Ruth Indeck compiled some answers to the question Guilio Palermo posted on the URPE listserv. I am adding my own reply — which I sent directly to him.
The original question, from Guilio Palermo, posted on Wed, Jan 20, 2010 at 8:57 AM was:
I have a problem with the understanding of Marx’s analysis of the causes of the limits of the rate of exploitation. Let’s keep the length of the working day constant. One way to increase the surplus value is to reduce the value of the labor power, for instance by means of innovations that reduce the amount of labor incorporated in wage goods. If I understand correctly (but probably not), only the increases of productivity within the sector that produce these commodities do in fact increase surplus value, whereas increases of productivity within other sectors leave the value of labor power (and the rate of exploitation) unchanged. Is this correct? Can you please give me concise references?
From Julio Huato, posted on Thu, Jan 21, 2010 at 5:30 AM:
But other sectors may produce means of production for the wage-goods sector. So if productivity there increases, it also winds up cheapening the production of wage goods. Because of the interdependence among the different sectors of production, virtually any increase in the productive force of labor anywhere may redound in a reduction of the value of labor power. I believe Marx mentions this in Capital, vol. I, on the section of relative surplus value production.
The two original responses that went onto the listserv:
From Anwar Shaikh, posted on Thu, Jan 21, 2010 at 10:30 AM:
The labor values of the commodity component of the normal standard of living of workers (the normal means of consumption) determines the value of labor power. For any given standard of living, the value of labor power falls if the labor values of the normal means of consumption fall. However the labor values of all commodities depend not only on the productivity of labor in the industries which produce them, but also on the productivity of the industries which produce the means of production of these industries, and of the means of production of the means of production of these industries, and so on. This is why Pasinetti refers to labor values as the “vertically integrated” productivity of labor.
Needless to say, the value of labor power changes when the standard of living changes, and these latter changes are often fought out on the issue of how the benefits of productivity increases are to be partitioned. A formal treatment of these issues appears in Chapter 4 of my book with Ahmet Tonak, Measuring the Wealth of Nations, 1994, Cambridge University Press. Empirical measures for the rates of exploitation of productive and unproductive labor in the US appear in Chapter 5, section 6, and in Appendix I. The Introduction to the book appears on my home page, as do other articles on this subject, at the address listed below.
Hope this helps. Best wishes for your research.
From Alan Freeman, posted on Thu, Jan 21, 2010 at 12:25 PM:
There are two issues. One is what happens in the branch of industry as a whole, and the second is what happens within the particular firm that introduces the innovation.
For the first issue, the best reference is Marx’s own discussion of Relative Surplus Value in Volume I.
For surplus profit, I think the best reference is Mandel’s Marxist Economic Theory and also his Late Capitalism, one of the few works I know containing a proper discussion of the role of surplus profit in Marx’s theory. In Marx, the relevant reference is chapter 10 in Volume III entitled ‘Equalisation of the General Rate of Profit through competition: Market Prices and Market Values: Surplus profit’. The issue is this: in the firm that introduces the innovation, its costs are lower than that of its rivals in the same branch. It therefore receives ‘a profit above the normal profit’ – superprofit or surplus profit. Marx I think argues that initially the price is not reduced and this advantage is not distributed to the rest of the economy.
However, the new technology actually establishes a new value – the average value of the producers in the industry, which is now lower, since a part of the product is being produced more efficiently. (see the initially unpublished Chapter 6 of Volume 1, the ‘Resultate’ which is published in the Penguin edition)
Soon, prices sink to levels close to this new value. At this point, the reduction in costs of the industry’s product will be passed on to the purchasers of this product.
If this product is itself a consumption good, the consequence will be that the value of the wage is reduced, since their price has fallen. Therefore, the rate of exploitation increases.
If it is a Department I (constant capital good), it might appear that this effect will not be realised. For example if the product is steel. But, actually, the good will enter into the value of other goods, produced by using steel (for example domestic kettles). Their value will then be reduced, since the value of the constant capital required for their production is also reduced.
As soon as the producers of kettles reduce the price of their product to its new value, the value of the wage will decrease in the same way as below.
An exception to this general rule may arise in the production of luxury goods, eg personal submarines, which are purchased out of revenues derived from profit. In general, productivity advances in the luxury goods sphere do not result in a reduction of the value of the wage, and therefore do not increase the rate of exploitation.
I can check out the exact references and will do so if it would be useful..
From Paddy Quick, posted on ??:
I am a little reluctant to intervene in an another wise self-contained theoretical discussion based on the assumption that all production consists of commodity production and all labor consists of wage labor. But I must point out that, while this is a useful assumption in theoretical discussions, it is indeed an assumption for theoretical purposes, and that the real world in which we live and struggle involves non-commodity production. Thus the reproduction of labor power requires the performance of household labor i.e. the performance of labor on the commodities purchased with the wage (wage-goods) – potatoes have to be cooked, clothes have to be washed, quite apart from the labor involved in the reproduction of the human beings – all of which is necessary for the reproduction of the capitalist mode of production. We distance ourselves from real world struggles if we ignore the non-wage labor of the working class of capitalist. An investigation of the circuits of capital can contribute little to the struggle for socialism if it is not presented in its dialectical relationship to the process of reproduction of the working class.
From Anwar Shaikh, posted on ??:
I was careful in my response to say that the value of labor power depends on the commodity portion of the normal consumption of workers, because this is what the capitalists have to pay for. What they do not have to pay for is not their problem, even though it is ours. In the cited book, we explicitly discuss household production of use-values (as distinct from commodities) on pp. 28, 30-31. The same effect obtains from non-commodity production within a village, which enters the standard of living of workers but need not have to be paid for by capital. This latter issue has a major role in the history of capital, and in its existence in the developing world. One concrete consequence of this is that as capital destroys non-capitalist production, some part of the missing use-value production has be replaced by commodity production. Thus labor becomes poorer because its overall standard of living has fallen, yet it appears more expensive to capital because the commodity component has risen.
I have always argued that Marx’s careful distinctions between use-values, commodities, and commodity-capital had very concrete underpinnings, and very concrete implications.
From Scott Carter, posted on ??:
There is actually an open question as to whether the value of labor power is determined by the commodity bundle, what Sraffa in his unpublished notes called the “inventory wage”, or whether it is determined according to the share workers command out of the value added. In his analytical work Sraffa at least had little need for the former conception in the questions he was looking at. The latter approach makes a lot of sense in a monetary economy wherein the wage bargain is carried out at the level of nominal wages. The New Interpretation also stresses this share conception of the VLP rather explicitly. The monotonic relation between the wage share and the rate of exploitation can then be simply posited.
Marx himself in various places adopted both approaches, the inventory or “commodity bundle” approach AND a share approach, without properly delineating the relationship between the two; Rosa Luxemburg actually makes a great deal of this with her “tendential fall in the relative wage” concept (see Rosdolsky’s reference to Luxemburg in an Appendix I think on the primitive accumulation). This has caused tremendous confusion and sectarian rift, all to the detriment of advancing Marxian wage theory. Personally I think that both approaches are relevant, albeit at different levels of abstraction.
Check out Marx’s WAGE LABOUR AND CAPITAL (written in serial form in 1849) and WAGES, PRICE, AND PROFITS (reply to Citizen Weston; 1865) where he speaks about both conceptions.