Risk, Uncertainty, and Profit
By Frank H. Knight
The text has been altered as little as possible from the original edition (
Risk, Uncertainty, and Profit, Frank H. Knight, Ph.D., Associate Professor of Economics in the State University of Iowa; Boston and New York, Houghton Mifflin Co., The Riverside Press, 1921).A few corrections of obvious typos were made for this website edition. However, because the original edition was so internally consistent and carefully proofread, we have erred on the side of caution, allowing some typos (such as for proper nouns and within references) to remain lest someone doing academic research wishes to follow up. We have changed small caps to full caps for ease of using search engines.Lauren Landsburg
Editor, Library of Economics and Liberty
First Pub. Date
Boston, MA: Hart, Schaffner & Marx; Houghton Mifflin Co.
1st edition. Based on award-winning dissertation essay.
The text of this edition is in the public domain. Picture of Frank H. Knight courtesy of Ethel V. Knight.
Change and Progress with Uncertainty Absent
Part II, Chapter V
We turn now to the third grand division of theoretical economics, the study of the use of resources in the increase of resources for the making of goods and in the refinement of wants alongside of and alternative to their direct use in making goods for consumption. The relations of these three theoretical problems are somewhat complex and confusions in regard to them have been a prolific source of error in economic thinking. The first problem is the use of
given goods in the satisfaction of given wants (with a given distribution of the goods to begin with, and free exchange) and its analysis and solution constitute the theory of market price. Market prices, besides determining the apportionment of given stocks of goods, the product of past industry, at the same time show the social estimate of the relative importance of different goods according to which the apportionment of resources under the second problem is worked out. In this first division, production goods do not enter at all, since costs already incurred have no bearing on price; as Jevons puts it, “bygones are forever bygones.”
The second problem deals with the use of
given productive resources in the production of goods to be used (always in accordance with market price principles) in the satisfaction of given wants; it has become known as the problem of the static society or “static state,” and has two aspects. The first phase relates to the value of productive services separately; the second, to the values of particular consumption goods, in relation to the values of the productive services which go into them, or their costs; this is the problem of the long-time or normal prices of consumption goods. In a sense it is, as Marshall suggests, a case of two classifications crossing each other. The first problem classifies on the basis of consumption goods, showing the equation of the value of a commodity to that of the
bundle of productive services entering into it. The second takes the productive service as a basis and shows the equation of the value of each unit of productive service to the value of the portion of each kind of consumption goods in whose creation it is used, for which it is responsible. The first is the
long-time “value” problem, the second is the
short-time “distribution” problem. The
changes in supply (and value) of consumption goods are studied in relation to
fixed conditions of production, including especially fixed supplies and methods of organization of productive resources.
The third general problem also relates to both value and distribution phenomena. Changes in the “fundamental conditions of demand and supply” of goods give rise to what Marshall calls “secular changes in normal price.” But the principal “fundamental conditions” subject to change are the supplies of the different productive services which evidently affect still more directly the prices of these services, the distributive shares. Our discussion, like Marshall’s, will be practically limited to this more simple and direct effect, the modification of the distribution situation, and its tendency toward an equilibrium.
First, let us try to formulate clearly and accurately what is involved in the problem of progress. What new variables come in for study? What is the exact content of the “general conditions of demand and supply,” or the “given resources used in the satisfaction of given wants,” which our previous analysis has assumed? And finally, what are the changes in these factors which call for consideration in order to bring our society into the closest possible approximation to reality? Marshall, whom the present study more closely follows than it does any other writer, seems to avoid, not to say evade, answering this question explicitly. He does at one point begin an enumeration of elements, but cuts it short at once with the blanket expression quoted above.
*82 A well-known explicit list of static state or dynamic factors to be excluded is that of Professor J. B. Clark, whose name is especially associated with the contrast between static and dynamic problems in this country. He gives these five elements of progress:
*83 (1) growth of population; (2) accumulation of new capital; (3) progress in technology; (4) improvement in methods of business organization; (5) development of new wants. Professor Seager modifies this list, and in the writer’s view greatly improves it, by combining the third and fourth factors and adding a new one, the impairment of natural resources or discovery of new natural wealth.
It will aid in clarifying the issues if we first consider separately the conditions of demand and of the supply of goods. Conditions of demand seem to include the following fundamental facts:
1. The population considered as consuming units; its numbers and physical composition as to age, sex, race, etc.
2. The psychic attributes of the population, its behavior attitudes toward the consumption of all sorts of goods, both inherited “instincts” (in whatever sense such things exist), and the “social inheritance” of habit, custom, tastes, standards,
mores, and what-not, including, of course, actual knowledge or beliefs as to the real characteristics of commodities. We must also include here any institutional facts as to the control of the consumption of some persons by other persons, such as authority of parents, sumptuary laws, etc.
Immediately, the money income of the population both as to aggregate amount
and distribution. Ultimately, in the equilibrium adjustment, the income and its distribution depend on the whole set of conditions of the supply of goods, especially the amount
and distribution of productive resources in the society. It is imperative to remember that the end result of the competitive adjustment depends on the initial facts in all these respects.
4. For completeness it is important, also, to consider the given facts as to the geographic distribution of the population as consuming units; this is determined, of course, by the distribution of productive resources and of environmental conditions affecting desirability of sites for habitation. Differences here would also produce effects ramifying throughout the whole organization.
Given conditions of supply include especially the supply of the factors of production, but there are other vital considerations. We may classify as follows:
1. The population considered as labor force, numbers, and composition.
2. The psychic or behavior attitudes, tastes, prejudices, etc., toward productive activities, inherited or acquired.
3. Immediately, money income and its distribution; ultimately, the distribution of ownership of productive resources of every kind. There is no difference between personal ability and productive property in this respect. It is obvious that income affects disposition to engage in productive activities and enters as a variable, independent of taste.
4. Although it belongs logically under number 3, or is at most a corollary from it, we specify separately the institutional situation as to the meaning and extent of private property. This includes all facts as to (a) control of the use of productive services and (b) of valid and enforceable rights to income. There is again no distinction to be made between personal powers and other productive facts.
5. The amount and form of material agents of production in existence. Under the static conditions hitherto discussed these can include only natural agents in the narrowest sense, or, what would amount to the same thing, implements inherited from past generations, and in either case subject to neither deterioration nor improvement.
6. The geographical distribution of productive agencies.
7. The state of the arts; the development of technology, business organization, etc.
Combining the two groups and removing duplication we find the following factors in regard to which change or the possibility of change must be studied:
1. The population, numbers and composition.
2. The tastes and dispositions of the people.
3. The amounts and kinds of productive capacities in existence, including
a. Personal powers.
b. Material agents.
4. The distribution of ownership of these, including all rights of control by persons over persons or things. (Impersonal control, by laws or
mores, is indistinguishable from number 2, tastes and dispositions.)
5. Geographic distribution of people and things. This stands in close relation to the facts of technology.
6. The state of the arts; the whole situation as to science, education, technology, social organization, etc.
Systematic completeness would call for a survey of possible changes in each of these elements and the relation of such changes to both value and distribution phenomena, the prices of consumption goods and of productive services (and in addition their relations to the capitalization rate, the sale prices of productive agencies). No such ambitious program can be entered upon, however. We shall merely point out some of the more important price bearings of changes and make such comments as seem especially significant in illuminating dark places in theory. The point for especial emphasis is that the really far-reaching effects of change are not the results of the fact of change itself, but of the uncertainty which is involved in a changing world. If any or all of these changes take place regularly, whether progressively or periodically or according to whatever known law, their consequences in the price system and the economic organization can be briefly disposed of. Through the machinery of the exchange of present and future values all of them will be fully ” discounted ” an indefinite time before they occur. They will not upset human calculations or destroy universal perfect equalization of alternatives. Hence, in particular, changes, if foreseeable, do not disturb the prerequisites of perfect competition for productive services, bringing about exact equivalence between costs and values, with absence of profit.
As a matter of fact the effects of changes in the general conditions of the production and consumption of goods upon the prices of consumption goods are either so obvious or so complicated and hopeless of practical prediction that it does not seem worth while to attempt systematic treatment of them. Our discussion will be confined almost entirely to the theory of distribution. In this field, also, let us note that
progressive changes can usually be fairly well foreseen and discounted and their effects are not generally important over short periods of time. They produce relatively little real disturbance in the competitive adjustment and are not a significant cause of profit. The significant disturbances and sources of profit are rather the short-period and erratic fluctuations, and the irregularities of progressive change, not the change itself. The increase in population and accumulation of new capital are not disturbing facts to any appreciable extent, and the disturbances arising from invention and improvement are due to the local and spasmodic way in which they originate, not to the general tendency.
In discussing the short-time theory of distribution (distribution under conditions of fixed supplies of productive agencies) we have repeatedly emphasized the absence of any valid ground for a general classification of productive agencies, either along the lines of the traditional three factors or along any other lines. That is, on the demand side they are alike or differ by innumerable imperceptible gradations, and for short-time problems the conditions of supply—given quantities in existence—are also obviously identical for all. The long-time point of view, however, brings in the new question of changes in supply, in regard to which there are real differences. These differences in the conditions of supply afford a basis for legitimate classification, somewhat along the lines of the tripartite division. It is superficially reasonable to recognize three categorically different conditions of supply. First we should have agencies whose supply is given once for all even over long periods, things not subject to increase or decrease, improvement or deterioration. The traditional definition of land fits this description. (We do not here raise the question whether anything exists to which the definition applies.) In the second place, some productive goods may be, and obviously are, freely reproducible in the same manner as consumption goods, under conditions in which supply becomes a definite function of the price of their services. The traditional view of capital gives it this character. (Again we make no assertions as to the correctness of the view.) And finally, the supply of still other agencies may be variable, but not a function of price, or not connected with price in an immediate or direct way. The traditional treatment of the long-time supply of labor (the merits of which are also reserved for later examination) differentiate it in this respect from other productive powers. This traditional classification is not accepted as valid, even from the long-time point of view, and will be criticized at length as we proceed. But the superficial basis for it and the fact that it is well established in the thought and terminology of the science may justify taking it as a starting-point.
The ramifications and interconnections of effects of any particular change are ultimately rather complicated, and may be followed out until nearly every aspect of the adjustment is modified in some way. This is obviously true of the first of the static characteristics named. Historically the population question has been considered with distribution in connection with wage theory through its relation to the supply of labor. Of course, an increase of population is an increase in the demand for goods and hence in the demand for all the productive services including labor itself. But the demand for any productive service depends finally upon two elements, the total output of industry and the relative importance of that service in increasing the output. In accordance with the law of diminishing returns and the specific productivity theory based upon that law, a relative increase in the supply of labor will increase the product of industry less than proportionally and decrease the relative productivity of labor. Both effects tend to lower wages per man. The same reasoning applies to any other productive service as well as to labor.
Much confusion has arisen in economic discussion through different meanings given to a distributive share. We may speak of wages, for example, as above, as wages per man, and similarly of other incomes in relation to the concrete agency which produces them. The problem of distribution from this point of view Cannan calls “pseudo-distribution,”
*85 seemingly an unfortunate term, for this is surely the phase of the subject in which we have the greatest and most direct interest. The classical economists themselves, led by Ricardo, usually centered their discussion around the fraction of the total social produce received by the “factor” under discussion. Another clearly possible meaning is the aggregate share of a “factor” measured in absolute terms.
The effect of an increase in a factor (meaning a large group of physically interchangeable productive units) on the fraction of the social income it will receive, depends on the rate of diminishing returns realized from the application of that agency to others in the vicinity of the proportions already in existence. If the increase in total production is nearly proportional to the increase in the factor (remembering that it cannot be equal or greater), its fractional share will rise; if much less, it will fall. The aggregate absolute share of income falling to the agency will increase unless the falling-off in product is in equal or greater ratio with the increase in the agency. Both points, however, are rather remote from the problem of immediate interest. If the income per unit is known, the relative and absolute shares of the factor can more naturally be determined indirectly.
Obviously a shift in the amount of any productive agency will, through its effect on incomes, react on the demands for goods, and ultimately affect nearly every feature of the organization of industry and of the price system. The resulting changes in the prices of consumption goods are what Marshall calls secular changes in normal price. It does not seem profitable, if indeed it is possible, to discuss these in the abstract. About the only general observation which seems worth making is that those goods in whose production any particular agency predominates will tend to fall in value as the supply of that agency increases, other things being equal.
The really difficult problem in the theory of progress relates not so much to the effects of particular changes. These effects, though complicated, can be traced out by the application of the principles of the market, the “laws” of supply and demand. The difficulty comes in the prediction of the changes themselves. What are the conditions of supply of the productive services? What changes in the supplies of the different services may be reasonably anticipated, and to what goals or equilibria do they tend? The question is of especial interest because it was in terms of these ultimate equilibrium levels that the classical theory of distribution was almost exclusively worked out. In our opinion the meaning of these equilibrium conditions was misconceived in classical economics and their significance perhaps somewhat overestimated. The early writers regarded the equilibrium condition as constantly at hand in a sense analogous to the normal price equilibrium between the production and consumption, cost and value, of consumption goods. Their “static state” was, if not the actual condition of society, a condition on which it constantly verged.
*86 It makes a great deal of difference in the theory when we recognize, as the facts require, that the equilibrium is an indefinite and usually a very great distance in the future. The condition must then be viewed as the theoretical result of a particular tendency only, which may be modified to any extent or reversed by the effect of other tendencies, or the conditions may be entirely changed by unforeseen developments long before any considerable approach to the equilibrium has been made. The equilibrium, then, in a particular case, is not a result actually to be anticipated; a concrete prediction of the future course of events must take into account all the tendencies at work and estimate their relative importance, and in addition must always be made subject to wide reservations for unpredictable influences. In fact, as we shall see, the interrelations of the various factors of progress are so complicated, and the functions themselves are so inaccurately known and are affected by so many unknown variables, that definite predictions extending any considerable distance into the future seem to be quite out of the question.
Turning now to the question of the conditions influencing the progress variables and of the changes to be expected in regard to each, we may begin with the factor of population once more and go through the list. The plan, of course, is not to investigate hypotheses at random, but to inquire seriously about the facts of the world we live in. The only arbitrary or unreal element in the procedure is the selection of the outstanding dominant features and their isolation with a view to ascertaining if possible their own inherent tendencies. The products of such an inquiry are, like all theoretical deductions,—all general principles,—partial truths which cannot be applied uncritically, but must be combined according to circumstances and supplemented with empirical data. Historic population theory, or Malthusianism, pictured laborers as analogous to a good supplied under conditions of constant cost. Wages were accordingly held to tend toward an equilibrium level equal to this cost, the (real or commodity, not money) cost of maintaining a static population. The premise was not, of course, that the production of laborers takes place from motives of pecuniary profit,
*87 but that in consequence of the physiological-psychological law of population, the supply varied in a strictly analogous way. The tendency of wages to the minimum of subsistence is indeed a natural and correct deduction from the tendency of population to press constantly upon the supply of the necessaries of life.
This early version of the theory of the cost of labor was immediately recognized as untenable and gave place to the standard of living theory which depends for its validity on the assumption that the standard of living will remain stationary when the wage level changes. The classical economists recognized that an increase in the supply of labor will increase the food supply, but insisted that the second increase would be at a smaller ratio (Malthus’s crude hypothesis of arithmetic
versus geometric progression being replaced in the later work, especially that of Mill, by the scientific principle of diminishing returns).
Mill also recognized that the standard of living
might not remain stationary if the wage level were raised, but was very pessimistic (much more so than Malthus in fact) about a permanent elevation of wages unless a wide gap could be produced and maintained for a generation between actual wages and the psychological standard controlling the population. The facts seem to be that if wages are suddenly raised through a general improvement in industry or the opening-up of extensive new natural resources, the population will increase, but the psychological standard which limits its increase rises at the same time. The new equilibrium should therefore be established with a wage level higher than the old. The historic facts are of this character. The modern industrial era began with the opening-up of vast new regions to European civilization, and the movement has gone on ever since, though recently at a slackening pace. The improvement of technology has perhaps accelerated in velocity clear down to the present. The world population of European stock has increased four or five fold, and the average standard of living (if definite meaning can be given to this concept) is also vastly higher. The relative amounts of the two changes could not be measured; the writer’s conjecture would favor a vindication of the Malthusian hypothesis on the whole. Certainly both changes are still in full swing.
The most serious omission in the classical reasoning was that already referred to, the neglect to allow for the length of time required for the long-time adjustment to work itself out. Not merely may innumerable “other things” interfere with the logical course of events, but it is a serious error to view the condition of equilibrium as an approximate description at any given time. The fact of the rapid increase in the population of the industrial world, still going on, proves that the wage level has been and is far above the psychological minimum standard. It would be idle to speculate as to the length of time which would be required to bring about the equilibrium adjustment even if other things were to remain equal. It is theoretically impossible to formulate the condition of equilibrium unless the amount of disparity between present wage level and psychological minimum is accurately known, and in addition the relative rates of change of the two, corresponding to this and all lesser differences between them.
Changes in the physical composition of a population do not call for detailed discussion in this brief survey. The principal facts to be noted would be differences between an increasing and decreasing population and changes due to immigration, emigration, and internal migration. If we abstract from all human interests which do not effectively manifest themselves in the market, and assume perfect intercommunication and freedom of movement, the migration factors would quickly come to an equilibrium.
The second of our progress variables is the psychological element, the dispositions and tastes of the people. Like the number and composition of the population, it affects conditions on both the consumption and production sides of the problem. Changes and great changes do, of course, take place in wants for consumption goods and in attitudes toward different lines of productive activity.
*90 Most of these changes cannot profitably be treated as functions of price and no conditions of equilibrium can be formulated for them. They remain in the class of external disturbing causes little subject to prediction, especially on the production side. Tendencies can often be noted, such as the “lure of the city” which now operates to increase industrial production at the expense of agriculture. In America the irrational preference for white-collar jobs has raised the wages of mechanics above those of clerical tasks calling for much more ability and education. Other preferences and vogues for particular kinds of work must be passed over with the mere pointing-out that they are part of the given conditions of the economic process and that changes in them have widely ramified effects. These considerations apply to uses of property as well as to personal powers, though in a much less degree.
On the consumption side there is a very important problem more amenable to scientific treatment, though still very treacherous to deal with. We refer to the familiar fact of the use of economic resources by private business to develop, create, or direct consumptive wants; i.e., the phenomenon of advertising.
*91 The increase of value through advertising, whether informative or merely persuasive, is quite parallel to any other form of production, or “creation of utilities.” Such values are largely transferred from other goods, but except in so far as they result from a positive disparagement of competing commodities they are to be regarded as merely an additional utility in the advertised commodity.
The business of want creation is, of course, very uncertain and aleatory or “risky”; but it is evident that, as with other changes, in so far as the results of action can be foreseen, competition will equalize gains with those in other fields. Costs will then be equal to values throughout the system, the conditions of profitless adjustment being present. Whether the creation of wants is subject to diminishing returns, the process consequently tending toward an equilibrium, where it would no longer take place, or whether it is inherently a perpetual cause making for continued change, is a matter we cannot discuss on its merits. The writer’s guess would favor the latter alternative.
In regard to the third progress factor, the amount of productive resources in existence, the first question relates to the classification of these resources from the standpoint of changes in supply. We have shown above that differences must be recognized somewhat along the lines of the conventional tripartite division, but we must emphasize that the differences have been much exaggerated and that definite classification along the traditional lines cannot be maintained.
The long-time conditions of the supply of labor consist of two elements: The first, the population, has already been discussed. The second is the factor of education, taken in the broad sense. Now training, which results in increased productive efficiency, is evidently similar to a material productive agency or capital good created by the diversion of resources from present consumptive uses. Even the population itself, as observed above, depends to a large extent upon considerations of pecuniary profit in the case of the social classes which subsist mainly by labor. The distinction between labor and capital thus shows a tendency to fade away. A degree of distinction, indeed, persists. Technical training cannot be sold or leased for use separate from its owner, and cannot in any direct sense be perpetuated beyond the owner’s working life. Capital is at least less attached to its owner’s personality (it is important to note that it is never absolutely detached) and may function in perpetuity. In addition the investment in education is more affected by other than profit-seeking motives, and in consequence is not so closely adjusted by effective competition to equality of return with other forms of investment.
*94 Investment in the improvement of human powers is rather a long-time proposition, yet does not look so far ahead as many other forms of investment; in other ways, however, it is subject to a very high degree of uncertainty. After all there seems to be as much difference between different cases or types of labor production and between different varieties of material productive goods creation as there is between the two classes of investment of resources as types. In so far as uncertainty is absent and competition obtains, it is clear that investment will distribute itself between the two fields and over all parts of each in such a way as constantly to equalize their net advantages. Which is to say (remembering that costs merely register competing attractions) that with uncertainty absent costs and values would be equal throughout the system; that is, there would be a perfect, profitless organization of production and exchange.
There is a fundamental similarity in the conditions of supply of all the productive services involving the investment of resources. In every case there is a diversion of productive power from use in making present consumption goods to the creation of sources of new consumption goods income. A discussion of the conditions of equilibrium for any of them will therefore be postponed until all can be dealt with together. The general theory of equilibrium in this case is in fact the long-run theory of interest.
The classical economist treated land, or natural agents, as
given in supply. This assumption was the basis for propounding a theory of rent different from the reasoning by which the other distributive shares were explained,
*95 and for positing a special relation between rent and cost. The definition given for land to make it fit the description of a fixed supply—the original and inexhaustible powers of the soil—is indeed drastic in its limitation. Later, this dogma of unconditional fixity of supply was made the basis for the single-tax propaganda. We cannot discuss this position at length, but must take space to remark quite briefly that it is utterly fallacious. It should be self-evident that when the discovery, appropriation, and development of new natural resources is an open, competitive game, there is unlikely to be any difference between the returns from resources put to this use and those put to any other. Moreover, any disparity which exists is either a result of chance and as likely to be in the favor of one field as the other, or else is due to some difference in psychological appeal between the fields; i.e., goes to offset some other difference in their net advantages. Viewing as a whole the historic process by which land is made available for productive employment, it must be said to be “produced”; i.e., to have its utility conferred upon it in a way quite on a par with that which holds for any other exchangeable good. This, of course, again abstracts from the factor of uncertainty. In real life a large speculative element is introduced; but this cannot be said to differentiate land generically from any other class of goods, though the results are met with on an especially large scale in the case of land.
A new form of productive resource has become of very great importance in modern society, consisting of special methods of production or exclusive technical processes, whether patented or kept secret, or merely not “yet” extended in use over the whole field of production. Such a process is a source of income like any other agent, and is produced in the first place in the same way, by the investment of present resources (in research and experiment). They are different from most capital goods, however, in that their cost of maintenance and multiple reproduction is so low
*96 that it is profitable to multiply them to the point of becoming free goods, except in so far as they inhere in the persons of their possessors. They thus tend to revert to the category of enhanced individual capacities, unless in some way “monopolized.” New productive processes are like natural resources in being produced under conditions in which the gambling element is large, but in so far as the results of operations can be foreseen they also tend to equality of return on investment in comparison with other fields.
We turn, therefore, to the ordinary and simple case of the investment of resources in the creation of new productive capacities; i.e., to the case of capital goods. In this connection we can conveniently discuss the general case, subsequently returning briefly to the problems of human powers, natural agents, and productive methods just mentioned. The argument will be closely related to, in fact may be said to take up and continue, the discussion in the last chapter on the subject of time preference and the purchase and sale of productive goods. We now have the further complication that our productive goods are no longer fixed in supply, but that opportunity exists for the indefinite creation of such goods through the diversion of resources from the production of present consumption goods. For it will be seen that to the individual the investment of present goods (their use to pay productive agencies while the latter, being liberated by the “advance,”
*97 devote themselves to the making of the new equipment) is equivalent to their exchange for productive services already in existence in the possession of others; it is an alternative method for securing the same result. The previous discussion of the motivation involved, therefore, applies to the present case; i.e., it fits the assumptions usually made as to the motives for capital formation. We would emphasize the importance of a new motive not present in the former hypothetical case, the opportunity to create, which we hold to be a motive on its own account very distinct from, or at least very much more than, the mere desire to possess the thing created. However, in this brief survey, it seems necessary to abstract from the complicating factors in the motive for saving and to treat new productive equipment as a perpetual value-income merely (with the possibility of cashing in by sale at any time, as in the previous case).
The demand for capital goods is, therefore, merely the demand for future income, already discussed. Assuming a static and universally known technology, all forms of such goods will necessarily be kept at a uniform level of productivity in relation to the investment necessary to create them, and they can be treated as a homogeneous class. The demand for capital goods in industry, like that for any other productive agency, is subject to the twofold law of diminishing productivity already familiar, and the more of such goods created the lower the value income they will yield, in terms of the goods themselves measured physically. But the base on which the investor figures is not the physical productive goods created. These are as non-existent to his calculation. He is interested exclusively in the relation between (a) the amount (i.e., value) of present goods he gives up and (b) the size of the value income which he receives. Hence, we have in this case a really fourfold law of diminishing effective demand: (1) The creation of producers’ goods involves a diversion of resources from the making of consumption goods, and this transfer takes place subject to diminishing physical returns. The sacrifice of a given amount and kind of consumption goods makes possible the creation of a smaller amount of any given kind of capital goods the further the process is carried.
*99 (2) Those productive goods which are more readily multiplied by the investment of resources must increase relatively to the other agents with which they are combined in production, and become subject to diminishing physical returns in their use. (3) To the extent that the relatively increased agencies enter into the production of certain commodities more than of others these commodities will have their supply relatively increased and will fall in price relatively to other commodities. (4) Finally, as present goods are progressively sacrificed to the creation of future income, the relative preference of the latter to the former must fall off as more of it becomes available.
Other things being equal, the investment of resources should ultimately be carried to a point of equilibrium at which the amount of value income and the amount of present value which must be sacrificed to create it become equal to every person in the system. As long as the income which can be produced by sacrificing a given amount of present goods has a sufficient appeal to induce new savings, the new savings must continue to be made and to reduce the amount of value income obtainable from a given amount of investment. A point must ultimately be reached at which the product of investment is just attractive enough to hold in existence capital already saved, without calling forth new savings. Of course some individuals may at any time be consuming capital previously saved, while others are saving and investing, provided the two offset each other.
The above is a brief statement of the “eclectic” theory of interest. The equilibrium ratio of the annual value income yielded by the capital goods created to the present value sacrificed in creating them—that ratio at which no further net conversion (saving and investment) takes place—is the theoretical long-time rate of interest. It is the magnitude toward which, as Marshall says,
*101 the interest rate constantly “tends.” Of course, “other things” must be assumed to be “equal.” But in the nature of the case other things are not and cannot be equal. As investment takes place, the new income derived from it makes the saving of any given amount constantly easier, thus progressively changing the conditions of supply of new capital. In addition it is inconceivable that wants and tastes, or even the state of the arts, should remain static while such an adjustment worked itself out. The theory is logically sound if correctly understood. It describes conditions under which the interest rate would not tend to change, and is of service in predicting the future movements of the rate: But it gives a very incomplete view of the facts which must be taken into account in an actual prediction. Changes in the other things—especially the psychology of spending and saving (partly a matter of the size of income)—in the given amounts of agencies not freely reproducible through investment, and the development of technology, not to mention wars and other catastrophes—do in fact commonly exert quite as much influence on the interest rate as does the tendency to equilibrium due to progressive saving and investment.
But the most serious criticism to be made of the eclectic theory as it is currently presented (e.g., in Marshall) is its failure to recognize the true meaning of the equilibrium, and its assumption that actual conditions at a given time approach that state. The contrary is true; the case is similar to that of population already discussed, but more striking and important. At a given moment in a society where new investment is taking place the rate of capitalization is the technical ratio of conversion of present goods into future income. It is the “productivity” ratio of new investment, the ratio between the annual value yield of the capital goods to be created
*103 and the value of the present goods sacrificed to create them. Where the possibility of conversion—of saving and investment or of consuming capital already in existence through inadequate maintenance—exists, it cannot be otherwise. The psychology of saving and spending can have no appreciable influence on the interest rate at a moment. The supply of capital is not for short periods a function of the interest rate, but a fixed physical fact. Changes in psychical attitudes may cause people to save (or consume) a little more or a little less, but the effect will be insignificant in comparison with the total supply and demand of capital in the society. The rate of time preference fixes the rate at which new capital accumulates, and influences the rate of interest at future times, but not at the moment. The possibility of conversion impels every individual to equate his time preference rate to the existing productivity rate, which is causal, by saving more or less of his income or consuming more or less capital already saved.
There are no limits to the time which may be requisite at any moment to bring about the equilibrium adjustment, even assuming all other things static. Throughout the modern industrial period the rate of interest has been above the equilibrium level, social conditions being as they are (including human psychology, the
mores, and especially the concentration of income in a few hands), as is proved by the fact that capital has constantly and rapidly accumulated. How long it would take to reach the equilibrium, if the demand for capital and other things remained constant, depends on the rate at which people save corresponding to any divergence between the actual interest rate and the equilibrium rate (allowance being made for the increase in income and reduction in the psychic cost of saving) and the rapidity of operation of the law of diminishing returns in the application of new capital to other productive agencies existing in society. Historically, of course, the other things have been so far from equal—especially the demand for capital has increased so rapidly through the increase of population and opening-up of new natural resources—that the interest rate shows an astonishing constancy. We should note, also, that improvements in technology generally tend to economize labor and land and relatively increase the demand for capital. The conditions of equilibrium we can formulate; the actual course of the events which are to bring about those conditions or the length of time they will occupy are probably matters of pure and unfruitful speculation. It is quite unnecessary to believe that there will really be any progress toward equilibrium, and it goes without saying that the failure of such progress to occur militates against neither the logical soundness nor the practical utility of the theory itself.
The above analysis does not refer to an interest rate in the ordinary sense of the term, but merely to a capitalization rate or ratio of exchange between present consumption goods and income property which is also the ratio of productivity of investment to the investment where the opportunity for investment is open. It is not clear whether the phenomenon of lending free capital at interest would be met with in a society where uncertainty was absent. The capital loan is an institution or device for separating the ownership of the value of a productive agent from the ownership of the concrete thing itself. The principal, if not the only significant motive for this separation, is the uncertainty as to future changes in the value of the agents. Where this value is not subject to change, or where it is variable, but the variations are predictable, the sale price of the agency will inevitably be such as to make it a matter of complete indifference to a prospective user whether he leases the agency or buys it with borrowed funds. The loan contract is an alternative to a rental contract. Producers borrow capital and invest it, converting it into productive goods by “advancing” it to laborers, landlords, and capitalists, who furnish the resources to make the new equipment. It is apparent that the original owner of the capital could just as well invest it himself and lease the agencies thus created as to lend the money. Investment would be a practically costless operation in a world where the future was perfectly foreknown. However, it may be reasonable to suppose that the inevitable minimum of care and trouble would be sufficient to specialize the investment function and separate it from the furnishing of the capital. If so, the capital loan and interest proper would appear, the rate of interest being, of course, the capitalization and productivity ratio just discussed (less pay for investment costs if these were appreciable).
After investment is once made we have already observed that the income is simply a matter of the value yield of the goods, and the value of the agency is determined by capitalization of this yield at the interest rate determined in the market for free capital. But with freely reproducible productive goods this value can never diverge appreciably from the cost of production. Capital goods in fact differ widely in the length of time required to adjust supply to changes in demand. If there are any agencies not subject to reproduction through investment at all, they conform to the classical description of land. It is the writer’s view that such agents are practically negligible and that in the long run land is like any other capital good. Investment in exploration and development work competes with investment in other fields and is similar in all essential respects to other production costs. The distinction between goods relatively flexible and those relatively inflexible in supply and the recognition of a special category of income (Marshall’s “quasi-rent “) for the latter is possibly expedient. With uncertainty absent such a distinction is, of course, irrelevant.
We must deal briefly with the remaining items in the list of factors assumed invariable in discussing the static state. The fourth was the distribution of ownership of productive services. The only points to be noted here are that the condition affects personal powers (labor) in precisely the same way as property, and that the facts depend
entirely on social institutions. It is only because we have been accustomed to it that we think in terms of rights to income from either inherited property or inherited ability. Nor is it any more inevitable that out-and-out ownership (nearly unlimited right of control plus right to entire income) should be conferred even for his own lifetime upon an individual who by the investment of present income has developed productive powers, whether in his own person, or in produced capital goods, or by the discovery and development of natural resources.
*104 That we should separate the two categories in our thinking, taking property rights for granted in the case of inherited personal powers and stigmatizing the yield of inherited material goods as “unearned income” seems to be quite inexplicable. Society will always have to find some way to encourage the development and serious, interested use of productive capacities of all sorts (as it may always have to recognize family relationships in securing continuity of control from one generation to another). But many other ways are conceivable for doing these things, though their practical availability is not a subject for discussion here. It is to be noted that society is now progressing rapidly in the limitation of ownership, on both the control and income sides; more and more restrictions are being thrown around the use of property and the conditions under which an individual may agree to work, and more and more income is being taken through taxation for “social” purposes.
In regard to geographical distributions—much might be said on this neglected topic, but space and the plan of this work do not permit. The question of mere concentration of population, irrespective of where it is concentrated, i.e., of city
versus country, is far-reaching and fascinating. Immigration and emigration and internal migration are obviously important and intricate problems. In this field also we can recognize the condition of an ultimate equilibrium wherein the advantages of all locations would be equalized; and here also progress toward the theoretical goal is slow in comparison with the interval which separates us from it at any particular time. Changes in wants, and activities directed to change wants from motives of private gain, are especially important in this connection. It is hardly too much to say that the political as well as economic history of America has been dominated by real estate speculation and by the cheap money controversy, largely an offshoot from the former. The actual distribution of population is, of course, largely determined by the distribution of natural productive resources and by the topography of the country in relation to transportation; partly also by mere desirability of locations for residential purposes. But it is interesting to observe that considerations of consumption and social motives alone would probably bring people together in groups of all sizes and degrees of compactness even in a world whose physical conditions were absolutely uniform.
Static conditions include finally static technology and knowledge in general, and this is one of the most treacherous concepts of all as a subject for scientific discourse. Activities directed to the increase of knowledge
may be very productive, but it is too great a strain on the imagination to try to think of their results as being predictable in a particular case. We have, however, an approach to predictability in large groups; in many fields research can even now be carried on more or less “intelligently” where the scale of operations is sufficiently large. It seems almost fanciful also to speak seriously of a condition of equilibrium where the rewards or chances of reward from further effort would no longer be adequate to entice productive energy into this field. But it is clear that even here, in so far as results can be foreseen, resources will be distributed so as to secure equality of return over the whole field of investment and under competition every value realized will be just equal to the cost incurred in creating it. In this field uncertainty is indeed an inevitable concomitant of progress. Yet there is an approach to predictability, a variation in the amount of unpredictability independent of variation in the amount of progress and the two factors must be separated in the causal analysis, for their effects are very different.
This completes the list of progressive changes. In every case the necessary and sufficient condition of a perfect, remainderless distribution of the product of industry among the agencies causally concerned in creating it, in addition to perfect competition itself, is that the change can be anticipated over the period of time to which producers’ calculations relate. Where the results of the employment of resources can be foreseen, competition will force every user of any productive resource to pay all that he can afford to pay, which is its net specific contribution to the total product of industry. No sort of change interferes with the no-profit adjustment if the law of the change is known.
Principles is almost entirely devoted to the long-time equilibrium tendencies of the distributive shares, hardly more than passing notice being given to the conditions of equilibrium from the standpoint of distribution at any given time or for short periods when the supply is to be taken as fixed. Nor does he identify or even explicitly connect the question of the long-time tendencies in distribution with that of secular changes in normal price, which are phases or points of view in the analysis of the same fundamental problem of social economic organization. In the writer’s view the problem of intelligible exposition and of fundamental comprehension of the price organization can be greatly lightened by the recognition and emphasis of these lines of relation. In addition, it is helpful to stress the close analogy in methodology of treatment between the short-time price theory of value and that of distribution, and similarly with respect to the two long-time or normal price theories.
In this connection it is interesting to compare Marshall with Professor J. B. Clark, who is especially known in connection with the use of the static hypothesis in this country. Clark’s organization is even more inadequate, and it is especially striking that he does not acknowledge the connection between his method and that of Marshall. The “static state” of Clark is the same problem as Marshall’s long-time normal price, while his economic dynamics corresponds with the secular changes in the field of value and the long-time tendencies in distribution. But Clark, under Austrian and German historical influence as Marshall was under English classical, gives us as
the theory of distribution the short-time analysis, and hardly goes beyond recognizing the existence of the problem of progressive change, the long-run results or conditions of equilibrium of which are Marshall’s almost exclusive concern. He is, indeed, much less satisfactory in this field than is Marshall in the short-time theory, for the latter does give, in passing, a very fair statement of the productivity analysis. It would, of course, be a serious error to confuse Clark’s “static state” with the “stationary state” of the classical economists. The stationary state of these writers was the
naturally static or equilibrium condition, which is the goal of progress or the subject matter of the third division of the study, not a state made static by arbitrary abstraction as a methodological device. It seems, however, that virtually all discussion of static conditions is vitiated by the failure to distinguish adequately between these two concepts. And we still lack a complete discussion of distribution which will give due weight to both the short-time and long-time problems; i.e., separate the assumption of fixed supplies of productive agencies from the assumption that supply is a function of price. A rough tabulation of the natural divisions of the theory may help to clarify their relations:
(i.e., consumption goods)
Given supplies of goods and given wants to be satisfied. (The situation at a moment.)
|Market price.||No problem of distribution involved.|
Given productive resources and given wants to be satisfied.
|Normal price. (Marshall’s long-time normal price.) Supply of each good a function of price.||Short-time or market price distribution theory. (Fixed supply of thing being priced.)|
Use of resources to increase resources and change wants as well as satisfy wants.
|Secular changes in normal price.||Long-time distribution theory. Supply a function of price.|
Principles of Economics, 6th ed., p. 379.
Principles of Political Economy, book IV, chap. IV, sec. 4.
Zeitgeist complicate the problem beyond measure. The great World War, in particular, has wrought changes in human attitudes about which it would be rash to say anything except that they are certain to be far-reaching.
The suggestion may seem fanciful, but I find it impossible to differentiate between elements in the physical form and appearance of a commodity which make no difference in its efficiency for the purpose intended (an agreeable color, decorative ornament often actually interfering with its uses, fancy containers, etc.), on the one hand, and on the other an element of appeal due to a high-sounding name or any other form of “puffing.” These things do make a difference in the commodity to the consumer and in an exchange system the consumer is the last court of appeal. If they are different to him, they are different; if he is willing to buy one sort in preference to the other, then the first is superior to the second; it contains “utilities” which the other does not have. I do not see that it makes any real difference whether these utilities are in the thing itself or in some associated fact.
Quarterly Journal of Economics, vol. V. It is not so generally recognized that in consequence it explains none of them. It is especially remarkable that the theory of distribution propounded by General Francis A. Walker, whose book was long a standard text in American colleges, amounted to nothing more than an elaboration of the proposition that each factor gets what is left after the others are paid. It is easy to show that the differential theory when stated in its significant form is identical with the specific productivity theory. Cf. A. A. Young,
Ely’s Outlines of Economics, 3d ed., pp. 415-16.
(Rent in Modern Economic Theory, p. 120) contends that an idea cannot be regarded as productive, because it is “its nature” to multiply itself indefinitely. It would simplify the problem of education if it were so! But perhaps we should wish some discrimination to be exercised in the extension of the quality to ideas generally! Even so, if the “natural” tendency is obstructed, the idea limited in application seems to be productive in the sense in which anything else is productive. (See below,
mores. Then we must emphasize the impulse to create. Probably the greatest single source of saving is the putting of income back into a business, because of sheer interest in the business and the desire to make it grow. That the desire for the increased income is not the dominant motive in much of this is proved by the fact that men invest as desperately in an enterprise never likely to be profitable as they do in the most prosperous concern, and by the further fact that much of the reinvestment in society is made by directors of corporations who will not get the fruits of the work for themselves at all. The truth is, we believe, that the real motives of human life, at least of those people who do big things, are idealistic in character. The business man has the same fundamental psychology as the artist, inventor, or statesman. He has set himself at a certain work and the work absorbs and becomes himself. It is the expression of his personality; he lives in its growth and perfection according to his plans.
“ce qu’on voit et ce qu’on ne voit pas.”
However, it is to be emphasized also, that the psychology of business is fundamental in the economic process and that it is a very complex, sensitive, even treacherous thing. It will not do to draw conclusions as to policy from mere cause-and-effect reasoning based on any simple or reasonable assumptions about human behavior. Bank loans may, after all, create more demand for capital than they supply. But it is outside our plan to enter into the intricate problem of changes in business conditions or the business cycle. Some interesting suggestions in this field may be found in a series of articles on “Commercial Banking and Capital Formation,” by H. G. Moulton and Myron W. Watkins,
Journal of Political Economy, 1918 and 1919.
The correct statement of the productivity theory as given in the text manifestly sidetracks the objection of Professor Fetter and the time discount school that the product of capital is not homogeneous with the capital, and that consequently no such ratio can exist until the capitalization process has been applied to the capital itself. Before the investment is made the capital and its anticipated product are quite homogeneous, and it is in the market for capital not yet invested that the interest rate is determined. Capital goods once created are, of course, valued by capitalization; this operation presupposes an interest rate, which is therefore in no wise affected by the relation between capital goods and their income.
Part II, Chapter VI.