ABSTRACT
The chain of reasoning by which the founders of the Austrian School in economics reached the notion of marginal utility was somewhat like this: they took as a starting point the empirical fact that when a man gets more of a certain good there is a certain feature in his state of mind that changes. In precise objective terms, it is difficult to explain what this change really consists of. But everybody has many times lived through such a process himself. Therefore it is possible to give some kind of description of the process, simply by referring to this internal experience and, so to speak, leave it to the reader or listener himself to supply the concrete meaning of the words. If, on this basis, we want a verbal definition of the process, we can say that the want or the craving for the good in question diminishes as one gets more of it. The satisfaction that is obtained by the consumption of the last unit of the good or, if one prefers, the utility created by this satisfaction, may be called the ‘marginal utility’. This definition has none of the objective criteria that we are accustomed to look for in the definitions of the natural sciences. It has a sense and is understood by other people only because these people themselves have had an internal experience, a want satisfaction, that corresponds to the definition given. It is a definition of the same kind as one we would obtain, say, by defining ‘force’ as ‘the strain exerted when heavy things are lifted’. In particular it should be noticed that this primitive utility definition does not establish marginal utility as a quantitative notion. Nevertheless, in the earlier days of marginal utility theory the notion was often handled as if it should have been quantitatively defined. It is easy to understand why it was handled this way. If marginal utility shall be of any use at all in
the theoretical structure of economics, it must be drawn into a quantitative reasoning. Those phenomena which marginal utility shall explain are, indeed, by their very nature quantitative. For instance: in some way or another one has to introduce the idea that when equilibrium is reached the prices are proportional to the marginal utility. Such an idea evidently implies that marginal utility is quantitatively defined. Something which shall be proportional to something else must necessarily be quantitatively defined. This procedure of drawing a non-quantitatively defined notion into a quantitative argument is, of course, entirely illegitimate. It really amounts to giving the notion a meaning by appealing to a certain mental association process in the reader or the listener, and then when a certain number of steps in the reasoning are passed and the attention of the reader or listener is dulled, to use the notion in a different meaning, to attribute to it features which do not lie in the given definition. As a matter of fact, this point was just one of those where the marginal utility theory in its original form was justly criticized. These criticisms started a series of attempts at rigorously defining utility in a quantitative way. These attempts proceed on two different lines which we may term the psychological and the behaviouristic. These terms are here used in a special economic-technical sense. One should not read into them the far-reaching implications of the controversy between the psychological and the behaviouristic schools in psychology, although the distinction here made is, of course, somewhat related to the point at issue in psychology. The exponents of the psychological and the behaviouristic approach to a quantitative definition of economic utility are Edgeworth and Fisher, respectively. Edgeworth believed that he could make the notion of marginal utility more precise by an appeal to experimental psychology.1 This was a natural avenue of approach if one would follow as closely as possible the original idea of the Austrian School. As a unit of utility, Edgeworth adopts the smallest additional utility that the individual can perceive. And on this basis he builds his hedonistic calculus. In this theory the individuals appear as essentially passive beings, rejoicing or suffering. Fisher’s point of view is different.2 He considers the individuals as active and acting. Fisher objects rightly against Edgeworth that Edgeworth’s theory is psychological, not economic. One could say that Edgeworth represents the psychological, Fisher the anti-psychological point of view. Fisher says:
The truth is, most persons, not excepting professed economists, are satisfied with very hazy notions. How few scholars of the literary and historical type retain from their study of mechanics an adequate notion of force! Muscular experience supplies a concrete and practical conception but gives no inkling of the complicated dependence on space, time, and mass. Only patient mathematical analysis can do that. This natural aversion to elaborate and intricate analysis exists in Economics and especially in the theory of value. The very foundations of the subject require new analysis and definition.
