[OPE-L] Theoretical issues concerning variable capital (was: if six turned out to be nine)

From: Jurriaan Bendien (adsl675281@TISCALI.NL)
Date: Wed Oct 04 2006 - 13:00:44 EDT

Rakesh wrote:

Wages are often paid after performance of labor but before commodity
output actually sold. Wages or variable capital thus appears to be


I was interested in this problem once, working with a friend on the New
Zealand economic data. He pointed out to me from anecdotal evidence (he had
been in business himself) that, as a fact of business accounting, wage
payments may also recouped during the year by a commercial business through
current revenues from ongoing sales, implying that the actual amount of
capital tied up in wage payments at any time during the year (the average
"stock level"), or if you like wage-payment funds held "in reserve", may be
only need to be e.g. one-tenth or one-seventh of the annual flow value.
Thus, the size of the variable capital outlay would depend partly on the
normal turnover time (rotation speed) of the total capital outlay (which
obviously will vary according to the kind of output of a business; some
businesses might indeed "advance" wages, insofar as it concerns a project
which comes on stream and yields profit only after a certain time interval).
I have never seen this possibility theoretically recognised in the Marxian
literature however. Perhaps it deserves an article.

That aside, the empirical measurement of a Marxian "variable capital
aggregate" confronts at least five other analytical difficulties:

1) The fact, that total labour costs to an employer normally - in advanced
capitalist countries - exceed (gross) wages paid (the excess being mainly
taxes, levies and benefits of various kinds). To some extent this is
reflected in the national accounts concept of "compensation of employees".
See an article I did on this, if interested:
http://en.wikipedia.org/wiki/Compensation_of_employees In the US social
accounts, a reasonable attempt is made to show the different entries to do
with disposable versus deferred labor-income, mainly I suppose because in
the US private employer-benefit pay-outs in respect of employees can be very
considerable. In many other countries, however, the components of
"compensation of employees" are not separately identified. Part of the
problem I guess is that there are many different factors influencing wage
payments made, making accurate wage statistics very difficult to compile.

2) The definition of capitalistically "non-productive" labour - on some
Marxist definitions that have been offered, about half of the nation's wage
payments (in advanced capitalist countries) reflect such non-productive
labour, thus, how the definition is fixed makes a very large difference to
the size of this capital outlay, and consequently to estimates of aggregate
profitability, S/V, and the OCC - especially since many Marxian authors
(e.g. Moseley, Shaikh/Tonak) include the gross wage-payments to employees
defined as "non-productive" in aggregate surplus-value, rather than treating
it e.g. as a portion of Cc.

3) Whether or not variable capital should really be valued on an accrual
basis, or not. In national accounts, it is typically valued on an accrual
basis, unlike the other inputs which are supposed to be valued at the time
they are actually *used* in production. Probably, this procedure raises the
value of labor-compensation in the account, insofar as at any time you
become entitled to money money than you actually receive, or that is
deferred in your favour.

4) The fact, that indirect taxes and profit-levies are imposed on worker's
consumption spending (such as VAT tax or GST tax, or various other levies).
Although this is not usually recognised in the Marxist literature, profits
are extracted not just "at the point of production", but also at the point
of workers' consumption. If for example you buy a muesli bar with your wage
income, you also pay the total profit and tax impost on that muesli bar,
which is part of its unit-price. Had Marx written about the sphere of
consumption (which he indicates as one of the four spheres of the economy,
in his introduction to the Grundrisse) he would no doubt have pointed this
out. It affects how we understand the value of labor power, i.e. the value
relation involved in the exchange between labor and capital. I referred
briefly to this issue in a short article here:
http://en.wikipedia.org/wiki/Rate_of_exploitation. More generally, we could
say that insofar as capitalist production is "the unity of the production
process and the circulation process", this involves both relations of
production and relations of distribution.

5) The fact, that employees spend a fraction of their own wages on directly
work-related activities, for which they may not be compensated. As I cite in
my CE article referred to in 1), "For example, British research showed the
costs associated with turning up for work each day reduce the average annual
wage among British workers by 2,300; The official average salary falls from
22,248 to 19,970 when the typical costs associated with having a job -
such as transport, snacks and clothes - have been deducted. A poll by
YouGov, sponsored by debit card group Maestro, showed workers typically
spent 120 extra a month on food, 50 on travel and 35 on work clothes. The
research found that the average worker spent 16 days a year getting ready
for and travelling to work. (source: The Guardian, 28 November 2005). Marx
might have pointed out that this expenditure is in such cases "gratis" to
the employer, in the same way as some forms of maintaining his productive
assets in the course of work are supplied gratis, and the surplus-labour is

In modern capitalism, wage-income and wage-related income are in reality
normally directly affected not by just two transactors (employers and
employees) but at least four (employers, employees, government institutions,
and private or semi-private/para-state benefit-paying institutions). Hence
the difficulties and the moral problems raised by the "just or fair reward"
for labor.

Had Marx written the "missing book on wage labour", he would no doubt have
covered all these issues from the point of view of the employee both as a
*producer* (who produces surplus value) and as a *consumer* (who incurs
additional costs due to appropriations of profit and tax imposts on what
s/he buys, as well as "incidental expenses" with respect to his work that
must be met from his own income). That is, the wage-worker would in this
book enter on the scene not just as a seller of labour-power, but also as a
"citoyen" who is a buyer of commodities to sustain himself. Frederick Engels
refers in his pamphlet "The Housing Question" to various possibilities of
rackets in this regard, which he calls "deductions from wages". But he does
not attempt to resolve the issues this raises for value theory.
He claims only that "When, however, the worker is cheated by his grocer or
his baker, either in regard to the price or the quality of the commodity,
this does not happen to him in his specific capacity as a worker. On the
contrary, as soon as a certain average level of cheating has become the
social rule in any place, it must in the long run be leveled out by a
corresponding increase in wages. The worker appears before the small
shopkeeper as a buyer, that is, as the owner of money or credit, and hence
not at all in his capacity as a worker, that is, as a seller of labour
power. The cheating may hit him, and the poorer class as a whole, harder
than it hits the richer social classes, but it is not an evil which hits him
exclusively or is peculiar to his class."
Engels sees the issue here as only one of "cheating", but the point is that
there will be a normal profit and tax impost quite independent of any
possible cheating.

I forget how Sraffa actually defines the payment of wages to employees, but
the general norm adopted in practice is that wages are paid after work is
done, and hence they are normally not "advanced" in that sense. The proviso
being, that a certain fund must often be held in reserve by the employer to
pay wages, particularly when an activity only begins to yield profit after a
certain interval of time.

It could be argued, in summary, that on those five grounds I mention alone,
the Marxian concept of variable capital is itself a theoretical
"abstraction" which identifies only an essential relationship, and which
would often need to be modified to encompass the empirical reality.

Arguably, the "true" measure of variable capital is either the disposable
income an employee actually receives in respect of time worked, over his
whole lifetime, in cash or in kind; or else the value of goods and services
actually appropriated by employees during their lifetime in respect of their
total financial claims issuing from their time worked. But a consistent
empirical valuation of such an aggregate is statistically speaking obviously
enormously difficult.

As a graduate student I did not delve greatly into the Physiocrats myself,
instead I looked at modern social accounting procedures. Possibly a dry
subject for a wet student, but it does point up the problems of very
abstract concepts, and the need to qualify them in making sense of
experience. For this reason too, the controversy between "Simultaneous
Valuation" and "Temporal Single System" adherents in the Marxist camp often
does not make much sense to me, or it seems a trifle scholastic.


This archive was generated by hypermail 2.1.5 : Tue Oct 31 2006 - 00:00:03 EST