This is my first substantive intervention - I must say that I had some
sneaking sympathy with Heiner's withdrawal; I have not really the time to
look at everything in the post, except very cursorily. But I decided to stay.
Alan (no 801) writes about 'non-dualism' as follows:
>One must fully theorise both value and price at the Volume I level of
>abstraction, that is, in terms of capital in general. Negatively put,
>it is not essential to theorise that goods exchange at their value, in
>order to have a theory of value. Negatively put another way, it is not
>necessary, and actually leads to error, to try and pass through the
>category of price of production in order to arrive at the category of
>market price.
>
>This clearly posits a theory of money in which the value of money can
>and does differ from the labour socially necessary to produce the
>money commodity.
>
>A logical completion is to recognise that the portion of the value-
>product appropriated by the worker (variable capital, V) is equal to
>the price of wage-goods, not their value: to put this precisely, V is
>equal to the value of the money used to pay the wage. If I have not
>misunderstood, this step is an essential part of the proposal of
>Duncan and Gerard.......
>Therefore as far as I can see the work of Duncan and Gerard was a
>crucial breakthrough. First because it is impossible to express these
>ideas without a concept of the value of money; second because the idea
>that the value of variable capital is represented by the money wage
>drew the attention of the marxist world to the possibility that the
>value of a purchase might be better represented by its money-cost than
>its embodied value.
>
I think that there is a lot that is not quite right here. In part the issue
turns on 'equal' or 'equivalent' exchange. (I know that there has been a lot
of post about this, but I have not had time to read it all yet; I may have
to revise what I now write.)
The reason why a notion of equal exchange is required is to theorize the
existence of exploitation; it is not good enough to say that if we have a
labour theory of value and we notice that there are profits, then
exploitation exists, because that exploitation might be attributable solely
to unequal exchange in the market-place. This in turn suggests regulatory
reforms which leave the mode of production unchanged. If exploitation is to
be considered a feature of the capitalist mode of production, inseparable
from it, then we have at least to begin with a notion of equal exchange.
This is what I take Marx to be doing in CI, chs. 1-6.
He proposes in Ch. 6 that labour-power (LP) be considered like any other
commodity, with its value determined by the value of the commodities
required to (re)produce it. (I take historical and moral elements for
granted in all that follows.) This is a theory of the VLP, proposing that it
is cost-determined but on a customary rather than a subsistence basis.
But how do we measure the VLP? Given equal exchange, we can measure it
either in terms of the wage, given a value of money, or in terms of the
commodities the wage purchases, since these are both equivalents by
assumption. The standard tradition has been to use the commodities the wage
purchases, and to treat this measure as the theory. This is wrong in general
because it identifies determination with measurement and then gets the
measure wrong. Let me elaborate.
We know that when the rate of profit is equalised (tendential, in the long
run, or whatever), we cannot in general have equal exchange: prices of
production are systematically different in general from money-values. So
that with unequal exchange, the value represented by the wage at the
prevailing value of money on the one hand, and the value of the commodities
which the wage purchases on the other hand cannot in general be the same.
But there is no other way to measure the VLP; it has to be one or other of
these measures. Further, we also know that the forces making for unequal
exchange because of the equalisation of the rate of profit do not apply to
the transaction whereby people sell their LP for a wage: LP is not a
produced commodity but an attribute of people, and people are not
(re)produced under capitalist relations of production. So there is no reason
not to take the wage as the equivalent of the VLP, given some value of
money. More precisely, the wage is the wage rate, and the VLP is per hour of
labour hired. (To convert this to total wages and variable capital, as Alan
appears to do above, we need additionally an account of what determines the
number of hours a worker is hired for, and what determines how many workers
are hired.)
So the VLP is measured by the wage rate, given a value of money, and is not
measured by the value of the commodities purchased by the wage. This
problematises Marx's theory of the determination of the VLP, because it is
not a theory to say that the VLP is whatever the wage can buy, unless we
also have a theory of the determination of the wage. All we have is class
struggle over what is to be regarded as customary, and this is hardly
satisfactory. So one rather major problem arises: how should we understand
the determination of wages and their evolution over time? This is a major
hole in the Dumenil/Foley interpretation, which all those who have pinned
their colours to this mast would, I think, recognise.
And now for some bitty remarks on some of Alan's points. Alan writes
>For me an equally vital parallel step is to recognise that the value
>transferred to the product by the means of production (constant
>capital, C) is equal to the price of these means of production, not
>their value: to put it precisely, the value they transfer to the
>product is equal to the value of the money used to purchase them, or
>the value they represent in exchange rather than the value with which
>they emerge from production, exactly analogously to V.....
However I think there is a conflict between the view that the value of
>variable capital is the price of its elements, and the assertion that
>the value of constant capital is the value of its elements. It seems
>simpler, more consistent and more fruitful to say there is a general
>result: the value transmitted to the recipient of any bundle of goods
>obtained on the market, be it capitalist or worker, is the value she
>or he parts with to obtain it. On the basis of Okham's razor, if no
>other. In a money economy this is in every case equal to the value of
>the money which represents the price of these goods.
>
>Once one makes this step, everything else falls into place.
>
If tendential equalisation of the rate of profit generates unequal exchange,
then it seems theoretically incoherent to follow Alan's proposal, because it
amounts to abolishing the labour theory of value. It is not surprising then
that everything falls into place, because there can be nothing out of place!
This further relates to Alan's remarks on 'sequentialism'
>Though I realise he does not accept the conclusions I draw, I think
>the 'iterative' solution to the transformation problem proposed by
>Anwar was a vital step in the development of this stream of thought -
>it was certainly the decisive influence on me - but I believe that in
>order to complete this evolution it is necessary to break altogether
>with the use of simultaneous equations, because these necessarily
>imply that the prices and values of inputs at time t must also be
>equal to prices and values of outputs at time t: what John calls
>'simultaneous valuation'.
>
>In this sense the difficulty may be greater than Duncan suggests: the
>problem is that simultaneous valuation is not treated in the
>literature as it should be and as Duncan proposes, merely as the
>limiting case of a difference equation; it is treated as the
>epistemological foundation of value and price.....
>
Where you get to in any iterative process must depend on where you start; I
think I have said enough to show that for me Alan is not starting in the
right place.
>1. The appropriate general mathematical instrument for formalising
>Marx's theory is the difference or differential equation, of which the
>simultaneous equation is only a special case. In particular, this
>special case excludes technical change or variations in price.
>
>2. The values and prices of outputs in period [t-1, t] are the values
>and prices of inputs in period [t, t+1]
I don't think anyone would disagree with this, although I don't think it is
as easy as Alan seems to, and if some substantive issues are raised by
'easier' special cases, it is surely better to deal with them at that level.
>3. The share of constant capital in the value of the gross product is
>equal to the value of the money paid for its elements.
>
>4. The share of variable capital in the value of the gross product is
>equal to the value of the money paid for its elements.
Net product not gross product, and even then I cannot see any analytical
reason why point 3 might be true.
>5. Surplus value is what remains after the value of constant and
>variable capital purchased at time t, have been deducted from the
>value of the gross product at time t+1 (that is, constant capital plus
>living labour less variable capital)
Is this not definitional?
>I would add a point 6 which may be more controversial:
>
>6. The special case of simultaneous valuation has been adequately
>studied and failure to go beyond it, at this point in the evolution of
>political economy, has become an obstacle to progress.
>
This is obviously controversial, unless we can agree on exactly what it is
that has been adequately studied.
>1. What actually is the value of money? Once technical change is
>introduced, two different magnitudes emerge: one may calculate a
>quantity we might call the stock-measure of the value of money, the
>ratio between the labour-time expression of the value of all
>commodities in circulation and the money-expression of the same
>collection of goods. This is only the same as what Duncan calls the
>value of money, and which we might term the flow-measure of the value
>of money, if there is no technical change and no inflation. It is
>typical of many such distinctions that they can only be made in a
>general, dynamic formalisation of the type Duncan suggests in [OPE
>787] - a sure indication that such formalisations are richer and more
>appropriate.
'All commodities in circulation' is somewhat ambiguous as between gross
output and net output. The Dumenil/Foley value of money is the ratio of the
labour expended in the production of gross output, (which is the same as the
labour value of net output) to the dollar measure of net output, or the
ratio of labour value added to money value added. This is not what Alan is
proposing above.
One last remark about equal exchange. Suppose that one were to understand
the whole of Capital I as an account at the level of abstraction of equal
exchange, and prices proportional to values. In a Dumenil/Foley
interpretation, the class distribution of value added is independent of any
principle of price formation, so that most of what Marx says in the section
on the accumulation of capital leading to the general law of capitalist
accumulation needs no revision. But since the commodity distribution of
value added is entirely dependent on how prices are formed, I think that any
other approach is going to find Marx's analysis problematic as soon as one
allows unequal exchange.
Finally, something apparently very different, but actually tied up with the
above, albeit at a lower level of abstraction. Without unequal exchange,
issues surrounding productive and unproductive labour become obscure. In
fact the productive-unproductive labour distinction really bothers me,
because there is not much consensus about it, and as soon as one starts to
look at empirical data, it becomes one of the central questions, but can
only be resolved with large amounts of arbitrariness. Maybe we have to be
satisfied with only being able to establish upper and lower bounds to the
time series we might want to construct. What does anyone else think?
Simon Mohun
Simon Mohun,
Dept of Economics,
Queen Mary and Westfield College,
Mile End Road,
London E1 4NS,
UK
Telephone: 0171-975-5089
Fax: 0181-983-3580