[OPE-L:3288] RE: TSS and value added

Duncan K. Fole (dkf2@columbia.edu)
Sat, 5 Oct 1996 07:22:45 -0700 (PDT)

[ show plain text ]

I've tried to summarize my position as clearly as I could in another post,
but I'd like to respond to a couple of Andrew's specific questions:

>A response to Duncan's ope-l 3249.
>I'm going to begin purely purely in labor-time terms, since I remain convinced
>that money magnitudes and discrepancies between them and labor-time
>magnitudes, are irrelevant to the value added/ revaluation/IVA discussion.

I think I mean the same thing by stipulating that the monetary expression
of value is given and constant in the examples we're discussing. However,
as I tried to make clear in my other post, the definition of money value
added plays a crucial role in converting the labor theory of value into a
theory of price and the money rate of profit in your examples.

>If 5 bushels of corn are purchased at their value, 500 labor-hours, at time 1,
>just before being used as means of production, and if they then "count" as
>anything other than 500 at time 2, then I would say that they are being
>revalued. I base this on my understanding of the word "revalued," and not on
>any concept of value added. Thus, when you point out again that the BEA does
>not count the 5 bushels as 500 in value in its value added/IVA calculations,
>I'll again admit that I'm very confused about this but happy to accept it as
>being true, but I'll say that the BEA revalues the corn.

This way of putting it isn't really very clear, because of the ambiguities
of the words "count" and "revalue". The historic labor content of the 5
bushels is 500 labor-hours, while the labor required to reproduce them
might be something else. Each of these concepts might be relevant to answer
different questions.

>If you want to say that valuing the corn as 500 labor-hours at time 2 *is*
>revaluing it, and that the BEA does not revalue it, I'm willing to accept that
>usage in this discussion (but I can't promise I can remember to use the terms
>this way). It is just a matter of terminology.

I don't want to say this, since I think it is very confusing. Let's use the
distinction between historical labor expended and reproduction labor

>What is not a matter of terminology is this. I'm still not sure whether you
>agree that a change in the *aggregate* value of this corn between times 1 and
>2 (say a change of -20 labor-hours, from 500 to 480) should not be counted as
>part of the *aggregate* value added by the extraction of living labor in
>capitalist production.

I do not think that the revaluation of stocks of inputs during the period
of production (in any accounting scheme) ought to be attributed to the
expenditure of living labor in production.

>Clearly the change in the *unit* value of the corn is
>affected by how much living labor is extracted *per unit of output*, but the
>question is this: if the workers work 100 hours between times 1 and 2, have
>they added a value of 100 labor-hours or a value of 100 - (-20) = 120

I agree that 100 social labor-hours adds the monetary equivalent of 100
social labor hours to the value of the inputs.

>I say the former, and add this 100 to the 500 to get 600
>labor-hours as the value of output at time 2.

The phrase "value of output" is ambiguous. It is the sum of the historic
labor embodied in the inputs and the living labor expended in production.
In my view it would not represent the money value of the output.

>Now, in arguing that I compute value added incorrectly,

I don't think I've used the word "incorrect", and if I have, let me
apologize and withdraw it. I am just trying to clarify the point that the
definition of Money Value Added you use to derive the price and money
profit rate paths in your examples differs from the NIA definition, and, as
I understand it at this point, includes the revaluation of the stocks of
inputs through the production period.

>you *in essence* took
>my value of output at time 2, 600 labor-hours, together with the input of 5
>bu. of corn at time 1 and the output of 6.25 bu. of corn at time 2, and
>calculated as follows: take the 600 labor-hours, subtract (5/6.25)*600 = 480
>labor-hours, and arrive at a value added of 120 labor-hours. You then imputed
>this 120 labor-hours to living labor expended. I disagree with this. I
>think that if the workers worked 100 hours, Marx's theory says they add a
>value of 100 labor-hours.

I don't recognize your characterization of my "calculation". The problem
here is that in my view this stage of the example has to involve explicit
reference to the monetary expression of value and its constancy over the
periods. I would calculate as follows: the 100 hours of living labor will
add $100 of value added, assuming the monetary expression of value is
$1/hour. This value added, calculated in NIA terms, is p(t)X(t) -
p(t)aX(t). I gather from your example that X(t) is 6.25 bu of corn, and
that a = 5bu/6.25bu = .8. This gives

p(t)(1-.8)6.25 = $100, or p(t) = $80/bu as the price which is consistent
with the maintenance of the monetary expression of value at $1/hour.

I agree with you that Marx's theory says 100 hours of labor added 100 hours
of labor worth of value to the inputs.

At this point Andrew begins another example involving a change in the
monetary expression of value, but I'd like to defer discussion of this type
of situation until we have better communication on the examples that assume
a constant monetary expression of value.

He then continues:

>There is one way and only one way your equation works --- you cannot accept
>any labor-time value of output as given, or, equivalently, you cannot accept
>any money-value of output *and* labor-time/money relation at time 2 as given.

I'm not sure exactly what you mean by this, but I hope the working out of
the example above makes my position clear. In my view the content of the
labor theory of value lies in the definition of the monetary expression of
value. If we take that as given, the LTV does determine prices, as I just
worked out.

>You *must* hold that if workers work 100 hours, and 5 bu. and 6.25 bu. of corn
>are invested and produced, respectively, the value of the output is 500
>labor-hours, no more, no less. This is, of course, the traditional
>simultaneist answer.

In the example I worked out above the 6.25 bu of corn at $80 a bu will have
a total price of $500. The value added is $100, corresponding to the
expenditure of 100 hours of labor time.

I think the word "simultaneist" is misleading in this context, since I'm
not assuming stationarity of prices.

>The point is this: you cannot take *my* labor-time value figures, or,
>equivalently, you cannot take *my* money-value of output and the
>money/labor-time relation I assumed in order to get the money figures (the
>latter was your actual procedure), and use your value added formula to expose
>an internal inconsistency in my interpretation of value determination in
>Marx's work.

I don't think I used the word "inconsistent" to characterize your model,
and if I did, I apologize and withdraw it. As I recall my post, I was at
pains to emphasize your correct derivation of the price and money rate of
profit paths on the basis of your assumptions. The issue is my mind is not
the consistency of the examples, but the source of qualitative divergence
in the predictions of the price and money rate of profit paths, which I
attribute to the difference in the definition of Money Value Added.

>Rather, it is your formula that is inconsistent with my
>interpretation and with any interpretation other than the simultaneist one.

I still don't think my position is "simultaneist", since I don't assume
prices are stationary.

>Your formula implicitly assumes that the latter is correct. If my value of
>output is correct, or any value of output other than l/(1-a) is correct, then
>your formula either gives the wrong labor-time/money relation or gives a value
>added that diverges from the living labor extracted, and it therefore can't be
>used to show internal inconsistency in my interpretation.
>Now, you may accept that the unit value of output must be l/(1-a) --- I'm
>still not sure. If so, we have a difference in interpreting Marx, and you and
>I will get different results for almost everything, the rate of profit, total
>value, the money/labor-time relation, constant capital, etc., everything
>*except* value added in labor-time terms (and perhaps variable capital and the
>rate of surplus-value, since I think we agree that variable capital isn't
>revalued [in my terms, i.e., doesn't change] over the course of the production

Sure, there's got to be a difference if we are lead to qualitatively
divergent predictions as to the path of prices and the money rate of profit
in the same model. In my view, however, this difference doesn't have to do
with "simultaneism" or with our interpretation of the labor theory of
value, but with the definition of Money Value Added.

>But as both Fred and Bruce recognize, in different ways (Fred's way is mine),
>whether I'm getting the wrong value added is not the issue. We both get the
>same value added in labor-time terms. The real issues are first, whether
>constant capital and therefore the value of output are what simultaneism says
>(400 labor-hours and 500 labor-hours in period 1 of my example) or what
>temporalism says (500 labor-hours and 600 labor-hours).

This remark seems to me to conflate two different issues. From the
capitalist's point of view the constant capital invested is $500, and the
money rate of profit will be calculated on that denominator. On the other
hand, the sales price of output consistent with the LTV and the NIA
definition of Money Value Added will be $500. The capitalists, as John
Ernst emphasizes, lose money on the stocks of inputs during the production
period due to the fall in prices, and as a result their money rate of
profit will be lower (as I tried to make clear in my other post). The point
at issue, however, is the diverging price and money profit rate path.

>Second, whether
>simultaneous determination can make any sense intertemporally, which I
>addressed in an earlier post today.

As I've said already, I don't view "simultaneism" in the sense of assuming
input and output prices as the same as relevant to our discussion.

>And third, which interpretation is
>consistent with Marx's results.

This is a big question, and one which might be very hard to achieve general
consensus on. But I don't recall any passage where Marx argues that a
capitalist economy with a constant real wage experiencing labor-augmenting
technical progress and a constant ratio of output to capital would
experience a fall in the money rate of profit to zero as your example


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