[OPE-L:2382] Re: Chapter 5 and Marx's method

Paul Cockshott (wpc@cs.strath.ac.uk)
Tue, 28 May 1996 01:49:58 -0700

[ show plain text ]

>Paul writes:

>>Conventionally one counts labour as person hours, and since power
>>is the ability to deliver work per unit time, this would imply that
>>labour power was measured as labour per unit time. Since
>>labour = persons x hours
>>we must have
>>labour-power = labour / hour = persons
>>so labour power must be measured in person-equivalents, in terms
>>of the number of workers of average ability required to complete
>>a task.

>1) I agree with what you've said here, based on Marx's treatment of value,
>subject to a small caveat that you may have intended anyway.  As you define
>its units, labor-power is a stock variable, such as a slaveowner would
>purchase.  But capitalists purchase labor-power as a flow variable, i.e. so
>many units per "transaction period" (let's say a day).  Thus, wouldn't it be
>preferable to represent labor-power in units of persons/day, and then
>proceed as you've indicated above?

Paul ---- I think that there is a problem with doing it this way, but your suggestion lower down may solve it.

But first a point of clarification do you really mean that you want to measure labour power as persons/day (persons per day) rathather than persons times days?

I would have thought that a capitalist puchases 100 people for 5 days rather than 100 people every 5 days, the latter option would betray a Molloch like appetite more appropriate to the concentration camp labour used by Krupps than normal capitalist business.

I will assume that what you mean is persons times days, but if we take this as the measure, then labour and labour power would have the same dimension. If we use this definition it becomes unclear just what the distinction between labour power and labour actually would be.

Your second point may provide a way round it:


>2) Now to a separate point.  There's something about Marx's measure of value
>that has puzzled me for a long time.  As corroborated by your remarks above,
>this measure is solely in hours.  However, there are two senses in which a
>worker expends labor, extensive (number of hours) and intensive (effort per
>hour).  These variables are substitutes in production.  Therefore the
>"labor" production coefficient, measured in hours, is in some sense an
>endogenous variable, i.e. a function of effort. No inconsistency here--it's
>what Marx says as well--but it does raise problems:
>1) If prevailing effort levels depend on such things as the unemployment
>rate (as in efficiency wage models), then labor values are in turn a
>function of such market variables as the wage rate.
>2)  Suppose capitalists manage to double the rate of *effort* estraction,
>all other things equal.  Then the rate of profit necessarily goes up, but
>the rate of exploitation stays the same (subject to some indexing variation
>I haven't thought about--but the statement certainly holds in the one-sector
>model).  This is perfectly in line with what Marx says of course, but seems
>to go against the spirit of the notion of exploitation.
I am not sure that this is really a problem. If effort is correlated
to output, then a doubling of effort, if applied throughout the economy,
would result in some increase in material output. If the real wage
remained constant, then a shorter portion of the working day would
reproduce the real wage, and the rate of surplus value must rise.


>These problems are avoidable.  Suppose we define "labor" in abstract
>terms--measured in ergs, perhaps--as "effective labor performed", understood
>as a function of both hours and effort per hour (a simple version of which
>would be hours times effort per hour).  Then problem (1) is solved because
>production coefficients are given by technology---actual abstract labor
>extracted being a function of social relations of production, as before.
>Problem (2) is solved, since under the stated conditions, s/v necessarily
I think that this is a very good question because it takes us
into the borders of the unknown, and allows us to problematise what is
apparently obvious in the concept of labour.

It forces us to look more closely at what labour is, and why it is possible.

What is involved in work?

What is the abstract property of work that links different types of work?

Why is it that work of humans can be replaced by machines?

Note that at this stage I am asking questions about work, as intercourse with nature, rather than any specifically social form it assumes.

In your proposal that labour, or work done, be measures as some product of Effort x Time, we have essentially Watt's solution to the problem. As a capitalist trying to sell mechanical power as a commodity he was faced with the practical problem of how to measure what he was selling.

The first solution, by Isaac Potter who installed 5 Newcomen engines at Windschacht by Cheminitz, had been to charge for the engine the savings over the previous pumping methods obtained by the machine in its first 5 years. When Potter was doing this the only competition was horses, so the comparison would have been between the feed and purchase of the horses and the wood for the engine.

When Watt came to apply the same sales formula in Cornwall, he was already in competition with installed Newcomen Engines, and could express the savings in terms of coal save. But applying the technology to the many areas where there was as yet no steam competition he needed to measure work in a way that was independent of the source of the power: human muscle, water, horses, or steam. His solution was to use a standardised measure of work as lifting capacity - so many foot pounds per second, a certain standard number of foot pounds per second was taken to be the power of a horse, and the steam engines were then rated as 5hp, 20hp etc.

Bolton and Watt charged customers 5 pounds per horse-power per annum. There then immediately arose the intensity problem you describe above. Suppose a factory owner installed a 20hp machine in the anticipation of his business growing, whilst at the start, ne needed but 5 hp. Should he be charged L100 or L25 per year?

Watt had in fact a secret means of measuring the work done by his machines - the indicator diagram - and could therefore charge at the actual delivered power rather than the rated power.

By proposing that labour be measured in ergs x hours, Gil is essentially adopting the Watt approach. The difference between Watt and the seller of labour power, is that the latter has no secret means of measuring his work output and specifying this in the contract of employment.

This would make sense of the distinction that Marx makes between labour and labour power, since the labourer sells a rated 1 man or woman power to the employer, who can then use it at above or below its rated capacity.

It should be recalled that these details of the commercial practice in the hire of engines would have been familiar to Marx. I would suggest that the distinction between labour power as the capacity to deliver work, and the actual work done, may have been suggested to him by the distinction between rated and delivered power of engines. Indeed the concept of power in the abstract, is a steam age concept. Smith, in a pre-steam economy could not have distinguished labour from labour power.

However, I am not happy with this formulation of work done in term of ergs. It would be socially realistic in an economy that was just in the process of transition from a muscle economy to a steam driven one, where muscle power was still the dominant source of energy. More importantly, under these circumstances, labour was labour in the sense of hard physical effort. So the delivery of power, performance of physical work, was what people were hired for. But in a developed capitalist economy, the predominant sources of energy are machines, and physical effort is a secondary aspect of labour performed compared to skill.

We then have to ask what it is about work, skilled work, that distinguishes it from muscle power, or horse power?

It is clearly something to do with the decision making power of the brain that is hired along with the muscles. Labour's ability to change the world, is a unity of this decision making ability with the physical ability to excert effort. The replaceability of humans by machines, stems from the possibility of machines having a 'self acting' or in non Germanic English, an automatic ability to make decisions and sequence their operations.

We then are left with the problem of measuring what it is that is being done. I would suggest that we have to see labour as a means of entropy reduction as well as a means of delivering power. Paul Cockshott

wpc@cs.strath.ac.uk http://www.cs.strath.ac.uk/CS/Biog/wpc/index.html