Rationaly Choice Thoery?-Reply

Justin Schwartz jschwart at freenet.columbus.oh.us
Tue Apr 4 23:23:53 MDT 1995

On Tue, 4 Apr 1995 fellini at keynes.econ.utah.edu wrote:

> But first I must confess that I still don't think that your
> argument and Wright's argument are quite different. >
> I think Wright, at least in those papers, argues that his argument
> requires a class analysis based on 'interests'. So, his argument
> must have something to do with class struggle.

Unlike G.A. Cohen, Wright, Levine, and Sober acknowledge that there may be
difference based on class interest in the development of more efficient
technologies. But basically they think that all comes out in the wash and
despite class conflict productive as well exploiting classes have
interests, if somewhat different ones, in increasing efficiency.
Resistance to domination plays no part on their argument. Increasing
efficiency plays no part in mine. So the arguments are quite different.

> Furthermore, Justin, I believe the difference between 'preference' and
> 'interest' in your argument is quite trivial. Rationality principle
> does not necessarily imply 'preference'; more generally it argues
> that individuals behave in their self-interest.

Au contraire. As you note below, the rationality assumption is that agents
choose means that maximize their expected utility, whether or not what
their preference schedules contain is in fact good for them. To talk of
interests is to appeal to an objective standard of what's good for someone
independent of their preferences. That's the point of talking about
interests. RCT has no place for the notion what is in fact good for
someone regardless of their preferences. Now in fact to get some of the
RCT results, notably the invisible hand results of the Arrow-Debrue
general equilibvrium theorem, RCT assume selfish preferences, denying us
interrsts in the interests of others. But I would argue that selfishness
is not in fact in our interests. My argumenyt appeals to interests of
people understood as members of groups, which requires intra-group
solidarity, and I would also argue that effective resustance requires
inter-group solidarity among subordinate groups. So the difference is
profound, not trivial.

. What we are talking about is 'means-end
> rationality', in that sense, rationality is a relation between
> _given_ means and _given_ ends. This is the form of RCT at least in
> Neoclassical economics: "for all individuals there is something
> that they optimize (depending on their constraints.)"  But this is
> a very general principle, it applies to almost everything: Like
> Gary Becker (Nobel winner in economics) did, you can use it to
> explain the behavior of ants on the one hand and marriages between
> homosexuals on the other. RCT does not distinguish between a saint
> and Hitler, for both are quite rational for their given ends (after
> all, nazis devised one of the most 'efficient' way for mass-
> murder.)

So, how can we explain the differences between them? (Here
> I am not going to mention the problem that why we should take both
> the means and ends as given. Is this a good way to understand human
> agency, which is an essential aspect of Marxist 'praxis'?) In this
> regard, I don't agree with Lisa in that "This is not intended to
> "explain everything" as some critics have said, no assumption can,
> but that is not what it is for." To my mind it was clearly intended
> to do so.

Well, I'm not using this in my argument.

> This is an important question, because most economists who use RCT
> framework do not, and cannot in my mind, consider historical
> differences and specifities of different societies/modes of
> productions etc. For them, ancient Greek society is as 'capitalist'
> as the contemporary societies. It seems to me that, a la Karl
> Polanyi, RCT is a way to look at non-capitalistic societies from
> the perspective of capitalism, for our minds are conditioned with
> the categories of capitalism and for we are forced to behave
> 'rationally'. This how I read Kevin Quinn's comments (am I right?):
I will ask what is supoosed to be wrong with choosing effective means to
our ends. We should choose ineffective means? No wonder socialism is in

> And I am afraid Justin, you have such an outlook (at least it
> appears so to me).

Well, apart from having some sympathy for the constrained use of RCT,
which I don't use in my argument here, I don't see this. Certainly I do
not deny the historical specificity of forms of domination and resistance.
I don't think that wage labor existed under slavery or that slaves had the
same interests workers do, once they are described specifically as slaves
and not just as members of a subordinate group.

> I agree with Howie; what I see in your argument is there are only
> two 'individuals', although they refer to 'groups'characterized by
> their interests and these two "classes" pursue their own interests.

This is pure Lukacs, as I now recognize.

> The struggle between them suggests a game-theoretic model,

There are game-theoretical aspects to the situation.

 and I am
> sorry but I fail to see the differences between your position and
> a Neoclassical one

Right. Neoclassical economics is full of stuff about dominant and
subordinate groups in asymmetrical power situations, class struggle, and
the like. You can't be serious.

NCE isn't the same as game theory. In fact, game theory, with its
prisoner's dilemmas and suboptimal consequences of invisible hand behavior
is a good RCT critique of NCE.

At the same time, while I don't explore the GT dimensions of the group
struggle, they are quite real.

 (If I remember correctly, Lukacs argued that
> Marxism is characterized by the dialectical method. Your model is
> very "dialectical" indeed!).

Since the term is undefined here I don't know what to make of this criticism.

> This brings about the other problem associated with RCT;
> methodological individualism. I am not going to pursue this issue
> here, but if you think that Marx was a methodological
> individualist, then I have nothing to say.

I'm in print denying it. See my paper on "Functional Explanation and
Metaphysical Individualism," Philosophy of Science June 1993. But I don't
think Marx's position settles the issue. I have doubts about whether MI is
coherent, but if it is it might be true. In that case Marx would be wrong.

> By the way, I am somewhat dogmatic about Elster or Roemer. Elster
> always pisses me off. I think D. Little is much more sophisticated
> than Elster. But still I don't like their "scientific"
> reconstruction of Marx for what they mean by science is
> Neoclassical economics.

That's not true. Elster and Roemer can be irritating in their
self-assurance, but they do not mean NCE by science. They mean, I think,
roughly speaking, empirically testible, precisely formulated hypotheses.
Roemer in particular, although more a fan of NCE than Elster, is extremely
brilliant, though, I think, mistaken in important ways. But his critique
of Marxian exploitation theory demands careful analysis and refutation. It
cannot be dismissed on general grounds.

 Further, in Little's book (Philosophy of
> Social Sciences) the chapter on 'materialist' conception of history
> is nothing but a technological determinist reading of historical
> materialism, whose merit in Marxism is quite debatable.
Well, Little and Elstertend to take Cohen as having got Marx's theory
exegitically right. Cohen's case _is_ impressive, and I think that he
establishes that Marx was intermittently commited to TD, although, I
think, not wholly consistently. Which is just as well, since TD is false.

Some time ago I had a debate with Jim Devine on this list--or may it was
the pen-l list--about Cohen and TD. I could dig up my part and repost it
if anyone is interested.

--Justin Schwartz

     --- from list marxism at lists.village.virginia.edu ---


More information about the Marxism mailing list