It is admitted by Marx that separate commodities exchange with each other either over or under their value according as the share of constant capital employed in their production is above or below the average. Stress is, however, laid on the fact that these individual deviations which take place in opposite directions compensate or cancel each other, so that the sum total of all prices paid corresponds exactly with the sum of all values. "In the same proportion in which one part of the commodities is sold above its value another part will be sold under its value" (III, 185).1
The aggregate price of the commodities I to V [in the table given by Marx as an example] would therefore be equal to their aggregate values, and would therefore be, in fact, a money expression of the aggregate amount of labor, both past and recent, contained in the commodities I to V. And in this way in the community itself — when we regard the total of all the branches of production — the sum of the prices of production of the commodities manufactured is equal to the sum of their values. (III, 188)
From this, finally, the argument is more or less clearly deduced that at any rate for the sum of all commodities, or for the community as a whole, the law of value maintains its validity.
Meanwhile it resolves itself into this — that by as much as there is too much surplus value in one commodity there is too little in another, and therefore the deviations from value which lurk in the prices of production reciprocally cancel each other. In capitalist production as a whole "the general law maintains itself as the governing tendency" only in a very complex and approximate manner, as the constantly changing average of perpetual fluctuations. (III, 190)
This argument is not new in Marxian literature. In similar circumstances it was maintained, a few years ago, by Conrad Schmidt, with great emphasis, and perhaps with even greater clearness of principle than now by Marx himself. In his attempt to solve the riddle of the average rate of profit Schmidt also, while he employed a different line of argument from Marx, arrived at the conclusion that separate commodities cannot exchange with each other in proportion to the labor attaching to them. He too was obliged to ask the question whether, in face of this fact, the validity of Marx's law of value could any longer be maintained, and he supported his affirmative opinion on the very argument that has just been given.
I hold the argument to be absolutely untenable. I maintained this at the time against Conrad Schmidt, and I have no occasion today in relation to Marx himself to make any alteration in the reasoning on which I founded my opinion then. I may content myself now with simply repeating it word for word. In opposing Conrad Schmidt, I asked how much or how little of the celebrated law of value remained after so much had practically been given up, and then continued:
That not much remains will be best shown by the efforts which the author makes to prove that, in spite of everything, the law of value maintains its validity. After he has admitted that the actual prices of commodities differ from their values, he remarks that this divergence only relates to those prices obtained by separate commodities, and that it disappears as soon as one considers the sum of all separate commodities, the yearly national produce, and that the total price which is paid for the whole national produce taken together does certainly coincide entirely with the amount of value actually embodied in it. (p. 51)
I do not know whether I shall be able to show sufficiently the bearings of this statement, but I shall at least attempt to indicate them.
What then, we ask, is the chief object of the "law of value"? It is nothing else than the elucidation of the exchange relations of commodities as they actually appear to us. We wish to know, for instance, why a coat should be worth as much in exchange as twenty yards of linen, and ten pounds of tea as much as half a ton of iron, etc. It is plain that Marx himself so conceives the explanatory object of the law of value. There can clearly only be a question of an exchange relation between different separate commodities among each other. As soon, however, as one looks at all commodities as a whole and sums up the prices, one must studiously and of necessity avoid looking at the relations existing inside of this whole. The internal relative differences of price do compensate each other in the sum total. For instance, what the tea is worth more than the iron the iron is worth less than the tea and vice versa. In any case, when we ask for information regarding the exchange of commodities in political economy it is no answer to our question to be told the total price which they bring when taken altogether, any more than if, on asking by how many fewer minutes the winner in a prize race had covered the course than his competitor, we were to be told that all the competitors together had taken twenty-five minutes and thirteen seconds.
The state of the case is this: to the question of the problem of value the followers of Marx reply first with their law of value, that commodities exchange in proportion to the working time incorporated in them. Then they — covertly or openly — revoke this answer in its relation to the domain of the exchange of separate commodities, the one domain in which the problem has any meaning, and maintain it in full force only for the whole aggregate national produce, for a domain therefore in which the problem, being without object, could not have been put at all. As an answer to the strict question of the problem of value the law of value is avowedly contradicted by the facts, and in the only application in which it is not contradicted by them it is no longer an answer to the question which demanded a solution, but could at best only be an answer to some other question.
It is, however, not even an answer to another question; it is no answer at all; it is simple tautology. For, as every economist knows, commodities do eventually exchange with commodities — when one penetrates the disguises due to the use of money. Every commodity which comes into exchange is at one and the same time a commodity and the price of what is given in exchange for it. The aggregate of commodities therefore is identical with the aggregate of the prices paid for them; or, the price of the whole national produce is nothing else than the national produce itself. Under these circumstances, therefore, it is quite true that the total price paid for the entire national produce coincides exactly with the total amount of value or labor incorporated in it. But this tautological declaration denotes no increase of true knowledge, neither does it serve as a special test of the correctness of the alleged law that commodities exchange in proportion to the labor embodied in them. For in this manner one might as well, or rather as unjustly, verify any other law one pleased — the law, for instance, that commodities exchange according to the measure of their specific gravity. For if certainly as a "separate ware" one pound of gold does not exchange with one pound of iron, but with 40,000 pounds of iron; still, the total price paid for one pound of gold and 40,000 pounds of iron taken together is nothing more and nothing less than 40,000 pounds of iron and one pound of gold. The total weight, therefore, of the total price — 40,001 pounds — corresponds exactly to the like total weight of 40,001 pounds incorporated in the whole of the commodities. Is weight consequently the true standard by which the exchange relation of commodities is determined?
I have nothing to omit and nothing to add to this judgment in applying it now to Marx himself, except perhaps that in advancing the argument that has just been under criticism Marx is guilty of an additional error that cannot be charged against Schmidt. For, in the passage just quoted from page 190 of the third volume, Marx seeks, by a general dictum concerning the way in which the law of value operates, to gain approval for the idea that a certain real authority may still be ascribed to it, even if it does not rule in separate cases. After saying that the "deviations" from value, which are found in the prices of production, cancel each other, he adds the remark that "in capitalist production as a whole the general law maintains itself as the governing tendency, for the most part only in a very complex and approximate manner as the constantly changing average of perpetual fluctuations."
Here Marx confounds two very different things: an average of fluctuations, and an average between permanently and fundamentally unequal quantities. He is so far quite right that many a general law holds good solely because an average resulting from constant fluctuations coincides with the rule declared by the law. Every economist knows such laws. Take, for example, the law that prices equal costs of production — that apart from special reasons for inequality there is a tendency for wages in different branches of industry, and for profits of capital in different branches of production, to come to a level, and every economist is inclined to acknowledge these laws as "laws," although perhaps there may be no absolutely exact agreement with them in any single case; and therefore even the power to refer to a mode of action operating on the whole, and on the average, has a strongly captivating influence.
But the case in favor of which Marx uses this captivating reference is of quite a different kind. In the case of prices of production which deviate from the "values," it is not a question of fluctuations, but of necessary and permanent divergences.
Two commodities, A and B, which contain the same amount of labor, but have been produced by capitals of different organic composition, do not fluctuate round the same average point, say, for example, the average of 50 shillings; but each of them assumes permanently a different level of price: for instance, the commodity A, in the production of which little constant capital, demanding but little interest, has been employed, the price level of 40 shillings; and the commodity B, which has much constant capital to pay interest on, the price level of 60 shillings, allowance being made for fluctuation round each of these deviating levels. If we had only to deal with fluctuations round one and the same level, so that the commodity A might stand at one moment at 48 shillings and the commodity B at 52 shillings, and at another moment the case were reversed, and the commodity A stood at 52 shillings and the commodity B only reached 48, then we might indeed say that in the average the price of both of these commodities was the same, and in such a state of things, if it were seen to obtain universally, one might find, in spite of the fluctuations, a verification of the "law" that commodities embodying the same amount of labor exchange on an equal footing.
When, however, of two commodities in which the same amount of labor is incorporated, one permanently and regularly maintains a price of 40 shillings and the other as permanently and regularly the price of 60 shillings, a mathematician may indeed strike an average of 50 shillings between the two; but such an average has an entirely different meaning, or, to be more accurate, has no meaning at all with regard to our law. A mathematical average may always be struck between the most unequal quantities, and when it has once been struck the deviations from it on either side always "mutually cancel each other" according to their amount; by the same amount exactly by which the one exceeds the average the other must of necessity fall short. But it is evident that necessary and permanent differences of prices in commodities of the same cost in labor, but of unequal composition as regards capital, cannot by such playing with "average" and "deviations that cancel each other" be turned into a confirmation of the alleged law of value instead of a refutation. We might just as well try in this way to prove the proposition that animals of all kinds, elephants and May flies included, have the same length of life; for while it is true that elephants live on an average 100 years and May flies only a single day, yet between these two quantities we can strike an average of 50 years. By as much time as the elephants live longer than the flies, the flies live shorter than the elephants. The deviations from this average "mutually cancel each other," and consequently on the whole and on the average the law that all kinds of animals have the same length of life is established!
This article is excerpted from Karl Marx and the Close of His System (1896).
- 1In the editor's introduction, Paul M. Sweezy writes, As editor of these works I have in general confined my efforts to making them more readable and more usable to present-day teachers and students of the social sciences. Style and spelling have been rendered uniform throughout. All quotations from Capital now refer to the Kerr edition, though the wording (except in the case of Hilferding’s quotations from Volume III) remains that of the translators of these works. The translations themselves have been altered in a few places which were unclear, or ambiguous, or dated, by checking back to the German originals. The Kerr edition is an English translation. It is available at econlib.org.
