9. Marginal Productivity and the Formation of Costs. The Static System

Of all our explanations, nothing is as likely to appear as “foreign to reality” as the principle of marginal productivity. The theoretical derivation of the principle, as we have presented for the so-called law of diminishing agricultural returns, will appear reasonable as a purely “theoretical construction.” It is very plausible that the increase of one of several cooperating factors of production will not lead to a proportionate increase in the output; this can only be expected from a corresponding increase in all factors of production. If one regards experience, however, then in fact the opposite relationship seems to exist in many cases. Consequently, would this be a case in which a doctrine—the doctrine of marginal productivity—is “theoretically correct, but wrong in practice?” To us the situation appears to be the same here as always whenever one believes oneself able to point out a contradiction between theory and practice: A theory can only be applied to experience as a whole. It would be wrong to believe that one could break off part of the theoretical structure and triumphantly refute it in practice. The assumptions, too, from which the reasoning set out always belong to a theory. And we must certainly keep this in mind, particularly with respect to the theory of marginal productivity Perhaps we must even strive to formulate what we have already presented in this regard more precisely.

Let us first, however, present the various possibilities which can result and in fact have actually occurred. We will have to distinguish between three cases.

1. Diminishing return or rising (marginal) costs: With a given combination of factors of production the increase of one of the factors of production employed results in an increase in output which remains behind the increase in this factor of production. Accordingly, a progressive withdrawal of units of one factor of production will bring about increasing losses in output with each withdrawal of a unit. This case corresponds to the assumption on which the doctrine of marginal productivity was based.

2. Proportional output or proportional costs: With an increase of one factor of production employed in a productive combination, the output will grow in the same proportion, just as with a decrease in the number of cooperating factor units it will go down proportionally.

3. Increasing output or falling costs: The increase of one of the factors of production working in a productive combination leads to an overproportional increase in outputs, just as the decrease in one factor of production leads to a subproportional drop in output.

These pure types can be found in various combinations. Most important will be the combination of rising and falling costs. Here, with an expansion of production the transition from rising to falling costs, as well as the transition from falling to rising costs, is conceivable. The case of proportional costs will be considered essentially as a link between these combinations. Not the entire cost curve will be relevant for the isolation of the marginal product of one cooperating factor, but rather only that section will be of importance which is relevant for all actual movements. It is in this section of the cost curve where that type which could not be integrated in the cost theory based on marginal productivity theory—the type of decreasing costs—must be located. The difficulty here lies not only in the fact that with this cost structure, marginal productivity in the sense intended here cannot be spoken of. For if all units of a cost factor whose employment is subject to this law of returns were paid in accordance with the increase in output produced by the marginal factor, this payment could be larger than the total proceeds from production. One would have to look for another principle to explain the formation of the price of a factor of production. But that is not all. For even if the price of factors of production could be explained in another way, the mechanism of the law of costs could not operate. Once loss prices arise, then according to the model of the law of costs, production should again become profitable by restricting production. If “reorganization” could only be started by entrepreneurs’ refraining from individual production processes—those which bring losses—thereby raising the price on the market and simultaneously lowering costs, then this would not occur. With every restriction in production, the entrepreneur will raise his own costs even more. His interest will not be in restricting, but rather in expanding production because only in this way can he reduce his costs. And since each restriction in production means that an entrepreneur raises his own costs, he leaves it to his competitors to supply the market at lower costs. One might deduce from this relationship that falling costs make the maintenance of firms impossible in free competition and that only an amalgamation of firms would be able to carry through those restrictions in production necessary for the adaptation of the market price to production costs. Production would have to be restricted until the increased product price surpassed the increase in costs connected with the restriction in production. And this kind of cost structure—falling costs—is characteristic of many modern firms, namely whenever one does not completely take advantage of one’s production capacity. It is considered a rule that an expansion of production at lower costs is possible if capacity is not fully employed, and hence that the use of additional factors of production will result in an overproportional output. Only once the firm has reached full capacity will a further expansion of production be possible, and only at increased costs. The problem thus arises in the area of the falling branch of the cost curve, and this difficulty will occur with great frequency. The reason for this kind of cost structure can be found in the large investment of fixed capital which, whenever production is restricted, results in the general expenses (“the cost of the firm’s readiness”) being divided up among a smaller production quantum. Consequently, a reduction in costs by increasing production is possible as long as these investments permit the proportional expansion without adding cost expenditures other than the costs for material and “productive” labor.

One must certainly admit that such a cost structure is very frequent. The question is only how under these circumstances the doctrine of marginal productivity can be applied. We will only be able to arrive at a satisfactory answer here if we reach clarity regarding a few points concerning the method of economic theorizing.

Let us assume, using a highly “construed” example that in a closed economy in which falling costs normally do not occur and in which even those firms that have a span of falling costs in their cost structure are employed in a region of rising costs—that is, in an economy in which otherwise the law of costs functions smoothly—there are ten large automobile firms that have falling costs. These firms function such that a further expansion of production would reduce their costs. These firms assume thereby that the prices are already loss prices and that precisely because of the structure of costs, no firm is in the position of restricting production. By limiting production, each firm would only increase its costs. The other firms would not limit their production, and each firm which limited production would only benefit a competitor and hurt itself. Now, for the sake of theoretical analysis, let’s make an assumption which can never exist in reality. We will imagine that these firms are suddenly transformed such that in each firm the principle of marginal productivity can take effect immediately. As impossible as this is, it is not difficult to see what would have to happen. For it is characteristic of each of these firms that they work at a loss, but that limiting those factors of production which are in fact variable, i.e., limiting the use of “productive” labor and the raw material iron (ignoring the others) cannot help. Hence this limitation must be attempted regarding other factors of production, such as invested capital, machinery, “previous labor,” and previously invested iron. Now it is technically impossible to withdraw these factors of production—the machines cannot be transformed back into iron, into unexpended labor—at least not so that these factors of production are available in the form in which they previously had been. But let us imagine that a miracle had transformed the invested factors of production so that this industry’s situation would immediately change. Old investments would be withdrawn from the firms, for in these productive combinations they do not bring any return⁴³; they operate at a loss, while elsewhere in the economy they could bring a return. The invested capital in particular could be used at the current interest rate (or with practically insignificant pressure on this interest rate) by other firms. Production in this “overcapitalized” branch of industry⁴⁴ would be changed by withdrawing fixed capital. Withdrawing previously invested capital is thereby possible with two different effects. Either the capital is withdrawn entirely from some of the ten automobile factories and fewer firms will then exist while others are dissolved, or in each of these firms a portion of the invested capital can be withdrawn so that all of the firms continue to operate on a more limited scale. Regardless of which of these paths is chosen, whether ten smaller firms or five large firms remain,⁴⁵ the result will be a reduction of the supply of capital of these firms up to the point where falling costs no longer exist. For as long as the costs are falling, the withdrawal of fixed capital must still be profitable. It follows from our assumptions that in the end a situation in which rising costs occur throughout will be reached, and thus with regard to all factors of production the principle of marginal productivity is effective.

Let us now draw some conclusions from this completely unrealistic example. One is immediately clear: Under any circumstance, for the relevant section of the cost curve, a structuring of production is possible in which increasing costs occur throughout for any single factor of production. It shall now be asked why in reality a smooth adjustment to the situation in which the law of costs based on the principle of marginal productivity takes effect does not occur; it shall be asked what the condition is which so often ties production to falling costs, in contrast to our example. It shall then also be asked whether something similar to that which the example illustrated will in the end happen in reality.

First, it is probably clear that the discrepancy between our example and reality lies only in one single condition: In the fact that the investment of free capital is a process which is physically carried out and hence cannot be reversed; in the fact that once invested factors of production have assumed a physical form they cannot be transformed unrestrictedly. If it were not for this obstacle of the physically restricted convertibility of products, if there existed unlimited variability of factors of production, the unrestricted possibility of transferring factors of production which have assumed the form of capital goods at any chosen stage of production from one employment to another, then the principle of marginal productivity could take effect without any friction.

But does not precisely the circumstance that fixed capital cannot be withdrawn from investment lead to the consequence that the principle of marginal productivity loses all meaning when considering a reality in which one finds a great number of production processes which are overcapitalized? Here we arrive at the second question that we brought up in connection with the presentation of our example.

The process of adapting the use of factors of production to a stratification corresponding to the principle of marginal productivity actually occurs in a real economy too. It cannot occur, as we presented it in our example, where we assumed the possibility of a retroactive transformation of investments that were made earlier. Even with frictionless movements it must occur more slowly, such that a successively progressing need for reinvestments brings about a reallocation of production factors in accordance with the law of costs. Once made investments can, of course, no longer be reversed.⁴⁶ But invested capital is never tied up for such a long period that such an investment can never be reversed. Every machine will be used up and must be replaced if production is to be maintained. However, maintaining capital investments which do not bring a return by continually introducing new free capital will not be possible. Somewhere in the economy the owner of capital who wishes to expend free capital will find a possibility for an investment that will bring a profit, contrary to the presently maintained one with falling costs. An investment of durable capital that operates with falling costs will no longer be renewed once it is used up. Insofar as the entrepreneur who owns such an investment can produce any renewal fund, he will not be able to invest this in his own firm if he wishes to achieve a profit. Consequently, capital will be withdrawn from the firm and invested elsewhere. And here we see that what could happen immediately with a free convertibility of already invested factors of production—the adjustment of investments to the principle of marginal productivity—will come about slowly in the real world of restricted convertibilities in the course of the successively arising need for reinvestments to replace exhausted factors. The transformation will occur because these reinvestments are not made. Thus, the economy will move towards a state whose structure is in accordance with the principle of marginal productivity and in which the law of costs immediately takes effect through changes in the employment of factors of production. As a result of the frequent tying up of factors of production in fixed investments, the law of costs will probably not operate such that it immediately brings about an adjustment of production. But there will be a tendency in the economy to bring about this adjustment. We can thus summarize: Tying up capital in durable investments, and hence frequently occurring falling costs, imply an important friction in the operation of the law of costs based on the principle of marginal productivity. This friction does not suspend the effect of this law, but rather only results in this law’s taking effect in a process which requires a longer period of time because it can only be effected through successive reinvestments.⁴⁷

Ever since economic science first mentioned a law of diminishing returns, it has been beyond doubt that this law is only valid rebus sic stantibus, and that the adoption of a new production technique interrupts the effectiveness of the law; and hence that there can be no explanation of the course of history, co-determined as it is by changes in technology in terms of the law of diminishing returns. Instead, there can only be an explanation of its effectiveness under the assumption of given data.⁴⁸ Since we have characterized the principle of diminishing returns more generally as a principle of the cooperation between economic goods, in particular as the cooperation between free capital and originary factors of production then the restriction of rebus sic stantibus must naturally also be significant here. The simplest formulation would then be: As a rule, lengthening a roundabout method of production brings about a diminishing increase in output, but technical progress can lead to a situation in which even a shortening of the roundabout method of production leads to an increase in returns. The distinction between the two possibilities for changing output does not mean that we wish to develop a classification system, which can be applied without difficulty in each individual case to explain experience, but it means instead that we wish to understand the constructive principle underlying and directing economic processes. Where there is a possibility of increasing the output without lengthening the roundabout production process, the economy will take advantage of this possibility. This is naturally not limited solely by our technical knowledge, but also by the profitability of individual production methods: The entrepreneur will not be able to employ even the technically most satisfactory method if no favorable balance in the relation between cost expenditures and revenues exists. Nonetheless, wherever a technically new production method means a lengthened roundabout method of production, the calculation of costs—and in particular the calculation of interest—will cause the adoption of a technique to be adjusted to the economic possibilities.⁴⁹ For us, however, it is significant that—entirely independent of the possibility of shortening roundabout production processes through technical inventions—with each given technology a lengthening of the roundabout production process with the effect of increasing output is possible. The problem of the structure of production, a problem which is of great importance for the economic process, lies in the limits of the economically possible length of the roundabout production processes; in the circumstances that restrict the economy in its possibility of utilizing the advantages of a lengthening of the roundabout production processes. Here lies the central significance of the problem of capital employment.⁵⁰

Here, too, it can be seen that one must consider one’s assumptions when applying the principle of the greater productivity of roundabout production methods to reality. It is possible to observe an increase in production with shortened roundabout methods,⁵¹ just as one can frequently observe falling costs in modern firms. A theoretical analysis of the production process must isolate those elements from the multifarious possibilities of reality which can be used in constructing a system. The system will be applicable to, and able to offer an explanation of actual events, if it is constructed in such a way that it sets out from the principles that represent the conditions for attaining economic success which must be fulfilled in the world of experience.

We have seen this clearly with regard to the principle of marginal productivity. It would be correct for one to believe that in each individual case—for the employment of any factor of production in each individual firm—a marginal product could be established. It is not this, but something else that is the issue there: That it is possible to structure the economy according to the principle of marginal productivity, and that a deviation from this structure must cause a tendency to adjust to this structure. And with respect to roundabout methods of production, it is not only that an extension of the roundabout methods of production can and does lead to an increase in output, but that this increase of returns is limited by the supply of capital in the economy.⁵²

If starting from general principles, economic theory draws a picture of a stationary course of an economy, then it does not provide a portrait of reality. It presents a picture in which prices, product quantities and the structure of production are determined by general laws and are integrated into one cosmos. It must recognize the fact that the economy of experience can never be a realization of this model; it must admit that in the world of experience, newly arising changes in the data always keep the structure of the economy in motion. Economic theory can only present a model towards which the economy strives without ever being able to actually realize it. The cosmos of economic theory is not reality, but the laws from which economic theory is constructed nonetheless determine the real economy. Not in the sense that the real economy could never be structured other than according to these laws, but in the sense that wherever the structure of an economy deviates from these laws, wherever an economy has organized the employment of goods differently than would be required for the given data according to economic laws, a change will be initiated which has as its goal an adjustment to these economic laws. Complete and certain knowledge regarding the totality of an economy is only possible through an understanding of the system. Should one do without a system because not everything in reality is structured in complete accordance with this totality? One thing in particular should keep the premature critic from doing so: Only an understanding of the system shows what the limitations of economic possibilities are and what adjustments must ensue if the economic structure deviates from this system. And once one has recognized the central importance of the doctrine of the function of capital in the structure of an economy, then one will not be able to close one’s eyes to the fact that this doctrine is also of the greatest practical significance. The structure of production is identical to the employment of capital. One can safely say that this is the most sensitive element in the entire economic system. Production pushes towards lengthening roundabout production processes, and the extremely sensitive measure of interest rates indicates the possible limits. In looking at the monetary economy, we will now see just how sensitive this instrument is and how easily it can be disturbed.

---

²¹Here this means the following: Consider a “stationary” economic system, i.e., an economic process in which the same steps are always repeated. With this, a constancy in the data is assumed. A further assumption, however, which shall not be further explained here, must be made regarding the temporal integration of economic goals: The economic subjects must desire a stable provision for the present and the future. We will have more to say about this later.

²²That there is no production in a technical sense here may not be an objection. The choice of this example should make it possible to abstract from the employment of several different kinds of factors of production (here we can ignore the “material”), and further, to ignore the so-called “advance payment” of wages by the entrepreneur, i.e., the payment of wages before the completion of the product in time-consuming roundabout methods of production.

²³Here we could also have spoken of an increase in any other proportion rather than doubling a cooperating factor of production. See on this the explanations on pages 85ff.

²⁴To a limited extent perhaps also by speeding up the tempo of the machine.

²⁵We will still see that a combination of freely movable factors of production, such that the increase of one factor of production brings an increasing return, can have no place in a rational economic plan.

²⁶On the question of the “coordination” of both previously mentioned equalities, see the explanation of Philip Wicksteed and John Hicks.

²⁷A decreasing output cannot be considered because we have assumed that labor is the sole factor of production.

²⁸Compare here the explanations on pp. 85ff.

²⁹Knut Wicksell formulates this as follows: “Capital is saved labor and saved soil energies; capital interest is the difference between the marginal productivity of saved (stored up) labor and soil energies and the marginal productivity of current (present) ones.” (Lectures, vol. 1, p. 154.)

³⁰Neither here nor in the following are we interested in the problem of the qualitative composition of free capital.

³¹Clearly, this expression is subject to misinterpretation as expressing a value-judgment (“just” wage). The more correct formulation would be: “economic successor of consumption sacrifices.”

³²One is reminded here of the famous formulation of Thünen. If we interpret the quantity a in the sense of marginal analysis as the representation of the smallest wage (support) for which those of the employed laborers who are subject to the lowest social pressure are still willing to work, and if—again in line with marginal analysis—we interpret the quantity p as the product of the “last” laborer still employed (marginal product of labor), then a is equal to p and the height of the wage is determined by each of these quantities or also by the formula

³³This is to say, ignoring that labor for which no “economically relevant” time-period passes between its employment and the attainment of a finished consumer good (see on this note 3). It should be clear that consideration of production then only excludes a relatively small sector of labor services from analysis.

³⁴If the laborer himself possesses this subsistence fund, then in this respect he himself is, of course, a “capitalist.” The theoretical analysis must, however, set out from a consideration in which the various functions are differentiated, as only then will the function of each factor of production—and in roundabout production one of these is free capital—for structuring production be recognized correctly.

This would also be valid if the “union of personality” between laborer and owner of the capital were of greater significance in practice than is actually the case today.

³⁵One thing must be repeated here: It is not necessary for a capitalist whose capital is invested, for example, in a roundabout method of production lasting two years, to actually wait two years for his capital to be freed up. He can sell the produced capital good and receive in return free capital. However, the sale of a pre-product is, of course, only possible if another capitalist can spare free capital in payment and—taking the first capitalist’s place, so to speak—keep his capital tied up until production is completed or until a new sale is made. If invested capital becomes free in this way before production is completed, then from the point of view which looks beyond the situation of the individual, this is nothing but an “interpersonal change in the position of liquidity.” That such a change recurs, in particular with synchronized production processes—namely every time a capital good moves on from one stage of production to the next carried out by a different entrepreneur—must not distract from the fact that even in synchronized production the payment for a capital good with free capital is only possible because earlier, in another process of production, consumer goods already had been created that assumed the function of free capital. Not recognizing that the synchronization cannot change the essence of roundabout production and that with synchronized production, too, the implementation of roundabout methods of production is only possible if “free capital” has been made available, has led repeatedly to grave errors.

³⁶The wage fund is thereby not only that part of the output of consumer goods which is newly saved, but also that part which was saved earlier and is now maintained.

³⁷In my article mentioned on page 165, Number 1, I have presented these relationships in a formula. If W is the size of the wage fund, 1 the number of laborers, r the quantity of rations into which the wage fund is divided, p the number of payments which occur during the roundabout method of production (for example, wage weeks), and finally, m the size of the marginal product, the following equations can be formulated:

W = l · r · p

r = m

If the wage fund is not equal to the quantity on the right side of the wage fund equation, all three quantities which are on this side change as a result of changes in the interest rate. If W is smaller, a rise in the interest rate will reduce the magnitude p (by shortening the duration of the roundabout production process), r, and perhaps also 1, until equality is reached. In contrast, a larger W will lead to a lower interest rate and increase the magnitude on the right side of the equation.

In this essay I have indicated a second possible tendency towards equalization: If the magnitude W is too small and the interest rate rises, then it can happen that greater saving increases the supply of free capital.

In this case, a fall in the magnitudes on the right side of the equation will contrast with a rise in the magnitude W, so that the equalization will be facilitated. The reverse can occur if the interest rate falls.

³⁸In the first case, the price on the free market would be OA. At this price the supply would be equal to the demand (OM). The price tax of the height OB causes the demand to fall to OM’, while at this wage the supply of laborers is equal to OM”. In the second case, the wage on the free market would be determined by the intersection point of the supply and demand curve, but at this price there would be a supply OM’ as opposed to a demand OM, because the supply curve for labor to the right of the point of intersection moves horizontally.

The labor supply MM’ cannot find employment at this wage because at this wage level the supply is greater than demand. In spite of this discrepancy between supply and demand, this supply cannot function to reduce wages because it is only willing to work for the wage price OA. Without a doubt, both of these cases of unemployment are possible. The question which of these two cases is more important in practice is a question of applying this scheme to reality, not a question of theory. Let it only be said here that the second case in particular will occur if in one country the productivity for some reason—for example, the disintegration from a more comprehensive system of interlocal division of labor or a relevant decrease in the supply of capital—has been reduced significantly. “Cyclical” unemployment will require special consideration later on.

³⁹Here the problem for economic theory is essentially no different than in the case in which a corresponding number of laborers is reduced—for example, by emigration. In the static system, a rise in wages is equivalent to a reduction in the number of laborers. Note in addition, that insofar as the unemployed are maintained “at the cost of the economy,” i.e., insofar as subsidies for the maintenance of the unemployed become production costs (the maintenance of the unemployed does not come out of other income), here, too, the incorporation of the situation into the static system is possible, although the rise in costs must again imply a restriction in production possibilities.

⁴⁰To continue with the previously introduced examples: Measures which have restricted the division of labor are eliminated; the supply of capital rises, but the chance of this occurring in this connection is probably not very great.

⁴¹A complete equalization of prices will perhaps not occur if the necessity of employing a non-increasable specific factor of production in the production of A restricts the expansion of production. It is well-known that in such a case, the more expensive good can come to be considered a luxury without “objective” justification because of its higher price.

⁴²There is, of course, also the substitution between labor and intermediary products: More expensive labor, or more labor, saves raw materials and vice versa.

⁴³Their discounted return value would be equal to zero, and insofar as another use cannot be considered for them and no later output can be expected, they would have to be considered worthless. In other words, insofar as no change can be expected, the stocks of a firm operating with falling costs could only represent the “liquidation value” of the investments. In practice, however, one only too often makes the mistake of calculating with cost values instead of with the value of the discounted return.

⁴⁴It is clear that there can be overcapitalization regarding a branch of industry, i.e., regarding a more or less large part of production, but never with regard to the whole production process. Overcapitalization means here that so much capital is invested in fixed equipment that full utilization of capacity, i.e., an expansion of production to the point where costs no longer fall, is not possible because in the entire economic system there is no cost covering demand, i.e., a demand which at this production level pays a price for every individual article thrown onto the market. Thus, here overcapitalization is an incorrect investment of capital in relation to the structure of demand. However, general overcapitalization is impossible as a result of the circulatory nature of the economy: Each productive achievement can expect a complementary return from the product and itself creates the demand for whatever it produces. It is only a question of whether what that demand is prepared to assume has been produced. That the product is often only finished long after the factor of production is employed plays no role here because with a “correct” structuring of production, a corresponding subsistence fund must be given for the interim. It can never become a problem that in general too much has been produced as long as an expansion of need satisfaction is possible. It is clear that the overcapitalization of an industry, of which we are speaking here, which can only be considered a relative one, may not be confused with an excessive tying up of free capital (overinvestment), i.e., with the direction of free capital into investments from which it cannot be freed in time, and hence, with the case where as a result of a lack of free capital, a production process cannot be completed.

⁴⁵Both cases only mean roundabout methods of production of different lengths unless a larger firm can simply be considered as a multiplication of the small firm (with an equally long roundabout production process). The length of the roundabout production processes must naturally-via the link of prices, in particular the interest rate—be adjusted to the general structure of production with consideration for the profitability of an expansion of the roundabout method of production, especially in this line of production.

⁴⁶In a private economy, an already made investment can occasionally be reversed by exchanging it for a liquid asset—for example, by selling individual machines—whereby in general significant losses will probably have to be incurred. With the dissolution of a firm, an “organizational value” is lost.

⁴⁷The economic policy which attempts to protect firms with falling costs does not realize that reinvestment in such firms means tying up capital in investments in which the return will be lower than elsewhere. Let it be pointed out here that it is characteristic of a specific stage of the business cycle that the possibilities for investing free capital are unusually limited. We will deal with this problem later. Here we are only concerned—as emphasized explicitly—with the general question of the possibility of structuring production according to the principle of marginal productivity.

⁴⁸This restriction finds its most important application in the law of population: An increasing population must lead to pressure on the food supplies because of the increase in production cost that results if production is expanded with the additional help of only one increased factor of production (human labor), unless technical progress makes an increase in output possible above and beyond the increase of this factor of production. Apart from technical progress the effect of the law of population naturally can be neutralized also by an increase in capital exceeding the size of the population increase. Here again, we have an example of the fact that a “correct law” of theory is only “applicable” if all of the theory’s assumptions are actually met.

⁴⁹Whether a new technical method—for example, the introduction of electric power—means shortening or lengthening the roundabout method of production is a question which theory cannot answer definitively in advance. The answer will depend on whether the new production method saves more capital or more labor. Consequently, the effect of new technology must not be considered only for one single stage in the vertical production structure, but for the entire course of the roundabout method of creating finished consumer goods out of originary factors of production.

⁵⁰One must refrain from confusing duration of production and length of the roundabout method of production. To again use a prior example, if an automobile factory is “modernized” with the effect of reducing the duration of production of an automobile from three months to a few days, then this is possible because machines are introduced to a greater extent. Hence, simultaneous to shortening the duration of production, an additional use of “previously done labor” takes place, and we will probably have to say that the roundabout method of production has been extended. This is so because it must be assumed that the attainment of an equivalent return with a reduced expenditure of labor has become possible because labor expenditures occur to a larger extent in preceding production stages. The temporal moving back of labor expenditure cannot be viewed solely in relation to the first finished product, but instead—with regard to the increased employment of more durable capital goods—also in relation to the products created later with this investment.

⁵¹It is hard to detect a shortening or lengthening of a roundabout production process in an individual case because it is difficult to evaluate the function of a single stage of production within the complex production process.

⁵²Here a brief summary is due. Whenever several factors of production cooperate, in principle various kinds of changes in the size of the output are possible by changing their combination. The possibility relevant from the point of view of economic theory, however, must be that one which corresponds to the law of diminishing returns. This follows from the fact that we are only considering factors of production that are scarce and that as a result of their scarcity must be economized. Insofar as a factor of production’s cooperation in production would be subject as a rule to the principle of increasing returns, no portion of the returns could be attributed to this factor of production. For even a decrease in the quantity of this factor of production would have to be irrelevant for production. Earlier we tried to present this principle of the cooperation of scarce factors of production as the foundation of the “law of diminishing agricultural returns.” The principle must be generally valid for the combination of different kinds of factors of production, but in particular also for the employment of free capital (decreasing returns with a lengthening of the roundabout methods of production). From an economic point of view, then, that which can enter into an epistemological system of the static economy is of primary relevance. Other formations of the data of an economic process can at the most be regarded separately as variations of the static course. From this point of view it was necessary for us to first consider the supply of labor in the form of an upward sloping supply curve. We tried to justify this assumption earlier.