[OPE-L:3185] Re: IVA and all that

Duncan K. Fole (dkf2@columbia.edu)
Mon, 30 Sep 1996 06:51:45 -0700 (PDT)

[ show plain text ]

Some comments on Jerry's methodological comments:

>While I certainly support Duncan's call in [OPE-L:3178] for us to
>attempt to develop a "catalogue of the things we seem to agree on",
>I am less clear than ever about what, if anything, we *do* agree on.
>I *thought* this discussion concerned, in part, technical change and
>the FRP. Am I correct?
>Yet Duncan, John, and Andrew want us to focus now on a "circulating
>capital 1-sector model." Duncan believes that "all of the
>methodological points can be made in this context."
>Can they?

Of course not all of the points relevant to a discussion of the FRP and
technical change can be made in the circulating capital model. As I see it,
we have a situation where a group (the TSSers) have put forward a novel
view of the FRP, based on their working out of a large number of numerical
examples. This view has some quite startling implications, such as that
technical change in and of itself, even when it augments both capital and
labor inputs, leads to a FRP. Some of the rest of us are having trouble
understanding how they get these results in these models on the basis of
assumptions that we share (such as that the value added is proportional to
living labor time expended.) It seems to me that the most fruitful path to
clarity (if not agreement) on where the TSS examples deviate from other
points of view lies in examining simplified models where the points of
departure can be uncovered transparently. This is, in my view, quite a
different exercise from considering the actual patterns of technical change
in a capitalist economy.

>* I thought that the discussion of technical change and the FRP by
> Marx was made in the context of a "level of abstraction" (pace
> Andrew K) where there was *competition*. How can you have competition
> in a 1-sector model?

I don't see why a 1-sector model precludes competition, since you could
imagine many captitalists employing essentially the same technology to
produce either a single output, or a spectrum of outputs all of which have
the same technical production conditions, and competing with each other in

>* I thought that the FRP concerned the tendency for the *general* rate
> of profit to fall. Clearly the LTGRPD was presented *after* the
> examination of the formation of a general rate of profit. Yet,
> now we are being asked to (for the time being) not consider
> different branches of production. Consequently, the possibility
> of having different branches of production with differing compositions
> of capital is assumed away through the methodological decision to
> employ a 1-sector model. Is this legitimate?

The logic, as I see it, goes as follows: the 1-sector model is a special
case of the general model. Therefore any general proposition claimed for
the general model has to hold in the 1-sector model, and any
counter-example in a 1-sector model disproves a conjecture for the general
model. On the other hand, proving that some property holds in the 1-sector
model does not establish it for the general model without further argument.
Thus any theory of the FRP that has problems in the 1-sector context will
have the same, and maybe worse problems in the general model.

>* It has long been known (since at least the mid-70's) that *all*
> 1-sector "growth models" exhibit certain pecularities akin to the
> "razor edge" in the Harrod-Domar model. Since we are discussing the
> "transition" from one period to another, *why* are we then using
> a 1-sector model? Yes, I know -- it simplifies the equations. But
> at what cost? One thing is certain: whatever "agreements" we come
> to in the case of where we are examining a 1-sector model, we can
> not assume that either the results or agreements will be the same
> when we examine a multi-sector model.
>* Why are we assuming a circulating-capital model? As far as I can tell,
> the only reason is because *Okishio* (not Marx) made this assumption.
> Of course, Roemer attempted to generalize the Okishio results in a
> model where there was "fixed capital." I thought (and John seems to
> agree) that in a discussion of the FRP in *Marx* _constant fixed
> capital_ is important. Yet we assume it away -- all in the name of
> simplifying the equations.

The TSS interpretation of the FRP has dramatic implications that are
sharply at variance with other understandings. These divergences are
already present in the simplest model cases. If, for example, the issue
turns out to be the definition of the value added and the application of
the LTV, then diverging interpretations of the 1-sector circulating capital
case are going to reappear in a more complex and more confusing form in the
more general models.

>* I find it more than a little ironic that with all of our discussions
> about the forms that technical change takes in the period of
> "modern industry" and after our extended discussions of straight
> and moral depreciation, we now assume a circulating capital model.

The distinction between "fixed" and "circulating" capital depends on the
period, anyway. If we consider a period of 10 years, 10-year lived capital
facilities become "circulating" in that time-frame.

>* At the very least, any suggestions that what we are discussing concerns
> technical change, depreciation, or the FRP in *Marx* should be
> dropped. Or, would anyone actually suggest that the process that
> Marx examined can be explained within the context of a 1-sector
> circulating capital model?

I would, on the grounds that the tendencies Marx talked about were
sufficiently fundamental that they ought to show up at any level of
abstraction, which is what is at issue here. If Marx's FRP can be
conceptualized only in a model with many competing capitals in many sectors
with different organic compositions of capital and complicated temporal
patterns of production, then it loses a lot of its generality and interest.
On the whole, however, Marx seems to have had an excellent instinct for the
fundamental issues, so his conjectures can be described at every level of

This issue of abstraction is worth thinking about (I suggest in the context
of concrete problems like the FRP), because a lot of wheel-spinning arises
when people don't locate the appropriate level of abstraction at which to
address the issue they're interested in.


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