Re: [OPE-L] empirical measurement of changes in the value of labour-power

From: Dave Zachariah (davez@KTH.SE)
Date: Thu Dec 13 2007 - 13:24:15 EST

Hi Jerry,

Let's use your example to proceed. Assume a rise in food prices and no
technical change nor changes in the nominal wage or the length of the
working day. The result is a reduction of the quantity of goods in the
weekly consumption basket of wage-labourers. Consequently there is a
fall in the labour-value of the real wage vector. However, according to
your use of "Value of Labour-Power", it has remained constant.

Now if we take this definition, the rate of exploitation is constant
too. But this must be wrong: not only have the capitalists in the food
sector received greater profits, but all wage-labourers across the
economy are spending less time to produce their -- now smaller --
consumption basket and consequently producing more surplus labour. The
rate of exploitation has risen.

The only tenable definition is to use the labour-value of the
consumption basket as the "necessary labour" or "value of labour-power"
in the analysis from Capital vol. 1. This also avoids the numerous
complications and qualifications that you mention.

You ask how to distinguish short-term fluctuations from the underlying
causes of the change in food prices? That is where the labour theory of
value comes in: one has to study the price to labour-value ratios and
unit labour-values of food items to see that.

You also ask whether it is worthwhile to have some sort of long-run
analysis of the reproduction of labour-power. I definitely think so, it
allows us to make important comparisons between different societies, but
the one Marx gives is deficient. A better alternative for this purpose
is "the total labour necessary to reproduce the capacity to work". Let
me elaborate briefly what this means.

1. It includes the social labour necessary to reproduce some
historically specific average consumption bundle  (i.e. your use of
VLP). But also domestic labour required for cooking, cleaning, child
rearing etc. In some social formations this distinction is blurred or
the latter hidden, and that is precisely why we must count the *total*

Take the case of socialized child care that exists in Sweden. Its
introduction lead to a rise in the amount of *social* labour necessary
to reproduce the next generation's capacity to work. However, this was
offset by the reduction in *domestic* labour, so that in fact the total
labour required for child rearing had fallen.

2. The "capacity to work", i.e. Marx's "labour-power", is even in the
abstract not a homogeneous entity. There is significant quantitative
difference between the capacity to work as a doctor and as a clerk: They
require different amounts of labour to reproduce.

//Dave Z

on 2007-12-13 03:32 GERALD LEVY wrote:
> > > Is the following the only reason you think so?
> > > > Would it not be better to explicitly distinguish:
> > > > (a) the labour-value of the real wage
> > > > (b) the total labour (social *and* domestic) necessary to reproduce
> > > > the capacity to work
> > Yes. I think the VLP was an ingenious theoretical invention by Marx but
> > it is also imprecise for the reason given above.
> > Take one wage-labourer, working for 1 month.
> ========================================
> Hi Dave Z:
> Hold on.  I don't think that this question should be framed in
> terms of "one wage-worker".
> This is a complicated question since a number of concepts
> (VLP, SNLT, abstract labour) are related to each other and
> because the period of time under consideration (and what happens
> during that time) is relevant.
> In general terms, I think that average wages (real and nominal) can be
> said to fluctuate around the VLP.  In other words, the VLP represents
> a kind of *average value*  which at a *given moment in time* can
> be assumed to be *given*.   It incorporates a "historical and moral
> element" associated with a *given* social formation: in other words,
> the VLP is different in different societies and is subject to change.
> But, it is essentially a *long-run* concept.  In the short-run,
> changes in  the supply and demand for labour power (over various
> phases of the business cycle) can cause wages to be greater than or
> lesser than the VLP.
> Now, is this a worthwhile distinction?  That also is a complicated
> question.  It depends on whether you think that there needs to be
> an analysis in which short-run variations can occur independently
> from long-run averages.  This is complicated because what happens in
> the short-run can cause there to be - *over time* - changes in
> standards of what constitutes SNLT and the VLP.  It is complicated
> also because it concerns issues of comparative statics and dynamic
> analysis. Indeed, it's complicated for a number of other reasons as well!
> All the more reason for us all to discuss it, I say!
> In the concrete question of a change in food prices, I would ask:
> what are the underlying causes of the change in food prices?  I
> think that short-term shortages, for example, need to be separated
> out from structural causes, like a major change in the productivity of
> labor in agriculture due to technological change.  The first would
> lead to a change in real wage (or what you call the real wage
> vector) and the second would cause a change over time in the VLP.
> I do recognize, though, that there are some difficulties in
> "operationalizing"
> the concept of the VLP, especially in periods of rapid technological
> change and intense class struggle, and this can lead to measurement
> difficulties.
> If I have spoken past some points which you want to see addressed,
> please speak up.
> In solidarity, Jerry
> ====================================
> > We have to distinguish:
> > 1. a (nominal) wage, i.e. the sum of money payed.
> > 2. a bundle of goods and services purchased, which I call a real
> > wage vector.
> > 3. the labour-value of that vector.
> > Thus I don't see a "fall of the wage below the value of labour-power",
> > rather I see a fall of the wage leading to a fall in the labour-value of
> > the real wage bundle. But the total labour necessary to reproduce the
> > capacity to work may be unchanged, and it is a matter of definition
> > whether one should call this or the labour-value of the real wage as
> > "VLP". But I think you run into problems if you don't distinguish them.

This archive was generated by hypermail 2.1.5 : Mon Dec 31 2007 - 00:00:04 EST