The Positive Theory of Capital

It has been almost a commonplace of economical teaching that the value of goods is regulated by the costs of their production. This doctrine has very seldom been questioned on grounds of theory,
*28 but very often its validity has been closely limited by the enumeration of exceptions, and insertion of all sorts of saving clauses. In this contracted sphere, however, it has held almost unquestioned authority down to our own times; it has a certain amount of support in practical experience, and, what is most serious, it seems to contradict the theory of value just put forward. For “Costs of Production” are nothing else than the sum of productive goods which must be used up in the making of a good—the concrete capital consumed, the labour expended, and so on. Now to the question as to the ground and amount of value which a good has, our theory answers: it depends on the marginal utility which a good is capable of rendering; that is to say, it depends on its
future employment. But the other theory answers: it depends on the value of the productive goods consumed in producing it; that is to say, on the conditions of its
origin. Putting aside this contradiction for a moment, and forgetting everything we have been taught as to costs, let us inquire impartially what our theory of marginal utility, logically carried out, has to say as to the value of productive goods, and as to “costs.”

For the sake of clearness it is desirable, before going further, to define with more exactness the object of our present inquiry, viz. Productive Goods. As compared with consumption goods
(Genussgüter), which directly serve to satisfy human wants, all productive goods have this common feature—they serve to satisfy human wants only indirectly. But they differ, again, from one another in the degree of indirectness. The flour, for instance, from which bread is baked, stands nearer the final satisfaction of want by several degrees than the field which grows the wheat. To express these degrees—which we shall find to be of importance both theoretically and practically—we shall avail ourselves of Menger’s division of goods into ranks.
*29 In the first rank we shall place consumption goods—those goods which serve immediately for the satisfaction of wants, such as bread. In the second rank we place those goods which assist in producing the goods of first rank—the goods which co-operate in the production of bread; as the flour, the oven, and the baker’s labour. In the third rank we place those goods which serve for the production of goods of second rank; as the wheat from which the flour is ground, the mill in which it is ground, the building materials of the oven, etc. In the fourth rank we put the means of production of goods of third rank; as the land which grows the corn, the implements used in cultivation, the labour of the agriculturist, the building materials of the mill, etc. And so on to the fifth, sixth, and seventh ranks, which embrace those goods, the useful service of which consists in producing goods of the rank immediately below them.

On the lines of our conception of value it must be self-evident that a productive good, like any other good, can only obtain value for us through our recognition that on its possession or non-possession depends our gain or loss of some one utility, of some one satisfaction of want. And it is equally self-evident that its value will be high when the dependent satisfaction is important, and low when it is unimportant. The only difference is that, in the case of goods for immediate consumption, the good and the satisfaction stand beside each other in a direct causal relation; while, in the case of productive goods, there is interposed, between them and the satisfaction finally dependent on them, a more or less lengthy series of intermediate members, their successive products. In this prolonged connection there is both matter and occasion for the development of new and legitimate relations, particularly between the value of means of production and that of their products. But the great law of value is neither destroyed nor disturbed by these relations. Exactly as in the analogous case of complementary goods it is only obscured, as it were, by a mass of details, to which the more ample development of the phenomena gives occasion. These details we have now to consider. To this end let us take a typical productive series.

A good for immediate consumption, which we shall call A, is made from a group of productive goods of second rank, which we shall call G
₂; this from a group of goods of third rank, G
₃; and this, finally, from a group of fourth rank, G
₄. For simplicity’s sake assume, first, that each of these productive groups passes without loss of time into the product which it creates, and that, at the same time, this particular employment is the only one of which it is capable. We have now to find out what is the relation of dependence between each member of the above series, and the wellbeing of its owner.

What depends on the final member, the good A, we already know. It is its marginal utility. Our inquiry, then, begins at the member G
₂. If we had not the group G
₂ we should not have its product A; that is to say, of the class of goods to which A belongs, we should have one fewer than we should otherwise have had. But, as we already know, one good less means one satisfaction less, and that the least satisfaction to which economically, one good of the stock would otherwise have been devoted. In other words, it means the loss of the marginal utility of the product A. On the group G
₂, therefore, exactly as on the final product A itself, depends the marginal utility of A. Looking now at the next member we find that, if we had not the group G
₃, we could not have the group G
₂ which is made from it; and, as consequence, we should lose, one good of the class A, or its marginal utility. On the group G
₃, then, depends exactly the same utility and importance for wellbeing as on the members which come after it in the production series. The same thing again follows in the case of the group G
₄. If it fails us, we, of course, lose one of the group G
₃, which otherwise might have been produced from it; we lose, further, one of the group G
₂, one of the class of good A, and, finally, the marginal utility of A. Thus we arrive at the following general proposition: On all groups of Means of Production of remoter rank which successively pass into one another, there depends one and the same gain to human wellbeing; that is, the marginal utility of their final product. No one will be surprised at this result. It is a foregone conclusion that a series of productions, which has no relation to our wellbeing except through its final member, can neither tend towards any other utility, nor condition any other utility, than that which this final member itself conditions. In every member of the chain successively we hold in our hand the condition of this final utility, sometimes at a further, sometimes at a nearer stage on the way to it.

From what has been said we may deduce the following general principles as regards the value of means of production. First, since on one and the same utility depend all the groups of means of production which successively pass into one another, the value of all these groups must be substantially the same. Second, the amount of this, their common value, is regulated for all, in the last resort, by the amount of the marginal utility of their finished product. I emphasise “in the last resort.” For, thirdly, the value of each group has its immediate measure in the value of its product, the succeeding group. In the
first instance, the utility and service of the means of production consist and exhaust themselves in the making of their product, and, naturally, the more important and more valuable the product is for us when made, the higher will be the estimate put on the importance of this utility, and of that which provides it. Substantially the third proposition is fully covered by the second, for, in the value of the goods of higher rank, the marginal utility of the final product is mirrored. From this marginal utility value is conducted to all the groups of means of production, but the conduction is done, as it were, by stages. First, and immediately, the amount of the marginal utility stamps itself on the value of the final product. This then forms the measure of the value of the group of goods from which this product comes. This again measures the value of the third group; and the third group, finally, the value of the last group, the goods of fourth rank. From stage to stage the name of the determining element changes, but, under the different names, it is always the same thing that acts—the marginal utility of the final product.

Although the second and third propositions, then, agree in substance, it is necessary to formulate the third explicitly. It is important as being a convenient abbreviated formula which we use in practical life much more frequently than the principal formula. If we are estimating what amount of wellbeing a productive instrument brings us, we look, naturally, first of all to the product which we get from it, and then, beyond that, to the wellbeing which that product brings us. If we do not know this, we must, I admit, go over the entire course of the conduction of utility, member by member, till we come finally to the marginal utility of the final member, the finished product. But very often this is not necessary. From previous consideration, or from experience, we meet with some opinion, already formed, on the value of the products, and, without further consideration, we make this the ground of our opinion as to the value of the means which produced them. A wood merchant, buying timber for cask staves, will not take long to consider the value of the wood to him. He estimates how many staves he can get out of the timber, and he knows what the staves are worth in the condition of the market at the time. Further than this he need not trouble himself.

Thus far we have formulated these principles as to the value of means of production on purely theoretical grounds; to some extent, as postulates of economical logic. If, now, we ask what experience says to these postulates, we shall find that it confirms them. Indeed we can appeal for confirmation to that very “law of costs” which is apparently so hostile to our theory of marginal utility. Experience shows that the value of most goods is equal to their “costs.” But “costs” are nothing else than the complex of those productive goods which have value—the labour, concrete capital, uses of wealth, and so on, which must be expended in the making of a product. The well-known identity of costs and value is only another form of expressing the identity of value between groups of goods of various ranks which pass into one another. I am quite aware, of course, that, as regards the cause of this identity, those who adopt the law of costs usually read it in the converse way. While we say that the value of means of production, and therefore the value of the costs, is regulated by the value of their products, the usual way of interpreting the law is to say that the value of products is determined by the value of their costs—that is, by the value of the means of production out of which they are made. Later on we shall have occasion to go thoroughly into this difference of opinion as to the cause of the identity. Meantime all I intend to do is simply to confirm the statement, that the asserted identity of value between groups of productive instruments which successively pass into one another—whatever be its cause,—is an actual empirical fact.

Of course this identity is not absolute, but approximate; we can only speak of a tendency towards identity of value. The divergences from absolute identity are of two kinds—partly irregular, partly normal. Both kinds arise from the fact that production costs time. In the long periods which often intervene while goods of sixth or eighth rank are passing gradually through all the transformation stages into the finished consumption good, both men and things may change. Wants may change; the relations between wants and their provision may change; and, not less important, the knowledge of these relations may change. With them, of course, changes the valuation of the goods at various stages on their way to the matured product. It is easy to understand that the fluctuations which proceed from this cause may be sometimes great, sometimes small, sometimes upwards, sometimes downwards; they are irregular fluctuations. But, besides these, we notice a divergence from complete identity which is constant and normal. It is a matter of observation that the total value of a complete group of remote rank lags somewhat behind the value of its product, and in a definite ratio; and that, indeed, the amount of this difference in value is graduated according to the time required to change the group of means of production into its product. If the value of the product, for instance, is £100, experience tells us that the total value of the labour, uses of land, fixed and floating capital spent in producing it, is something less than £100—perhaps £95 if the production process lasts a year; perhaps £97 or £98 if it lasts only half that time. This difference of value is the crease, as it were, in which Interest is caught. Its explanation is a subject by itself, with which we shall have enough to do in following chapters. It would be very far from advisable to mix it up with our present inquiry, where we are dealing with the general relation between the value of means of production and that of their products, and for the moment we shall therefore entirely disregard the existence of this particular difference of value.

Up to this point we have expounded the law which governs the value of productive goods under the simple hypothesis that each group of productive instruments permits of only one quite definite employment. But in actual life the cases in which this hypothesis corresponds with facts are very limited. It is, indeed, characteristic of productive goods that they admit of an infinitely more various use than consumption goods. The vast majority of them are adapted to several productive uses, while many of them, like iron, coal, and, above all, human labour, are adapted to thousands of different uses. In theoretical research we must, of course, take note of these actual circumstances, and see whether they do not involve some modification of our law, that the value of a group of goods of remote rank is determined by the value of its product.

Suppose, then, we vary the assumptions of our typical illustration. A man possesses a great stock of groups of productive instruments of second rank (G
₂). From one such group he can, at will, make a finished commodity of the kind A, or one of the kind B, or one of the kind C. Naturally he will provide for his various wants harmoniously, and will therefore, by means of different parts of this stock, produce simultaneously finished goods of all three classes according to the measure of his requirements. In a scheme of provision that was really harmonious, the amounts produced would be so regulated that, in each kind, wants of something like the same importance would depend on the last sample of the kind, and the marginal utility of every sample would therefore be approximately equal.
*30 Nevertheless there will be differences, and even considerable differences, of marginal utility, because, as we already know,
*31 the gradation of the concrete wants in any kind of want is not always uniform and unbroken. One fireplace in a room, for instance, will give me a very considerable utility—which I may represent by the figure 200—while a second fireplace would not be of any further use to me. Naturally, in providing for my wants, I shall therefore, in any case, stop at fireplaces when I have one fireplace with its marginal utility of 200, even if in other branches of wants the provision goes down, on the average, as low as a marginal utility of 100 or 120. To make our typical illustration true to nature, therefore, we must assume that the marginal utility of one sample is of different amount in the three kinds A, B, and C—say 100 in A, 120 in B, 200 in C. The question now is, In these circumstances what is the value of G
₂?

After the practice we have had in drawing distinctions of a similar kind, we can give the answer without hesitation—the value will be equal to 100. For if one of the available groups were lost the owner would naturally shift the loss to the least sensitive part; he would neither limit the production of the kind B, where he would lose a marginal utility of 120, nor of the kind C, where he would lose a marginal utility of 200. He would simply produce one less of the kind A, whereby his loss of wellbeing would be only 100. To put it generally: The value of the productive unit adjusts itself to the marginal utility and value of that product which possesses the least marginal utility among all the products for whose production the unit might, economically, have been employed. All the relations which we found to hold as regards the value of means of production and of their products under the simple hypothesis of the single employment, hold, therefore, generally between the value of means of production and their
least valuable product.

And how does it stand with the value of the remaining classes of products, B and C? This question brings us to the source of the “law of costs.”

If, under all circumstances, the marginal utility attainable within the kind itself were to decide,, the kinds of goods B and C would possess a value diverging, as well from the value of the kind A, as from the value of its costs G
₂. B would have a value of 120, C a value of 200. But this is one of those cases where, through substitution, a loss occurring in one kind of goods is shifted to another kind, and consequently the marginal utility of the latter becomes the standard for the former.
*32 That is to say, if one of the kind C gets lost there is no occasion to give up the marginal utility of 200, which it would have directly afforded; we can and will immediately procure a new C out of a productive unit G
₂, and we shall prefer to produce one less of that kind of good in which the marginal utility, and with it the loss of utility, is least. This, in our illustration, is the kind A. In virtue of the opportunity of substitution offered by production a good of the kind C is therefore valued, not at its own marginal utility 200, but at 100, the marginal utility of the least valuable cognate product A. The same holds, of course, of the value of kind B, and would hold, generally speaking, of every kind of good which is “cognate in production”
*33 with A, and has at the same time an immediate marginal utility greater than that of the kind A.

This leads to several important consequences: First of all, in this way the value of goods which have a higher individual marginal utility is put on a level with the value of the “marginal product”—as we shall call that product which has the least marginal utility—and thus with the value of the means of production, from which both in common come; the theoretical identity of Value and Costs, therefore, holds in this case also. But it is well worthy of notice that here the agreement between value and costs is brought about in a way essentially different from the agreement between costs and marginal product. In the latter case the identity was brought about by the value of means of production adapting itself to the value of the product; the value of the product was the determining, that of the means of production the determined. In the present case, on the contrary, it is the value of the product that must adapt itself. In the
last resort, of course, it adapts itself only to the value of another product, the marginal product of the cognate production; but, in the
first instance, it accommodates itself also to the value of the means of production from which it comes, and which are mediated by the substitutionary connection with the marginal product. Here the conduction of value describes, as it were, a broken line. First it goes from the marginal product to the means of production and fixes their value; then it goes in the opposite direction, from the means of production to the other products which may be made from them. In the end, therefore, products of higher immediate marginal utility get their value from the side of their means of production. To translate this from the abstract formula into practice. If we are considering what a good B or C (generally speaking, a product of higher immediate marginal utility) is worth for us, we must say first of all: It is worth exactly as much as the means of production from which we could replace it at any moment. Then if we examine further how much the means of production themselves are worth, we come to the marginal utility of the marginal product A. But very often, indeed, we may save ourselves this further inquiry, as we already know the value of the goods that make up the cost without having to begin at the foundation and follow it from case to case; and in all such cases we measure the value of the products in an abbreviated form, both accurate and convenient—that is to say, simply by their costs.

Here, then, we have the whole truth about the celebrated Law of Costs. As a fact people are right when they say that costs regulate value. Only they must always be conscious of the limits within which this “law” holds, and the source from which it gets its strength. It is, first, only a particular law. It holds only in so far as it is possible to obtain, at will and at the right time, substitutes through production. If there is no opportunity of substitution the value of every product has to be measured by the immediate marginal utility of its own kind, and its agreement with the value of the marginal product, and with the intermediate means of production, is disturbed. Hence the well-known empirical proposition that the law of costs holds only as regards goods “reproducible at will;” or “freely produced,” and that it is simply an approximate law which does not bind the value of the goods that come under it with slavish exactitude to the level of costs, but—according as production for the moment comes short of demand or runs beyond it—permits of fluctuations now on one side, now on the other.

But it is still more important to emphasise, in the second place, that, even where the law of costs holds, costs are not the final but only the intermediate cause of value. In the last resort they do not
give it to their products, but receive it from them. In the case of productive goods which have only a single employment this is perfectly clear. That Tokay is not valuable because there are Tokay vineyards, but that the Tokay vineyards are valuable because Tokay has a high value, no one will be inclined to deny, any more than that the value of a quicksilver mine depends on the value of quicksilver, the wheat field on the value of wheat, the brick kiln on that of bricks, and not the other way about. It is only this many-sided character of most cost goods—their capacity of being employed in many different uses—that gives the appearance of the contrary, and a little consideration shows this to be an appearance and nothing more. As the moon reflects the sun’s rays on to the earth, so the many-sided costs reflect the value, which they receive from their marginal product, on to their other products. The principle of value is never in them, but outside them, in the marginal utility of the products. The law of costs is not an independent law of value; it only forms an incidental case inside the true universal law of marginal utility. It is simply the great counterpart to the law of Complementary Goods. As the latter disentangles and explains those relations of value which result from the temporary and causal
collocation—the simultaneous co-operation of several goods to a common useful end; so does the Law of Costs for the value relations of those goods which act in temporary and causal
sequence—the working of goods after one another and through one another to the same final goal. If we think of the value relations of goods that work into one another as a much-tangled net, we might say that the former law disentangles the meshes in their length and breadth, while the latter disentangles them in their depth; but both fall under the all-embracing law of Marginal Utility, and are nothing but special applications of that law to special problems.