On the Value of Software - A Response

The following document is a reply by RV to Sander's document On the Value of Software, together with comments from Graham Seaman.

I finally sent your text "On value of software" to Graham S. with this introduction:

I still have difficulties to deal with that question and I would appreciate to know if some of the reasons for which you think that "normal software production does not generates value" coincide with the arguments given in this text.

Graham answered:

Yes, this is extremely close to my own argument. But although we reach the same conclusions I think my route to them is simpler and consistent with social/technical history, where Sander starts more from the 'surface' phenomena.I'll try to write something more formal about this,but basically:

In manufacturing, design (technical, artistic etc) becomes separated from production (I would have written 'capitalist manufacturing' but this clearly applies also eg. to the Roman slave factories manufacturing Samian pottery).

The design element is a prerequisite for production; but it does not increase with the volume of production (unlike labour, raw materials,consumed fixed capital, etc).

Bourgeois economics simplifies the analysis of such things by treating them as 'sunk costs', which have no impact on price (their main importance is that they create a 'barrier to entry' for small firms wishing to enter a market but which cannot afford the investment in these sunk costs).

Marx accepted many simplifications from bourgeois economics (eg. perfect competition, which he uses as an assumption wherever convenient,especially in Book 3[of Capital). The treatment of sunk costs is another one, which is why there is no treatment of them in Capital. Sunk costs do not add to the value of commodities.

Computers are multi-purpose machines. The design of the machine is changed by changing the program (a computer is a television, or an arcade machine,or an assembly line controller, or a calculator, or....) with no production process required at all (after the initial production of the computer, clearly).

In striving to reduce production costs, capital pushes for the widest possible use of computers or computer-like equipment.

Writing programs is a modern equivalent of the design element in the general production process; it is a sunk cost which creates no value. The value of programs comes only from the material basis needed for distribution. So the price of software CDs in the Moscow or Hong Kong markets closely reflects their true value, which is the cost of the plastic CD, labour required for duplication etc.

Value (like capital) is not a thing but a relationship. The idea that sunk costs do not create value is not just a theoretical one found in Marx, but also a real part of people's social relationships. People accept that taking something with value is theft. They accept that paying for commodities is normal, provided there is an approximate proportion between the labour required to create the commodity, the amount they are paid for their own labour, and the price of the commodity. They do not in general accept that copying a program is theft, or that paying for something they can copy themselves without labour is normal.

It is not tolerable for capital to have expanding areas of production without value; it therefore needs to try to reimpose the form of value on software. The problem is in three parts: making people perceive software as a commodity ('piracy is wrong'); finding legal and technical means to force software into the mould of a commodity (the software can no longer be physically copied without using illegal equipment); and establishing prices for licenses to use software (the last one is interesting, and there is a lot to be said about this, although the actual fights are around the first two).

Reimposing the form of value in these ways does not actually create value. Where firms succeed in imposing the form, the result has the same effect is a surplus profit created through monopoly: surplus value is moved from the companies which have genuinely created surplus value to the software companies. Currently, this also generally means to the US: software is currently a means for imperialism to extract surplus value from other countries. If the form of value cannot be thus artifically reimposed, the final alternative is to move to providing support services for software. This is no longer production of a commodity. In this case the software is provided at value (ie. near zero price) but the user is charged for support services. This is a normal part of the service industries which do not themselves create value; prices are determined by the general rate of profit. This generally works only where such payments are seen as an expected deduction from revenue; ie, for commercial, not personal, use of software. For capital in general (rather than individidual capitals) to accept this alternative would be impossible. OK, Like Sander I'm seeing this as 'a conflict between capitalism and the law of value which the former is doomed to lose'. The tension comes from the productive forces as driven by capitalism. But I think it's simpler to see it this way round, instead of starting from monopoly and working backwards as Sander does.



Things are still not clear for me. My first problem is about the idea that "the service industries ... do not themselves create value".Today, in the US, about 60% of the personal consumption expenditure consists in services (40% in 1959, how much in Marx's time?) (See page 5 of http://www.bea.doc.gov/bea/ARTICLES/2005/12December/D-pages/1205DpgD.pdf ) [IP Note - This article is currently unavailable on-line] (1)

Why would all that amount of work intended to allow the reproduction of the work force have no value? Producing bread would create value, but health caring not? In Marx's times services were almost insignificant in the reproduction of the work force. He often (not always) assimilated services to faux frais. I don't think this is correct today, at least for a big share of services.

More... later.


1. In another order of things, it is also interesting to see on that site the historical data of "The profit margin, non-financial corporations", in page 6, sort of proxi for the rate of profit).

Home IP Archive Texts Discussion IP's French site Links