[OPE-L:3374] Re: Marxian Emprical Research / Skilled labor & surplus value

Steve Keen (s.keen@uws.edu.au)
Sat, 12 Oct 1996 14:31:37 -0700 (PDT)

[ show plain text ]

Allin Cottrell wrote:
> I owe Steve a reply on skilled labour.
> What is the basis of the objection here? Is it something
> like, "We know that a doctor produces value at 5 times the
> rate of an ordinary worker, because the doctor is paid 5
> times as much; but that squares with the LTV only if the
> training time stands in relation to the doctor's working
> time as 4:1, which is obviously false"? If so, Hilferding's
> response does not seem right to me: the obvious response is
> that doctors don't "produce value" at 5 times the standard
> rate, and if they're paid 5 times as much, a substantial
> whack of that is rent.

I agree that part of the response is the one you give--I made the same
elaboration in my JHET paper on the subject, that the wage of a skilled
worker is likely to include a component reflecting rent. But that
doesn't override the main argument which Hilferding makes, that you can
explain how a skilled worker can be much more productive (as well as
cost much more) once you apply the use-value/exchange-value analysis to
training and effectively de-couple the cost of training (its
exchange-value) from its impact on the value-productivity of trained
labor (its use-value).

On the next issue:

> There ought to be a non-arbitrary relationship between the
> cost of training and the "ability to create additional value
> it imparts". Take a schematic case where we can actually
> compare productivities. Training method X enables a worker
> to produce 10 widgets per hour, while untrained, she can
> only produce 5. So the training doubles a worker's "ability
> to create value". Now suppose that when the labour-time
> cost of the training is figured (and properly depreciated
> over the trainee's working life), it stands in much less
> than 1:1 relative to the trainee's working time. Well,
> doesn't something follow? The training + trained labour
> dominates the untrained; and in a rational system of
> labour-time accounting nobody would be left untrained.
> Under capitalism, though, the result may fall short of
> rationality. Who is to pay for the training? If the
> employer, he may not do so for fear that the trained workers
> may go off and work for somebody else (they're not slaves);
> and if the worker, she may simply not have the resources.
> So, there are some complexities here, but I don't find
> Hilferding's answer at all satisfactory. It short-circuits
> a proper account of how the cost of training may diverge
> from the _apparent_ "ability to create value" of the trained
> workers.
> Allin Cottrell

The example you give should have some more realistic numbers attached to
it. Sweezy (as a counter example!) gave the following instance in
support of the "skilled labor produces a higher value of output because
it costs more value to produce it" argument. The skilled laborer

"expends in production not only his own labor ... but also indirectly
that part of the labor of his teachers.... If the productive life of a
worker is, say, 100,000 hours, and if into his training went the
equivalent of 50,000 hours of simple labor (including his own efforts in
the training period), then each hour of his labor will count as one and
a half hours of simple labor." (Sweezy 1942, p. 43)7

Now the best that case can argue for is a 1.5:1 ratio of skilled to
unskilled labor productivity. But how does it tally with the most
intensive form of training one can undergo, a one-on-one apprenticeship?
Not well: With a 48 week year and a 40 hour week, total training hours
for both trainer and apprentice sum to 15,360. If the average working
life was 40 years, the educated apprentice would clock up a further
76,800 hours of labor. This results in a pitiful skilled labor to
unskilled ratio of 1.2 to 1. Including the equipment depreciated in the
training isn't going to increase this a great deal.

This is the bind which Marx's use-value/exchange-value analysis, as
applied by Hilferding, lets a Marxian analysis of skilled labor escape.

I therefore disagree that the "ability to create value" of the trained
workers vis-a-vis unskilled is "apparent": it's real. This factored
ability to create value reflects the use-value of the training; the
additional cost of skilled labor reflects (a) the exchange-value of the
training (b) the monopoly-rent factors you note above.