RE: [OPE] Constant returns to scale - IRS

From: Michael Williams <>
Date: Sun Oct 05 2008 - 14:54:48 EDT

My remarks are not directed at anyone in particular. They , obviously, spring from some comments of Jerry’s.


That perfection competition was not conceived by its founders and developers ‘simply as part of a larger project, etc.’ is not relevant to the points I was making.


Comparative statics doesn’t, per definition, treat of dynamic processes – it abstracts from them. More sophisticated, but still entirely

 orthodox, models do attempt to capture such processes. Comparative static abstractions can help in the difficult task of interpreting dynamic models.


The standard information assumption of most game theory is ‘common knowledge of rationality’ and perfect information about the pay-offs to all combinations of strategies. However, learning processes can also be modelled, including strategic ploys to hide real and disseminate misleading information. What GT explicitly does NOT assume is that players know in advance what other players are going to do (except in the special, and generally uninteresting, case of the existence of a dominant strategy that it is clearly the rational thing to do whatever your rivals might do. And then it is not an assumption but a deduction.)


I am not defending orthodox economics as a basic theory of capitalism, but as a contribution towards the theory of markets. Having said that, there are, of course, orthodox theories of money and credit.




Dr Michael Williams, BA, MSc, PhD


Mob +447906172655

Home tel +4423 80768641

P Help save paper - Do you need to print this email?


From: [] On Behalf Of Gerald Levy
Sent: 05 October 2008 15:05
To: Outline on Political Economy mailing list
Subject: Re: [OPE] Constant returns to scale - IRS


Hi Michael W:


I guess your questions are directed at me [?].




1a. Rather than capturing something real about competition,

the concept of perfect competition obscures reality. It is at

base a fanciful ideology (like its corollary, consumer sovereignty)

which obsures actual relations rather than moving in the direction

(via) further concretization of comprehending the relations.

[NB: Let's not forget when perfect competition was first conceived

of and how long it was before Chamberlin and J. Robinson

presented similar theories of imperfect competition. This

should tell you that at the time that perfect competition was

first put forward it was _not_ put conceived of simply as a part of

a larger project which analysed/concretised competition more



1b: In the 'move towards the actual concrete' the use of

comparative statics leads to a 'now you see it, now you

don't' slight of hands (it's magic!) situation rather than understanding

competition as a _process_.


2: See 1b. Also, how can you reasonably model strategic

behavior employing the assumption of perfect information?

With that assumption (and the assumption of rationality),

every 'player' knows what the other players are going to do

before they do it.


I agree with Alejandro that Austrian theory poses less problems

than Walrasian theory. At least, they have money in their

theory - unlike GET. Or, is it OK - Michael - to not include

money in one's basic conception of capitalism with the understanding

that one can introduce it later because it (allegedly) is a 'concretisation'?


In solidarity, Jerry



Re. Alejandro’s excerpt from Jerry,

1):a) Why can I not reconstruct this argument in terms of abstraction and concretisation? Perfect competition is an abstraction that thus may capture something real about the meaning of competition; imperfect competition models are also abstractions, but emerge in the presentation as conceptual transcendences of contradictions at the more abstract level of perfect competition; they are concretisations.

b) Comparative statics is an abstraction associated with perfect competition that captures something about the real world, whilst abstracting from the actual dynamic mechanisms that have to be conceptualised in order to move towards the actual concrete.


Re: Alejandro’s excerpt from Jerry, 2):

But this is simply untrue. Economies of scale/down-sloping LRAC conceptualise the competitive advantage that certain kinds of technology reflected in costs structures provide; then profit-maximising capitalists will exploit those technologies to grow bigger and to shield themselves from competition (strategic barriers to entry etc.); economies of scale and strategic entry deterrence also motivate mergers and takeovers. These dynamic processes can be modelled, inter alia in game theoretic models. This kind of theory of capitalists behaviour also grounds competition (anti-trust) policies by most capitalist states, but can also be extended to argue that a non-capitalist state would be more likely to pursue such policies rigorously. Etc.


1) Jerry L: *Well, I think a part of the problem is the entire way in which 'competition'
is conceived of in mainstream economic theory. Most commonly, they create an 'ideal' which doesn't (and can not) conform to real-world conditions (perfect competition) and then contrast that to other 'market structure models' (monopoly, monopolistic competition, oligopoly). Among the many problems in so doing is that they take the *dynamic* subject of competition and try to analyze it simply in the realm of *comparative statics*.


I agree!


2) Jerry L: *Where this theory most dramatically fails is in its inability to explain how you get from one market structure to another. That is, there is no explanation of how the dynamic process of competition *itself* brings about concentration and oligopolization.*


Yes, but on this respect Austrian Economics is very good and has overcome the handicaps of Neoclassical Economics. So the problem seems not to be *marginalism* but deeper problems arising from the untractable phenomena that Mathematical Economics simply can’t handle.

No virus found in this incoming message.
Checked by AVG -
Version: 8.0.173 / Virus Database: 270.7.5/1708 - Release Date: 04/10/2008 11:35

ope mailing list
Received on Sun Oct 5 15:01:59 2008

This archive was generated by hypermail 2.1.8 : Wed Dec 03 2008 - 15:12:03 EST