[OPE-L:3476] RE: Marx and historical costs

andrew kliman (Andrew_Kliman@msn.com)
Fri, 18 Oct 1996 11:32:28 -0700 (PDT)

[ show plain text ]

A reply to Fred's ope-l 3460.

Fred: "Andrew argues that the STOCK of constant capital (the total price of
the means of production; the denominator in the rate of profit) is valued
differently from the FLOW of constant capital (the value transferred from the
means of production)."

Yes, *if* one is referring to an historical cost measure of the profit rate.
Again, I think other measures of the profit rate are valid for some purposes,
and I think Marx also employed something like a "replacement cost" measure of
the profit rate in some contexts. I think, however, that the rate of profit
which he says will fall due to increases in the TCC/OCC and rising labor
productivity is an historical cost rate.

Fred: "In the absence of evidence to support this dual interpretation of
constant capital, there would seem to be a general presumption that the stock
and the flow of constant capital would be valued in the same way. And since
the evidence regarding the valuation of the flow of constant capital in
current costs is so strong, this presumption of similar valuation suggests
that the stock of constant capital is also valued at current costs."

I don't accept this. Why should it be presumed they are valued in the same
way? Why can't I claim, on the contrary, that we should presume they are
valued differently, so that no evidence is needed to support this presumption
but that evidence is needed if it is to be challenged?

Actually, I don't want to make that claim. Rather, I want to say that
*external* evidence is relevant --- evidence about how capitalists think and
keep their accounts --- and that this evidence creates an a priori presumption
that, for *some* purposes, the stock was not revalued by Marx. The evidence
concerns rates of return. Historical costs are used. If, for instance,
someone buys a perpetual bond for $1000, and gets interest of $100 p.a. on it,
his/her rate of return is 100.a. If the price of this bond plummets in the
market to $400, this does not make him/her better off. It does not raise
his/her rate of return to 25%. On the contrary. S/he has suffered a $600
capital loss (whether or not s/he sells the bond, BTW). The same goes for
the internal rate of return on productive investments. This is the way
capitalists think and calculate, this is the way the system operates (or at
least seems to) and therefore, in the absence of evidence to the contrary, it
should be presumed that Marx also refrained from revaluing the investments
when discussing rates of return.

There is however, additional, internal textual evidence to support my
interpretation. To my knowledge, Marx never wrote explicitly anything like
"[t]he flow of constant capital is valued at current costs, but the stock of
constant capital is valued at historical costs" or "technological change will
reduce the flow of constant capital, but will not reduce the stock of constant
capital," but many things he wrote presuppose this, IMO. The lack of explicit
statements is readily explained by noting that, since capitalists do measure
their returns against their actual investments, Marx probably thought these
points were obvious.

So what is my evidence? Off the top of my head, this is what I have come up
with; there may be more:

(a) the concept of capital as "advanced"
(b) the concept of the rate of profit as an external measure of the
"self-expansion of capital"
(c) the concept of "moral depreciation"
(d) Marx's differential treatment of used up constant capital and the
constant capital advanced in his account of the transformation of values into
production prices
(e) my interpretation replicates Marx's theoretical results concerning the
rate of profit (and much else)

I'll refer to (c) now, in connection with the next point in Fred's post, and
come back to the rest at the end.

Fred: "According to Marx, the magnitude of the flow of constant capital (the
value transferred from machines) (c) in any
given accounting period is equal to the quotient of the stock of constant
capital (the total price of machinery) (C) and the expected number of periods
the machines are expected to last (n): i.e. c = C / n. Therefore, assuming
that the expected lifetime of the machines remains constant, the flow of
constant capital can change only if the stock of constant capital changes. If
technological change occurs, this will change the stock of constant capital,
which will in turn change in flow of constant capital."

I have no problem with this formula when used to discuss turnover, which is
how Marx used it, if I remember. But I don't consider it evidence that the
sum of value transferred throughout time must be equal to the stock of
constant capital.
The reason is that ---and here is my first and most important piece of
positive internal evidence --- this (Fred's) interpretation precludes the
possibility of MORAL DEPRECIATION (MD).

As I understand Marx's concept of MD, the following is true by definition:

value of capital stock = total value transferred throughout time + total
MD throughout time

To relate this to the formula above, assume the capital is acquired in period
0 and lasts n periods. Then I would write

value of capital stock = C(0)

and therefore

C(0) = sum of {c(t)*n} + sum of {MD(t)}

where t = 0, 1, ..., n.

Therefore there will be MD iff the sum of value flows is *less* than the value
of the stock:

sum of {MD(t)} > 0 <====> sum of {c(t)*n} < C(0).

Fred, on the other hand, would write

value of capital stock = sum of {c(t)*n}

and therefore

sum of {c(t)*n} = sum of {c(t)*n} + sum of {MD(t)}

so that sum of {MD(t)} = total MD throughout time = 0.

Hence, MD results precisely from a difference between the sum of flows and the
stock. If the *only* measure of the capital stock is one that equates it,
always, at every moment, to the sum of value transferred, there can be no MD.
In other words, there can be no *losses* of capital-value due to falling
values. And consequently, there can be no crises of devaluation. As Michael
P. wrote in ope-l 3429 on Tuesday:

"Now, think about a sequence of rapid technological changes, each of which
wipes out much of the historical costs accross an industry. After a few
iterations, the industry [or on a grander scale, business in general] faces

"This is the logic that links historical costs to the frp."

Fred quotes a passage in Ch. 8 of Capital I:
"As the value of the raw materials may change, so too may that of the
instruments of labor, the machinery, etc. employed in the process; and
CONSEQUENTLY that portion of the value of the product transferred to it from
them may also change. If, as a result of a new invention, machinery of a
particular kind can be produced with a lessened expenditure of labor, the old
machinery undergoes a certain amount of depreciation, and THEREFORE transfers
proportionately less value to the product. C.I. 317-18; emphases added)"

and concludes from it that Marx held that "[t]echnological change reduces the
stock of constant capital (the value of machinery, etc.), and CONSEQUENTLY
also reduces the flow of constant capital (the value transferred)."

But note also that Marx does NOT say the *constant capital-value* invested
depreciates, only that the *old machinery* does. As we know, there's a
conceptual difference between the value of the means of production and the
value of constant capital. My reaction is the same concerning the two other
passages Fred quotes (Econ. Mss. of 1861-63, and Ch. 6 of Capital III): Marx
is referring to technological change reducing the value of the machines, not
the value advanced as capital. When Fred interprets the former as being "in
terms of the value of the machinery, etc., i.e. is about the stock of constant
capital," I maintain that the first phrase isn't identical to the 2nd.

Marx certainly doesn't say, in the above passage, that part of the firm's
investment vanishes into thin air without even bringing about a loss on the
investment, so that its rate of return can be computed as if the lost value
never existed.

But even if Marx had been referring to the value of the capital stock in these
passages, I would still not consider them evidence against my interpretation
because, again, I do not deny that *one* measure of the stock of constant
capital is affected by technological change. I only deny that this is Marx's
*sole* measure, and that it is not the measure to which his law of the FRP

Now to the rest of my evidence that Marx didn't always equate the stock of
capital-value to the sum of value transferred. First, he repeatedly writes
about the capital-value being "advanced." This term seems to come from the
French; _avant_ means "before." The capital advanced is the sum of value
actually laid out, invested. It simply cannot be revalued.

Closely related to this, second, is Marx's concept of the "self-expansion of
capital" and the rates of surplus-value and profit as the internal and
external measures of the rate of self-expansion. In the latter, surplus-value
is related to the value of the capital advanced. The measure gives us the
rate of growth of the capital, the relative size of the increment
(surplus-value) to the capital originally invested, i.e., the relative extent
to which the capital-value advanced has brought forth an addition to itself.
A simultaneous measure of the rate of profit is certainly not a measure of the
rate of self-expansion. In the example of mine that we've been discussing,
the simultaneous calculations tell us that the farmer casts out a value of 500
on the waters, and 500 returns at the end of the period. The rate of
self-expansion is 0, even though the simultaneist "profit rate" is 25%.

Third is the moral depreciation issue, discussed above.

Fourth is Marx's differential treatment of used up constant capital and the
constant capital advanced in his account of the transformation of values into
production prices. As we know, he says several times that the "cost-price" of
commodities is affected by the difference between production price and value,
so that the cost-price should not be understood as the sum of the *value* of
the used up means of production and subsistence, but their *price*. Yet in
none of these passages does he ever say that the constant capital advanced,
i.e., the constant capital as a whole including the part not consumed, is
affected by the transformation. Why not? Why is the constant capital-value
advanced not raised (lowered) if the price of means of production is greater
(less) than its value? Because it was already advanced. It cannot change in
the next period. Thus, importantly, the rate of profit is based on the
*given* capital-value advanced, to measure the rate at which capital
self-expands. Now subsequent investments, additions to the capital stock,
will be valued at the prices reigning when the investments occur, but the sum
already invested will not.

Fifth, my interpretation replicates Marx's theoretical results concerning the
rate of profit. The profit rate falls precisely *because* labor becomes more
productive if the TCC/OCC increases. The rate of profit in luxury production
affects the general rate. The distribution of profit across firms/branches
does not affect the general rate. The determinants of the profit rate are
value sums and are irreducible to technology and the real wage. With a
replacement cost measure of the profit rate, none of these results hold.
Therefore Marx was not speaking of a replacement cost rate in these contexts.

Now, it is true that one could maintain that the whole stock of capital is
valued at its PRE-production cost of reproduction and get these results. But
a fall in the profit rate due to labor-saving and productivity-increasing
technical change would be much, much less likely, hardly something one would
want to state as a law. And for the other reasons I've mentioned, especially
the connection between devaluation of capital, the FRP, and crisis, it seems
clear to me that only historical valuation of the capital stock is plausible
in connection with the law of the FRP.

Andrew Kliman