[OPE-L:3045] Re: Okishio and mathematical Economics

Paul Cockshott (wpc@cs.strath.ac.uk)
Wed, 18 Sep 1996 01:53:26 -0700 (PDT)

[ show plain text ]

>Here is the quote from Paul C:
>"I doubt that this is necessary. One could assume initially a random
>distribution of behaviours and show that the profit maximising behaviour
>was what Maynard Smith calls the evolutionary stable strategy, in that
>in a mixed population the profit maximising sub-population would displace
>the others."
>Well, I think he is right in a methodological individualist framework. As
>I, and others, use MI, this implies a reduction of social causality to some
>pre-given social atoms that are homogeneous to their type. Notably, in and
>of itself, MI does not entail rationality, although both MI and rationality
>are often used together. Now, if we assume all firms operate with the same
>norm, e.g., maximization, and maximize over the same definition of profit,
>and over the same time frame, then eliminationist models leading to an
>evolutionary stable strategy of profit maximizing firms might occur. But,
>as soon as one admits firms with different norms of behavior, different
>goals, different measures of profit, and different time horizons, then, in
>evolutionary terms, we have a problem with the unit of selection. A
>variety of maximizing and non-maximizing firms can survive in the economy
>(population). And, what people as different as Sidney Winter and Cutler,
>Hindess, Hirst and Hussain have argued is that you don't get the
>eliminationist outcome in such a world populated by heterogeneous firms. I
>agree with this and I am arguing that a move away from homogeneous firms,
>qua social atoms, that underlie the ERP condition as argued above, is pari
>passu a move away from MI. I hope I am making clear that I consider a MI's
>argument to be more than simply one which focuses on individuals, whether a
>person or a firm, but rather MI is specific form of reductionist social
>explanation, reducing complexity to homogeneity (and again, rationality is
>a separate issue).

The Maynard Smith argument, as do all Darwinian arguments, assumes a hetrogenous
population of the feature under selection, in this case: behaviours.

Paul Cockshott