Why is the issue of whether or not a particular activity creates value important? The answer varies. For Marx, I think there were two main answers: it was important in his critique of older economists (the discussion of productive versus non-productive labour in Theories of Surplus Value), and it was important in his analysis of the circulation of capital (the sections on sales and accountancy in Book 2 of Capital). Both these are quite theoretical reasons; there was no particular political immediacy to the question for him, mainly because he concluded that the cases where there was doubt were marginal to the economy as a whole (eg. singers or doctors) or were residues from older economic formations (eg. peasant farmers). Now, the issue does have political importance, both because of the relative weight of the 'doubtful' cases in the economy as a whole, and because those cases are the result of new technologies, and so signs of the future rather than relics from the past. In the first aspect we have the debates over the decline in the rate of profit - if service industries and the new technologies are value-creating then they might perhaps now have reached a scale where they can offset any shrinkage in the surplus value created in commodity prodution. In the second aspect there is the attempt to impose the value-form on digital information; if this somehow inherently has value - even if it is small - then the success of laws to enforce property rights over digital information is inevitable (without a revolution), while if it has no value, then we have the beginnings of the basis of a new system with which property laws are quite incompatible.
I don't have such strong opinions about the first aspect of the question, though I do know it has been argued over endlessly and it's unlikely I could say anything original about it. I also agree with Sander's point that not all 'service industries' are the same, but still think it is possible to make some statements about services in general. It seems to me that most of the people who actually try to empirically measure the rate of surplus value believe that the service industries as a whole do not produce value; for example:
The debate on OPE-L that started somewhere around here: http://ricardo.ecn.wfu.edu/~cottrell/OPE/archive/9801/0061.html (this is hard to follow in places because some of the links are broken). Here Juriaan Bendien, who was working on the New Zealand economy, argued against service industries creating value.
or the paper by Mohun in
where Mohun shows that the figures from the US economy can only be reconciled with a Marxist approach on the assumption that Marx's view in Capital II was correct (ie. services in circulation create no value).
Similarly, the old USSR excluded the service industries from calculation of GDP. They found that there was no way to assign any monetary value to the additional use-value created for example by a teacher, other than by treating the teachers wage as the measure. Which would then lead to the GDP increasing when teachers have a pay rise, rather than when the teacher improves her teaching (Alec Nove, The Soviet Economy, p.285). (I'm assuming that the USSR was state capitalist, and that the CP was trying to run a state capitalist system using Marx's description of capitalism, which is why this is relevant here).
Sander points out that Marx quite definitely insisted that service workers could be productive for capital, which makes it sound like there is a clash between Marx's theory and mere 'empiricists' working on the raw data (though also between the Marx of Theories of Surplus Value and the Marx of Book 2 of Capital). I think there is a simple solution to this: a service worker may be productive for a particular capital without being productive for capital as a whole. In that case, he is simply helping to redistribute surplus value create elsewhere to increase the capital he works for. On the other hand, a worker who is productive for capital as a whole must be producing value, which most service workers do not. Marx is quite explicit, for example, that doctors do not create value (Raoul's favourite example):
"The doctor's services belong to the faux frais of production. They can be counted as the cost of repairs for labour-power. Let us assume that wages and profit fell simultaneously from whatever cause... If in such situation capitalist and workman wanted to consume the same amount of value in material things they did before, they would have to buy less of the services of the doctor, schoolmaster, etc. And if they were compelled to continue the same outlay for both these services then they would have to restrict their consumption of other things. It is therefore clear that the labour of the doctor and the schoolmaster does not directly create the fund out of which they are paid, although their labours enter into the production costs of the fund which creates all values whatsoever, namely the production costs of labour-power" (TSV, chapter 3, section 4, para 310)
This is perfectly consistent with saying that the labour of the doctor who works for a private health company is productive for that company. When a doctor employed by a private hospital treats a patient, the hospital receives a payment (from patient or more likely insurance company) sufficient to cover the labour costs, material costs, and provide the general rate of profit - more, if as is likely the hospital has some kind of monopoly over the treatment. The doctor's labour has been productive for the hospital, in terms of increasing its capital, but no value has been created at any point - the hospital has simply 'captured' some value created elsewhere.
However, it seems clear that even the productivity of services for a particular capital is highly limited unless combined with some kind of monopoly. A service, by definition, is the provision of labour-power as the commodity purchased. Unlike physical production, where it is possible to make huge relative gains in profit by being the first to introduce new productive machinery, the profit (without monopoly) from services will be decided only by the relative scarcity of the particular type of labour power. I know from having been a teacher that managers of services always have fantasies about automating the service - in this case education - but also that in practice it always turns out that this just moves the labour requirement elsewhere (eg. from standing in lectures to answering emails), and I think this is in the nature of services - any such gains will be small and temporary.
Going on to the value of software (and other digital information, though personally I'm mainly interested in software). Software is not mainly provided as a service, though it can be - the main case being companies which survive purely through support for free software. The turnover of this type of company is tiny compared with the license-selling software industry, and since what they are selling is labour-power (for installation, customization and support for the software), the profits are also quite limited: in spite of the promises/threats of the last 30 years programming has not been automated, and additional profits come mainly not from productivity differences but from monopoly of knowledge of particular software (which is limited in duration if the software is free software) and relative exploitation of the workers (which may be self-exploitation; free software companies are typically small or tiny). (Companies like salesforce.com which provide web-based software are sometimes also described as service companies, but I think it makes more sense to see them simply as using a rental version of the proprietary software approach).
The more interesting area is the companies which realise huge profits through software. Here I believe that the software per se has no value at all, although there is always some (small) value in the software as delivered for use, since the delivery medium does have value. Sander says he disagrees with this ('I do not think that the labor spent on R&D is not value creating') but then also says that he does agree ('since no new R&D cost must be incurred to copy it, the value of the copy will be no higher than the value of the labour time needed to copy it'). I do not think it is possible to have your cake and eat it like this. Labour is spent on creating the original, but spending labour is not equivalent to creating value. Even the legal process of marketing software confirms this: software is never sold by distributors, since it has no exchange value; instead, they sell a license to use the software. Using labour is not sufficient to create value; value is realised through a market, and software is not sold on a market (and is not a commodity). Just ask Microsoft if they will sell you Word ;-)
Sander says I am using Marx's authority as the sole justification for this argument, but it is not entirely true: not only Marx, but NO economist, pre-Marxist, Marxist, Sraffian, marginalist etc has ever tried to make fixed, one-off costs which do not scale with production (ie. sunk costs) part of the value (or for non-value based economic theories, part of the sale price). Of course sunk costs have an importance, but the importance is not in determining value: it is in acting as a barrier to entry to a market, which small companies may not be able to afford to cross, and in social costs (not only the labour spent in programming - whether paid or for pleasure - but also the system of education and training - official or social - which produces the programmers).
Whether you take my approach ('software has no value, but copies have some') or Sander's ('software has value, but copies of software do not') makes no difference to our view of the source of the profits of companies such as Microsoft: almost entirely due to the use of monopoly to move surplus value from other sectors. But I think there are other, political considerations which go along with each view. If software is simply an extreme case of the general tendency to reduce the direct labour component of commodities and increase the scientific component, the part played by the 'collective worker', then it introduces nothing fundamentally new. If on the other hand it is the first case of a considerable amount of social labour time being directed, within capitalism, into a product which is valueless then it introduces completely new considerations. On our side, I believe that the majority of people actually do not agree with Sander (though on considerably less theoretical grounds ;-) and believe software has no value, so that copying it is not theft. This is a pre-political position which can be potentially expanded to become a political one which embraces other products. Certainly the threat of being arrested for copying a valueless product could be quite a politicising one! On the other side, forcing a non commodity into the value form will have major effects on all of us, as the first stage of a move towards a system even worse than capitalism (I don't believe it could actually exist as a full-blown system without a complete counter-revolution, but elements of it clearly can). In this system, even the rights we had won under capitalism are taken away. When we pay money for something, we no longer own it (you cannot resell software licenses you have paid for). We cannot give something we have paid for to a friend. To stop us from doing so will need more and more complete control over all lives.
After all that rambling, I think the three of us are actually quite close in our positions. Sander seems to think my position is based on a distinction between productive and unproductive consumption, which isn't the case at all - I hadn't come across that argument before, but as far as I understand the contents of the debate between Sander and Goldner I actually agree with Sander on this point... which is why I avoided mentioning it above. Anyway, if you two are happy to go on discussing this, I am too.
13 January 2006
|Home||IP Archive||Texts||Discussion||IP's French site||Links|