An Essay on the Nature and Significance of Economic Science

CHAPTER IV

THE NATURE OF ECONOMIC GENERALISATIONS

1. We have now sufficiently discussed the subject-matter of Economics and the fundamental conceptions associated therewith. But we have not yet discussed the nature of the generalisations whereby these conceptions are related. We have not yet discussed the nature and derivation of economic laws. This, therefore, is the purpose of the present chapter. When it is completed we shall be in a position to proceed to our second main task—investigation of the limitations and significance of this system of generalisations.

2. It is the object of this essay to arrive at conclusions which are based on the inspection of Economic Science as it actually exists. Its aim is not to discover how Economics should be pursued—that controversy, although we shall have occasion to refer to it en passant,1 may be regarded as settled as between reasonable people—but rather what significance is to be attached to the results which it has already achieved. It will be convenient, therefore, at the outset of our investigations, if, instead of attempting to derive the nature of economic generalisations from the pure categories of our subject-matter,2 we proceed rather by examining specimens drawn from the existing body of analysis.

The most fundamental propositions of economic analysis are the propositions of the general theory of value. No matter what particular “school” is in question, no matter what arrangement of subject-matter is adopted, the body of propositions explaining the nature and the determination of the relation between given goods of the first order will be found to have a pivotal position in the whole system. It would be premature to say that the theory of this part of the subject is complete. But it is clear that enough has been done to warrant our taking the central propositions as established. We may proceed, therefore, to inquire on what their validity depends.

It should not be necessary to spend much time showing that it cannot rest upon a mere appeal to “History”. The frequent concomitance of certain phenomena in time may suggest a problem to be solved. It cannot by itself be taken to imply a definite causal relationship. It might be shown that, whenever the conditions postulated in any of the simple corollaries of the theory of value have actually existed, the consequences deduced have actually been observed to follow. Thus, whenever the fixing of prices in relatively free markets has taken place it has been followed either by evasion or by the kind of distributive chaos which we associate with the food queues of the late war or the French or Russian Revolutions.3 But this would not prove that the phenomena in question were causally connected in any intimate sense. Nor would it afford any safe ground for predictions with regard to their future relationship. In the absence of rational grounds for supposing intimate connection, there would be no sufficient reason for supposing that history “would repeat itself”. For if there is one thing which is shown by history, not less than by elementary logic, it is that historical induction, unaided by the analytical judgment, is the worst possible basis of prophecy.4 “History shows”, commences the bore at the club, and we resign ourselves to the prediction of the improbable. It is one of the great merits of the modern philosophy of history that it has repudiated all claims of this sort, and indeed makes it the fundamentum divisionis between history and natural science that history does not proceed by way of generalising abstraction.5

It is equally clear that our belief does not rest upon the results of controlled experiment. It is perfectly true that the particular case just mentioned has on more than one occasion been exemplified by the results of government intervention carried out under conditions which might be held to bear some resemblance to the conditions of controlled experiment. But it would be very superficial to suppose that the results of these “experiments” can be held to justify a proposition of such wide applicability, let alone the central propositions of the general theory of value. Certainly it would be a very fragile body of economic generalisations which could be erected on a basis of this sort. Yet, in fact, our belief in these propositions is as complete as belief based upon any number of controlled experiments.

But on what, then, does it depend?

It does not require much knowledge of modern economic analysis to realise that the foundation of the theory of value is the assumption that the different things that the individual wants to do have a different importance to him, and can be arranged therefore in a certain order. This notion can be expressed in various ways and with varying degrees of precision, from the simple want systems of Menger and the early Austrians to the more refined scales of relative valuations of Wicksteed and Schönfeld and the indifference systems of Pareto and Messrs. Hicks and Allen. But in the last analysis it reduces to this, that we can judge whether different possible experiences are of equivalent or greater or less importance to us. From this elementary fact of experience we can derive the idea of the substitutability of different goods, of the demand for one good in terms of another, of an equilibrium distribution of goods between different uses, of equilibrium of exchange and of the formation of prices. As we pass from the description of the behaviour of the single individual to the discussion of markets we naturally make other subsidiary assumptions—there are two individuals or many, the supply is in the hands of a monopoly or of a multiplicity of sellers, the individuals in one part of the market know or do not know what is going on in other parts of the market, the legal framework of the market prohibits this or that mode of acquisition or exchange, and so on. We assume, too, a given initial distribution of property.6 But always the main underlying assumption is the assumption of the schemes of valuation of the different economic subjects. But this, we have seen already,7 is really an assumption of one of the conditions which must be present if there is to be economic activity at all. It is an essential constituent of our conception of conduct with an economic aspect.

The propositions so far mentioned all relate to the theory of the valuation of given goods. In the elementary theory of value and exchange no inquiry is made into the conditions of continuous production. If we assume that production takes place, a new set of problems arises, necessitating new principles of explanation. We are confronted, e.g., with the problem of explaining the relation between the value of the products and the value of the factors which produced them—the so-called problem of imputation. What is the sanction here for the solutions which have been put forward?

As is well known, the main principle of explanation, supplementary to the principles of subjective valuation assumed in the narrower theory of value and exchange, is the principle sometimes described as the Law of Diminishing Returns. Now the Law of Diminishing Returns is simply one way of putting the obvious fact that different factors of production are imperfect substitutes for one another. If you increase the amount of labour without increasing the amount of land the product will increase, but it will not increase proportionately. To secure a doubling of the product, if you do not double both land and labour, you have to more than double either one of the factors. This is obvious. If it were not so, then all the corn in the world could be produced from one acre of land. It follows, too, from considerations more intimately connected with our fundamental conceptions. A class of scarce factors is to be defined as consisting of those factors which are perfect substitutes. That is to say, difference in factors is to be defined essentially as imperfect substitutability. The Law of Diminishing Returns, therefore, follows from the assumption that there is more than one class of scarce factors of production.8 The supplementary principle that, within limits, returns may increase, follows equally directly from the assumption that factors are relatively indivisible. On the basis of these principles and with the aid of subsidiary assumptions of the kind already mentioned (the nature of markets and the legal framework of production, etc.), it is possible to build up a theory of equilibrium of production.9

Let us turn to more dynamic considerations. The theory of profits, to use the word in the rather restricted sense in which it has come to be used in recent theory, is essentially an analysis of the effects of uncertainty with regard to the future availability of scarce goods and scarce factors. We live in a world in which, not only are the things that we want scarce, but their exact occurrence is a matter of doubt and conjecture. In planning for the future we have to choose, not between certainties, but rather between a range of estimated probabilities. It is clear that the nature of this range itself may vary, and accordingly there must arise not only relative valuation of the different kinds of uncertainties between themselves, but also of different ranges of uncertainty similarly compared. From such concepts may be deduced many of the most complicated propositions of the theory of economic dynamics.10

And so we could go on. We could show how the use of money can be deduced from the existence of indirect exchange and how the demand for money can be deduced from the existence of the same uncertainties that we have just examined.11 We could examine the propositions of the theory of capital and interest, and reduce them to elementary concepts of the type we have been here discussing. But it is unnecessary to prolong the discussion further. The examples we have already examined should be sufficient to establish the solution for which we are seeking. The propositions of economic theory, like all scientific theory, are obviously deductions from a series of postulates. And the chief of these postulates are all assumptions involving in some way simple and indisputable facts of experience relating to the way in which the scarcity of goods which is the subject-matter of our science actually shows itself in the world of reality. The main postulate of the theory of value is the fact that individuals can arrange their preferences in an order, and in fact do so. The main postulate of the theory of production is the fact that there are more than one factor of production. The main postulate of the theory of dynamics is the fact that we are not certain regarding future scarcities. These are not postulates the existence of whose counterpart in reality admits of extensive dispute once their nature is fully realised. We do not need controlled experiments to establish their validity: they are so much the stuff of our everyday experience that they have only to be stated to be recognised as obvious. Indeed, the danger is that they may be thought to be so obvious that nothing significant can be derived from their further examination. Yet in fact it is on postulates of this sort that the complicated theorems of advanced analysis ultimately depend. And it is from the existence of the conditions they assume that the general applicability of the broader propositions of economic science is derived.

3. Now of course it is true, as we have already seen, that the development of the more complicated applications of these propositions involves the use of a great multitude of subsidiary postulates regarding the condition of markets, the number of parties to the exchange, the state of the law, the minimum sensible12 of buyers and sellers, and so on and so forth. The truth of the deductions from this structure depends, as always, on their logical consistency. Their applicability to the interpretation of any particular situation depends upon the existence in that situation of the elements postulated. Whether the theory of competition or of monopoly is applicable to a given situation is a matter for inquiry. As in the applications of the broad principles of the natural sciences, so in the application of economic principles we must be careful to enquire concerning the nature of our material. It is not assumed that any of the many possible forms of competitive or monopolistic conditions must necessarily always exist. But while it is important to realise how many are the subsidiary assumptions which necessarily arise as our theory becomes more and more complicated, it is equally important to realise how widely applicable are the main assumptions on which it rests. As we have seen, the chief of them are applicable whenever and wherever the conditions which give rise to economic phenomena are present.

Considerations of this sort, it may be urged, should enable us easily to detect the fallacy implicit in a view which has played a great rôle in continental discussions. It has sometimes been asserted that the generalisations of Economics are essentially “historico-relative” in character, that their validity is limited to certain historical conditions, and that outside these they have no relevance to the analysis of social phenomena. This view is a dangerous misapprehension. It can be given plausibility only by a distortion of the use of words so complete as to be utterly misleading. It is quite true that in order fruitfully to apply the more general propositions of Economics, it is important to supplement them with a series of subsidiary postulates drawn from the examination of what may often be legitimately designated historico-relative material. It is certain that unless this is done bad mistakes are likely to be made. But it is not true that the main assumptions are historico-relative in the same sense. It is true that they are based upon experience, that they refer to reality. But it is experience of so wide a degree of generality as to place them in quite a different class from the more properly designated historico-relative assumptions. No one will really question the universal applicability of such assumption as the existence of scales of relative valuation, or of different factors of production, or of different degrees of uncertainty regarding the future, even though there may be room for dispute as to the best mode of describing their exact logical status. And no one who has really examined the kind of deductions which can be drawn from such assumptions can doubt the utility of starting from this plane. It is only failure to realise this, and a too exclusive preoccupation with the subsidiary assumptions, which can lend any countenance to the view that the laws of Economics are limited to certain conditions of time and space, that they are purely historical in character, and so on. If such views are interpreted to mean merely that we must realise that the applications of general analysis involve a host of subsidiary assumptions of a less general nature, that before we apply our general theory to the interpretation of a particular situation we must be sure of the facts—well and good. Any teacher who has watched good students over-intoxicated with the excitement of pure theory will agree. It may even be conceded that at times there may have been this degree of justification in the criticisms of the classical economists by the better sort of historian. But if, as in the history of the great methodological controversies has notoriously been the case, they are interpreted to mean that the broad conclusions springing from general analysis are as limited as their particular applications—that the generalisations of Political Economy were applicable only to the state of England in the early part of the reign of Queen Victoria, and such-like contentions—then it is clearly utterly misleading. There is perhaps a sense in which it is true to say that all scientific knowledge is historico-relative. Perhaps in some other existence it would all be irrelevant. But if this is so, then we need a new term to designate what is usually called historico-relative. So with that body of knowledge which is general economics. If it is historico-relative, then a new term is needed to describe what we know as historico-relative studies.

Stated this way, surely the case for the point of view underlying the so-called “orthodox” conception of the science since the time of Senior and Cairnes is overwhelmingly convincing. It is difficult to see why there should have been such fuss, why anybody should have thought it worth while calling the whole position in question. And, of course, if we examine the actual history of the controversy it becomes abundantly clear that the case for the attack was not primarily scientific and philosophical at all. It may have been the case that from time to time a sensitive historian was outraged by the crudities of some very second-rate economist—more probably by some business man or politician repeating at second-hand what he thought the economists had said. It may have been the case sometimes that a pure logician has been offended by an incautious use of philosophical terms on the part of an economist, anxious to vindicate a body of knowledge which he knows to be true and important. But in the main the attacks have not come from these quarters. Rather they have been political in nature. They have come from men with an axe to grind—from men who wished to pursue courses which the acknowledgment of law in the economic sphere would have suggested to be unwise. This was certainly the case with the majority of the leaders of the younger Historical School,13 who were the spearhead of the attack on international liberalism in the Bismarckian era. It is equally the case to-day with the lesser schools which adopt a similar attitude. The only difference between Institutionalism and Historismus is that Historismus is much more interesting.

4. If the argument which has been developed above is correct, economic analysis turns out to be as Fetter has emphasised,14 the elucidation of the implications of the necessity of choice in various assumed circumstances. In pure Mechanics we explore the implication of the existence of certain given properties of bodies. In pure Economics we examine the implication of the existence of scarce means with alternative uses. As we have seen, the assumption of relative valuations is the foundation of all subsequent complications.

It is sometimes thought, even at the present day, that this notion of relative valuation depends upon the validity of particular psychological doctrines. The borderlands of Economics are the happy hunting-ground of minds averse to the effort of exact thought, and, in these ambiguous regions, in recent years, endless time has been devoted to attacks on the alleged psychological assumptions of Economic Science. Psychology, it is said, advances very rapidly. If, therefore, Economics rests upon particular psychological doctrines, there is no task more ready to hand than every five years or so to write sharp polemics showing that, since psychology has changed its fashion, Economics needs “rewriting from the foundations upwards”. As might be expected, the opportunity has not been neglected. Professional economists, absorbed in the exciting task of discovering new truth, have usually disdained to reply: and the lay public, ever anxious to escape the necessity of recognising the implications of choice in a world of scarcity, has allowed itself to be bamboozled into believing that matters, which are in fact as little dependent on the truth of fashionable psychology as the multiplication table, are still open questions on which the enlightened man, who, of course, is nothing if not a psychologist, must be willing to suspend judgment.

Unfortunately, in the past, incautious utterances on the part of economists themselves have sometimes afforded a pretext for these strictures. It is well known that certain of the founders of the modern subjective theory of value did in fact claim the authority of the doctrines of psychological hedonism as sanctions for their propositions. This was not true of the Austrians. From the beginning the Mengerian tables were constructed in terms which begged no psychological questions.15 Böhm-Bawerk explicitly repudiated any affiliation with psychological hedonism; indeed, he went to infinite pains to avoid this kind of misconception.16 But the names of Gossen and Jevons and Edgeworth, to say nothing of their English followers, are a sufficient reminder of a line of really competent economists who did make pretensions of this sort. Gossen’s Entwicklung der Gesetze des menschlichen Verkehrs certainly invokes hedonistic postulates. Jevons in his Theory of Political Economy prefaces his theory of utility and exchange with a theory of pleasure and pain. Edgeworth commences his Mathematical Psychics with a section which urges the conception of “man as a pleasure machine”.17 Attempts have even been made to exhibit the law of diminishing marginal utility as a special case of the Weber-Fechner Law.18

But it is fundamentally important to distinguish between the actual practice of economists, and the logic which it implies, and their occasional ex post facto apologia. It is just this distinction which the critics of Economic Science fail to make. They inspect with supererogatory zeal the external façade, but they shrink from the intellectual labour of examining the inner structure. Nor do they trouble to acquaint themselves with the more recent formulations of the theory they are attacking. No doubt this has strategic advantages, for, in polemics of this kind, honest misconception is an excellent spur to effective rhetoric; and no one who was acquainted with recent value theory could honestly continue to argue that it has any essential connection with psychological hedonism, or for that matter with any other brand of Fach-Psychologie. If the psychological critics of Economics had troubled to do these things they would speedily have perceived that the hedonistic trimmings of the works of Jevons and his followers were incidental to the main structure of a theory which—as the parallel development in Vienna showed—is capable of being set out and defended in absolutely non-hedonistic terms. As we have seen already, all that is assumed in the idea of the scales of valuation is that different goods have different uses and that these different uses have different significances for action, such that in a given situation one use will be preferred before another and one good before another. Why the human animal attaches particular values in this sense to particular things, is a question which we do not discuss. That is quite properly a question for psychologists or perhaps even physiologists. All that we need to assume as economists is the obvious fact that different possibilities offer different incentives, and that these incentives can be arranged in order of their intensity.19 The various theorems which may be derived from this fundamental conception are unquestionably capable of explaining a manifold of social activity incapable of explanation by any other technique. But they do this, not by assuming some particular psychology, but by regarding the things which psychology studies as the data of their own deductions. Here, as so often, the founders of Economic Science constructed something more universal in its application than anything that they themselves claimed.

But now the question arises how far even this procedure is legitimate. It should be clear from all that has been said already that although it is not true that the propositions of analytical economics rest upon any particular psychology, yet they do most unquestionably involve elements which are of a psychological—or perhaps better said a psychical—nature. This, indeed, is explicitly recognised in the name by which they are sometimes known—the subjective or psychological theory of value; and, as we have seen, it is clear that the foundation of this theory is a psychical fact, the valuations of the individual. In recent years, however, partly as a result of the influence of Behaviourism, partly as a result of a desire to secure the maximum possible austerity in analytical exposition, there have arisen voices urging that this framework of subjectivity should be discarded. Scientific method, it is urged, demands that we should leave out of account anything which is incapable of direct observation. We may take account of demand as it shows itself in observable behaviour in the market. But beyond this we may not go. Valuation is a subjective process. We cannot observe valuation. It is therefore out of place in a scientific explanation. Our theoretical constructions must assume observable data. Such, for instance, is the attitude of Professor Cassel,20 and there are passages in the later work of Pareto21 which permit of a similar interpretation. It is an attitude which is very frequent among those economists who have come under the influence of Behaviourist psychology or who are terrified of attack from exponents of this queer cult.

At first sight this seems very plausible. The argument that we should do nothing that is not done in the physical sciences is very seductive. But it is doubtful whether it is really justified. After all, our business is to explain certain aspects of conduct. And it is very questionable whether this can be done in terms which involve no psychical element. It is quite certain that whether it be pleasing or no to the desire for the maximum austerity, we do in fact understand terms such as choice, indifference, preference, and the like in terms of inner experience. The idea of an end, which is fundamental to our conception of the economic, is not possible to define in terms of external behaviour only. If we are to explain the relationships which arise from the existence of a scarcity of means in relation to a multiplicity of ends, surely at least one-half of the equation, as it were, must be psychical in character.

Such considerations would be decisive so long as it were taken for granted that the definition of the subject-matter of Economics suggested in this essay was correct. But it might be urged that they were simply an argument for rejecting that definition and substituting one relating only to “objective”, observable matters, market prices, ratios of exchange, and so on. This is clearly what is implied by Professor Cassel’s procedure—the celebrated Ausschaltung der Wertlehre.

But even if we restrict the object of Economics to the explanation of such observable things as prices, we shall find that in fact it is impossible to explain them unless we invoke elements of a subjective or psychological nature. It is surely clear, as soon as it is stated specifically, that the most elementary process of price determination must depend inter alia upon what people think is going to happen to prices in the future. The demand functions which Professor Cassel thinks enable us to dispense with any subjective elements, must be conceived not merely as relating to prices which prevail now, or which might prevail, on present markets, but also as relating to a whole series of prices which people expect to prevail in the future. It is obvious that what people expect to happen in the future is not susceptible of observation by purely behaviourist methods. Yet, as Professor Knight and others have shown, it is absolutely essential to take such anticipations into account if we are to understand at all the mechanics of economic change. It is essential for a thorough explanation of competitive prices. It is indispensable for the most superficial explanation of monopolistic prices. It is quite easy to exhibit such anticipations as part of a general system of scales of preference.22 But if we suppose that such a system takes account of observable data only we deceive ourselves. How can we observe what a man thinks is going to happen?

It follows, then, that if we are to do our job as economists, if we are to provide a sufficient explanation of matters which every definition of our subject-matter necessarily covers, we must include psychological elements. They cannot be left out if our explanation is to be adequate. It seems, indeed, as if investigating this central problem of one of the most fully developed parts of any of the social sciences we have hit upon one of the essential differences between the social and the physical sciences. It is not the business of this essay to explore these more profound problems of methodology. But it may be suggested that if this case is at all typical—and some would regard the procedure of theory of prices as standing near the limit of proximity to the physical sciences—then the procedure of the social sciences which deal with conduct, which is in some sense purposive, can never be completely assimilated to the procedure of the physical sciences. It is really not possible to understand the concepts of choice, of the relationship of means and ends, the central concepts of our science, in terms of observation of external data. The conception of purposive conduct in this sense does not necessarily involve any ultimate indeterminism. But it does involve links in the chain of causal explanation which are psychical, not physical, and which are, for that reason, not necessarily susceptible of observation by behaviourist methods. Recognition of this does not in the least imply renunciation of “objectivity” in Max Weber’s sense. It was exactly this that Max Weber had in mind when he wrote his celebrated essays.23 All that the “objective” (that is to say, the wertfrei, to use Max Weber’s phrase) explanation of conduct involves is the consideration of certain data, individual valuations, etc., which are not merely physical in character. The fact that such data are themselves of the nature of judgments of value does not necessitate that they should be valued as such. They are not judgments of value by the observer. What is of relevance to the social sciences is, not whether individual judgments of value are correct in the ultimate sense of the philosophy of value, but whether they are made and whether they are essential links in the chain of causal explanation. If the argument of this section is correct, this question must be answered in the affirmative.

5. But now the question arises whether the generalisations of economics, in addition to being based on this fundamental assumption of relative valuations, do not also depend upon a more general psychological assumption—upon the assumption of completely rational conduct. Is it not correct to describe the subject-matter of Economics as the rational disposal of goods?24 And in this sense cannot Economics be said to depend upon another, and more contentious, kind of psychological assumption than any we have yet examined? This is a matter of some intricacy which deserves attention, not only for its own sake, but for the light it casts upon the methods of Economics in general.

Now in so far as the idea of rational action involves the idea of ethically appropriate action, and it certainly is sometimes used in this sense in everyday discussion, it may be said at once—there will be more to be said about it later—that no such assumption enters into economic analysis. As we have just seen, economic analysis is wertfrei in the Weber sense. The values of which it takes account are valuations of individuals. The question whether in any further sense they are valuable valuations is not one which enters into its scope. If the word rationality is to be construed as in any way implying this meaning, then it may be said that the concept for which it stands does not enter into economic analysis.

But in so far as the term rational is taken to mean merely “consistent”, then it is true that an assumption of this sort does enter into certain analytical constructions. The celebrated generalisation that in a state of equilibrium the relative significance of divisible commodities is equal to their price, does involve the assumption that each final choice is consistent with every other, in the sense that if I prefer A to B and B to C, I also prefer A to C: in short, that in a state of perfect equilibrium the possibility of advantage from further “internal arbitrage operations” is excluded.

There is a wider sense, too, in which the conception of rationality as equivalent to consistency can be understood as figuring in discussions of the conditions of equilibrium. It may be irrational to be completely consistent as between commodities, in the sense just described, just because the time and attention which such exact comparisons require are (in the opinion of the economic subject concerned) better spent in other ways. In other words, there may be an opportunity cost of “internal arbitrage” which, beyond a certain point, outweighs the gain. The marginal utility of not bothering about marginal utility is a factor of which account has been taken by the chief writers on the subjective theory of value from Böhm-Bawerk onwards. It is not a recent discovery. It can be taken into account in a formal sense by permitting a certain margin (or structure of margins) of inconsistency between particular valuations.

It is perfectly true that the assumption of perfect rationality figures in constructions of this sort. But it is not true that the generalisations of economics are limited to the explanation of situations in which action is perfectly consistent. Means may be scarce in relation to ends, even though the ends be inconsistent. Exchange, production, fluctuation—all take place in a world in which people do not know the full implications of what they are doing. It is often inconsistent (i.e., irrational in this sense) to wish at once for the fullest satisfaction of consumers’ demands, and at the same time to impede the import of foreign goods by tariffs or such-like obstacles. Yet it is frequently done: and who shall say that economic science is not competent to explain the situation resulting?

Of course there is a sense in which the word rationality can be used which renders it legitimate to argue that at least some rationality is assumed before human behaviour has an economic aspect—the sense, namely, in which it is equivalent to “purposive”. As we have seen already, it is arguable that if behaviour is not conceived of as purposive, then the conception of the means-end relationships which economics studies has no meaning. So if there were no purposive action, it could be argued that there were no economic phenomena.25 But to say this is not to say in the least that all purposive action is completely consistent. It may indeed be urged that the more that purposive action becomes conscious of itself, the more it necessarily becomes consistent. But this is not to say that it is necessary to assume ab initio that it always is consistent or that the economic generalisations are limited to that, perhaps, tiny section of conduct where all inconsistencies have been resolved.

The fact is, of course, that the assumption of perfect rationality in the sense of complete consistency is simply one of a number of assumptions of a psychological nature which are introduced into economic analysis at various stages of approximation to reality. The perfect foresight, which it is sometimes convenient to postulate, is an assumption of a similar nature. The purpose of these assumptions is not to foster the belief that the world of reality corresponds to the constructions in which they figure, but rather to enable us to study, in isolation, tendencies which, in the world of reality, operate only in conjunction with many others, and then, by contrast as much as by comparison, to turn back to apply the knowledge thus gained to the explanations of more complicated situations. In this respect, at least, the procedure of pure economics has its counterpart in the procedure of all physical sciences which have gone beyond the stage of collection and classification.

6. Considerations of this sort enable us to deal also with the oft-reiterated accusation that Economics assumes a world of economic men concerned only with money-making and self-interest. Foolish and exasperating as this may appear to any competent economist, it is worth some further examination. Although it is false, yet there is a certain expository device of pure analysis which, if not explained in detail, might give rise to strictures of this nature.

The general absurdity of the belief that the world contemplated by the economist is peopled only by egotists or “pleasure machines” should be sufficiently clear from what has been said already. The fundamental concept of economic analysis is the idea of relative valuations; and, as we have seen, while we assume that different goods have different values at different margins, we do not regard it as part of our problem to explain why these particular valuations exist. We take them as data. So far as we are concerned, our economic subjects can be pure egoists, pure altruists, pure ascetics, pure sensualists or—what is much more likely—mixed bundles of all these impulses. The scales of relative valuation are merely a convenient formal way of exhibiting certain permanent characteristics of man as he actually is. Failure to recognise the primacy of these valuations is simply a failure to understand the significance of the last sixty years of Economic Science.

Now the valuations which determine particular transactions may be of various degrees of complexity. In my purchase of bread I may be interested solely in the comparison between the bread and the other things in the circle of exchange on which I might have spent the money. But I may be interested too in the happiness of my baker. There may exist between us certain liens which make it preferable for me to buy bread from him, rather than procure it from his competitor who is willing to sell it a little cheaper. In exactly the same way, in my sale of my own labour or the hire of my property, I may be interested only in the things which I receive as a result of the transaction; or I may be interested also in the experience of labouring in one way rather than another, or in the prestige or discredit, the feeling of virtue or shame in hiring out my property in this line rather than in that.

All these things are taken into account in our conception of scales of relative valuation. And the generalisations descriptive of economic equilibrium are couched in a form which explicitly brings this to the fore. Every first-year student since the days of Adam Smith has learnt to describe equilibrium in the distribution of particular grades of labour in terms of a tendency, not to the maximisation of money gains, but to the maximisation of net advantages in the various alternatives open.26 As we have seen already, the theory of risk, too, and its influence on the capital market depends essentially on assumptions of this kind. But sometimes for purposes of exposition it is convenient to start from the first approximation that the valuation is of a very simple order, and that, on the one side is a thing desired or offered, and on the other is the money to be got or given in exchange for it. For the elucidation of certain complicated propositions, such as the theory of costs or marginal productivity analysis, it permits an economy of terms. It is not in the least difficult, at the appropriate stage, to remove these assumptions and to pass to analysis couched in terms of complete generality.

This, then, is all that lies behind the homo œconomicus—the occasional assumption that in certain exchange relationships all the means, so to speak, are on one side and all the ends on the other. If. e.g., for purposes of demonstrating the circumstances in which a single price will emerge in a limited market, it is assumed that in my dealings in that market I always buy from the cheapest seller, it is not assumed at all that I am necessarily actuated by egotistical motives. On the contrary, it is well known that the impersonal relationship postulated is to be seen in its purest form when trustees, not being in a position to allow themselves the luxury of more complicated relationships, are trying to make the best terms for the estates they administer: your business man is a much more complicated fellow. All that it means is that my relation to the dealers does not enter into my hierarchy of ends. For me (who may be acting for myself or my friends or some civic or charitable authority) they are regarded merely as means. Or, again, if it is assumed—which in fact is usually done for purposes of showing by contrast what the total influences in equilibrium bring about—that I sell my labour always in the dearest market, it is not assumed that money and self-interest are my ultimate objects—I may be working entirely to support some philanthropic institution. It is assumed only that, so far as that transaction is concerned, my labour is only a means to an end; it is not to be regarded as an end in itself.

If this were commonly known, if it were generally realised that Economic Man is only an expository device—a first approximation used very cautiously at one stage in the development of arguments which, in their full development, neither employ any such assumption nor demand it in any way for a justification of their procedure—it is improbable that he would be such a universal bogey. But of course it is generally thought that he has a wider significance, that he lurks behind all those generalisations of the “Laws of Supply and Demand” better described as the theory of comparative statics, whose elucidation so often is inimical to the desire to be able to believe it to be possible both to have your cake and to eat it. And it is for this reason that he is so furiously attacked. If it were Economic Man who barred the gates of Cloud-cuckoo-land, then it might well seem that a little psychology—it would not matter much of what brand—might be expected to burst them open. What prestige, what repute for really deep insight into human motivation might be expected to accrue from so spectacular an exposure!

Unfortunately this belief rests upon misapprehension. The propositions of the theory of variations do not involve the assumption that men are actuated only by considerations of money gains and losses. They involve only the assumption that money plays some part in the valuation of the given alternatives. And they suggest only that if from any position of equilibrium the money incentive is varied this must tend to alter the equilibrium valuations. Money may not be regarded as playing a predominant part in the situation contemplated. So long as it plays some part then the propositions are applicable.

A simple illustration should make this quite plain. Let us suppose that a small bounty is granted in respect of the production of an article produced under conditions of free competition. According to well-known theorems there will be a tendency for the production of that commodity to increase—the magnitude of the increase depending upon considerations of elasticity into which it is not necessary for us to enter. Now upon what does this generalisation depend? Upon the assumption that producers are actuated only by considerations of monetary gain? Not at all. We may assume that they take into account all the “other advantages and disadvantages” with which Cantillon and Adam Smith have made us familiar. But, if we assume that before the bounty was granted there was equilibrium, we must assume that its institution must disturb the equilibrium. The granting of the bounty implies a lowering of the terms on which real income is obtainable in this particular line of enterprise. It is a very elementary proposition that if a price is lowered the demand tends to increase.

There is perhaps one refinement of this conclusion which needs to be stated explicitly. It may quite well be that, if the change contemplated is a very small one, no primary movement will take place.27 Is this in contradiction with our theory? Not at all. The idea of scales of valuation does not assume that every physical unit of any of the things which enter into the range of effective valuation must necessarily have a separate significance for action. In the assumption of the hierarchy of alternatives we do not ignore the fact that, for change to be effective, it must attain the minimum sensibile.28 Changes in price of a penny or twopence may not affect the habits of a given economic subject. But this is not to say that changes of a shilling will not be effective. Nor is it to say that, given limited resources, the necessity of spending more or less on one thing does not inevitably affect the distribution of expenditure, even if in the line of expenditure directly affected it leaves the quantity demanded unchanged.

7. In the light of all that has been said the nature of economic analysis should now be plain. It consists of deductions from a series of postulates, the chief of which are almost universal facts of experience present whenever human activity has an economic aspect, the rest being assumptions of a more limited nature based upon the general features of particular situations or types of situations which the theory is to be used to explain.

It is sometimes thought, however, that such a conception is essentially statical in nature, that it relates only to descriptions of final positions of equilibrium, variations being essentially outside its scope. Since the world of reality is not in a state of equilibrium, but rather exhibits the appearance of incessant change, it follows that knowledge of this sort has little explanatory value. This belief, which apparently is very widespread, needs further examination.

Now it is quite true that the elementary propositions of economic analysis are descriptions of stationary equilibrium. We start by examining, not conditions of complete rest, as in the Statics from which by analogy the name of this part of our subject is sometimes taken, but conditions in which the various “flows” of activity exhibit no tendency to change, or change only in a recurrent cycle.29 Thus we may take the conditions of a simple market in which the fundamental conditions of supply and demand are unaltered from day to day and enquire under what conditions would the quantities exchanged day by day remain invariable, even though the parties to the exchange were free to vary their bargains. Or we may consider the case in which the production takes place, but in which the fundamental data—that is, the valuations of the economic subjects, the technical possibilities of production and the ultimate supplies of the factors—are unchanged, and enquire under what condition there would be no tendency to change in the rate of flow of products. And so on. There is no need to rehearse the whole list of possibilities; any of the more rigorous textbooks on the subject—for instance, Wicksell’s Lectures on Political Economy, or Walras’ Elements—provide examples of the sort of thing under discussion.

But it is quite wrong to suppose that our investigations are limited to these essential preliminaries. Once we have thoroughly investigated the conditions of constant flows, and hence learnt by contrast to understand the conditions in which the flows will be tending to alter, we may push our investigations further and consider variations.

We may do this in two ways. In the first place, we may compare the equilibrium positions, assuming small variations in the data. Thus we may assume the imposition of a tax, the discovery of a change in technical methods, a change in tastes, and so on. And we may endeavour to ascertain in what respects one equilibrium position differs from the other. The so-called classical analysis, imperfect as a full description of final states of equilibrium, provides a great variety of useful comparisons of differences of this sort. This part of our theory has sometimes been called the theory of comparative statics.30

But we may go beyond this. Not only may we compare two final states of equilibrium assuming given variations, we may also endeavour to trace out the path actually followed by different parts of a system if a state of disequilibrium is given. This, of course, is the significance of Marshall’s “period” analysis. Into this category falls also much of what is most significant in the theory of money and banking. And in doing all this we make no assumption that final equilibrium is necessary. We assume that there are operative in different parts of the system certain tendencies which make for the restoration of an equilibrium in respect to certain limited points of reference. But we do not assume that the composite effect of these tendencies will necessarily be equilibrating. It is easy to conceive of initial configurations of the data, which have no total tendency to equilibrium, but which rather tend to cumulative oscillation.31

In all this, as will be obvious to anyone acquainted with the procedure of economic analysis, our knowledge of the statical foundations is fundamental.32 We examine change by comparing small differences of equilibrium or by comparing the effects of different tendencies to equilibrium; it is difficult to see what other procedure could be adopted. But it should be equally obvious that we study these statical problems not merely for their own sake, but in order to apply them to the explanation of change. There are certain propositions of economic statics which are significant and important in themselves. But it is hardly an exaggeration to say that their chief significance lies in their further application in economic dynamics. We study the laws of “rest” in order to understand the laws of change.

But now the question arises, Can we not even transcend all this? Will the dynamic operations described so far relate to the study of the effects of given variations in the data, or the consequences of given disequilibria? Can we not go outside all this and explain changes in the data themselves? This raises questions which can be treated more conveniently in another chapter.

• 1“Economic Science in Recent Discussion”, American Economic Review, vol. xxiv., pp. 225–238.
• 2Unsettled Questions of Political Economy, p. 120.
• 3“Nicht die ‘sachlichen’ Zusammenhänge der ‘Dinge’ sondern die gedanklichen Zusammenhänge der Probleme liegen den Arbeitsgebieten der Wissenschaften zugrunde” (Max Weber, Die Objectivität sozialwissenschaftlicher und sozialpolitischer Erkenntnis, Gesammelte Aufsätze zur Wissenschaftslehre, p. 166).
• 4See Cannan, Review of Economic Theory, pp. 1–35, and Schumpeter, Epochen der Methoden- und Dogmengeschichte, pp. 21–38.
• 5See Chapter II., Section 5, especially the footnote on p. 42, for further elaboration of this point.
• 6Wealth, 1st edition, p. 17.
• 7Principles, 8th edition, p. 1.
• 8Cours d’Economie Politique, p. 6.
• 9Essentials of Economic Theory, p. 6. See also Philosophy of Wealth, ch. i. In this chapter the difficulties discussed below are explicitly recognised, but, surprisingly enough, instead of this leading to a rejection of the definition, it leads only to a somewhat surprising attempt to change the significance of the word “material”.
• 10But see Chapter II. below for an examination of the validity of this interpretation.
• 11In this connection it is perhaps worth while clearing up a confusion which not infrequently occurs in discussions of terminology. It is often urged that scientific definitions of words used both in ordinary language and in scientific analysis should not depart from the usages of everyday speech. No doubt this is a counsel of perfection, but in principle the main contention may be accepted. Great confusion is certainly created when a word which is used in one sense in business practice is used in another sense in the analysis of such practice. One has only to think of the difficulties which have been created by such departures in regard to the meaning of the term capital. But it is one thing to follow everyday usage when appropriating a term. It is another thing to contend that everyday speech is the final court of appeal when defining a science. For in this case the significant implication of the word is the subject-matter of the generalisations of the science. And it is only by reference to these that the definition can finally be established. Any other procedure would be intolerable.
• 12Cannan, An Economist’s Protest, p. 49.
• 13Wealth of Nations (Cannan’s ed.), p. 315.
• 14Theories of Production and Distribution, pp. 18–31; Review of Economic Theory, pp. 49–51.
• 15It is even arguable that the reaction has gone too far. Whatever its demerits, the Smithian classification had a significance for capital theory which in recent times has not always been clearly recognised. See Taussig, Wages and Capital, pp. 132–151.
• 16The Nature of Capital and Income, ch. vii.
• 17This is the title of ch. ii. of Wealth (1st edition).
• 18There are other quarrels which we might pick with this particular definition. From the philosophical point of view, the term “material welfare” is a very odd construction. “The material causes of welfare” might be admitted. But “material welfare” seems to involve a division of states of mind which are essentially unitary. For the purposes of this chapter, however, it has seemed better to ignore these deficiencies and to concentrate on the main question, namely, whether the definition can in any way describe the contents of which it is intended to serve as a label.
• 19It is perhaps worth emphasising the significance of this qualification. The application of technically similar means to the achievement of qualitatively similar ends at different times constitute alternative uses of these means. Unless this is clearly realised, one of the most important types of economic action is overlooked.
• 20This may seem an unnecessary refinement, and in the first edition of this essay I left it out for that reason. But the condition that there exists a hierarchy of ends is so important in the theory of value that it seems better to state it explicitly even at this stage. See Chapter IV., Section 2.
• 21Cp. Schönfeld, Grenznutzen und Wirtschaftsrechnung, p. 1; Hans Mayer, Untersuchungen zu dem Grundgesetze der wirtschaftlichen Wertrechnung (Zeitschrift für Volkswirtschaft und Sozialpolitik, Bd. 2, p. 123).
• 22It should be clear that there is no disharmony between the conception of end here employed, the terminus of particular lines of conduct in acts of final consumption, and the conception involved when it is said that there is but one end of activity—the maximising of satisfaction, “utility”, or what not. Our “ends” are to be regarded as proximate to the achievement of this ultimate end. If the means are scarce they cannot all be achieved, and according to the scarcity of means and their relative importance the achievement of some ends has to be relinquished.
• 23Cp. Menger, Grundsätze der Volkswirtschaftslehre, lte Aufl., pp. 51–70; Mises, Die Gemeinwirtschaft, pp. 98 seq.; Fetter, Economic Principles, ch. i.; Strigl, Die ökonomischen Kategorien und die Organisation der Wirtschaft, passim; Mayer, op. cit.
• 24On the distinction between analytical and classificatory definitions, see Irving Fisher, Senses of Capital (Economic Journal, vol. vii., p. 213). It is interesting to observe that the change in the conception of Economics implied by our definition is similar to the change in the conception of capital implied in Professor Fisher’s definition. Adam Smith defined capital as a kind of wealth. Professor Fisher would have us regard it as an aspect of wealth.
• 25See his Objekt und Grundbegriffe der theoretischen Nationalökonomie, 2 Aufl. The criticisms of Schumpeter and Strigl on pp. 110–125 and pp. 165–156 are particularly important from this point of view. With the very greatest respect for Professor Amonn’s exhaustive analysis, I cannot resist the impression that he is inclined rather to magnify the degree of his divergence from the attitude of these two authors.
• 26See Mises, Die Gemeinwirtschaft, pp. 94–138. In his Economic Planning in Soviet Russia, Professor Boris Brutzkus has well shown the way in which this difficulty has been exemplified in the various phases of the Russian experiment.
• 27See Strigl, op. cit., pp. 23–28.
• 28Professor Cassel’s dismissal of Crusoe Economics (Fundamental Thoughts, p. 27) seems unfortunate since it is only when contemplating the conditions of isolated man that the importance of the condition that the scarce means must have alternative uses if there is to be economic activity, which was emphasised above, leaps clearly to the eye. In a social economy of any kind, the mere multiplicity of economic subjects leads one to overlook the possibility of the existence of scarce goods with no alternative uses.
• 29The objections outlined above to the definition suggested by Professor Amonn should be sufficient to indicate the nature of the objections to those definitions which run in terms of phenomena from the standpoint of price (Davenport), susceptibility to the “measuring rod of money” (Pigou), or the “science of exchange” (Landry, etc.). Professor Schumpeter, in his Wesen und Hauptinhalt der theoretischen Nationalökonomie, has attempted with never to be forgotten subtlety to vindicate the latter definition by demonstrating that it is possible to conceive all the fundamental aspects of behaviour germane to Economic Science as having the form of exchange. That this is correct and that it embodies a truth fundamental to the proper understanding of equilibrium theory may be readily admitted. But it is one thing to generalise the notion of exchange as a construction. It is another to use it in this sense as a criterion. That it can function in this way is not disputed. But that it throws the maximum light on the ultimate nature of our subject-matter is surely open to question.
• 30Wealth (1st edition), ch. i.
• 31On the various types of equilibrium contemplated, see Knight, Risk, Uncertainty and Profit, p. 143, note; Wicksell, Lectures on Political Economy, vol. i; and Robbins, On a Certain Ambiguity in the Conception of Stationary Equilibrium (Economic Journal, vol. xl., pp. 194–214).
• 32See Pareto, Manuel d’Economie Politique, p. 147; also my article on Production in the Encyclopædia of the Social Sciences. In the first edition of this essay I subsumed the theory of comparative statics and the theory of dynamic change under the single heading, “Theory of Variations.” I now think it it better to make explicit the two types of variation theory. For further elucidations see below, Chapter IV., Section 7.