[OPE] The Rate of Profit as a Random Variable

From: Jurriaan Bendien (adsl675281@tiscali.nl)
Date: Sat Jun 28 2008 - 09:06:50 EDT

To be more specific, "I think" the dissertation is brilliant. Why? 

Its choice of topic and problem formulation in this controversial field, the method it devises to explore and test theoretical propositions empirically, and the economical presentation. I didn't say I personally completely agree with the interpretations taken, or with a number of measurement and theoretical issues, because obviously I don't - as often happens with mathematically inclined people, some assumptions which turn out to be really important remain implicit, rather than being made explicit. But I have very high regard for Julian taking up this analytical challenge and pursuing it. 

In the field of statistics, the theory of probability was not my own forte  (in this regard, for my academic degrees I just had to know the basics of the analysis of variance), and I was much more concerned with interpreting time series and economic aggregates - later my professional interest became the conceptual issues regarding what is to be measured, and how or why you would measure it (the goals and means of quantification). The distribution of profit rates might be "something like" a gamma distribution, but it is possible for instance that there might be other parameters. 

I already previously indicated a number of points of difference I have with Farjoun/Machover, although I have not yet written up a paper about their rather meandering, discursive book. Effectively, in their theory, Marx's concept of "prices of production" as a regulator of product market-prices is abandoned altogether, because it is argued such prices presuppose a uniform profit rate on capital invested, and if that rate doesn't exist, it logically follows that the very notion of production prices is meaningless and irrelevant. 

But I think a sophisticated rebuttal of that reasoning is possible, provided: 

(1) you do careful justice to Marx's own concept of production prices (rather than making do with neo-Smithian Marxism, a la Fred Moseley, Ajit Sinha, Paul Cockshott, Farjoun/Machover etc.), 
(2) you are prepared to go beyond Marx in specifying which type of production prices exactly are actually relevant in regulating price levels, 
(3) you properly grasp the economic dynamism that results out of the clash between the law of value and the law of capitalist competition. 
(4) you do not confuse stochastic distributions of recorded price observations with processes of social causation.

I have not yet written that up in a paper. Personally I have rarely entered into the fray in the transformation problem controversy, mainly because - leaving aside that I am not a professional economist by training - I thought it was futile. Futile, because: 

(1) Marx's own draft manuscript is badly misinterpreted in every possible way, 
(2) because Marx's own (unfinished) idea is interpreted through the prism of other theories which are really alien to it, 
(3) the discussions are theoretically and conceptually weak in defining what the key categories involved operationally mean, and the causal relationships they imply. 

As a digression, the rigour of economic theory is supposed to lie in its mathematization, which is itself based on the fact that prices are numbers (quantities or countable units). Indeed, economics is often assumed to be about money-prices. 

Once you have some numbers, you can calculate, and prove definitely that if certain assumptions are made, certain conclusions quantitatively do or do not follow - the dispute that remains is then over why we would group certain numbers in preference to other ways of grouping them. But it turns out that economic theory often lacks rigour, because it cannot adequately explicate and defend why certain assumptions/hypotheses are chosen in the first place for inclusion in a model. The defence is inadequate, because there is no adequate theory of social causation behind them, and indeed the very meaning of a model (a likeness, analogy or isomorphism) is that such an adequate theoretization is not yet available. 

The temptation of the economist is then to say that quantitative inference will sort out which conceptual distinctions ought to be drawn, but in reality quantitative analysis is only an aid for this purpose, and not the whole story. Thus, the mathematical triumphalism often masks theory-choices which are themselves undefended. Then, when novel experience proves the assumptions wrong, new assumptions are mooted, but the very process and history of mooting successive assumptions itself is never questioned, and economics is not very self-reflective about the specific patterns in this process, so that little that is new is learnt from experience, at least not via economics. 

Ironically, economists therefore repeat arguments which were made two or three centuries earlier, without being aware of it, and without any notion that two or three centuries of human experience should be considered in evaluating the merit of those arguments. 


ope mailing list

This archive was generated by hypermail 2.1.5 : Mon Jun 30 2008 - 00:00:16 EDT