[OPEL:6210] RE: Historical, real and current costs: general

Duncan K. Foley (dkf2@columbia.edu)
Sun, 22 Feb 1998 21:58:53 -0500 (EST)

On Andrew's "Historical, real and current costs"

Andrew continues the discussion of the definition of the Monetary
Expression of Labor Time, making several points, and posing some questions
in the form of examples.

Andrew contrasts what he calls the "TSS" definition of the MELT as the
ratio on a given date of the money value of the commodity stocks on that
date (including stocks of fixed assets) to the labor time embodied in those
commodity stocks, with the "New Interpretation" MELT, which is the ratio of
the flow of value added to the flow of labor time expended.

Andrew then gives three "examples" raising questions about the New
Interpretation MELT.

This post comments on the general points Andrew raises.

Andrew's general points, with my comments:

>Duncan writes:
>"From my point of view, the difficulty with Alan's stock definition is
>that it
>will not in general lead to the quantitative equivalence of gross profit and
>unpaid living labor time. In other words, if you divide the money measure of
>gross profit over the year by Alan's definition of the monetary expression of
>labor time [MELT], you will get a number of labor hours that is not equal to
>the unpaid labor time over the year."


>First, however, I want to clarify some terms for non-initiates -- others can
>skip to the next section.

>What Duncan calls Alan's "stock definition" is what I'll call the temporalist
>interpretation. It is an instantaneous measure of the relationship between
>value expressed in money and value expressed in labor-time, a measure of that
>relationship at a particular "point" in time. For instance, in Alan's
>example, the MELT on 1 January 1997 is the ratio on 1 January 1997 between
>money-price of the stock of all commodities in circulation at that time, and
>the labour-time value of the stock of all commodities in circulation at that

This has the advantage of making explicit an alternative definition of the
MELT, which I was wondering about in earlier posts on this topic.

While it's not theoretically critical, it's worth pointing out that this
definition of Alan's bristles with difficulties of measurement, and is
probably not operational in terms of accounting data that we will ever
have. The "money price of the stock of all commodities in circulation" on
Jan 1, 1997, on the one hand, is very difficult to pin down. (I assume "in
circulation" just means "in existence".) This stock includes the stock of
fixed capital. As John Ernst has so eloquently pointed out, there is no
unambiguous way of valuing the stock of fixed capital, because we don't
know the "true" depreciation on it. Furthermore, some stocks of commodities
are revalued in money terms all the time through speculative commodity
markets. Which exact prices are we going to use?

The "labour-time value of the stock of all commodities in circulation" also
raises a number of difficult issues. The stock of commodities in existence
on Jan 1, 1997 were produced at widely varying times. In the case of some
fixed capital, they may have been produced decades ago. To begin with, we
have to keep track of the "vintages" of all the stocks, in order to trace
their labour-time value. But even if we have done this, it's not clear what
sense there is in adding together labor times from widely different
periods, say, from 1987 and 1996, or what the appropriate method of
weighting past labour-times would be in order to reach the denominator of
this MELT.

Andrew continues:

>The alternative "flow definition" (?) is what I'll call the simultaneist
>interpretation. It measures the relationship between value expressed in
>and value expressed in labor-time over some "interval" of time. For
>the MELT on 1 January 1997 is the ratio between the money-value-added of the
>period 1 January 1997 to 1 January 1998 , and the current labour of that


I think it's accurate to characterize the New Interpretation MELT as a
"flow" conception. (This is why it escapes the measurement problems I've
just mentioned. You can measure value added from actual sales of goods
produced in a period, and as long as the period is short enough, you can
aggregate the labor expended in it.) There's no reason, however, why you
couldn't in principle calculate the NI MELT at a single instant of time by
taking the limits of the MELT over smaller and smaller intervals covering
the instant. Furthermore, calculating the MELT by taking the ratio of a
yearly value added to yearly labor expended is at best only an
approximation to the MELT at any point in time, since it involves some time
aggregation. Quarterly would be better, and daily better still, but we
don't have daily measures of value added.


>A further distinction concerns the valuation of materials (and fixed capital
>depreciation, but I'll ignore that for expository simplicity). According to
>the temporalist interpretation, they are valued at their cost (in money
>and in
>labor-time terms) when they enter production, while the simultaneist
>interpretation holds that they are valued at their end-of-period "replacement
>cost." Note that this difference gives rise to two different interpretations
>of "profit," and of "total price" and "total value."
>The temporalist interpretation of the MELT is held by proponents of the
>temporal single-system (TSS) interpretation of Marx's value theory. The
>simultaneist interpretation is held by proponents of the "New Interpretation"
>(or "Solution") and, if I'm not mistaken, by all proponents of the
>simultaneous single-system interpretations (Chai-on, Bruce, Antonio, Fred,


The reason for being interested in the NI MELT is because it is a
mathematical theorem that it is the _only_ definition of the MELT which
always assures that value added is proportional to living labor time, wages
to paid labor time, and gross profit to unpaid labor time. In fact, you can
show with elementary algebra that the NI MELT is exactly equivalent to
assuming these proportionalities (which is why some critics have dismissed
it as trivial and tautological). I hope Alan and Andrew consider this point
carefully. I don't think anyone questions that Marx insisted on these
proportionalities in his treatment of the relation between labor values and
prices, whatever he may have said about "total value" and "total price" and
the average rate of profit in value and price terms.

I'd therefore reiterate the point I originally made, which is that once we
sort out exactly how to calculate the "stock" MELT, it will either turn out
to be equivalent to the NI MELT, or it will sometimes fail to respect these


Duncan K. Foley
Department of Economics
Barnard College
New York, NY 10027
fax: (212)-854-8947
e-mail: dkf2@columbia.edu