Re: On Utility Theory

["Alan G. Isaac" <AISAAC@xxxxxxxxxxxx>]

On Sat, 12 Nov 1994 15:42:06 -0700 James K. Galbraith said:
>   You write that I "have avoided saying that the folk-psychological concept
>of action as the manifestation of desire and belief is useless." True, I
>have avoided saying this. The folk psychological concept is not useless, as
>folk psychological concepts go. But it isn't a universal solvent either. And
>to mix metaphors I don't see why I need to genuflect before this particular
>icon every time I pass the church.

Because no one has found a way to talk about human action and
simultaneously discard these notions. (Indeed, philosophers
of psychology are more determined to _elucicate_ them.)

>
>   You then write "if utility theory at its most abstract simply embodies
>this concept, why will you not grant utility theory equal usefulness?"
>
>   My response is two-fold. First, granting your premise ("if..."), I'll
>grant utility theory equal usefulness to the folk-psychological concept.
>But then Occam's razor tells me that while equally useful, under the
>conditions you state utility theory is also redundant. We can start with the
>folk-psychological, end with the folk-psychological, be no worse off than we
>were, and avoid trudging through a lot of uninformative and pointless
>abstractions, mathematical and geometric.

In my view, the usefulness is that the formal apparatus
undergirds empirical explorations and _exposes_ the
presumptions of the work to relatively precise criticism.

>
>   This leads to the second dimension of response, which is to deny your
>premise. Utility theory at its most abstract does not simply embody a
>folk-psychological notion of action. Indeed, lacking a dimension of time (at
>its most abstract, yet!) utility theory is caught in a (futile, in my view)
>effort to draw a distinction between "choice" and "action." Rational
>economic man is forever "choosing," never "acting." As Veblen wrote, he
>"lacks both antecedent and consequent."  Frozen between a fixed field of
>preferences and an equally fixed budget constraint, the theory says nothing
>of action, development, or change. Nor can it. To talk about those things,
>we need a different theory, which we never seem to get to because of
>conversations like this.

OK, I'll shut up after this post. But whatever the weaknesses
of the neoclassical treatment of time (and there are many,
I'll agree), Veblenian critiques of utility theory as static
seem quaint and anachronistic. There are mainstream models
of learning, preference change, preference discovery through
sampling, and the emergence of new preferences in response
to the development of new goods. Perhaps you would exclude
these from utility theory, but I would not.

>
>   What then was the purpose of introducing utility theory, and why do
>economists stick to it?

I'm unpersuaded by the physics envy explanations of the introduction,
but the more interesting question is why economists stick to it.
I believe the answer is the _normative_ appeal of utility theory.
While the neoclassical characterization of rational behavior
is of course extremely restrictive, it is still the case that
we understand quite well the sense in which it is rational for
an individual to do the best s/he can in achieving hir current
goals. This has been the basis of the paradigm's attraction to
psychologists and sociologists as well.

I know of no other explicit theoretical structure with this normative
force. Do you?

The question of why a normative theory should interest empirically
oriented social scientists, and of what domains of empirical
application would be appropriate for such a theory, is a
methodological question that I find very interesting. We
could take it up if you're interested.


>Are the puzzles it poses, which seem so dreary and
>repetitive to me, truly fascinating to other people? <Readers and
>intervenors: examples welcome!>  Or is the explanation better drawn from
>Mirowski, who cites the authority of an allegedly higher form of science
>lending its structure to utility theory, or perhaps from Foucault, who
>would, I think, interpret the neoclassical structure as a throwback to
>classical modes of thinking in a world otherwise virtually taken over by
>modern evolutionary constructs and, more recently, paradigmatic pluralism.
>

I hope my posts have made clear that I find utility theory to
have limited domains of useful application. The same thing is
true of evolutionary paradigms, of course.

>   To my mind intellectual activity of all kinds, including the
>further development of the neoclassical theory of the consumer, should be
>welcome. What troubles me is the insistence, by so many who are committed to
>that way of looking at things, that everyone else should be equally so
>committed.

Here we agree.

>Insistence on paradigmatic uniformity is, in my view, driving
>professional economists into an intellectual ghetto inhabited only by
>themselves.

The tendency I see is the opposite: the tendency for the
neoclassical paradigm to exercise hegemony in the social
sciences. Either outcome is regrettable, of course.


>   I hope this can stand as a last word on this so we can move on to other
>topics.

Fine.

It should be clear that we are largely in agreement as to
the limitations of utility theory. My persistent
beef is with claims that the obvious limitations of utility
theory render it useless. Instead, I have argued that the
limitations i. stimulate research and ii. simply limit the
domain of useful application of utility theory. You seem to
feel that attachment to the limitations stifles research,
and that the domain of useful application is nearly empty.
Unless we expend energy on specific examples and counterexamples,
and perhaps even if we do, there is unlikely to be much
shift in position.

--Alan


[][][][][][][][][][][][][][][][][][][][][][][][][][][][][][][][][][]
[]   ** *     ** **  *       Alan G. Isaac, Associate Professor   []
[]  *.* *    *.* * * *       Economics, The American University   []
[] *  * *** *  * *  **       Washington, DC 20016-8029   U.S.A.   []
[] FAX:  (202)885-3790       Internet: aisaac@xxxxxxxxxxxx        []
[] Phone:(202)885-3785       Bitnet:   aisaac@xxxxxxxxxxxxxxxxx   []
[][][][][][][][][][][][][][][][][][][][][][][][][][][][][][][][][][]