[OPE-L:792] Re: Still More Digression

Duncan K Foley (dkf2@columbia.edu)
Tue, 16 Jan 1996 06:14:40 -0800

[ show plain text ]

On Mon, 15 Jan 1996, John R. Ernst wrote:

> Duncan,
> I agree with what you say below and take note of
> what J. Levy wrote concerning this. The problem
> is how we deal with technical change and valuation
> as we move from one period of production to another.
> To be sure, we could start with a model that resembles
> Marx's reproduction schemes. Yet, when we introduce
> technical change and assume some initial drops in
> price as Marx does, we are confronted with problems.
> a. How do we deal with the loss of capital value
> as prices decrease?

There are a number of models that deal with this problem, including some
I looked at in "Money, Accumulation and Crisis" in the context of the
Marxian circuit of capital. In a competitive economy existing capital
goods are systematically devalued by technical change, and their owners
experience a capital loss.

> b. Are these loses, to some extent, anticipated by
> capitalists as they figure "moral depreciation"
> into their pricing strategies?

I think so, as a matter of practical fact. Most studies of business
investment decisions show that they use relatively primitive calculations
of rate of return, mostly along the lines of payback times which are
typically very short. This may reflect businesspeople's caution about the
uncertainty of future technical change. The real estate industry is
particularly prone to disasters involving devalutation of assets.

> c. Given we answer "a" and "b" in a way that takes
> into account decreases in prices and values, how
> do we move to model that applies to a modern
> economy where there is a conscious effort to
> prevent nominal price decreases?

I don't understand the phrase "conscious effort to prevent nominal price
decreases". In a competitive economy there's nothing, by definition, a
producer can do to prevent competitive price decreases. If a firm, like a
computer maker, has some monopoly power, they can use it dynamically.
There's a substantial literature on this problem in the mainstream journals.

> Thus, I see the overall project as two-fold. First,
> the effort has to be to understand Marx in the
> context of the time he lived -- a world with price
> decreases as productivity increased. Second, once
> we have a better idea of Marx's outlook, we can,
> hopefully, develop a view of the economy of today
> in which prices generally do not fall and, indeed,
> more often, increase.

Of course it's not true that prices don't fall, as the computer industry
and a number of other highly oligopolistic and technologically
fast-moving cases shows. I also would be careful about distinguishing
between nominal prices, which have an element of valuation of the
government's liabilities in them, and relative prices, which reflect real

> Thus, my problem with simultaneous valuation is that
> it loses much of whatever Marx was trying to say and
> leaves us in a different world without the tools he
> developed.

Most of the discussion about the transformation problem takes place at a
pretty abstract level, where the main concern is about the consistency of
certain theoretical frameworks and interpretations. I assume that people
tend to argue these issues through on the assumption of an economy
without technical change and in long-run equilibrium because that is the
simplest case analytically, and the theoretical issues are all present
there without the need to grapple with the complexity of more realistic
models. In fact, I think it is a sound methodological procedure to
require someone to explain their interpretation, say, of the labor theory
of value in this context before one can understand how it might apply to
more complex, and more realistic, cases.


> Duncan says:
> If you write down an explicitly disequilibrium model with time subscripts
> differentiating commodities at different times, thus allowing for prices
> of inputs and outputs to differ, one solution will be the equilibrium
> prices where the inputs and outputs have the same (relative) prices. This
> is also the easiest solution to analyze, and its existence is a good
> indication that the equations make sense. As an historical aside, this
> seems to be the approach Marx took, for example, in his work on
> reproduction schemes.
> John