From: Paul Cockshott (wpc@DCS.GLA.AC.UK)
Date: Tue Mar 22 2005 - 17:40:19 EST
As Co Author, let me put my 'penny's worth' into this. 1. Our paper was deliberately restricted to the issue of price changes in commodity money systems. We decided not to go into the implications for fiat money since the neo-classicals who are the target of the critique do not systematically theorize this. What we show, albeit by implications, is that the equilibrium point of an economy after technical change depends on what appears to be something arbitrary - the numeraire. Assume you have two possible numeraires - call them gold and silver. After an a technical change the price of some commodity, let us say sewing machines, will rise in terms of silver but fall in terms of gold. So if the currency were based on a gold standard, then sales of sewing machines would be expected to rise. If we had a silver standard sales would be expected to fall. - This is the immanent critique of neoclassical theory. 2. Personally, and I can not speak for Ajit on this, I think that the commodity theory of money is at the very least inadequate. As long term list members will know, I then to follow Wray's State theory of money. In this context the result of the indeterminacy of the direction of price movements with respect to the numeraire can also be read as an immanent critique of commodity money theory. 3. On the relation between Sraffa's work and my own researches into the labour theory of value. Empirical work by Allin and I, and also by David Zachariah indicates that the actual price vector seems to lie between that predicted by a Sraffian model and that predicted by a simple labour theory of value. The Sraffian model of prices can thus not be treated as an adequate predictive model nor is it significantly superior to the simple labour theory of value as a predictor of prices. This empirical result is an extrinsic critique of the work of Hodgson and Steadman who had argued that Sraffa rendered the labour theory of value scientifically redundant. Conversely however, our work does indicate that the Sraffian model does have some independent explanatory value independently of what the labour theory of value predicts. There is some partial transformation of values into prices of production. A model which treated prices as being the result of some linear combination of labour values and Sraffian prices would predict market prices better than either of the theories by themselves. 4. I consider that Sraffa should not be treated as being identical with his interpreters. I am unsympathetic to the polemical use made of him by Steadman and Hodgson, but I think that The Production of Commodities by Means of Commodities, was one of the greatest works of 20th century critical political economy. With the Standard Commodity and Basic System it introduced the notion of a set defined by transitive closure into economic thought. The sort of recursive definition here is conceptually on a par with other 20th century logical innovations - Russell's barber paradox, or Turing's proof of the impossibility of a solution to Hilbert's decision problem. The Basic System he identified is critical to analysis of: a) the maximal growth rate of an economy b) the effective planning of a socialist economy c) and here I speculate, it provides an underlying model for the growth constraints on tumours and bacterial colonies. I also argue in a joint paper with David Zachariah to appear shortly in Science and Society, that it provides the key to giving a scientifically rigorous definition to the concept of productive labour. 5. On the lack of dynamism in Sraffa's system. In science Occam's razor is a good guide. Do not complicate things beyond what is necessary. If you can get a reasonable simulacrum of reality with a simple model, accept this it for what it is good for. I think that Sraffa's model already went just over the edge in terms of added complexity relative to gain in predictive power - for which I would cite the fact that it has little predictive edge over the somewhat simpler labour theory of value. I think that to go beyond this and demand a fully dynamic theory of value is to ask too much. Constructing a genuine dynamic model of value is very hard, especially if it hopes to be a specific in its predictions as Sraffa's. I do not regard the work of the Kliman school as being a serious competitor in this regard. I suspect that it is better to abandon full deterministic models and restrict ourselves to more parsimonious stochastic models.
This archive was generated by hypermail 2.1.5 : Thu Mar 24 2005 - 00:00:02 EST