Eugen von Böhm-Bawerk, “The Error in the Marxist System” (1896)

Note: This extract is part of The OLL Reader: An Anthology of the Best of the OLL, the table of contents of which can be found here. It is from "Part X: The Critique of Socialism and Interventionism".

For more information on socialism see:

Eugen von Böhm-Bawerk (1851–1914)


English trans.: Eugen von Böhm-Bawerk, Karl Marx and the close of his system, a criticism. Translated by Alice M. Macdonald with a Preface by James Bonar (London: T. Fisher Unwin, 1898). Revised slightly by Paul M. Sweezy (1949) and David M. Hart (2018).


The text is in the public domain.

The Text

Chapter Four. The Error in the Marxist System—its Origin and Ramifications

Section I

THE evidence that an author has contradicted himself may be a necessary stage, but it cannot be the ultimate aim of a fruitful and well-directed criticism. To be aware that there is a defect in a system, which may possibly be accidental only and peculiar to the author, requires a comparatively low degree of critical intelligence. A firmly rooted system can only be effectually overthrown by discovering with absolute precision the point at which the error made its way into the system and the manner in which it spread and branched itself out. As opponents we ought to study the beginning, the development, and the final issue of the error which culminates in self-contradiction as thoroughly, I might almost say as sympathetically, as we would study the connection of a system with which we were in agreement.

Owing to many peculiar circumstances the question of self-contradiction has, in the case of Marx, gained a more than ordinary importance, and consequently I have devoted a considerable space to it. But in dealing with a thinker so important and influential as Marx it is incumbent upon us to apply ourselves to the second and, in this case I think, the actually more fruitful and instructive part of the criticism.

We will begin with a question which will carry us straight to the main point: in what way did Marx arrive at the fundamental proposition of his teaching—the proposition that all value depends solely upon incorporated quantities of labor?

That this proposition is not a self-evident axiom, needing no proof, is beyond doubt. Value and effort, as I have stated at length in another place, are not ideas so intimately connected that one is forced immediately to adopt the view that effort is the basis of value. "That I have toiled over a thing is one fact, that the thing is worth the toil is another and a different fact, and that the two facts do not always go hand in hand is far too firmly established by experience to admit of any doubt. It is proved by all the labor which is daily wasted on valueless results, owing either to want of technical skill, or to bad speculation, or to simple misfortune; and not less by each of the numerous cases in which a very little toil has a result of very great value."1

When therefore it is affirmed that a necessary and natural correspondence between value and effort exists in any quarter, it behooves us to give ourselves and our readers some grounds in support of such a statement.

Now Marx himself advances proofs of it in his system; but I think I shall be able to convince my readers that from the outset his line of argument is unnatural and not suited to the character of the problem; and further that the evidence which Marx advances in his system is clearly not the same as that by means of which he himself arrives at his convictions, but was thought out subsequently as an artificial support for an opinion which was previously derived from other sources; and finally—and this is the most decisive point—that the reasoning is full of the most obvious faults of logic and method which deprive it of all cogency.

Let us examine this more closely.

The fundamental proposition which Marx puts before his readers is that the exchange value of commodities—for his analysis is directed only to this, not to value in use—finds its origin and its measure in the quantity of labor incorporated in the commodities.

Now it is certain that the exchange values, that is to say the prices of the commodities as well as the quantities of labor which are necessary for their reproduction, are real, external quantities, which on the whole it is quite possible to determine empirically. Obviously, therefore, Marx ought to have turned to experience for the proof of a proposition the correctness or incorrectness of which must be manifested in the facts of experience; or in other words, he should have given a purely empirical proof in support of a proposition adapted to a purely empirical proof. This, however, Marx does not do. And one cannot even say that he heedlessly passes by this possible and certainly proper source of knowledge and conviction. The reasoning of the third volume proves that he was quite aware of the nature of the empirical facts, and that they were opposed to his proposition. He knew that the prices of commodities were not in proportion to the amount of incorporated labor, but to the total cost of production, which comprises other elements besides. He did not therefore accidentally overlook this, the most natural proof of his proposition, but turned away from it with the full consciousness that upon this road no issue favorable to his theory could be obtained.

But there is yet another and perfectly natural way of testing and proving such propositions: the psychological. We can by a combination of induction and deduction, much used in our science, investigate the motives which direct people in carrying on the business of exchange and in determining exchange prices on the one hand, and on the other hand which guide them in their co-operation in production; and from the nature of these motives a typical mode of action may be inferred through which, among other things, it is conceivable that a connection should result between the regularly demanded and accepted prices and the quantity of work necessary for the production of the commodities. This method has often been followed with the best results in exactly similar questions—for instance, the usual justification of the law of supply and demand and of the law of costs of production, and the explanation of ground rents, rest upon it. And Marx himself, in a general way at least, has often made use of it; but precisely in dealing with his fundamental proposition he avoids it. Although, obviously, the affirmed external connection between exchange relations and quantities of work could only be fully understood by the discovery of the psychological links which connect the two, he foregoes all explanation of these internal connections. He even once says, incidentally, that "the deeper analysis" of the two social forces, "demand and supply"—which would have led to this internal connection—"is not apposite here" (III, 223), where the "here" refers only to a digression on the influence of supply and demand on the formation of prices. In reality, however, nowhere in the whole Marxist system is a really "deep" and thorough analysis attempted; and the absence of this analysis is most noticeable where he is preparing the ground for his most important leading idea.

But here again we notice something strange. Marx does not, as might have been expected, pass over this second possible and natural method of investigation with an easy carelessness. He studiously avoids it, and with a full consciousness of what the results of following it would be, and that they would not be favorable to his thesis. In the third volume, for instance, he actually brings forward, under their roughly collective name of "competition," those motives operative in production and exchange, the "deeper analysis" of which he foregoes here and elsewhere, and demonstrates that these motives do not in reality lead to an adjustment of the prices to the quantities of labor incorporated in the commodities, but that, on the contrary, they force them away from this level to a level which implies at least one other co-ordinating factor. Indeed it is competition which, according to Marx, leads to the formation of the celebrated average rate of profit and to the "transfer" of pure labor values into prices of production, which differ from them and contain a portion of average profit.

Now Marx, instead of proving his thesis from experience or from its operant motives—that is, empirically or psychologically—prefers another, and for such a subject somewhat singular line of evidence—the method of a purely logical proof, a dialectic deduction from the very nature of exchange.

Marx had found in old Aristotle the idea that "exchange cannot exist without equality, and equality cannot exist without commensurability" (I, 68). Starting with this idea he expands it. He conceives the exchange of two commodities under the form of an equation, and from this infers that "a common factor of the same amount" must exist in the things exchanged and thereby equated, and then proceeds to search for this common factor to which the two equated things must as exchange values be "reducible" (I, 43).

I should like to remark, in passing, that the first assumption, according to which an "equality" must be manifested in the exchange of two things, appears to me to be very old-fashioned, which would not, however, matter much were it not also very unrealistic. In plain words, it seems to me to be a wrong idea. Where equality and exact equilibrium obtain, no change is likely to occur to disturb the balance. When, therefore, in the case of exchange the matter terminates with a change of ownership of the commodities, it points rather to the existence of some inequality or preponderance which produces the alteration. When composite bodies are brought into close contact with each other new chemical combinations are produced by some of the constituent elements of one body uniting with those of another body, not because they possess an exactly equal degree of chemical affinity, but because they have a stronger affinity with each other than with the other elements of the bodies to which they originally belonged. And as a matter of fact modern political economists agree that the old scholastico-theological theory of "equivalence" in the commodities to be exchanged is untenable. I will not, however, dwell any longer on this point, but will proceed to the critical investigation of the logical and systematic processes of distillation by means of which Marx obtains the sought-for "common factor" in labor.

It is these processes which appear to me to constitute, as I have before said, the most vulnerable point in the Marxist theory. They exhibit as many cardinal errors as there are points in the arguments—of which there are not a few—and they bear evident traces of having been a subtle and artificial afterthought contrived to make a preconceived opinion seem the natural outcome of a prolonged investigation.

Marx searches for the "common factor" which is the characteristic of exchange value in the following way: He passes in review the various properties possessed by the objects made equal in exchange, and according to the method of exclusion separates all those which cannot stand the test, until at last only one property remains, that of being the product of labor. This, therefore, must be the sought-for common property.

This line of procedure is somewhat singular, but not in itself objectionable. It strikes one as strange that instead of submitting the supposed characteristic property to a positive test—as would have been done if either of the other methods studiously avoided by Marx had been employed—Marx tries to convince us that he has found the sought-for property, by a purely negative proof, by showing that it is not any of the other properties. This method can always lead to the desired end if attention and thoroughness are used—that is to say, if extreme care is taken that everything that ought to be included is actually passed through the logical sieve and that no mistake has been made in leaving anything out.

But how does Marx proceed?

From the beginning he only puts into the sieve those exchangeable things which contain the property which he desires finally to sift out as "the common factor," and he leaves all the others outside. He acts as one who urgently desiring to bring a white ball out of an urn takes care to secure this result by putting in white balls only. That is to say he limits from the outset the field of his search for the substance of the exchange value to "commodities," and in doing so he forms a conception with a meaning narrower than the conception of "goods" (though he does not clearly define it), and limits it to products of labor as against gifts of nature. Now it stands to reason that if exchange really means an equalization, which assumes the existence of a "common factor of the same amount," this common factor must be sought and found in every species of goods which is brought into exchange, not only in products of labor but also in gifts of nature, such as the soil, wood in trees, water power, coal beds, stone quarries, petroleum reserves, mineral waters, gold mines, etc.1 To exclude the exchangeable goods which are not products of labor in the search for the common factor which lies at the root of exchange value is, under the circumstances, a great error of method. It is just as though a natural philosopher, desiring to discover a property common to all bodies—weight, for instance—were to sift the properties of a single group of bodies—transparent bodies, for instance—and after passing in review all the properties common to transparent bodies were to declare that transparency must be the cause of weight, for the sole reason that he could demonstrate that it could not be caused by any of the other properties.

The exclusion of the gifts of nature (which would never have entered the head of Aristotle, the father of the idea of equality in exchange) is the less to be justified because many natural gifts, such as the soil, are among the most important objects of property and commerce, and also because it is impossible to affirm that in nature's gifts exchange values are always established arbitrarily and by accident. On the one hand, there are such things as accidental prices among products of labor; and on the other hand the prices in the case of nature's gifts are frequently shown to be distinctly related to antecedent conditions or determining motives. For instance, that the sale price of land is a multiple of its rent calculated on an interest usual in the country of sale is as well-known a fact as that the wood in a tree, or the coal in a pit, brings a higher or lower price according to differences of quality or of distance from market, and not by mere accident.

Marx also takes care to avoid mentioning or explaining the fact that he excludes from his investigation a part of exchangeable goods. In this case, as in many others, he manages to glide with dialectic skill over the difficult points of his argument. He omits to call his readers' attention to the fact that his idea of "commodities" is narrower than that of exchangeable goods as a whole. He very cleverly prepares us for the acceptance of the subsequent limitation of the investigation to commodities by placing at the beginning of his book the apparently harmless general phrase that "the wealth of the society in which a capitalist system of production is dominant appears as an immense collection of commodities." This proposition is quite wrong if we take the term "commodity" to mean products of labor, which is the sense Marx subsequently gives to it. For the gifts of nature, inclusive of the soil, constitute a by no means insignificant, but on the contrary a very important element of national wealth. The ingenuous reader easily overlooks this inaccuracy, however, for of course he does not know that later Marx will give a much more restricted meaning to the term "commodity."

Nor is this made clear in what immediately follows. On the contrary, in the first paragraphs of the first chapter we read in turns of a "thing," a "value in use," a "good," and a "commodity," without any clear distinction being made between the last and the three former. "The usefulness of a thing" it says on page 42, "makes it a value in use"; "the commodity ... is a value in use or good." On page 43 we read, "Exchange value appears ... as the quantitative proportion ... in which values in use of one kind exchange with values in use of another kind." And here let it be noticed that it is just the value in use = good which is still directly indicated as the main factor of the exchange phenomenon. And with the phrase "Let us look into the matter more closely," which surely cannot be meant to prepare us for a leap into another and a narrower field of research, Marx continues, "a single commodity, a quarter of wheat, for instance, exchanges in the most varying proportions with other articles." And "Let us further take two commodities," etc. In the same paragraph the term "things" occurs again, and indeed with the application which is most important for the problem, namely, "that a common factor of equal amount exists in two different things" (which are made equal to each other in exchange).

On the next page (p. 44), however, Marx directs his search for the "common factor" only to the "exchange value of commodities," without hinting, even in the faintest whisper, that he has thereby limited the field of research to a part only of the things possessing exchange value.1 And immediately, on the next page (p. 45), the limitation is again abandoned and the results just obtained in the narrower area are applied to the wider sphere of values in use, or goods. "A value in use, or a good, has therefore only a value because abstract human labor is stored up or materialized in it."

If Marx had not confined his research, at the decisive point, to products of labor, but had sought for the common factor in the exchangeable gifts of nature as well, it would have become obvious that work cannot be the common factor. If he had carried out this limitation quite clearly and openly this gross fallacy of method would inevitably have struck both himself and his readers; and they would have been forced to laugh at the naïve juggle by means of which the property of being a product of labor has been successfully distilled out as the common property of a group from which all exchangeable things which naturally belong to it, and which are not the products of labor, have been first of all eliminated. The trick could only have been performed, as Marx performed it, by gliding unnoticed over the knotty point with a light and quick dialectic. But while I express my sincere admiration of the skill with which Marx managed to present so faulty a mode of procedure in so specious a form, I can of course only maintain that the proceeding itself is altogether erroneous.

But we will proceed. By means of the artifice just described Marx has merely succeeded in convincing us that labor can in fact enter into the competition. And it was only by the artificial narrowing of the sphere that it could even have become one "common" property of this narrow sphere. But by its side other properties could claim to be as common. How now is the exclusion of these other competitors effected? It is effected by two arguments, each of a few words only, but which contain one of the most serious of logical fallacies.

In the first of these Marx excludes all "geometrical, physical, chemical, or other natural properties of the commodities," for "their physical properties only come into consideration in so far as they make the commodities useful—make them values in use, therefore. On the other hand, the exchange relation of commodities evidently involves our disregarding their values in use"; because "within this relation [the exchange relation] one value in use is worth exactly as much as every other, provided only it is present in proper proportions" (I, 44).

In making clear what this argument involves I may be permitted to quote from my Capital and Interest (p. 381):

"What would Marx have said to the following argument? In an opera company there are three celebrated singers, a tenor, a bass, and a baritone, each with a salary of £2,000. Someone asks, 'What is the common circumstance on account of which their salaries are made equal?' And I answer, 'In the question of salary one good voice counts for just as much as any other, a good tenor for as much as a good bass or a good baritone, provided only it is to be had in proper proportion. Consequently in the question of salary the good voice is evidently disregarded, and the good voice cannot be the common cause of the high salary.' That this argument is false, is clear. But it is just as clear that Marx's syllogism, from which this is copied, is not an atom more correct. Both commit the same fallacy. They confuse abstraction from the genus, and abstraction from the specific forms in which the genus manifests itself. In our illustration the circumstance which is of no account as regards the question of salary is evidently only the special form in which the good voice appears, whether as tenor, bass, or baritone, and by no means the good voice as such. And just so is it with the exchange relation of commodities. The special forms under which the values in use of the commodities may appear, whether they serve for food, shelter, clothing, etc., is of course disregarded, but the value in use of the commodity as such is never disregarded. Marx might have seen that we do not absolutely disregard value in use, from the fact that there can be no exchange value where there is no value in use—a fact which Marx is himself repeatedly forced to admit."1

The second step in the argument is still worse: "If the use value of commodities be disregarded"—these are Marx's words—"there remains in them only one other property, that of being products of labor." Is it so? I ask today as I asked twelve years ago: is there only one other property? Is not the property of being scarce in proportion to demand also common to all exchangeable goods? Or that they are the subjects of demand and supply? Or that they are appropriated? Or that they are natural products? For that they are products of nature, just as they are products of labor, no one asserts more plainly than Marx himself, when he declares in one place that "commodities are combinations of two elements, natural material and labor." Or is not the property that they cause expense to their producers—a property to which Marx draws attention in the third volume—common to exchangeable goods?

Why then, I ask again today, may not the principle of value reside in any one of these common properties as well as in the property of being products of labor? For in support of this latter proposition Marx has not adduced a shred of positive evidence. His sole argument is the negative one, that the value in use, from which we have happily abstracted, is not the principle of exchange value. But does not this negative argument apply equally to all the other common properties overlooked by Marx? And this is not all. On page 44, in which Marx has abstracted from the influence of the value in use on exchange value by arguing that any one value in use is worth as much as any other if only it is present in proper proportion, he writes as follows about products of labor: 

"But even as the product of labor they have already changed in our hand. For if we abstract from a commodity its value in use, we at the same time take from it the material constituents and forms which give it a value in use. It is no longer a table, or a house, or yarn, or any other useful thing. All its physical qualities have disappeared. Nor is it any longer the product of the labor of the carpenter, or the mason, or the spinner, or of any other particular productive industry. With the useful character of the labor products there disappears the useful character of the labor embodied in them, and there vanish also the different concrete forms of those labors. They are no longer distinguished from each other, but are all reduced to identical human labor—abstract human labor."

Is it possible to state more clearly or more emphatically that for an exchange relation not only any one value in use, but also any one kind of labor or product of labor is worth exactly as much as any other, if only it is present in proper proportion? Or, in other words, that exactly the same evidence on which Marx formulated his verdict of exclusion against the value in use holds good with regard to labor? Labor and value in use have a qualitative side and a quantitative side. As the value in use is different qualitatively as table, house, or yarn, so is labor as carpentry, masonry, or spinning. And just as one can compare different kinds of labor according to their quantity, so one can compare values in use of different kinds according to the amount of the value in use. It is quite impossible to understand why the very same evidence should result in the one competitor being excluded and in the other getting the crown and the prize. If Marx had chanced to reverse the order of the examination, the same reasoning which led to the exclusion of the value in use would have excluded labor; and then the reasoning which resulted in the crowning of labor might have led him to declare the value in use to be the only property left, and therefore to be the sought-for common property, and value to be "the cellular tissue of value in use." I think it can be maintained seriously, not in jest, that, if the subjects of the two paragraphs on page 44 were transposed (in the first of which the influence of value in use is thought away, and in the second labor is shown to be the sought-for common factor), the seeming justness of the reasoning would not be affected, that "labor" and "products of labor" could be substituted everywhere for "value in use" in the otherwise unaltered structure of the first paragraph, and that in the structure of the second paragraph "value in use" could be substituted throughout for "labor."

Of such a nature are the reasoning and the method employed by Marx in introducing into his system his fundamental proposition that labor is the sole basis of value. In my opinion it is quite impossible that this dialectical hocus-pocus constituted the ground and source of Marx's own convictions. It would have been impossible for a thinker such as he was (and I look upon him as an intellectual force of the very highest order) to have followed such tortuous and unnatural methods had he been engaged, with a free and open mind, in really investigating the actual connections of things, and in forming his own conclusions with regard to them; it would have been impossible for him to fall successively by mere accident into all the errors of thought and method which I have described, and to arrive at the conclusion that labor is the sole source of value as the natural outgrowth, not the desired and predetermined result, of such a mode of inquiry.

I think the case was really different. That Marx was truly and honestly convinced of the truth of his thesis I do not doubt. But the grounds of his conviction are not those which he gives in his system. They were in reality opinions rather than thought-out conclusions. Above all they were opinions derived from authority. Smith and Ricardo, the great authorities, as was then at least believed, had taught the same doctrine. They had not proved it any more than Marx. They had only postulated it from certain general confused impressions. But they explicitly contradicted it when they examined things more closely and in quarters where a closer examination could not be avoided. Smith, in the same way as Marx in his third volume, taught that in a developed economic system values and prices gravitate towards a level of costs which besides labor comprises an average profit of capital. And Ricardo, too, in the celebrated fourth section of the chapter "On Value," clearly and definitely stated that by the side of labor, mediate or immediate, the amount of capital invested and the duration of the investment exercise a determining influence on the value of the goods. In order to maintain without obvious contradiction their cherished philosophical principle that labor is the "true" source of value, they were obliged to beat a retreat to mythical times and places in which capitalists and landed proprietors did not exist. There they could maintain it without contradiction, for there was nothing to restrain them. Experience, which does not support the theory, was not there to refute them. Nor were they restrained by a scientific, psychological analysis, for like Marx they avoided such an analysis. They did not seek to prove—they postulated, as a "natural" state, an idyllic state of things where labor and value were one.1

It was to tendencies and views of this kind, which had acquired from Smith and Ricardo a great but not undisputed authority, that Marx became heir, and as an ardent socialist he willingly believed in them. It is not surprising that he did not take a more skeptical attitude with regard to a view which was so well adapted to support his economic theory of the world than did Ricardo, to whom it must have gone sorely against the grain. It is not surprising, too, that he did not allow those views of the classical writers which were against him to excite any critical doubts in his own mind on the doctrine that value is wholly labor, but considered that they were only attempts on their part to escape in an indirect way from the unpleasant consequences of an inconvenient truth. In short, it is not surprising that the same material on which the classical writers had grounded their half-confused, half-contradictory, and wholly unproved opinions should have served Marx as foundation for the same assumption, believed in unconditionally and with earnest conviction. For himself he needed no further evidence. Only for his system he needed a formal proof.

It is clear that he could not rely simply on the classical writers for this, as they had not proved anything; and we also know that he could not appeal to experience, or attempt an economico-psychological proof, for these methods would have straightway led him to a conclusion exactly opposite to the one he wished to establish. So he turned to dialectical speculation, which was, moreover, in keeping with the bent of his mind. And here it was a question of using any means at hand. He knew the result that he wished to obtain, and must obtain, and so he twisted and manipulated the long-suffering ideas and logical premises with admirable skill and subtlety until they actually yielded the desired result in a seemingly respectable syllogistic form. Perhaps he was so blinded by his convictions that he was not aware of the monstrosities of logic and method which had necessarily crept in, or perhaps he was aware of them and thought himself justified in making use of them simply as formal supports, to give a suitable systematic dress to a truth which, according to his deepest convictions, was already substantially proved. Of that I cannot judge, neither is it now possible for any one else to do so. What I will say, however, is that no one, with so powerful a mind as Marx, has ever exhibited a logic so continuously and so palpably wrong as he exhibits in the systematic proof of his fundamental doctrine.

Endnotes to Chapter 4

1 Capital and Interest, p. 377.

1 Karl Knies makes the following pertinent objection against Marx: "There is no reason apparent in Marx's statement why the equation, 1 quarter wheat = a cwts. wild-grown wood = b acres of virgin soil = c acres of natural pasture-land, should not be as good as the equation, 1 quarter wheat = a cwts. of forest-grown wood" (Das Geld, 1st edition, p. 121, 2nd edition, p. 157).

1 In a quotation from Barbon, in this same paragraph, the difference between commodities and things is again effaced: "One sort of wares are as good as another, if the value be equal. There is no difference or distinction in things of equal value."

1 For example, p. 48: "Lastly, nothing can be a value without also being an object of use. If it is useless, the labor contained in it is also useless; it does not count as labor [sic!], and therefore creates no value." Knies has already drawn attention to the logical fallacy animadverted upon in the text (see Das Geld, Berlin, 1873, pp. 123 ff.; 2nd edition, pp. 160 ff.). Adler (Grundlagen der Karl Marxschen Kritik, Tubingen, 1887, pp. 211 ff.) has strangely misunderstood my argument when he contends against me that good voices are not commodities in the Marxist sense. It did not concern me at all whether "good voices" could be classed as economic goods under the Marxist law of value or not. It only concerned me to present an argument of a logical syllogism which showed the same fallacy as that of Marx. I might for this purpose just as well have chosen an example which was in no way related to the domain of economics. I might, for example, just as well have shown that according to Marx's logic the common factor of variously colored bodies might consist in heaven knows what, but not in the blending of various colors. For any one combination of colors—for example, white, blue, yellow, black, violet—is as regards variety worth just as much as any other combination, say green, red, orange, sky-blue, etc., if only it is present "in proper proportion"; we therefore apparently abstract from the color and combination of colors!

1 The position which is taken by Smith and Ricardo towards the doctrine that value is wholly labor I have discussed exhaustively in the Geschichte und Kritik, pp. 428 ff. and have there also shown especially that no trace of a proof of this thesis is to be found in the so-called classical writers. Compare also Knies, Der Kredit, 2nd section, pp. 60 ff.