Risk, Uncertainty, and Profit

Frank H. Knight
Knight, Frank H.
Display paragraphs in this book containing:
First Pub. Date
Boston, MA: Hart, Schaffner & Marx; Houghton Mifflin Co.
Pub. Date
1st edition. Based on award-winning dissertation essay.

Part II


Part II, Chapter III

The Theory of Choice and of Exchange


We turn now from historical and critical considerations to the real work of construction. We have seen that the historic body of economic theory rests upon the assumption of perfect competition, but that the precise character of this assumption has been partially implicit and never adequately formulated. We do not criticize the older economists for making abstract assumptions in order to simplify and analyze their problem, but contend that the assumptions actually made and their implications need to be brought to the surface and emphasized. To display these implicit premises of theoretical reasoning is, we have argued, to explain the problem of profit, the absence of which is the essential distinction between theoretical and actual economic society. This explanation will immediately take the form of a general inquiry into "Uncertainty," the presence or absence of which will appear as the most important underlying difference*1 between the conditions which theory is compelled to assume and those which exist in fact. The present chapter and the two next following will be taken up with the attempt to define and analyze perfect competition. The argument is to be regarded as a condensed summary of classical economic theory, with especial reference to and emphasis upon those premises and implications which have not been adequately emphasized in the theory itself and have been liable to escape the observation of its readers. Aside from this special emphasis the argument will differ not a great deal from that of J. S. Mill and very little from Marshall's "Principles."


Economics is a human science; its foundations are laid in the principles of human behavior, and consequently we must begin with some observations on the psychology of human conduct which controls economic life. Economic analysis may be truly said to deal with "conduct," in the Spencerian sense, of acts adapted to ends, or of the adaptation of acts to ends, in contrast with the broader category of "behavior" in general. It assumes that men's acts are ruled by conscious motives; that, as it is more ordinarily expressed, they are directed toward the "satisfaction of wants."*2 At the very outset the science is thus subjected to notable restrictions, since it is only to a limited extent that our behavior, even our economic behavior, is of this character. Much of it is more or less impulsive and capricious. The conclusions of economic theory must in general be admitted subject to the qualification, in so far as men's economic activities are rational or planned.


This limitation is far more sweeping in its scope and import than is easily imagined. It raises the fundamental question of how far human behavior is inherently subject to scientific treatment. In his views on this point the writer is very much of an irrationalist. In this view the whole interpretation of life as activity directed toward securing anything considered as really wanted, is highly artificial and unreal. To be sure, this characterization seems to hold good for an individual at a given time and place, if the time is short enough. It is the way we think of ourselves as acting, not for the sake of the action or experience itself, but in order to some ulterior object. If, however, the object is merely accidental and temporary, such "wants" are of little service in interpreting an economic process which must look far forward. It is the writer's belief that this view of behavior, even though it is the view taken by the subject himself, is superficial at best. It appears that a relatively small fraction of the activities of civilized man are devoted to the gratification of needs or desires having any foundation beyond the mere fact that an impulse exists at the moment in the mind of the subject.


Most human motives tend on scrutiny to assimilate themselves to the game spirit. It is little matter, if any, what we set ourselves to do; it is imperative to have some objective in view, and we seize upon and set up for ourselves objectives more or less at random—getting an education, acquiring skill at some art, making money, or what-not. But once having set ourselves to achieve some goal it becomes an absolute value, weaving itself into and absorbing life itself. It is just as in a game where the concrete objective—capturing our opponents' pieces, carrying a ball across a mark, or whatever it may be—is a matter of accident, but to achieve it is for the moment the end and aim of being. And, as in a game again, so with life generally, the social situation furnishes much of the driving power, though again there are many who can become intensely interested in solitaire.


The basis of a science of conduct must be fixed principles of action, enduring and stable motives. It is doubtful, however, whether this is fundamentally the character of human life. What men want is not so much to get things that they want as it is to have interesting experiences. And the fact seems to be that an important condition of our interest in things is an element of the unanticipated, of novelty, of surprise. We must beware of the temptation to judge the nature of our conduct by the way in which we think about it. To think about it is, of course, to rationalize it, at least to "think" in the scientific sense, which has pretty well preëmpted the word. Logical thought is instrumental in character, a device for controlling and using the environment. It is, perhaps, a vice of Western civilization that the habits of thought which condition our wonderful material achievements tend to be carried over into the sphere of our personal lives. The writer ventures to surmise that this sort of thing is approaching, if it has not already reached, a climax. The fever of achievement in an external sense which now dominates our attitude toward life may be expected to give place to a saner, more epicurean view. Men will think more in terms of thought, beauty, and joy for their own sakes and less in terms of what things are good for, what can be done or gotten with them.*3


Economics, as we have observed before, is the science of a certain form of organization of human activities. The fact of organization still further limits the scope of the discussion to the rationalistic view of activity as directed to the satisfaction of wants conceived as given and permanent entities. Conduct itself is necessarily forward-looking, but organized conduct is still more so. Any machinery of organization implies relatively much taking thought, since it requires time for its development and time for its operation. A most essential feature of economic organization as it exists is its anticipation of the wants of the consumer over a long and ever longer period of production; and this anticipation implies stability in the character of the wants themselves.


A clear view of what we are doing demands special emphasis on this character of economic theory as the science of a system of organization. Human activity might be relatively unorganized or it might be organized in many different ways. History, and especially modern history, is largely the story of progressive organization and its changes in form. Organization is nearly synonymous with division of labor. In organized activity individuals perform different tasks, and each enjoys the fruits of the labor of others. The two fundamental problems of organization are the assignment of tasks and the apportionment of rewards. In unorganized action each person performs all the tasks by whose performance he benefits, and his reward is the immediate, physical benefit of his own work. But when men work together some machinery must be provided to give each his special work and to determine the amount of the results of others' effort which he shall obtain and the amount of his own product which he shall give up to others.


Modern industrial society, the "existing economic order," performs this twofold task chiefly through free agreement and voluntary exchange between individuals themselves. Economic theory is the analysis of this mechanism, viewed for the scientific purpose of simplification as the only form of human relation. Going back to mediæval times or to the American frontier, we find relatively little joint activity, except for the division of labor between the sexes and in the family. Such organization as existed for war, religion, etc., was not along free exchange lines. But there was always some commerce with different regions, and this has always been worked out largely through exchange. As time passes we find that the greatest change is in the development of organization, and especially of the voluntary, free exchange type, though, to be sure, the functions of the political state develop also. We can imagine that industrial progress might have taken a very different form. The problems of the apportionment, of tasks and rewards might be solved for a complicated, technical civilization by an autocratic, theocratic, or militaristic giving of orders and rationing of produce in which the individual would have no voice in the least detail either of his work or his enjoyment.*4 Or, again, we might have any one of numerous forms of democratic socialism. Some (the anarchists) have imagined that organization might be carried out without either exchange relations or a centralization of authority, simply by general consent. But it has been and is done principally through competitive free agreement, and our task is to study this mechanism and not any other.


The first essential of the existing system is that it solves its two fundamental problems together, as one. It is individualistic; it apportions tasks through the apportionment of rewards; it is an automatic system, in which the interrelations of individuals are determined by self-seeking on the part of each. The foundation of the process is the private ownership of productive resources—a synonym for individual freedom. There is (as we shall see more at length as we proceed) no difference in principle between the ownership of one's own powers and the ownership of other productive resources. The essence of ownership is the association or union of these two facts: (1) control of the agency, and (2) the right of disposition over its product. Modern society (on the economic side) is organized on the theory that the owners of productive resources will find their best use and place them in it, because in that way they can procure the largest returns for themselves. This system, therefore, involves the assumption that even in a complex organization the separate contribution of each separate productive agency can be identified, and that free competitive relations tend to impute to each agency its specific contribution as its reward for participation in productive activity. And to the extent that the system works at all, that we have an economic order and not chaos, this assumption must be justified.


From another point of view we may envisage the task of organization in three steps or stages:


1. Society as an organized entity must decide the relative importance of different lines of consumption as a basis for the guidance of production. Closely connected with this task, and worked out together with it, is the apportionment of existing stocks of goods, the product of past industry, in the satisfaction of existing wants. This twofold problem is worked out in the consumption goods market from day to day. The study of the process constitutes the first main division of economic science, the theory of market price.


2. Society must actually organize production. Every available productive agency is, so far as the system is successful, to be assigned to that task, and grouped with others in that way which will enable it to make the greatest possible contribution to the social dividend (of goods equated quantitatively according to the value scale established in the consumption goods market). The machinery for the direction of productive resources to their different uses is organized in the market for productive resources. The study of its workings is the second fundamental division of the science. It falls into two subdivisions, short-time distribution theory and long-time value theory.*5 For the purpose of this study the supplies of productive resources must be taken as fixed, as well as the demand which they are to satisfy. Both the prices of consumption goods and the distributive shares are in fact much affected by the third general problem cutting across both the others.


3. At the same time that society is employing existing resources to satisfy existing wants it is also setting aside a portion of its existing resources to increase the supplies of those resources themselves, to improve the effectiveness of their use by working out better methods of production, and to increase its own membership in numbers and quality by providing for an excess of births over deaths and through education and refinement. There is thus another aspect to the problems of relative importances and of organization. Decision must be made as to how much of society's income is to be diverted from present consumption and to be used for the purpose of furthering social progress, and the diverted income must be applied to this purpose as effectively as may be. The first part of the problem is solved in the market by competition between present goods and the prospective fruits of their investment, giving rise to a rate of capitalization or of interest; and the second part is solved by competition for savings between different opportunities for their use.*6


The fact that theoretical reasoning must take a large, long-run view of life leads to a difficulty in the treatment of wants which has been the source of much confusion. Our wants have the character of intermittence and recurrence; in any short period of time they are satisfied with a relatively small amount of what the want calls for, and we turn to the satisfaction of some other want. But if it is a true fundamental want it comes back again, and from a long-run point of view they all, with their satisfactions, take on the character of continuity. The periodicity, alternation between desire and satisfaction in the case of any one and dominance of different wants in succession, drops out if we look ahead a considerable distance so as to include a number of "complete cycles," so to speak. This long-run point of view is the one necessarily taken by a planned program of satisfying wants; it is evident that our activities at a moment are not predominantly affected by the thing we happen to be "hungry" for at that moment. When we go into a store to make our purchases we do not consult the momentary state of appetite or satiety in respect of any particular need, but its long-run importance in our existence viewed as a continuous process.


The problem of want-satisfaction is, therefore, a problem in proportions, or relative rates. The question is not how much absolutely of this or that, but how much—i.e., how large a share—of our time or income is to be devoted to each need or line of activity, how much per year or some other period long enough to get rid of the fluctuations. We can get the point of view by imagining that we had to plan our lives for a year on the first of January and live out the plan in detail. Economic discussion in terms of "quantities" of effort or satisfaction or choice between alternatives, under the influence of motives as immediate desires, is therefore elliptical, and more or less dangerous. The quantities of economics are properly rates, the motives not desires immediately present to consciousness, but detached judgments of need or value.


A fundamental fact about wants is their habit of conflicting among themselves. In fact, conflict seems to be essential to the very nature of conscious desire. It is questionable whether wants, as conscious motives to conduct, ever exist unless we are in a position of having to choose, to adopt one line of conduct and renounce another. Wants must be distinguished from needs which do not enter into our planful ordering of life. We "need" iodides and vitamines, and an infinite number of things of whose existence the race at large has been blissfully ignorant; but we do not "want" them, because they give rise to no conflicts and hence no "conduct." The common basis of conflict, and we may say of the existence of wants at all, is the limitation in the means of gratifying some impulse or need. When some means of satisfaction is limited in amount so that we have to plan its use and plan to increase its supply, then it enters into the field of conduct and we have a want. The most common and fundamental conflicts are between claims for our own time and energy, and after these upon some limited material agency or means employed as an aid in satisfying ourselves. Our personal powers are, of course, limited absolutely, and limited in fact still further, conditionally, by the tendency of exertion to become disagreeable, giving rise to a "want" to avoid it.*7 The confusion to be avoided is that between a want, proper, as related to consciously planned action, the weighing of alternatives, and such things as supposed needs or metaphysical explanations of the immediate fact.


The power of things to satisfy conscious wants, or quality of being wanted, is utility in the economic sense, which is equivalent to "power over conduct." Utility, of course, must have the same fundamental properties or dimensions as want; it is not, therefore, a quantity in any simple sense, but a quality having intensity, or a rate. We speak of the utility of a given amount of a thing, but this again is elliptical; the psychological variable is in fact a degree of utility of a certain rate of consumption of the good. And as want is a correlate of conflict, utility is a correlate of limitation; intensity of want and rate of supply of means of satisfying it are strictly connected, each varying inversely as the other; that is to say, as a good is supplied for the satisfaction of any want at higher rates it loses degree or intensity of utility in that use and gains (degree of) utility in the conflicting employment.*8 The confusion between a want and a need or hypothetical reason for having the want is manifest in the field of utility in ascribing economic utility to "free" goods, goods that exist in superabundance. This is a pernicious error. Such goods have no causal relation to conduct and no place in a science of conduct. The confusion has doubtless arisen from the fact that there are many things like air and water which under some circumstances do come to have power over conduct, or utility, though ordinarily they do not. This fact brings home to our consciousness their "potential" utility, the fact that they would have great utility if cut off or subject to limitation; but they have utility only when not free.


Diminishing utility is the scientific designation for the general fact that as any want is satisfied relatively to others*9 it diminishes in intensity, or, from the point of view of the means of satisfaction, that the one loses in utility and the other gains. The essential relation of conflict and relativity of utilities is somewhat obscured by the existence of intermediate "means" of satisfaction, and even of series of such. But the further course of the analysis will show that without significant exception there is always in question a diversion of the ultimate means from one use to another; it is a matter of alternatives, and the ground of one want or satisfaction being alternative to another is the dependence on a common, limited means of satisfaction.


The intermittence of wants, with wave-like alternation of desire and satisfaction, tends to give a false conception of diminishing utility. It is beside the point to talk of boys eating successive oranges or other "dinner-table" illustrations as is so commonly done. The serious error resulting from this method is that it gives the impression that there is a difference between the utilities of different portions of supply. This also is fatal to clear thinking, as will be seen if the contrast between such a situation and that of laying in supplies for a long time in advance (or even an ordinary shopping trip) is considered for a moment. The utility of any one unit is, in its effect on conduct, which is the only relevant consideration, exactly like that of any other; the essential fact is that as there are more units relatively, the utility per unit or utility of any unit is relatively less.


The fact of relativity is important, because easily and commonly lost sight of. Every valuation is a comparison; we have no conception of an absolute utility or an absolute standard of utility. The notion of value is meaningless except in relation to alternatives of choice. Not only is utility measured by another utility,—all things are measured by things of their own kind as standards, but its existence is conditioned by that of the alternative; it is like a force in the physical world; action and reaction are equal, a force cannot be imagined separate from an equal and opposite force or resistance.


The case of conflict of utilities most crucial in economic analysis is the familiar alternative of enjoying utilities at the expense of effort vs. sacrificing the utility for the sake of freedom from the exertion. "Labor" is usually thought of in an inverted, positive sense as a disutility. It is important to see that there is sufficient practical reason for this usage, but also that there is really no exception to the general principle of alternatives without distinction of kind. The point is that "labor" is really the sacrifice of some desirable alternative use of one's time and strength. If there is no alternative there is no sacrifice, nor any motivation, valuation, or "problem" of any kind. In truth, there is no distinction for conduct between a pain and the absence of a pleasure; it is all a matter of choice between alternatives, of "preference." The pleasure-pain question belongs exclusively in the field of the inner consciousness, and has no bearing on problems such as those of economics.*10 The valid reason for the distinction between kinds of alternatives, for fixing our attention on something chosen in one case and something avoided in another, is, as will be shown more at length later on, that we are interested in measuring the alternatives, and we can come nearer a satisfactory quantitative determination of time and effort than we can of the indeterminate uses that would have been made of them if the labor of producing the (measurable quantity of) goods had not been performed.


The whole theory of conduct may now be summed up, as far as it is relevant for our purposes, in a comprehensive "Law of Choice": When confronted with alternative, quantitatively variable lines of action or experience, we tend to combine them in such proportions that the physically correlated amounts or degrees of each are of equal utility to the person choosing.*11


A somewhat different statement of the principle of choice may better emphasize the basis of the alternative character of the alternative lines of conduct, the fact that not only must one give up more of the one to get more of the other, but, that this is true in a quantitative sense, that a definite amount of one is given up in return for a definite amount of the other. The reason for this fact we have found in the circumstance that the two kinds of satisfaction are both dependent on some common "means" or "resource." Accordingly we may restate the fundamental law of conduct in this way: In the utilization of limited resources in competing fields of employment, which is the form of all rational activity in conduct, we tend to apportion our resources among the alternative uses that are open in such a way that equal amounts of resource yield equivalent returns in all the fields.


This formulation makes it possibly a little more obvious that the principle is a true statement of the goal of rational planning. For, clearly, if a given unit of a given resource is yielding in one use a want satisfaction preferable to that which a similar unit is yielding in another, the yield of that resource can be increased by transferring some of it from the second use to the first until the importance of the one is increased and of the other decreased to the point of equivalence.*12


It will be apparent that utility curves, as commonly drawn, representing diminishing utility and increasing sacrifice as absolute and independent magnitudes, and ascribing varying utility to successive units of commodities (and of disutility of exertion), require considerable modification or reinterpretation if the foregoing reasoning is valid. If utility is relative and in its essence a comparison, such a curve can only represent one variable measured in terms of the other, or each curve presupposes the other already drawn. The rôle of money in the process tends to complicate and confuse the exposition still further.


The principles above stated in general terms can be brought into relation with current treatments of the subject and with concrete fact if we begin by taking up a simple case of choice between alternatives such as is constantly dealt with in economic analysis. Let us take Marshall's*13 example of a boy gathering and eating berries, but with the stipulation that some re-wording would be necessary to make the exposition accurately fit the case of choice between (i.e., combination of) alternatives in a comprehensive, long-time, plan of conduct. We can hardly suppose that the boy goes through such mental operations as drawing curves or making estimates of utility and disutility scales. What he does, in so far as he deliberates between the alternatives at all,*14 is to consider together, with reference to successive amounts of his "commodity," the utility of each increment against its "cost in effort," and evaluate the net result as either positive or negative, either of a character to prompt the combined action of production and consumption of that unit, or not of this character. The "cost in effort" is evidently in fact the sacrifice of some alternative use or uses of the effort. Even that nondescript conduct called merely idling is still conduct, an alternative motive, and subject to the law of diminishing utility or relative proportions like any other. However, while to the eye of critical scrutiny there is no "logical" distinction between an increasing disutility experienced and an increasing utility foregone, a "psychological" difference must be admitted; there is no difference for conduct, but there is one for consciousness, to our pecuniarily sophisticated consciousness at least.


If it is desired to represent the situation graphically without the misleading implications of a comparison of separate absolute variables, it can be done by omitting the commodity axis as in the accompanying figure. The line OY is merely directed in Graph III.1space to show that "preference" increases in a vertical direction. Quantities of commodity are measured by a scale as shown, but the "utilities" are not fitted to any scale at all. If we call the curve U which represents the desirability of the commodity, and the other E for exertion, the one will show a (relative) fall in value and the other a (relative) rise as the production and consumption of the commodity increases. It is a matter of indifference whether the ascending curve is thought of as a sacrifice or a positive pain, whether the growing motive to divert energy from the use in question is imaged as an attraction or a repulsion. The intersection shows that at a certain point (on the commodity scale) the diversion will take place.


Beyond this point the curves have still less meaning for the reason that the E curve really represents nothing definite, but merely any alternative whatever; as drawn they indicate a rapidly increasing pressure against this particular line of activity. The curves indicate no absolute values of any sort; the vertical distance between them alone has meaning, each being the "base" for the other; this distance shows what might be called the "net utility" of picking and eating the successive increments of berries, as compared with all possible alternatives of conduct.


A still simpler and less ambiguous way to represent the facts would be to draw on a Cartesian plane a single curve of "net utility," as in the Graph III.2accompanying sketch. This curve will cut the X or commodity axis at the point where some other alternative becomes preferable, and then fall away rapidly into the "negative utility" field. It will be seen that the Y values of the curve have only the vaguest quantitative character. The boy not only does not ask how much sacrifice is how many berries worth, but merely, are these berries worth the sacrifice; he does not even ask, "by how much" are these berries worth "the" sacrifice. There is no true psychic quantity involved; only the commodity is measured or measurable. Still, there is a certain feeling of quantitative variability in the degree of preference, and such a curve is not utterly false to the facts of consciousness. The only point of clearly determinate locus on the curve is the zero point, and it is questionable whether that is to be interpreted as a quantitative equality between opposite incentives to action or merely the absence of incentive altogether.*15


It follows at once from the non-quantitative or indefinitely quantitative character of the psychic variables*16 that the "surpluses" which have cut so much figure in economic discussion are very shadowy and elusive things, if not altogether unreal. If the ordinates of the curves discussed above mean nothing definite, of course the areas under the curves mean no more. The fallacious notion of the surplus follows naturally from the confusion between momentary satiety and the correct standpoint, the estimation of relative importance of things in planning ahead, commented on above. The illicit use of "dinner-table " illustrations in the exposition of diminishing utility shows the same error. We cannot insist too strongly upon the point that men do not determine the expenditure of their income, generally speaking, on the basis of a comparison of momentary cravings for things for instantaneous consumption. A child in a candy store would not do that. From such a viewpoint there is a psychic difference in different units of a commodity, and it might be possible to substantiate a surplus doctrine. But this is not the viewpoint of economic reasoning, because in so far as men plan at all, they do not expend their incomes and so fix the prices of things and determine the utilization of social resources and the whole structure of the competitive economic system, on the basis of that sort of calculation.*17 If we take a rational attitude toward the problem of value—as, for example, by the device, previously suggested, of placing ourselves in the position of one who had to determine the apportionment of his resources for a year or five years in advance—we shall get a different view of it. Then the earlier units are no different from the later ones, on either side of the balance; up to a certain point the balance is positive, then it suddenly becomes negative, and when the balance is struck the debits and credits are equal. There is a sort of Emersonian principle of Compensation applicable to every item; each is worth what it costs, but also costs what it is worth.


It does not at all follow that we have proved the pleasures of life just equal to its pains. That question is irrelevant to our problems, and our analysis has nothing to say about it. It is not the province of economics to determine the value of life in "hedonic units" or any other units, but to work out, on the basis of the general principles of conduct and the fundamental facts of the social situation, the laws which determine the prices of commodities and the direction of the social economic process.*18 It is therefore not quantities, nor even intensities, of satisfaction with which we are concerned (though the limitations of language compel the use of these terms at times), or any absolute magnitude whatever, but the purely relative judgment of comparative significance of alternatives open to choice. Now, for conduct, it is self-evident that the importance of anything is the effort or sacrifice necessary to get it. Two things, each of which can be obtained at will by the sacrifice of the other, cannot conceivably have any other than equal importance from this point of view, and it is meaningless to speak of a surplus. The situation is especially clear in an exchange system which fixed prices where things can be converted at will at known rates by purchase and sale. We submit that it is clearly impossible, in such a situation, to conceive of things serving as motives to action in any other than the established ratios of conversion or substitution.


For understanding the psychology of valuation, the two points are equally important: (1) that, logically, choice is a matter of comparing alternatives and combining them according, to the law of rational procedure above formulated,*19 and (2) that there is none the less a practical difference between two kinds of alternatives in an ordinary situation. This difference is perhaps connected with the distinction between our feelings of painfulness and pleasantness, but in its essence it relates to the quantitative character of the alternatives (in their physical aspects, not the psychic states involved). In the case just considered, of the boy and berries, the difference is evident from the fact that we use the berry alternative to measure the leisure alternative. We speak of a certain quantity of berries and the sacrificed alternatives corresponding to them, not of a certain quantity of alternative independently determined. The "trouble," "exertion," or what-not is not quantitative on its own account, it is measured by the berries; it is "the" amount of exertion, etc., connected with a specified amount of the measurable commodity. This result is inevitable because, as remarked above, "the" alternative is not in fact some particular alternative, but any alternative; it is not merely not measurable, but is heterogeneous and wholly indeterminate. It is this fact which throws us back on the conception of "resources" for rationalizing the deliberative process, making of it a quantitative comparison; it is this fact which gives its great importance to the "time" measure of effort. Time does not in any true sense measure the alternative or sacrifice, and, as we have seen, its employment in any use is a sacrifice in the first place only because there are other uses for it, which are the real sacrifice; but it is measurable, and our intelligence, forced to have something quantitative to feed upon, like the proverbial drowning man catches at any straw.


In spite, therefore, of the purely relative character of pain and pleasure and of the essential parity as motives of all alternatives of conduct, it is pragmatically necessary to distinguish in productive activity between the incoming "economic" utility and the sacrificed (resources, representing) non-economic, unspecified alternatives in general, between utility and disutility, or commodity and cost. "Cost," in this sense, is "pain cost," or "opportunity cost," as one prefers; there is no real difference in meaning between the two.


From this long but apparently necessary discussion of the fundamentals of valuation of psychology, we may proceed to consider a somewhat more complicated situation, as an approach to the study of the principles as manifested in the field of exchange relations. We will suppose an individual choosing between the production and consumption of a large number of "commodities," in addition to the alternative of not producing any of them, but of putting his time, etc., to "noneconomic" uses. This is the situation of Crusoe on his island, of which many economists have made use. The same law of choice will hold as before; between any two alternatives or among all that are open, the man will choose such amounts, or divide his time and "resources" among them in such proportions, that the physically alternative or correlated quantities of all are to him equally desirable. The only difference is that the alternatives are more complicated than in the case of the boy and his berries, and of a somewhat different character; in particular, the presence of a number of economic alternatives, involving concrete, measurable sources of satisfaction, is important.


In Crusoe's mind there would undoubtedly be built up something of the nature of a price system or value scale, if he seriously attempted to get the maximum of satisfaction out of the conditions of his environment. For an "intelligent" use of his opportunities can be arrived at in no other way. He must ascertain the ratios in which different goods are to be obtained for subjectively equivalent sacrifices in "effort," and similarly form judgments of their relative subjective importance to him, and attempt to bring the two sets of ratios into coincidence. But a set of equivalence ratios or scale of equivalent amounts of things is the essence of a price system. Exchange is a means by which things may be conveniently converted into or sacrificed for each other in determinate amounts, and substantially the same result follows from choosing between different lines of production in a Crusoe economy. It is sufficiently evident that the quantities involved in such a calculation are quantities of things and not of satisfaction or any psychic magnitude.


The rôle of the "resource" idea and the concept of "cost" will also take on characteristic form in the Crusoe case. The mental labor of evaluating everything in terms of everything else must force recourse to a crude measurement of "effort" as the common standard of value or "medium of exchange" (it is almost like that) for mediating the comparisons. It is clear that this is an "instrumental" but none the less very important device. "Really," it is purely a question of combining alternatives, among which are those indefinite, "non-economic" occupations, exploring the island, chatting with the parrot, sport or recreation of any appealing kind, or "loafing and inviting the soul." But the indefinite, heterogeneous, and uncertain character of these last, and the convenience of "time" as a rough basis for an approximate evaluation of the stuff they are made of, make it a matter of economy to resort to its use as a common denominator of alternatives. It will not be true that all things produced in equal times will be equated, for there are elements of "irksomeness," etc., which have to be taken account of. Crusoe's value scale will probably be based on time as a "first approximation" with mental allowances for the other factors to be considered.


Measurement relations will be reciprocal, in this case as always. The use of effort to measure other things amounts to an evaluation of effort in terms of other things. Thus we get the concept of a quantitative outlay cost meaning something more than merely any sacrificed alternative. As pointed out before, in stating in terms of "resources" the general law of choice among alternatives, this concept of cost has no very substantial independent meaning; "when pressed" we reformulate our resource or effort (or money) costs in terms of positive alternatives we might have had; but as a mediating, instrumental idea, it is none the less a useful and universally used notion. There is, however, no occasion to speak of a possible divergence between outlay cost and value return, of anything like a "profit" from operations.


There are many intermediate stages in the successive complication of alternatives which might be discussed, and which would shed light on various phases of economic relations; but for present purposes it is best to pass at once to the case of a group of people producing goods for exchange in a free market. The relations among the want-satisfying activities of a plurality of persons are based upon another "conflict," the conflict between similar wants of different individuals, to a large extent dependent on common, immediate means of satisfaction, while these immediate goods are almost entirely dependent upon a common fund of ultimate productive resources. The effect of the possibility of exchange is vastly to multiply and complicate the alternatives open to any individual. He is now free, not merely to make any possible combination of commodities for production and consumption, but to combine the production of some with the consumption of any combination—on terms afforded by an established set of exchange ratios, the investigation of which is the principal problem before us. In order to study first the most essential features of exchange relations, it will be necessary to simplify the situation as far as possible by a process of "heroic" abstraction. We therefore explicitly make the following assumptions as to the characteristics of our imaginary society:


1. The members of the society are supposed to be normal human beings in essential respects as to inherited and acquired dispositions, differing among themselves in the ways and to the degrees familiar in a modern Western nation—a "random sample" of the population of the industrial nations of to-day.


2. We assume that the members of the society act with complete "rationality." By this we do not mean that they are to be "as angels, knowing good from evil"; we assume ordinary human motives (with the reservations noted in the following paragraphs); but they are supposed to "know what they want" and to seek it "intelligently." Their behavior, that is, is all "conduct," as we have previously defined the term; all their acts take place in response to real, conscious, and stable and consistent motives, dispositions, or desires; nothing is capricious or experimental, everything deliberate. They are supposed to know absolutely the consequences of their acts when they are performed, and to perform them in the light of the consequences.


3. The people are formally free to act as their motives prompt in the production, exchange, and consumption of goods. They "own themselves"; there is no exercise of constraint over any individual by another individual or by "society"; each controls his own activities with a view to results which accrue to him individually. Every person is the final and absolute judge of his own welfare and interests.*20


4. We must also assume complete absence of physical obstacles to the making, execution, and changing of plans at will; that is, there must be "perfect mobility" in all economic adjustments, no cost involved in movements or changes. To realize this ideal all the elements entering into economic calculations—effort, commodities, etc.—must be continuously variable, divisible without limit. Productive operations must not form habits, preferences, or aversions, or develop or reduce the capacity to perform them. In addition, the production process must be constantly and continuously complete; there is no time cycle of operations to be broken into or left incomplete by sudden readjustments. Each person continuously produces a complete commodity which is consumed as fast as produced. The exchange of commodities mint be virtually instantaneous and costless.


5. It follows as a corollary from number 4 that there is perfect competition. There must be perfect, continuous, costless intercommunication between all individual members of the society.*21 Every potential buyer of a good constantly knows and chooses among the offers of all potential sellers, and conversely. Every commodity, it will be recalled, is divisible into an indefinite number of units which must be separately owned and compete effectually with each other.


6. Every member of the society is to act as an individual only, in entire independence of all other persons. To complete his independence he must be free from social wants, prejudices, preferences, or repulsions, or any values which are not completely manifested in market dealing. Exchange of finished goods is the only form of relation between individuals, or at least there is no other form which influences economic conduct. And in exchanges between individuals, no interests of persons not parties to the exchange are to be concerned, either for good or for ill. Individual independence in action excludes all forms of collusion, all degrees of monopoly or tendency to monopoly.


7. We formally exclude all preying of individuals upon each other. There must be no way of acquiring goods except through production and free exchange in the open market. This specification is really a corollary from numbers 2 and 3, which exclude fraud or deceit and theft or brigandage respectively, but it deserves explicit mention.


8. The motives for division of labor and exchange must be present and operative. These have never been adequately treated in the literature of economics in spite of the fact that the subject has been discussed more or less by countless writers on social problems from Plato down. The principal condition is diversification of wants associated with specialization of productive capacities or dispositions, or with physical restrictions on the range of productive activity. An important fact in this connection in the real world is the space distribution of the different resources of the earth and the limitations on human mobility. In addition the physical nature of the production process frequently calls for the simultaneous prosecution of a number of operations. For simplicity we shall assume that the first two conditions alone are sufficient to restrict each individual to the production of one single commodity at any given time. (Cf. number 11.)


9. All given factors and conditions are for the purposes of this and the following chapter and until notice to the contrary is expressly given, to remain absolutely unchanged. They must be free from periodic or progressive modification as well as irregular fluctuation. The connection between this specification and number 2 (perfect knowledge) is clear. Under static conditions every person would soon find out, if he did not already know, everything in his situation and surroundings which affected his conduct.


The above assumptions, especially the first eight, are idealizations or purifications of tendencies which hold good more or less in reality. They are the conditions necessary to perfect competition. The ninth, as we shall see, is on a somewhat different footing. Only its corollary of perfect knowledge (specification number 2) which may be present even when change takes place, is necessary for perfect competition. In addition to these differences in degree only from actual life, we must lay down for the special purpose of the immediate analysis two further suppositions quite contrary to the facts.


10. The first is that for the present there is to be no productive property in the ordinary sense in the society. Every productive agency or capacity is an inseparable part of the personal endowment of some member of the society. Material implements of production may be used provided they are either superabundant, and consequently free goods, or else are absolutely joined to their owners (not subject to lease or sale) and not subject to increase or decrease. The last characteristic, if not that of inseparability, is, of course, really implied in the specification of static conditions. We must also observe explicitly that personal powers themselves are similarly fixed in amount and character. The social consequences of the transfer of productive goods between individuals, and especially of their increase by "investment," will call for extended discussion later, and must be isolated by a preliminary study of a society in which they are absent.


11. The second "analytic" assumption is also contained in the preceding "idealizing" group. Under number 8 we declared that division of labor was to be carried to the point where each individual produced a single commodity. In modern industrial life it is, of course, carried vastly farther. But it is important to study separately a society where production is organized through the exchange of finished products only.*22 At a later stage we can then discuss the special problems of that further stage of organization called secondary division of labor.


This isolation is of especial importance in view of the fact that the distribution of products is very much complicated when the agencies of production coöperate in the production of a single commodity, the product of a single agent being then no longer immediately identifiable. The problem of isolating the product of a single agency, where a number work jointly, is, of course, the familiar problem of "imputation" or distribution in the technical sense, which has been the greatest single center of controversy in economic discussion.


The above list of assumptions and artificial abstractions is indeed rather a formidable array. The intention has been to make the list no longer than really necessary or useful, but in no way to minimize its degree of artificiality, the amount of divergence of the hypothetical conditions from those of actual economic life about us. For the most part these same assumptions, especially the first eight, and to a considerable extent the ninth, are really involved at one point or another in a large part of the discussion of economic literature. If they are present, and necessary, and when present whether necessary or not, there will be no disparaging the importance of having their abstract and unreal character brought conspicuously to the surface.


Our next task is to form a picture of such a society in action, and to discover the conditions of equilibrium or natural results of the operation of the forces and tendencies at work in it. We are therefore to imagine such a population, set down in such an environment as described, starting out de novo in the business of satisfying their wants. Each person, on taking in the situation in its essential outlines, will enter upon the production of some commodity, with a view, through exchange with others, of securing the means of satisfying his varied wants. After a brief interval of time has elapsed, each will have accumulated a small stock of his particular good, and we may think of them all as meeting in a central market to exchange their wares.


The situation now presented is the familiar one in economic discussion, of a group of individuals with given stocks of goods which have to be disposed of,*23 and we need not dwell upon the process by which fixed rates of exchange among all commodities will be established.*24 When the process is finished the whole mass of commodities will have been reduced to a single homogeneous fund of exchange equivalence or value. Nor do we need to concern ourselves with the mode of expressing and handling this fund; in practice it would be inevitable that some sort of standard exchange medium would be set apart; but it is immaterial for present purposes whether there is some one kind of money or as many kinds as there are different commodities.


If intercommunication is actually perfect, exchanges can take place at only one price.*25 We may imagine it to be determined all around what the ratios are to be through the medium of inquiries. Every individual, knowing the worth of the thing he possesses in terms of everything else, is in substantially the same position as a person spending a given money income in a market where selling prices are fixed by the seller and placarded. The good in his hands represents exchange power, a "resource," and he will apportion it among the possible uses according to the law of choice, so that each unit of it purchases equivalent utilities, want satisfactions, or "importances."


To show just how the price scale itself results from the fact that individuals act according to the law of choice in apportioning their purchasing power in a situation where the prices are given, is the task of that branch of economics known as the theory of market price. At any given price (ratio of sacrificing one good for the other) the more purchasing good is expended for any one commodity the less becomes the amount of want satisfaction purchased with each unit (relatively to the want-satisfying capacity either of the good given up or of any other good for which it might have been exchanged). From this it follows that the higher the price of any good (relative to others, including the purchase good), the less of it will be purchased by any individual.*26 It is therefore theoretically possible to construct a schedule, or curve, of the amounts of any good that will be taken by any individual at every price in terms of other goods, and by adding these amounts for all individuals, to construct a similar schedule for the society as a whole. But there is a fixed amount of each good available in any given short space of time to be disposed of, and it must all be sold at one price. Therefore, in a perfect market each commodity will command a definite price, which is the highest uniform price at which the entire existing stock can be disposed of (including taking out of the market by present owners).


The diagrammatic representation of the market-price equilibrium is simple and obvious. The utility relations involved in the figures and analysis for the boy-and-berries situation above*27 are applicable. The exchange situation is shown in the accompanying sketch. Graph III.3The horizontal base line is a scale of prices. The "demand" curve D shows the potential purchases at each price, for any individual or for the society as a whole, according to the scale used. The amount for sale is independent of price, a fixed physical quantity, and is represented by a horizontal line cutting the vertical or commodity axis at the proper point. The horizontal value of the intersection point gives the market price under the conditions.*28


It is especially to be observed that all the quantities involved in this whole analysis are physical and not psychic. If utility in the individual consciousness is not a true, measurable magnitude, as argued, it is still more evident that utility in any social sense, involving a sublimation of individual utilities into a "social" estimate is a wholly inadmissible supposition. The concept of social utility is in fact a mere substitute for analysis. The whole problem is precisely this of showing how an objective and uniform price results from palpably subjective and variable individual preferences. This must be done by exhibiting the interactions of individual offers and bids in the actual market.*29 We in fact know nothing about any absolute utility to any individual or about absolute amounts purchased by any one. All that can be said about the adjustment which results from perfect competition is comprised in three statements: (1) Under the conditions (the price alternatives as they are fixed) each individual achieves the goal of rational action, maximizing the want satisfaction procurable with his given resources (whatever they are) in purchasing power, by distributing them among the alternatives according to the law of choice; (2) the conditions themselves, the prices or exchange ratios being the same for all individuals, and the relative utilities adjusted to equality with these, it follows that the relative utilities of all goods (which any individual purchases at all) are the same to every individual; (3) the exchange ratios will be so adjusted that at those ratios no individual will wish to exchange anything in his possession for anything in the possession of any one else.


The emphasized expressions are so treated because of current ambiguous or actually confused conclusions in regard to the beneficence of the results of ideal competition. To call this result socially ideal or the best possible, involves assuming in addition to all the theoretical conditions as to the workings of the process itself*30 that the initial situation, the distribution of goods before the exchanges commenced, was the best possible (i.e., either absolutely ideal or absolutely beyond human power to modify). All that is true (and stated baldly it is little better than a truism) is that free exchange tends toward that redistribution of goods which is the most satisfactory all around of any that can be obtained by voluntary consent all around.


It is self-evident that in ideal exchange the quantities exchanged are equal in value terms, and there is no chance for anything like a "profit" to arise.


The main condition of perfect exchange not realized in real life is that of "perfect intercommunication," which is to say perfect knowledge of what they are doing on the part of all exchangers.*31


In our actual system middlemen fix a price which in the absence of monopoly is their best estimate of the theoretical price—which would just enable the visible supply to be disposed of—and change it from time to time as the rate of sales indicates it to be too high or too low. It is a familiar fact that in consequence of imperfect intercommunication appreciably different prices for the same commodity may obtain at different points in the general market area. Certain factors aggravate the effect of uncertainty in disturbing the theoretical adjustment: (1) Inertia or inflexibility of prices, due to habit, indifference, rounding off of figures, etc.; (2) variations in the "commodity" (and fraudulent representations of variations which do not exist); and this both in the crude physical ware, and still more in by-perquisite utilities, convenience or fashionableness of place of sale, ornamental containers, trade names, personality of vendor, etc.; (3) consumers' speculation; consumers do not buy continuously for their current needs, but lay in supplies or hold off, according to their prognostications of the market.


When terms are properly defined and allowances made for real commodity differences (which include all the factors under number 2 above) the tendency toward a definite and uniform price for similar goods is strong and conspicuous, and a fair approximation to this result is generally reached. There is, of course, the greatest difference in commodities in respect of this standardization, from wheat and cotton at one extreme to artistic products at the other.


When in our imaginary perfectly competitive society the exchanges are finished and the goods consumed, everybody will again start out to engage in production. But occupations will not be chosen as before; there will now be an established scale of prices of every good in terms of every other, and in accordance with this price scale every one will direct his effort and gauge its intensity, conforming, of course, to the Law of Choice in making his decision. The commodities produced will be thought of simply as purchasing power over goods in general, and the immediate alternatives are simply producing "wealth" and not producing it, which means doing something, or nothing (which is also doing "something") entirely outside the scale of quantitative comparisons, and this now means outside the market sphere. Every man will, therefore, like Crusoe, or the boy in the berry patch, carry his exertions to the point where utility and disutility—"really" sacrificed utility, but of an unspecified and non-quantitative sort—are of equal importance in the amounts which are alternative to each other.


As production goes on and goods accumulate in the hands of our "homines œconomici," they will be exchanged as before, distributed among the exchange possibilities in accordance with the Law of Choice; and the exchange possibilities will continuously be modified by the same process so as to be kept constantly at that point where momentarily the utility ratios of every one can be brought to equality with the price ratios. But this process of adjustment and readjustment also tends toward an equilibrium; the investigation of this tendency toward a condition in which production and consumption of all commodities would go forward at unvarying rates falls in the province of the second grand division of economic theory, one branch of which is the theory of normal price.*32


In a situation such as we have described, with the production, exchange, and consumption of commodities going on continuously, the value scale or system of quantitative equivalences of commodities, becomes much more objective and definite than it could ever be in the economy of an individual Crusoe. The constant presence of the published scale of exchange ratios and the working-out of the whole organization in terms of it must have a tremendous influence in "rationalizing" the economic activity, in impressing its quantitative features on men's minds, and enforcing precise calculations and comparisons. The result is that all goods are reduced to a homogeneous aggregate or fund of value units. This fund of value, as the medium of solving the problems of alternatives, naturally divides the economic process for each individual into two parts or stages fairly distinct in his thought. The goods he produces being thought of merely as so much value in exchange, the problems of combining alternatives in production is separated and simplified by the necessity of considering but two alternatives, as we have noted above. Similarly, the problem of consumption is considered independently, taking the form of the problem of expending value in exchange, which is worked out on its own account in accordance with the principle of rational choice or distribution of resources among competing uses. Thus value in exchange on the expenditure side, becomes like the concept of exertion to Crusoe; it is an instrumental idea, with no ontological content, but extremely useful in solving the problem of choice. The separation of the two halves of the economic problem is much heightened in real life by the storing-up of value in exchange, and the production of it for the purpose of storing it up, against unknown contingencies, with no thought of any particular use to be made of it. The separation is still further heightened by the tendency of the production of wealth to lose all connection with the notion of consuming utilities and take on the form of a competitive contest in which value in exchange becomes a mere measure of success, a counter in the game.


The further establishment and objectification of the value-system will also involve a more definite evaluation of productive sacrifices or "exertion," really the "non-economic" alternative occupations given up to perform productive labor. This evaluation being in terms of value in exchange, productive labor is in this sense brought into the general value fund, though under the conditions we are now discussing (independent individual production only) it would not actually come into the market and be exchanged. The evaluation of productive effort, i.e., its measurement in terms of an established scale of equivalences of economic alternatives, furnishes a correspondingly substantial content for the notion of "outlay cost" in a quantitative or value sense, and men's minds would undoubtedly work largely in terms of this concept.


Now it is especially important to note that at this point in the hypothetical construction we have first arrived at a set of conditions where the outlay cost of a particular good is not necessarily and axiomatically equal to the value of the good itself. For, while the readjustment toward normal price or equilibrium conditions is taking place, the "value" of the labor will be determined in the market price situation at one moment, while the value of the good which it yields will be determined at a slightly later time, and there will typically be some difference between the two. The value of the productive effort is that which the good it produces has previously had, while the value of the good it does actually produce will when it comes on the market be something else. The difference, positive or negative, between the value of a good and the (value of) its cost is analogous to "profit." Its occurrence is manifestly due to the fact that men must base their acts on past conditions, or on uncertain inferences as to the future based upon past conditions, and not on the actual future conditions to which they really relate. As soon as men find out accurately what goods are going to be worth after they are produced, they will employ their productive energy accordingly, and the profit differential will disappear. And since this is what they constantly strive to do, with some measure of success, the system will tend toward that equilibrium adjustment in which no profit exists.


The theory of the normal price adjustment is precisely analogous to that of market price, since there is no difference in principle (but only one in complication) between the purchase of a good by the sacrifice of another in exchange and its"purchase" by the sacrifice of the production of another good in its production. Both normal price and market price theories are little more than corollaries from the single fundamental Law of Choice.


On the production side of the twofold alternative, the utility or importance of any good is its purchasing power, and the higher the price the more of it will be produced, for the same reason that Crusoe would Graph III.4produce more of a more wanted good or an individual in a market purchase more of a similar one. But the higher the price of any good the less of it can be disposed of. Now since the amounts produced and disposed of are axiomatically the same, the price will move toward the point at which the natural amounts of production and sales at that price are the same. Diagrammatically, taking again a scale of prices as a horizontal basis, an ascending curve will represent the (rate of) production or supply at different prices (in terms of other goods), while a descending curve will represent the (rate of) sales or demand. The intersection of the curves gives the price point.


A slightly different way of viewing exactly the same facts will make clearer the individual motivation and show the bearings of the idea of value-cost. The demand curve, viewed from the other direction, or with the axes interchanged, is in fact a cost of production curve. The amount produced (in unit time, the rate of production) at any price is the amount that can be produced at that price without either profit or loss. For if any given price yields a profit, resources will be diverted to, and if a loss, from the production of that good; the real meaning of profit is simply that resources being used to produce other goods (and valued in the other uses) will yield more in the production of the good in question; while similarly, loss means that resources producing the good in question are worth more in other uses (their value being determined by that of the best use). From the present point of view the demand curve shows the possible selling prices of different sizes of supply, and the condition of equilibrium is that cost and selling price shall be equal. The intersection of the curves then shows on one axis the equilibrium rate of production and consumption, and on the other the equilibrium price. The character of the whole analysis as an easy deduction from the Law of Choice is clear enough without further elaboration.*33


Space does not permit us to give more consideration to these first fundamentals, and we must allow the above brief and perhaps somewhat dogmatic treatment of controverted issues to stand. It is difficult in the light of such an analysis to see any real meaning in such questions as the causal relation between cost and value, and others about which controversy has raged. Under competitive conditions a value involves an equal cost and a cost an equal value, so directly and obviously (since it is all a purely relative matter of choosing between alternatives in such a way as to equate them) that the two are but little more than different words for the same phenomenon viewed from different standpoints. Cost is the value of the resources embodied in a thing, which is to say the value of some use for them; it may be an "economic" or a "non-economic" (measurable and marketable or the opposite) use, but if there is not a competing attraction of some sort the "resources" will not be "resources" at all, just as if the thing itself is not wanted somewhere else it will not have (exchange) value, and we should say not even utility if the word is properly defined.


The whole argument is merely an elaboration of the Law of Choice (the correct form of the principle of utility), that preference ratios between alternatives will by combining the alternatives in the requisite proportions be made equal to the externally given physical equivalence ratios, first in the market and then in production. That "goods" are largely alternative to each other in production (involving the use of the same ultimate resources) is the condition of our having an economic order, an organization of want-satisfying activities based on free production and exchange. We turn now to consider the further complications of the competitive situation arising from the organization of a plurality of productive agents in the making of a single commodity.

Notes for this chapter

Outside of monopoly considerations. But see chapter VI.
This is intended as a statement of historic fact, not a dogma of necessity or desirability. To the extent that in behavior of any other sort principles may be discovered of a sufficiently general applicability to enable useful conclusions to be drawn from them, there is no reason why such principles should not be incorporated in the premises of pure theory. On the other hand, it is indisputably legitimate to begin, as an early approximation to reality, with the assumption that all the behavior of which we treat is of the character which certainly belongs to a great part of it. In any case we have to separate fundamental tendencies by such a process of analysis (i.e., abstraction) if we are to know anything about them individually. Here we are not concerned to inquire into the possibilities of an economics of instinct and reflex, much less to build up the science; we rest on the fact that the historic body of speculation has dealt with that section of behavior which we call "conduct," and, in line with our leading aim, point out the corresponding limitations of the conclusions from the reasoning. It would be futile to insist further (for those who have not grasped the point already) that limitations are no valid objection to a theory,—may even be a condition of its having any worth,—but the limitations must be recognized and appreciated.
It is impossible to follow out this line of thought to the length that its importance really justifies. Considerations somewhat along the line suggested are ably put forward in a lecture on John Ruskin as an Economist, by Patrick Geddes (The Round Table Series); also by Professor H. W. Stuart in his essay on "The Phases of the Economic Interest," in the volume by Dewey and others entitled Creative Intelligence. Cf. also Wesley C. Mitchell, "Human Behaviour and Economics," Quarterly Journal of Economics, vol. XXIX, pp. 1 ff.

At the opposite extreme a presentation of economics uncritically rationalized and devitalized to the point of approximate chemical purity may be found in the writings of Professor T. N. Carver. The old economists employed the concept of an economic man deliberately and intelligently; for Carver he is literally the man in the street.

The extinct civilizations of Mexico, and especially of Peru, are alleged to have been largely of this character.
For fuller statement see below, chapter V.
We must by no means be understood to assert or assume that these things are done ideally or even in the best practicable manner by the free exchange system of organization. In the first and third problems in particular, the formation of the social value scale and the use of resources in furthering progress, its methods and results, are open to severe criticism. But again we do not assert that there is any better method or solution practically available. It is our business simply to analyze and describe the workings of a purely voluntary, individualistic, competitive system in relation to the fundamental tasks of organization.
It is outside our purpose to attempt a detailed classification of wants. We may notice in passing the difficulty of distinguishing between really different wants and different means of satisfying the same want. For example, we may speak of the want for food, or wants for different foods; one can supply the place of another within limits, but only within limits, and finally the desire for variety itself becomes a want. In our view wants must be classified for the purposes of economic science in accordance with the actual market classification of goods. Nor shall we pretend to go into the psychological problem of the basis of desire. Our discussion deals with things in relation to conduct, and it is a matter of no concern whether we want the things or the conscious states we expect to derive from them, or what, so long as the relation between the acts themselves and the material changes toward which they are directed is clear.
There seem to be and perhaps are exceptional cases where this description does not fit the facts; there seem to be, that is, absolute wants, based on absolute limitation and not on limitation due to conflicting demand for the means of satisfaction. These are certainly of negligible importance in economics, however, and on scrutiny they have a tendency to lose the character of "wants" altogether. It is hard to see how a science can deal fruitfully in a constructive way with utterly capricious phenomena; of course it must deal with them in the sense of recognizing that they exist and form a limitation on the completeness of theory, but they can hardly be taken account of in the theory itself.
We carry some wants to complete satiety because it takes less effort than would be required to calculate accurately the most desirable place to stop when this point would be near the absolute satiety limit, as in the case of eating bread, for example. The fact may serve to illustrate the fundamental "irrationality" of a perfectly "rational" attitude to life. One of our most significant "wants" is freedom from the bother of calculating things or making close estimates. Cf. J. M. Clark, "Economics and Modern Psychology," Journal of Political Economy, vol. 26, nos. 1 and 2.
Even "for consciousness" the difference between pleasure and the absence of pain and conversely, though real, is of an "accidental" and very elusive character; we cannot formulate a difference between the two series or classify experiences between them. It is too obvious to call for discussion that the same event will be a pleasure to one person and a pain to another, and even pleasurable to the same person at one time and painful at another, according to circumstances, and, especially, expectations. The difference fades out on scrutiny. An inheritance of a hundred thousand, which is a pleasure to one to whom it is a surprise, may be an intense grief if he has expected and made his plans for ten million. A prison sentence is undoubtedly a source of joy to a man who counted on being hanged, and it is ridiculous to say that it is "really" only an escape from a worse pain, or the inheritance a deprivation of a greater pleasure. The comparison of alternatives and fact of preference is the real thing; pleasure and pain are accidental and arbitrary matters.
The phrase "equal utility," as we shall presently see, should be taken to refer merely to the fact of indifference in choice, and not a comparison between quantities in the true sense at all. We avoid the expression "marginal" utility, because of its implication that there is a difference in the significance of different portions of the same supply. In speaking of the utility of a supply, however, it is sometimes useful to have some word to distinguish between the utility per unit and the utility of the supply as a whole. When it seems advisable we shall use the expression "specific utility" to indicate utility per unit.

The general method of taking the principle of choice as the starting-point of economic reasoning and treating "diminishing utility" in a comparative sense has been used with especial clearness and force by Wicksteed (Common Sense of Political Economy), and is also adopted by Fetter in his recent work (Economic Principles). Economists generally have been coming to recognize that the psychology of the subject is properly behavioristic; that an economist need not be a hedonist (Jevons and Edgeworth notwithstanding), and that he does not need even to consider the issue between rival psychologies of choice. See Mitchell, "The Rôle of Money in Economic Theory," Proceedings, Twenty-Eighth Annual Meeting of the American Economic Association. The principle of relativity of utility and value holds in the same way under any theory of motivation. B. M. Anderson, Jr. (Social Value, and Value of Money, chap. I) advocates a theory of absolute social value, defining value, as we have done, as power to motivate conduct. It is hard to explain his failure to see that this notion is as relative as any other, is in fact the most obviously relative of all. Motivation of conduct means of "this" conduct rather than some other, and is obviously inconceivable apart from a situation presenting alternatives between which comparison and choice must be made. Davenport, also (Economics of Enterprise, chap. VII), while insisting on the importance of relative utility in economic reasoning, treats utility itself as an absolute magnitude. The present writer finds it impossible to conceive such an entity.

Close scrutiny makes it appear doubtful just how much real explanatory value the viewpoint of the utilization of resources adds to the bare principle of combining alternatives. It seems that what we call a "resource" is such, not on its own account, but solely because of the uses to which it can be put, and its quantitative aspect, how much resource there is, is still more evidently determinable only in terms of the use. But at least the resource idea helps us to mediate in thought the fact of the quantitatively alternative character of the opposed lines of utilization, as is shown by the fact that we habitually make use of it. The form of the unsophisticated psychosis in regard to sacrifices or "costs" is in fact a bit puzzling. If we ask what a thing has cost, we seem inclined to answer first in terms of money or effort, etc., i.e., of "resources"; but when pressed, we are likely to go back of the latter and evaluate the resource in turn in terms of some other utility which might have been had for it. The "ontologizing" of the notion of resources seems to be an illustration of an "instrumental concept," but one which it would be difficult to get along without.
Principles of Economics, book V, chap. II, sec. 1.
Which, to be sure, is not very far. Nor is this any criticism of the boy. Quite the contrary! It is evident that the rational thing to do is to be irrational, where deliberation and estimation cost more than they are worth. That this is very often true, and that men still oftener (perhaps) behave as if it were, does not vitiate economic reasoning to the extent that might be supposed. For these irrationalities (whether rational or irrational!) tend to offset each other. The applicability of the general "theory" of conduct to a particular individual in a particular case is likely to give results bordering on the grotesque, but en masse and in the long run it is not so. The market behaves as if men were wont to calculate with the utmost precision in making their choices. We live largely, of necessity, by rule and blindly; but the results approximate rationality fairly well on an average.
The discussion assumes that the quantitative relation between the alternatives themselves remains unchanged, that one is sacrificed for the other in the same ratio throughout, or "resources" converted into both at the same rate, In practice this is only exceptionally possible; in general not only the relative importance of given quantities of alternative goods will change as the supply changes, but in addition the amount of one which must be sacrificed to obtain a given amount of the other will increase as the supply of the first increases; i.e., a "law of diminishing productivity" (likewise a law of proportions merely) becomes operative in addition to the law of diminishing utility (and works in the same direction).

Professor Patten has raised the objection to the utility analysis that consumption also requires time, which must be saved out of the productive operations. (See Annals, Amer. Acad. 1892-93, pp. 726-28. Cf. also Edgeworth, Mathematical Psychics, p. 68, where the energy as well as time required for consumption is considered.) It seems logically more accurate, however, to include in production everything except the actual experience of satisfaction, and if this is done the objection loses its force. In our method of approach to the problem, viewing it as a matter of choice between (i.e., combination of) alternatives, and taking the alternatives simply for whatever they may be in the facts of the case, the whole issue loses its relevance.

This may be expressed in technical phrase by saying that they are "ordinal" rather than "quantitative"; they are variable, but not measurable, can be ranked, but not added. The nature of this attribute will lose its mystery if any simple sensation, as a sensation, is considered for a moment. It is easy to tell when one light is brighter than another, impossible to tell how much brighter. The intensity of light is indeed "measured" by science, but it is done by a method analogous in principle to the discussion of utility above. One light is removed to such a distance that it becomes equal in intensity to the standard, and the distance is measured. Obviously this does not involve the measurement of sensation at all. Similarly, a thermometer does not measure the sensation of heat, or a balance that of weight. A better illustration of "ordinal" variables is furnished by the field of æsthetics (another form of "value," of course). We can tell that one poem or picture is better than another, but no one would seriously propose measuring the superiority. To be sure, in school and in contests we may go through the motion of "grading" such things (even deportment!) on a percentage scale, but no one whose opinion is entitled to respect attaches any particular weight to the results of this make-believe.
That to a considerable extent purchases are based on momentary impulse and not on an estimation of relative long-time significance, is, of course, true, and perhaps increasingly so with the development of the "anti-social" arts of window-dressing, display advertising, and salesmanship. This is one of the important "allowances" that has to be made in applying economic theory to actual fact, until the progress of the science reduces the phenomena to general laws and incorporates them into the deductive system. (Cf. above, p. 52 [II.III.2-3—Ed.], and note; also p. 61, note. [II.III.17, nn8—Ed.]) Effects balance out to approximate rationality under the law of large numbers.
The doctrine of the surplus is one of the few points where the writer is compelled to disagree with Marshall on a fundamental matter of doctrine. (See Principles, 6th ed., pp. 125-33, esp. p. 129, note.) The question relates to "scope and method," however, rather than to fact or logic. I simply cannot see any use for the notion in understanding human conduct or explaining economic phenomena, and am convinced that the confusion of viewpoint which underlies putting it to the fore has led to serious error and the drawing of wholly irrelevant conclusions from economic reasoning. Moreover, an appeal to "unsophisticated common sense" seems to fail utterly to substantiate the existence of the phenomenon. A man might pay, say, a thousand dollars for the "first" loaf of bread (whichever one that is) rather than do without it, but it does not follow and is not true that when he gets it for a dime he gets $999.90 worth of free satisfaction. Various thinkers have perceived the mythical character of these alleged surpluses; it is hoped that the argument above will suggest the source of the error and so render it more easily identified and avoided.
Pages 64, 65 [II.III.22-24—Ed.].
Dependent members of the society must be completely dependent on some particular individual in it. The wants of any dependent person will then operate only through wants on his behalf felt by his sponsor, and we need not consider them at all. We need simply regard the independent members of the society as having normal solicitudes in regard to families, etc., but each person enters into economic life on an absolute equality with others or not at all.

The meaning of the above assumptions is not necessarily that they form a complete description of the people and their relations. This is but an emphatic way of saying that we here consider only their market behavior, which is assumed to conform to these specifications.

It goes without saying that our imaginary society is "isolated." Every individual who has anything at all to do with it is in it and of it on a par with all the rest.
We might characterize such a society as a "handicraft" system in contrast with "enterprise," in which the operative has lost his responsible status and lives, not by the production and sale of a commodity, but by the sale of productive services to an entrepreneur.
We treat the entire stock as for sale without reserve. The demands of present owners for their own goods, which underlie any possible reservation prices, are in fact no different from the demand of other persons, and the situation as a whole is most truthfully and significantly represented as given quantities of goods over against given dispositions to own them, since the question of whose disposition it is has nothing to do with the price that will be established. We must, of course, include the demand of present owners in the demand for every good; that it is "backed up" by the good itself instead of some other good in hand has nothing to do with the result. (Cf. Davenport, Economics of Enterprise, chap. V, pp. 48 ff.)
The problem of a perfect market is best treated mathematically (i.e., symbolically) and has been well handled by mathematical economists. See Edgeworth, Mathematical Psychics, pp. 40 ff., and Marshall, Principles, Appendix F, and Mathematical Appendix, note XII bis.
Easily proved by disproving the contrary. If exchanges be thought of as taking place at different prices the buyer at the higher price and seller at the lower will get together at an intermediate figure.
These two propositions are often treated as equivalent in economic discussion, but the relation between them is not so simple as that. To prove the second from the first, suppose that at any given price the individual has determined upon the proper amount to purchase. (For the sake of similarity with the pecuniary situation let us leave the purchase good out of account and think of a comparison between two commodities being bought with money which has no commodity value.) Now let the price of one commodity rise, relatively to that of another. If the commodity which has risen in value is a very important one, it is probable that the individual will spend as much of his resources for it as before, quite possibly even more. But he will not buy as much of the commodity, measured in physical units. For to do so he would have to spend correspondingly less resources for the alternative good, and buy less of it. But if he buys the same amount of one good as before, and less of the other, the utility ratio between the two is upset (since it was in equilibrium), and a given amount of resources is buying less utility in the good of which relatively more is purchased; resources will therefore be diverted from this good to the other. That is, he will buy less of the good which has risen (relatively) in price. Q. E. D.
Pages 66 ff [II.III.26 and nn13—Ed.].
It is also possible, but complicates matters needlessly, to plot the demand of others than present owners of the good, only, in the demand curve, and draw an ascending curve to represent the sales at different prices taking account of the present holders' reservation prices. The same data will give the same price point whichever method is used, and the one described in the text is the more significant description of the situation, since there is no practical difference in the causes or motives back of reservation prices and demand prices.
Seligman's treatment (Principles of Economics, pp. 179 ff. and 198 ff.) is a particularly glaring instance of the organism fallacy. B. M. Anderson, Jr.'s Social Value involves the same error. Anderson palpably confuses social influences back of individual judgments and preferences with social judgments and preferences in any proper sense. Of course the individual is a social product, but consciousness is still an individual phenomenon, and the conduct with which economists are concerned no less so. It is individual purchases and sales which fix prices, not social, unless in a socialistic state or one organized in some other way than through free exchange between individuals, the kind economics deals with.
See above, pp. 76-80 [II.III.39-50—Ed.].
The use of money does not affect the theory at all, and the use of circulating credit not in any way that vitiates the argument, if it does not change in value.

In one respect the actual situation is very much simplified as compared with the theoretical, and the disparities which would otherwise arise mitigated. The continuity of the process and the constant existence of published prices means in general that sellers will not come into the market at all unless they are willing to take the quoted price (or more) and buyers not unless they are willing to pay that or anything less. It is then easy to see how an excess of goods offered or an excess of purchase offers will move the price downward or upward to the equilibrium point. The real, practical problem, that is, relates to price changes, not to the establishment of price, and is vastly less complicated than the latter.

The other branch is the theory of distribution under static conditions, but under our present assumptions there is no such problem since joint production is absent.
It will be noticed that our cost curve is one of increasing costs. This is the only case to be considered from the present point of view. The question of decreasing costs comes in at a later stage of the analysis under more complicated conditions. It is obvious that to increase the production of any good involves the diversion of resources from producing other goods, which will raise their value while lowering that of the good first considered, and since resources are valued according to the best available use, this means increasing cost with increased output. At the present stage of the argument there is no problem as to the cost of any unit of commodity or yield of any unit of productive agency, since only one kind of agency is used in making any one good.

Part II, Chapter IV.

End of Notes

Return to top