----------------- HES POSTING ----------------- 
One of the difficulties here, as with many technical terms in economics, is 
that the term is invariant while its meanings (or assumptions) vary over 
time. In the Anglophone literature alone, the meaning of "utility" has 
varied with changing theoretical commitments with respect to the nature of 
utility, where utility resides and whether utility is measurable. 
 
Consider, in particular, 
 
1) Does "utility" denote a theoretical view of what is good, and, if so, 
what is that view? 
2) Is utility an objective property of things, or a property of an 
individualís subjective view of things? 
3) Wherever it resides, in things or minds, is utility scientifically 
measurable?  And, if not, is welfare economics possible? 
4) Are the different kinds of utility a person experiences (measurable or 
not) commensurable? 
 
The classical economists generally thought of utility as an objective 
property (roughly akin to "usefulness") of things ñ consistent with their 
labor theory of value.  Jeremy Bentham originally used the term to refer 
the capacity of a thing to produce good, such as benefit, well being, 
pleasure, happiness, etc. When applied to actions rather than things, 
Benthamís principle of utility refers to the tendency of actions to produce 
good. Bentham famously viewed good in hedonic terms, as the balance of 
pleasure over pain; he was committed to the scientific measurement of 
utility, and regarded all utilities as commensurable: pushpin is as good as 
Pushkin. 
 
1) Yes, there is a theoretical view of what is good, 
2) utility resides in things, 
3) utility is potentially measurable, and 
4) different kinds are commensurable. 
 
By the late 19th century "utility" had come to be identified not with the 
tendency of an object or an action to produce good, but with the good 
itself.  Thus a person's utility is not her usefulness in promoting good 
around her, but her own good. 
 
1) Yes, there is still a theoretical view of what is good, but 
2) utility now resides in minds. 
3) Utility is potentially measurable, but, for some, 
4) different kinds may not be fully commensurable, as with J.S. Millís 
utilitarian apostasy. 
 
20th-century rational-choice theory uses "utility" as a name for a 
mathematical function that represents a single, complete, transitive 
ordering of preferences over consumption bundles. The term "utility" does 
no analytical work -- "preference function" would be more precise -- and 
is, in fact, confusing in that when the function is maximized, the consumer 
is said to be maximizing utility, as if utility were some quantity of well 
being, pleasure, happiness, etc.  But consumers don't chose x over y 
because U(x) > U(y), that is, because more good is produced by x than by y. 
It's the reverse.  A utility function is chosen because x is preferred to y 
-- the preferences are the analytical primitives.  Thus,  
 
1) No, modern utility theory eschews a view of what is good, save for the 
(mostly implicit) idea that it is good to satisfy one's preferences, i.e., 
it is good to get what you prefer. (This last point introduces another 
potential confusion, the conflation of satisfying preferences -- getting 
what you want -- with the idea that utility should seen as a synonym for 
satisfaction). 
2) Utility resides in minds; 
3) utility is not measurable, which makes policy (if not voluntary trade) 
problematic, but 
4) different kinds of utility are presumed commensurable, at least in the 
standard theory that assumes a single (and complete) ordering of 
preferences. 
 
P.S.  Expected utility theory (choice under risk) of the VN-M variety 
manages to partly reintroduce (or at least smuggle back in) the cardinality 
that modern utility theory had sought to banish.  Unlike choice under 
certainty, we can express the ratio of EU(x) and EU(y) numerically, 
producing a magnitude that at least implies some quantity of good. 
 
Thatís quite a semantic journey for one term: from the usefulness of 
things, as measured by their ability to produce good, to the good that 
things produce, to the name of a maximand in a theory that disavows any 
concern with good. 
 
(Many of these ideas can be found in John Broome, Utility Economics and 
Philosophy) 
 
 
Thomas (Tim) Leonard 
 
------------ FOOTER TO HES POSTING ------------ 
For information, send the message "info HES" to [log in to unmask]