Does Marx Need to Transform? - users.wfu.edu

1 does marx Need to Transform? W. Paul Cockshott and Allin F. Cottrell Revised, May 19951 IntroductionAlmost all of the voluminous literature on the Marxian transformation prob-lem is predicated on the assumption that, whether or not he succeeded, whatMarx wastryingto do in Part II ofCapital, volume III namely, derive aset of prices consistent with the equalisation of the rate of profit across allcapitals was correct. Those neo-Ricardians who argue that there is reallyno transformation problem as such (on the grounds that labour values aretheoretically redundant see Steedman, 1977) most emphatically share rare exception to this orthodoxy is Farjoun and Machover sLaws ofChaos(1983). Farjoun and Machover, like Steedman, conclude that thereis really no transformation problem, but for a very different reason, namelythat the assumption of a tendency toward the equalisation of the rate ofprofit is both empirically false and theoretically untenable.

2 Rather, theyclaim, the predictions of the simple labour theory of value, as in Volume IofCapital, are in better accord with the aims in this paper are to explain this claim, to present some em-pirical data by means of which the claim may be assessed (based mainlyon analysis of the UK input output tables), and to offer some (tentative)thoughts on the economic mechanisms that might be responsible for gen-erating the observed data. We should point out thatifthe transformationproblem is conceived as a purely logical exercise a matter of showing how Turing Institute, Glasgow, and Department of Economics, Wake Forest University,respectively. This paper was prepared for the conference onKarl marx s third volume of Capital : 1894 1994, University of Bergamo, December 15 17, 1994, and published inR.

3 Bellofiore (ed.)Marxian Economics: A Reappraisal, volume 2, Basingstoke and NewYork: Macmillan and St. Martin s, aggregative labour theory of value can be reconciled with the assump-tion of an equalized rate of profit then our paper has nothing to say aboutit. Our belief is that the transformation from labour values to prices ofproduction was a live issue for marx because he thought there really wasa strong tendency for the rate of profit to be equalised, so that the simplelabour theory of value would yield seriously counterfactual predictions. Wewill argue that marx was wrong on this begin with a brief examination of the logic of the standard equalised-profit The tendency towards an equalised rate of profitSupporters of the assumption of an equal rate of profit for the purposes of thetheoretical analysis would surely admit that rates of profit, in any particulareconomy at any particular time, show quite a wide dispersion.

4 Their claimis not that there actually exists a single rate of profit, but that there exists adefinitetendencyto produce equalisation, and that for theoretical purposesit is legitimate to assume that this tendency is fully what exactly is the status of such a tendency? On this theory,should we expect to see the dispersion of rates of profit narrowing over timein actual capitalist economies? If that is the idea, it seems to be empiricallyfalse. Farjoun and Machover produce evidence that the empirical frequencydistribution of profit rates is broadly stable over time, with no observabletendency to collapse towards degeneracy. The alternative is to claim that thetendency towards equalisation is something inherent in the process of com-petition among capitals, but that it is masked by the continuous occurrenceof external shocks or disturbances.

5 This theory relies on a partitioning ofthe causes operating on the dispersion of profit rates. Internal to the logic ofthe system is a competitive process that drives towards equalisation, whilethe dispersion-enhancing disturbing factors are exogenous. What are thelatter factors? If they were sunspots, hurricanes, earthquakes and so on, thetheory would be coherent (but even so, if the net result of the endogenousequalisation process and the exogenous shock process is the maintenance ofa roughly steady degree of dispersion, the equalised-rate assumption wouldnot be very useful for analysis of real economies). But surely the mostsignificantfactors making for increased dispersion of profit rates are justas endogenous to the process of capitalist competition, or rivalry, as theequalizing factors: the development and application of new technologies;the development of new products; the exploitation of new markets or newsources of supply of labour or raw the classical analysis shared by Smith, Ricardo and marx the pri-mary force working towards equalisation is the mobility of capital betweensectors of the economy in response to profit-rate differentials.

6 If industryXis showing above-average profit, capital will move in, increasing the supplyof the product and hence driving down both price and profit-rate. If In-dustryYshows below-average profit, capital will tend to exit the industry,reducing supply and hence raising price and profit-rate. This mechanismmakes sense in itself,1but it represents only one aspect of capitalist com-petition, understood broadly as the restless search for the greatest possibleprofit. Admit the other aspects of inter-capitalist rivalry (alluded to above),and it becomes an empirical question whether competition produces (a) anactual tendency towards equalisation, (b) a tendency towards ever greaterdispersion, or (c) a roughly stable probability distribution for the rate ofprofit.

7 As we have noted, the available data favour conclusion (c).Why, then, does the equalised-profit assumption exercise such a holdover theorists? It may be that there is a temptation to think of competitionamong productive capitals on the model of arbitrage in financial this model is very misleading. The equalisation of returns on finan-cial assets comes about almost instantaneously via revaluation of securities,while the equalisation of returns on industrial capital is at best a very slowprocess, dependent upon on the rate of depreciation and the speed withwhich new production facilities can be financed, built, and brought intoproduction. There is also the syndrome of looking for one s keys underthe lamp post.

8 Suppose the equalisation assumption is false all the same,howelseis one supposed to derive determinate theoretical results? If oneassumes a non-equalised set of profit rates, how can one reach any conclu-sions? Indeed, will this not undermine the simple labour theory of valuejust as severely as it undermines the theory of prices of production? Theproblem here is the restriction of the search todeterminateresults: asto-chasticversion of the labour theory of value can manage quite well withoutan equalised profit rate, and still generate interesting and testable predic-tions regarding the laws of motion of capitalism. Farjoun and Machover(1983) show how this can be so. In chapters 5 and 6 of their book theyexplain why it might be that prices tend to proportionality with labour con-tent for broadly defined groups of commodities, in a context where the rateof profit is far from equalised, while in chapter 7 they offer an interesting1 Although it raises the question of the conditions required for such migration to producestable convergence on an equalised rate of profit, on which topic see Steedman (1984),Dum enil and L evy (1993).

9 3discussion of the dynamic-historical law of decreasing labour-content .3 The stochastic approachFarjoun and Machover make a distinction between the realm of production,where matters are relatively determinate, and the realm of price-formationand profits, where the anarchy of the market prevails and the relevantmagnitudes must be thought of as random variables. The search for the correct determinate linkages between these variables is displaced by ananalysis of the relevant probability distributions, their respective degrees ofdispersion and their interconnections. In this spirit we offer below a listof the most important distributions to be examined in order to assess therelative merits of the simple labour theory of value and the theory of pricesof production (either Marxian or Sraffian).

10 1. The distribution of ratios of market prices to labour values,f( ),where =P/ . (Pdenotes market price and denotes embodiedlabour-time.)2. The distribution of rates of profit,f(r), wherer=S/(C+V). (Asusual,S,CandVdenote, respectively, surplus value, constant capitaland variable capital.)3. The distribution of ratios of market prices to prices of production,f( ), where =P/ . ( denotes prices of production.)4. The distribution of organic composition of capital,f(o), whereo=C/(S+V).5. The distribution of rates of surplus value,f(s), wheres=S/(S+V).A word on the definitions of these distributions. First of all, we shouldemphasise that the magnitudesS,CandVare all expressed in money while we refer to the ratioS/(S+V) as the rate of surplus value forthe sake of brevity, it should properly be called themoney-rate of surplusvalue.