[OPE-L:3413] RE: Marx and historical costs

andrew kliman (Andrew_Kliman@msn.com)
Mon, 14 Oct 1996 23:12:10 -0700 (PDT)

[ show plain text ]

A reply to Fred's ope-l 3411. I'm going to deal with the issues Fred poses
somewhat pedantically, because I think that doing so may help clear up some

Fred wrote: "there seems to be an inconsistency in Andrew's interpretation of
the valuation of constant capital: with respect to the 'transformation
problem' Andrew agrees that constant capital is NOT valued in historical
costs, but with respect to the 'falling rate of profit', he argues that
constant capital is valued in historical costs."

In the same post, Fred defines historical costs as "the actual costs at which
the means of production were purchased at various times in the past." He
distinguishes this from "constant capital ... valued at the prices prevailing
at the BEGINNING of the CURRENT production period."

Using this terminology (which is the terminology I myself use), my
interpretation of Marx's value theory is this: the constant capital-value
preserved by production and transferred to the value of the product is
determined by the prices prevailing at the *beginning* of the *current*
production period. I agree with Fred ONE HUNDRED PERCENT that all the textual
evidence, and there is a whole lot of it, in many different texts by Marx,
indicates that the amount of constant capital-value transferred is NOT
determined by the historical costs of the means of production, as defined

I employ this interpretation not only to the value/production price
transformation, but also to the FRP. Whether for one-sector (YES!) or many, I

[1/e(t+1)]p(t+1)Q(t) = [1/e(t)]p(t)*A(t) + N(t) (total price)

[1/e(t+1)]v(t+1)Q(t) = [1/e(t)]p(t)*A(t) + N(t) (total value)

e(t) and e(t+1), both scalars, are the monetary expressions of value at the
start and end of period t. p(t) and p(t+1) are the vectors of unit input and
output *money* prices of period t. v(t+1) is the vector of unit *money*
output values of period t. Q is output (a vector), A is circulating means of
production plus physical depreciation of fixed capital (a vector). N is
living labor (a vector). (A and N are absolute, not per-unit, amounts here.)

These equations deflate money magnitudes by the appropriate MEV to get
labor-time sums. One can multiply by e(t+1) to get the money prices and
values of output.

Now note that value transferred, measured in labor-time, is [1/e(t)]p(t)*A(t).
So it depends on p(t), the *current* money price of the means of production
at the beginning of the period. These means of production could have been
produced, not at the beginning of this period (= end of last period), but at
the beginning of the period before that, or 10 periods before, or 100. It
doesn't matter. They transfer the value they have at the beginning of *this*

With respect to the rate of profit, note that total price equals total value.
Therefore, although total price can be *written* as a function of the rate of
profit, it is *determined* as outlined above. And therefore, however the
profit rate may be computed/measured, different measures of the profit rate do
NOT and can NOT affect the total value or the total price of output.

Hence, if I compute a rate of profit in which capital (fixed and circulating,
constant and variable) is valued at historical cost (as I do), this does not
mean that the VALUE TRANSFERRED is computed at historical cost. On the
contrary, it is computed at its current cost of reproduction at the start of
the period, [1/e(t)]p(t)*A(t). So there is no inconsistency in the way I
formalize the determination of value. I do it the same way for the
transformation as for the FRP.

Now, in addition to profit divided by historical cost, I can also compute a
measure that values capital at the current prices at the start of the period,
as well as a measure that values capital at the current prices at the end of
the period ("replacement costs"). I am happy to call them *all* rates of
profit, though they each measure something different and mean something
different. None of these different computations have *any* effect on the
determination of value. The value and price of output will always be
determined as above, on the basis of the current prices of means of production
at the start of the period.

Is there any inconsistency here? I think not. The value of output is always
current, and profit is always current. Yet it is entirely legitimate to
*express* profit in relation to the current cost of the means of production,
etc. advanced, but also to *express* it, alternatively, in relation to the
capital actually advanced. None of these expressions imply that the numerator
is being determined by the denominator, any more than per capita GDP implies
that the magnitude of GDP is determined by population size or any more than
the unemployment rate implies that the number of unemployed is determined by
the size of the labor force. The number of unemployed remains the same even
if we measure unemployment in relation to population instead of to the labor

Again, I maintain that total price (= total value), and therefore profit, are
determined independently of the amount of capital advanced, and are instead
determined as outlined above, in Marx's theory.

As Fred knows, I think Marx employs more than one measure of the profit rate.
I think he has an historical cost measure and a replacement cost measure, at
least. I maintain that his law of the tendency of the rate of profit to fall
refers to the historical cost rate.

I will also note that capitalist firms, and governments, measure a variety of
profit rates without committing themselves to any theory --- and without
caring one whit about any theory --- of how total price is determined. All of
these rates are legitimate; they just measure somewhat different things. One
of these is the rate of return on investment. It is an historical cost rate,
because it equates the actual amount invested, not the replacement cost of the
resources invested, to the discounted stream of returns.

Andrew Kliman