In Smith it appears in the confused discussion of “capital” and “revenue” in book II, chapter iii (cf. also book I, chapter viii, on revenue and stock, p. 71). Ricardo refers to it more explicitly several times, and in his first edition (op. cit., pp. 32-3) he even set up an annuity to provide for replacement. As to Senior, there are few more puzzling passages in all the literature than his argument as to how to classify the replacement fund (pp. 94-5). Strange to say, Mill seems hardly to refer to the point directly, though he mentions capital as a perpetual fund (p. 72) and discusses its “perpetual consumption and reproduction” (pp. 74-5) and his definition of production (productive labour) restricts this to additions to total wealth (p. 49).

The maintenance of social capital was of the profoundest concern to all the writers and this interest underlies the doctrine of productive and unproductive labour. This, however, involves the fallacy of failing to impute product. The labourer who does not directly produce capital produces whatever he does produce, which is presumably worth his wages to whoever pays them, and worth at least as much as any other possible product. Failure to maintain capital means non-productivity of capital, not of labour. The reasons for not deducting the maintenance of the labour from apparent gross product will be considered presently.

¹³ Treating capital growth as income involves a serious theoretical confusion, for it represents double counting (as Professor Fisher in particular has emphasized and as we have indicated). The wealth as a quantity is a capitalized service value, the “present worth” of a future income stream; logically its production is the production of this income stream. Yet it is unquestionably necessary in practice to treat an addition to wealth made in any interval as product and the activity as production in that interval, and to treat the use of the wealth to produce consumed service as production in later time. This is especially because the wealth rarely, if ever, in an organized economy, either yields its return or maintains itself automatically, but requires constant management. The student must simply be careful to recognize at all times that such product and income do not represent consumption, and reflect accounting convenience rather than economic reality.

¹⁴ The measurement of service income is subject to limitations of a theoretically serious nature, and accurate measurement is impossible practically. In fact, the very notion of a total of value, in connection with either service flow or capacity to render service, is subject to serious conceptual limitations, value being relative. But these difficulties cannot be gone into here. It is necessary to distinguish analytically between larger and smaller (perpetual) service income streams, whatever the margin of error in objective discrimination may be. The character of the problem may be suggested by noting that what is involved is the definition of money of unchanging purchasing power. (This does not necessarily imply “ideal” money in an ethical or practical sense.)

¹⁵ See above, p. 4.

¹⁶ (1) Discounting Method: Total Cost = Present Worth of Yield

Let: Cost of Capital good be $S per year for c years, and its yield be $R per year for l years

For continuous compounding

Amount of $1 for n years at i% = eⁱⁿ

Then eⁱⁿ replaces (1+i) ⁿ

Formula becomes: Se^il (e^ic −1) =R(e^il −1).

¹⁷ This is not the place for any extended discussion of interest theory; but it is in order to note that it is not properly a part of distribution theory at all. The special problem of interest is that of the rate, which is entirely a matter of the valuation of the wealth item, or its quantification as wealth. The yield, either in perpetuity or as a known duration (and, more generally, the shape) of income stream, is a datum; in other words, the distribution problem must be solved before the interest problem can be attacked. We shall later have occasion to notice the peculiar fallacy of the classical theory, still generally persistent to-day, that rent and interest are different distributive “shares” arising from different sources or in connection with different economic functions.

The equality of the two expressions suggested above constitutes an explanation of the interest rate (and capital quantity) in the conditions obtaining in a capital market at a moment of time, assuming freedom of investment, perfect competition, and full knowledge of all circumstances and conditions affecting the prospective yield of investments to the infinite future, on the part of all parties, interested in capital transactions. In such a market, the rate of interest is simply the demand price for capital, set by the productivity of the increment of investment about to be made. Further discussion of the problem of interest would deal with the possible changes in demand and supply conditions as investment goes forward. That analysis cannot, however, be given the form of demand and supply curves or functions in the ordinary sense, because the changes have a time dimension, which the curves for demand and supply of a commodity do not have. The situation at a moment can be represented by supply and demand curves, but they have practically no meaning; for, as further analysis would indicate, the demand for capital is indefinitely (at a moment, infinitely) elastic, and the supply, even over time, is practically inelastic—if there is any functional relation between saving and the interest rate, which is questionable. For a brief discussion of the matter, see the article, “Interest”, in The Encyclopaedia of the Social Sciences. See also the writer's article, “Capital, Time and the Interest Rate”, in Economica, August, 1934. On “abstinence” as subjective or pain cost in relation to interest, see below pp. 23ff.

In viewing interest as the marginal productivity of capital, as above, it must be recognized that capital quantity itself involves the interest rate. But the rate can be derived very simply in terms of perpetual income without reference to capital quantity. Whether income is paid continuously (if that is conceivable) or in instalments on any plan, the rate of interest, per cent, per annum, is the ratio of the additional perpetual income creatable by investing a one-year segment of any given perpetual income to the given income itself. Of course, both incomes must be realized on the same plan in relation to accrual, and investment made instantly (presumably at the highest rate which available investment opportunities offer).

¹⁸ It is true that at various points in the classical writings (not to mention modern expositors) we find explicit or implicit reference to a supply function for labour, and a cost of production. Also, that in recent history efforts have been made (cf. Fisher, Huebner, and many others) to define and measure the capital value of human beings. These notions cannot be given critical examination here. Our conclusion must be that it is “impossible” to separate in the expenditure upon human beings the elements which represent respectively (a) want-satisfaction or consumption proper, (b) maintenance, and (c) growth of productive capacity, hence that it is “necessary” to treat all such expenditure as representing current final consumption.

The question whether the creation of human productive capacity is to be treated as “production” is explicitly discussed only by Mill of the four writers considered here, and he takes both positions (cf. pp. 46 and 48 vs. 51-2). The kinship of labour capacity with capital is referred to by Smith in various connections (e.g., vol. I, pp. 103, 123). In his general analysis of capital (book II, ch. i, pp. 264-5) Smith explicitly includes in fixed capital “the acquired and useful abilities of all the inhabitants or members of the society”. But these “abtiliies” are certainly not included in the “value” product which distinguishes productive labour in the definition with which chapter iii of book II begins (p. 313). Senior makes much of the relation, analysing nominal wages into “all three” types of return: wages proper (payment for direct sacrifice), profit (payment for training acquired at a cost), and rent (costless surplus) (see pp. 128ff., also pp. 61, 69). Ricardo gives no discussion of production—in classical terms productive vs. unproductive labour—but his allusions to the topic show that he accepts Smith's position (see especially chaps, vii and xx).

¹⁹ As to whether land “produces” or not, the writers take contradictory positions. The crucial point is that they had no clear notions of production or the (symmetrical) relations between agencies in production, because they had none of mechanical causality. See above, part I, on the fifth aberration, and the argument of part III, below, passim.

²⁰ In a physical analysis, as we have seen, labour power is also produced at a cost, and it has to a considerable extent the other properties of capital. The classification of types of productive agents and of productive capacity is a technological problem, and also a sociological and “human” one; but economic theory is concerned only with the fact that there are different kinds, which co-operate in production, and with general characteristics, not with descriptive classification. See below, pp. 22ff.

²¹ It is common usage in economics to refer to phenomena of growth or change in the given general conditions of a stationary economy as “dynamic” and to this general branch of study as “economic dynamics”, in contrast with “economic statics”. This is misleading, because entirely different from the meaning of the terms statics and dynamics in the field of theoretical mechanics, from which they are taken. Dynamics deals with motion, and the working of a stationary economic system involves the analogue of motion. The process of establishing a moving equilibrium under stationary conditions might more reasonably be called economic dynamics. Changes in these given conditions (wants, resources, and technology) are analogous with nothing dealt with in the science of mechanics, and should be designated by some other term, such as “historical” or “evolutionary” economics.

²² This term was not used by the British classical writers, and is said to come from J. B. Say.

²³ See above, pp. 16-7.

²⁴ As regards “natural resources”, the political interest is a further source of confusion. There is no contradiction between complete equality between value and economic cost of all such resources within any “country”, and the greatest difference between countries is the effectiveness with which existing productive capacity can be employed in “producing” these types of “capital goods”.

²⁵ The classical treatment of the entrepreneur function and the corresponding type of income (or lack of such treatment) will call for notice in part III.

²⁶ See Ricardo, preface, first paragraph; Senior, p. 88, heading. Mill's reference to classes is more casual and uncertain in meaning (cf. pp. 353, 417). Smith speaks of the different “ranks” of the people, in his heading to book I, and of “ranks and conditions of men” in the brief “Introduction and Plan” (pp. 4, 2). None of the writers had any notion of a class as scientifically defined by sociologists.

²⁷ See Senior, pp. 59, 89. Note the inconsistency in the definitions of abstinence, one calling it an agent, the other a kind of conduct. The germ of the abstinence idea is visible in Ricardo's repeated references to the time required to bring products to market and to the effects of the durability of capital on product value. In Economica for February, 1933, Mr. Victor Edelberg ingeniously argues that Ricardo expounded a “sound” theory of capital, à la Böhm-Bawerk. The present writer is unable to find in Ricardo's discussion very much either of Böhm-Bawerk's analysis or of a sound one, which to him mean very different things (cf. below, note 29).

²⁸ The paradox is identical with that raised, but not analysed, by Marshall, in meeting Nicholson's objection to consumers' surplus (Principles, p. 127, note). The choice involved in saving is analogous to that of moving from Africa to England in Marshall's illustration. If one lives in Africa on an income of £1,000 per year, spending it rationally, one's total utility and its subjective cost is £1,000 per year, and in England, it is the same. Yet there may obviously be a gain in moving, at a cost, because of the difference in the composition of the utility stream purchasable with the same money. Similarly, there may be a sufficient incentive to save in the privilege of securing in the future an enlarged utility stream at an equally increased cost (though that is not in any important degree the real motive of saving).

The notion of maximizing “total utility” (defined as whatever men do maximize in behaving rationally) will serve to explain the expenditure of a given income among given alternatives of choice. But any attempt to use it to explain movement from one economic situation to another seems to run inevitably into a “double-counting” paradox. For a present act must be explained by a present motive; and psychologically, it seems that the books actually do balance from moment to moment—and not merely “at the margin”, either. If we give up using quantity of utility to explain changes taking place over time, we are released from assuming that consumers' surplus is “real”, and our reasoning harmonizes with the fact that no one in a routine situation experiences the free satisfaction pictured by the theory.

²⁹ The main argument of this part should suffice to make it clear that there is no relation between the quantity of capital used in a society and the length of any “production period” or interval between production and consumption. Real production is simultaneous with consumption, while the time taken to build up the capital of society is its entire period of historical growth. The production period and service life of concrete things have no definite relation to the quantity of capital. Cf. also articles by the writer in Economic Essays in Honour of Gustav Cassel and in Economica for August, 1934, and one scheduled to appear in The Economic Journal for March, 1935; and also, part III of the present essay.