Posted on the Shoe list on 27 January 2020.
Is “Utility” the Most Suitable Term for the Concept It is Used to Denote?
by Irving Fisher
American Economic Review, volume 8 (1918), pp. 335-7.
In all sciences, and particularly in one like economics, which appeals to the general public and which uses concepts and terms already at least partially familiar, it is a matter of some practical importance to select a suitable terminology.
The concept called “final degree of utility” by Jevon’s, “effective utility”, “specific utility”, and “marginal efficiency” by J. B. Clark, “marginal utility” and “marginal desirability” by Marshall, Gide and others, “Grenznutzen” by the Austrians, “Werth der letzten Atome” by Gossen, “rareté” by Walras, and “ophélimité” by Pareto, seem still in need of really satisfactory terms by which to express it.
Marshall improved greatly upon Jevons’ phrase when he substituted the term “marginal” for “final degree of”, and this improvement has been very generally recognized and accepted.
But, as yet, no generally accepted substitute for “utility” has been found. The term is a heritage of Bentham and his utilitarian philosophy. It is misleading to every beginner in economics and to the great untutored and naïve public who find it hard to call an overcoat no more truly useful than a necklace, or a grindstone than a roulette wheel. Economists cannot with impunity override the popular distinction between useful and ornamental, much less that between useful and useless, without confusing and repelling the man in the street.
In the last few years a new source of confusion has arisen from the use, in a special sense, of the phrase “public utilities.” This phrase must itself be used by economists who now find themselves discussing the marginal utility of a public utility! — and distinguishing between the marginal utility “in the economic sense” representing the esteem of the political ring or other powers that be for that public utility (which marginal utility imparts economic value to said public utility), and utility “in the popular sense” representing the real social serviceableness of that public utility!
Genuine utility for social service must, as Pareto says, be more and more studied by economists as they fulfill their task of working out plans for economic and social betterments. He therefore suggested that we should not abandon the term utility but reserve it to express the genuine article and employ in its place in price analysis the term “ophelimity” — as it has been anglicised — to express the value-making quality.
It is true that coined words have the great advantage of breaking away from the misleading associations which cling to terms already in popular use. But the difficulty has been with “ophelimity” as with most coined words, that, just because it has no association to introduce it, it would not and could not dispossess the old term.
The term “desirability” comes very near the required mark and I have used it in most of my books; but, unfortunately, like utility it carries with it to some extent an ethical connotation. Usage seems to imply that a desirable object is one which ought to be desired, rather than one which simply has the potentiality of being desired. We are force to call the most undesirable articles and services, such as whiskey and prostitution, economically “desirable” in price analysis.
It has occurred to me that the term really needed may be built on the good old economic term “want”. Long before the days of “marginal utility” economists spoke of “human wants”. Wants include wants for purposes of ornamentation as well as for purposes of real utility; wants for what is trivial or useless as well as for what s important, useful and desirable; wants for evil as well as for good purposes. So far as the influence on price is concerned the essential fact is that an object is actually wanted, or rather that it is capable of being actually wanted under stated circumstances. Whether it ought to be wanted, or whether it is wanted for a proper purpose is immaterial. It must merely have the capacity for being wanted, it must be wantable, it must have wantability. Ordinarily the short term “want” will suffice. We can speak of a marginal want for whiskey, and if we prefer a phrase in which “of” replaces the “for”, we can speak of the marginal “wantability” of whiskey. The two terms “want” and “wantability” might well be used alternatively, affording welcome variety in expression.
The more technical term of the two, “wantability”, is only half coined. It is sufficiently coined to serve notice on the reader that he must learn, not assume, its meaning; while the association of ideas it carries, leads the mind along the right path without paradox, contradiction or confusion. It is readily recognized when seen and easily recalled when wanted. In short, it bears its meaning on its face. As hinted above, it could be piloted into use by speaking of “the marginal want for” as an alternative to “the marginal wantability of.”
Another advantage is that these terms afford the means for coining an expression, to me at least much needed, for a unit of “wantability”. Such a unit might be called a “wantab”. In this case we have a free field for a coined word and no term in use to dispute possession. If, as I anticipate, the science of measuring human wants is to be developed in the future a convenient term for this unit will be needed.
No equally suitable term for a unit of “desirability” or “utility” or “ophelimity” seems available; although in my doctor’s thesis of 1891 on “mathematical investigations in the theory of value and prices” I made an attempt. The appearance last year of a French translation of this little essay has renewed my interest in a better terminology and, together with the opportunity to secure the necessary data which the war seems to promise, has led me to hope for a statistical measurement of marginal “wantability”.
Before attempting to launch any new terms for this concept, I should be glad to receive expressions of approval or disapproval from other economists.
This was David Colander’s comment on posting the article:
Many early neoclassical economists followed classical thought and distinguished between theoretical discussions and ethical policy discussion. Policy involved ethical judgements, which they were quite willing to make, and scientific theory related to what they could potentially measure. This led them to distinguish the measurable desirableness of a product (along revealed preference lines–a concept Pareto called ophelimity and which Fisher wanted to call “wantabes”) from a different and broader concept of utility in which value judgements entered in. Individuals maximized wantabes, not the utility that was relevant for public policy. The goal of public policy was to maximize utility, not to maximize ophelimity, which was what was potentially measurable. The distinction was never fully developed, but I think it sheds some light on Colin’s question.
Below is a short piece the Fisher wrote to drum up support for his concept of utility. It demonstrates how he uses the distinction to blend his social policy into his scientific analysis. That distinguish between ophelimity and utility approach lost favor in part because the ethical judgements Fisher made were questionable, as they involved support of eugenics among other things. In my book with Craig Freedman on Where Economics Went Wrong, I argued it should be brought back, and policy analysis should be distinguished from scientific analysis more than it currently is.