[OPE-L:4562] Surplus value and capitalist consumption

Alan Freema (a.freeman@greenwich.ac.uk)
Thu, 27 Mar 1997 06:44:37 -0800 (PST)

[ show plain text ]

Regarding Ajit's [4556]:

I think it is a fair comment that you have a difficulty
responding on two fronts. However, I suffer also since I
must respond both to you and to 'SSS' (TSS without the T).
And they in turn must respond to both of us. So we have a
triangle, which is always stormy. You and SSS agree that
values and prices are determined simultaneously. We and SSS
agree that the value of constant and variable capital are
determined by the labour-time expressed in the money paid
for these. As for you and me, we agree on nothing but at
least we are talking. You say:

============================================================
If we construct two parallel hypothetical situation where
everything in the two economies remain the same except that
composition of capitalist consumption differ, i.e. there is
difference in allocation of labor but no difference at the
level of production...their solution will give you two
different rates of exploitation in the two systems.
============================================================

I recall you making a similar point at the ASSA. I am sure
there is a case to answer. I do not want to be dismissive of
this point and will excuse myself for responding at some
length to give you a clear target:

It seems a valid criticism of SSS, but I cannot respond on
their behalf. They might reply that in these two
hypothetical societies the levels of output in different
sectors would be different, although the proportions are the
same. I may have misunderstood your argument but it is not
clear to me how the capitalists can consume more jackets
without making more jackets. So although this might not make
the outcome any more acceptable to you, there are perhaps
productive differences between these societies as well as
differences in circulation.

As for us, we prefer one reality to two hypotheses. So I do
not think this is such a problem for us, but I will return to
this. I begin with a little word that crops up in your next
question:

============================================================
How do you determine commodity values? And how is the "value
of money" determined?
============================================================

My question is, what do you mean by 'determine'? Anwar
Shaikh in his 1981 article 'A Poverty of Algebra' in Ian
Steedman's 'Value Controversy' (Verso) introduced a
fundamental distinction between real and conceptual
determination. I do think they are different. I have to ask,
therefore, which meaning you have in mind?

If you ask how I should *calculate* (conceptual
determination) commodity values then I have your answer:
their value is equal to the sum of the dead labour-time
expressed in the money paid for the commodities used to
produce them, and the new labour-time expended in producing
them. Since this has been questioned by others, bear with me
while I lay it out in full.

Initial givens
==============

Suppose initially at some point in time the value of all
commodities in society is known in labour-time terms. I hope
I am allowed to call this a 'given'. It is the only one.

This value, a definite magnitude in hours, is expressed in
another definite magnitude in money, which we can observe,
namely the total money price of all these commodities (Marx's
'total price'). From this we can calculate the new monetary
expression of labour-time (inverse of the 'value of money'
in the above sense) which in general changes from one period
to the next. It is simply the ratio of one measure to the
other. It represents the purchasing power of money over the
accumulated dead labour of society or, as Marx puts it,
the monetary expression of this labour.

With this calculated, we may determine values in the next
period. Given the initial condition this calculation is
'determinate'. Suppose for example that the commodities in
society contain initially 10,000 hours of labour which is
expressed in (priced at) 20,000 and suppose the capitalists
then consume goods worth 10,000 in constant capital.

Then we know that this constant capital represents one-half
of the aggregate value in circulation, and hence its value
is 5,000 hours.

If we know how many hours the workers work - let us say 5000 -
then we know that in the next period the gross value output
will be

5,000 + 5,000 = 10,000 hours.

Now suppose the capitalists spend 5,000 on wages and
2000 in frivolities.[1]

We know how many hours society consumed in this period: it
is the sum of the C, the V and frivolities:

5,000 + 2,500 + 1,000 = 8,500 hours

The new total value of commodities in society is then the
old 10,000, less the 8,500 consumed, plus the 10,000 newly-
created goods, that is 11,500 hours:

10,000 - 8,500 + 10,000 = 11,500

Note that the increase in value of the commodities in
society is exactly equal to the surplus value of 2,500, less
the frivolities. This represents accumulation.[2]

The calculation can now be repeated for the next period;
once we observe the new prices we can calculate a new
monetary expression of labour-time.

At this point since in circulation new relative prices are
determined, some will lose value while others gain.
However this is a pure redistribution. Since circulation may
neither destroy nor create value, which (I agree) arises
entirely in production, it follows that in fixing relative
prices, the market only redistributes value which already
exists as a result of the prior phase of production.[3][4]

What is the MELT?
=================

The conservation principle enunciated above is the
justification for the calculation of the MELT which is
not an external assumption but an expression of this
axiom. It is a ratio between the form of appearance of
value (money prices) and its essence (historically-
accumulated, that is, dead, socially necessary labour
time).

This is my main difference with Duncan who argues, I think,
that it expresses the value-creating capacity of live labour.

I would say that MELT is like an exchange-rate, an
instantaneous relation between two measures of the same
stock. I think Duncan sees it more as a relation between two
flows, the living labour of a definite period and the
money value added in the same period.

Our two calculations are equivalent in the stationary state
but differ under technical change. We have to explore the
consequences of these two different measures, which give
quite different quantitative results.

The MELT thus calculated is distinct both quantitatively and
qualitatively from the neoclassical conception which treats
money as purchasing power over use-values, and measures the
price-level accordingly. I think the main reason we differ
from our critics is that either consciously or unconsciously
they confuse the MELT with the neoclassical purchasing
power of money, thus confusing Marx's determination of the
magnitude of value from labour-time with the determination
of the magnitude of value from use-value, which I am sure
you would agree is not Marx's conception.

The initial given and estimation errors
=======================================

The only unknown quantity from which error might arise is
the initial estimate of the labour-time represented by the
very first purchase of the capitalists, when we began our
calculation. As with any temporal process, some assumption
must always be made about the initial condition. However it
should be clear from the above calculation that if we make
an error in this initial condition it will decay
exponentially with an index equal to the ratio between input
and output measured in money, which means within a few years
it will be insignificant.

Hence we have no trouble *calculating* both the quantities
to which you refer, to any desired degree of accuracy.
Moreover the calculation of value is not, as has been
supposed by some of our critics, dependent on the monetary
expression of labour-time. Quite the reverse. When we
calculate the labour-time embodied in the constant capital,
what we actually do is to multiply the value output in the
last period (in hours) by a coefficient, being that portion
of the commodities in circulation that are consumed as input
in the present period. In this coefficient, the MELT appears
in both numerator and denominator and hence cancels out.

Thus I hope I have also answered your objection to Andrew
that our results depend on a specific assumption concerning
the MELT. In the calculation outlined above the MELT
is a calculated ratio, involves no arbitrary assumption
and there are no constraints on it that I can see.

Surplus value, capitalist consumption, and determination
========================================================

So if by 'determination' you mean 'calculation' then I think
that I have answered your question, though I am sure not to
your satisfaction.

If, however, as I suspect, your question really means
something else, that is, what 'causes' these prices and
these values, then I cannot answer your question because I
think it is unanswerable. Since every causal factor in the
universe bears on the price which jackets are sold for, I am
constantly surprised that economists say a mere 100-by-100
production matrix contains within it such a perfect summary
of all these causes that it can predict the price of
jackets.

Indeed, precisely because we live in an alienated society,
almost by definition we cannot predict the outcome of
circulation from the circumstances of production and this
leads me to view with suspicion any theory which claims to
do this.

I think it is more for you to say why you find it so
important that the conditions of production suffice
to predict the prices that goods exchange at. I do not see
that either Marx or Sraffa thought this. I think what Marx
gave was an invariant of circulation, namely, whatever these
prices are, exploitation is the same. To put this the other
way around, only production can create surplus-value, which
is merely the obverse of the statement that circulation
cannot create surplus-value. Marx in my view axiomatically
separated production from circulation, defining the production
of value to be the expenditure of human labour-power on the
creation of new use-values for sale.

Now I return to the parallel societies. In my view, the
manner in which production 'determines' value is in the
normal sense of this word, that is, production comes first
and circulation comes second, and since nothing can be
determined by what comes after it in time, it remains only
to be shown that the results of the 'immediate process of
production' as Marx terms it, are fully determinate before
circulation commences and do not require circulation for
this determination.

In your parallel societies there is an additional third
constraint which is stated as an assumption but plays a
determining role, namely the requirement that goods should
exchange at the prices that will sustain a steady state.
This is, I think, a hidden third cause, neither production
nor circulation but a purely algebraic constraint that
exists only in the heads of the economists. My view is the
same as Anwar's: I think this is an idealist conception of
determination and I think it comes from neoclassical
economics. I don't think there is evidence that Marx had
this notion of determination and this is why I dissent
from the view that the rational kernel in Marx is the
surplus approach, since I think the surplus approach
requires such a concept of determination. But maybe there
is a way of expressing surplus approach ideas that do not
require this conception of determination. In the early
work of Hawkins, Simon and Leontieff I think an alternative
was present, but it has been lost, or rather, suppressed.
Maybe you can do something to recover it. I don't think
Sraffa would have dissented.

I would turn your question around and ask: what happens to
your determination of the rate of exploitation if the
capitalists in the first parallel, in the process of
changing their consumption habits to move to the second
parallel, drive prices out of the steady-state level? In
that case our 'determination' of values remains valid and
fully determinate but I have no idea how either you or Fred
could determine anything. Perhaps one of you can enlighten
me.

I'll speak about transformation separately because this is
already too long.

Alan

Notes
=====

[1]This leaves 3,000 in fixed capital whose value cannot be
altered during production, but which may of course lose or
gain value in the next phase of circulation, the only point
in the circuit during which value can be transferred from
one capital to another

[2]Consequently value is fully conserved. The whole of the
accumulation is exactly equal to the difference between the
value added by living labour and the value of what was
consumed as revenue (wages + frivolities) during the last
phase of production

[3]It is always possible that the fixed capital that was
once worth 3,000 may fall in price because newer and
cheaper goods of the same type have arrived on the market.
In that case, value will be redistributed between these
stocks and the newer and more competitive producers. Their
capital will decline in value but not, however, immediately
to the value of the newly-produced goods of the same type.
They will experience this as moral depreciation.

[4]Capitalists who continue production at the same physical
level may find they need less or more money than the value
consumed in production, depending on whether the goods that
they purchase have risen or fallen in value. As a result of
this release or tie-up of capital discussed by Marx in
Chapter 6 of Volume III they may find themselves with a cash
shortage or surplus.