[OPEL:6190] Re: Historical Costs (Fred)

John R. Ernst (ernst@PIPELINE.COM)
Sat, 14 Feb 1998 14:16:22 -0500 (EST)

Hi Fred,

Thanks for your quick response. My last post was a bit too hurried
but I do think we may be getting somewhere on this. Let me see if
I can combine our ways of depicting the matter in terms of the
notation each of us used.

You wrote:


T1 is the beginning of some production period in the PAST (could be the
previous period, or five years ago, etc.) If the past period is the prior
period, as in your framework, then T1 would correspond to your 1B.

T2 is the BEGINNING of the CURRENT period. If the past period is the
prior period, as in your framework, then T2 would correspond to your 2B.

T3 is the END of the CURRENT period. Which would correspond 2E in your

Technological change is assumed to take place between T1 and T2 and
between T2 and T3.

The point that you make that I would emphasize is that the value of
the capital in your period 2 is not simply the historical cost minus the
actual depreciation, but also is devalued as a result of technological
change (moral depreciation). In my framework, this means that the value
of the capital in the current period is determined at T2, not at T1, as I
said in my original post.

I comment:

Here, we have a bit of unclarity. I do not completely
understand your statement (Pardon the repetition.):

"The point that you make that I would emphasize is that the value of
the capital in your period 2 is not simply the historical cost minus the
actual depreciation, but also is devalued as a result of technological
change (moral depreciation)."

I'd agree with this if you stated that the moral depreciation to which
you refer is taken into account in the 1st period.

Let me see if I can use your T1,T2 and T3 to think about this.

Value at the beginning Value at the
of the period end of the period.
1 C(T1) C(T2)
2 C(T2) C(T3)
3 C(T3) C(T4)

You might say I am mushing things a bit but let me explain. For the
sake of clarity, let's start with the purchase of the machine at
the beginning of period 1. This initial advance (say a purchase of
a new machine) takes place immediately prior to its use in
Period 1. (Here I simply want to avoid going 5 periods back.)
At the end of the first period, the machine has a value C(T2).
The initial value C(T1) is decreased by the amount of depreciation,
natural and moral, in that first period. If we call the depreciation,
X(1), then we can write:

(1) C(T1)-X(1)=C(T2).

If there are better and/or cheaper machines available at the end of
the 1st period, the X(1) would include moral depreciation. But you
want to talk about technical change in Period 2. Fine. At the
beginning of Period 2, the depreciated machine is worth what it is
at the end of Period 1. At the end of Period 2, C(T3) is the
value of the machine and, again, we see that

(2) C(T2)-X(2)=C(T3).

We could keep going till the machine's life ends but I think we need
to linger a bit. You refer to Period 2 as the current period. So
let's say we are in Period 2. If at the end of that period, better
and/or cheaper machines show up the now "twice-used" machine will
suffer moral depreciation. In (2), X(2) increases. What the
capitalist advanced as fixed capital at the beginning of the period
is what the machine was worth at that time. There's no going back
in theory or in practice. But the greater X(2), the less C(T3).
The fixed capital has suffered devaluation as those better/cheaper
machines become available. Less value is advanced at the start of
Period 3, C(T3). But note that C(T2) is unchanged. Devaluation has
occurred since C(T3) < C(T2).

At the start of Period 1, there is one technique available for use
in that period. The possibility of an improved technique appears
at the end of Period 1. This decreases the value of C(T2) and does
not affect C(T1). Why is C(T1) unchanged?

It represents the investment at the beginning of Period 1. The
moral depreciation only appears at the end of the period when a
new technique becomes available. You cannot go back in time and
change something that already happened.

Constant capital has been devalued and we could write:

(1) C(T1)-X(1)=C(T2)

Let's return to what you call the current period, Period 2.
In looking at it, I think what you might do is to reduce
the figure for C(T2) based on the new technique that becomes
available at the end of the Period(2). But our poor capitalist
saw his investment at the start of Period 2 as C(T2) and
nothing less. Note that these new machines that become available
at the end of Period 2 are not used in competition with the
machine he used in Period 2. Why then must he revalue the
amount he invested at the start of the period -- it has already
been devalued from C(T1) due to the moral depreciation in
that first period?

I hope this is clear. I have read most of the passages you
cite in Marx in this way. (Your compilation is quite
extensive so I make no claim that I've gone over all of it.)
Again, I see the devaluation of fixed capital due to
moral depreciation. If one were to focus on a single
period of production with no reference to that which
comes before or after it, one is forced into what we
might call "current cost" valuation of inputs and outputs.
There is simply no alternative. Note that this is
precisely the procedure used by the "linear production
approach." It focuses on a single period and ignores
Marx's notion that the capitalist processes of production
and reproduction must be seen as sequence of temporal
phases. Further, with that approach, we have prices and
values that in no way correspond to what the capitalist sees
as he invests and computes his rate of profit. Indeed, only
the capitalists seem to know that investments in fixed capital
mean that his returns span many periods of production. The
theorist is in the dark, or, rather, stuck at a point in time.
He is forced to assume that the new machine that became
available at the end of any period was actually used in production
during that period. It is as though technical change must stop
at a moment in time so that we can derive prices and rates
of profit solely from that momentary data. But, as Marx himself
observed, technical change is more or less continuous;
hence, if we are to "model" Marx, continuous technical
change should be at least possible. It is not when we
focus solely on some magic moment, known only to economists.

Hopefully, this clarifies things a bit. Let me deal with another
issue that, so far, I have put on the back burner -- historical costs.
For me, the fixed capital advanced at the beginning of Periods
1 and 2 is C(T1) and C(T2), respectively. If we were to compute
the total value of the advanced fixed capital for any given
period, say Period 2, we would have to add up the values of
all fixed capital at the start of that period no matter how
old it is. In so doing, I would be ignoring the value
of the various hoards that represent funds collected as
depreciation by the capitalists. Thus, tacitly, I have been
assuming that they are minimized and used for further accumulation.
However, what if the capitalists were required to hold onto these
funds until the machine must be replaced?

We could then say that the amount of capital advanced initially
represents the figure advanced in all periods. At first, the
entire advance would be in the value of the machine. As it
depreciates from period to period, less value is in the machine
and more in the form of hoarded funds. Hence, the capital
advanced in each and every period would often be unchanged.
(Assuming that moral depreciation were not so large that this
hoard would experience insufficient growth.) I do not think
of the matter in this way and thus simply assumed that we
could start our discussion by assuming that C(T1) was the
capital advanced historically or in the first period of the
machines use. Ignoring hoards, I computed C(T2) solely on
the basis of the value of the machine at the start of
Period 2.

Fred, as always, this discussion is helpful. I hope to hear
from you soon.