Marxian empirical research

Fred B. Moseley (
Tue, 13 Jan 1998 15:54:51 -0500 (EST)

I have a few brief comments on the recent discussion
initiated by Alan of a collaborative project to derive
comparable Marxian national income accounts for as
many nations as possible.

1. In general, I am sympathetic with this proposal.
However, I agree with Duncan that it would require an
enormous amount of work (I can say this having now
attempted to do this work for two countries). And I have
not tried to estimate taxes on wages or the "social wage",
as do Shaikh and Tonak. Nor have I inverted matrices to
convert money estimates into labor-time estimates, again
as Shaikh and Tonak do (although these labor-time
estimates are less important to them than to others).
And there are so few of us and so few resources. And
there is so much other work to be done. Therefore, I am
not sure that this project (dear as it is to my heart)
should become such a priority that it "crowds out" other
important and necessary tasks. Who is going to do it?
How will it be organized?

2. I certainly agree that, when we do Marxian empirical
research, we should try to make our estimates as
compatible as possible with estimates in other countries
in order to make international comparisons possible. And
that we should present alternate estimates based on
different assumption as much as possible to clarify the
effects of different assumptions on the results. I have
tried to do these comparisons extensively in my own work.

However, there are limits (at least for me). I myself don't
want to try to estimate the "social wage" and I don't want
to invert matrices to derive labor-time estimates (nor do
I want to try to separate the income of the self-employed
into variable capital and surplus-value), both because I
think these are conceptually incorrect interpretations of
Marx's theory and also because they require a whole lot
of work. Someone else who thinks that these are
conceptually correct interpretations can of course do the
work to derive these estimates, but not me.

3. Alan, what practically do you suggest as the next step?
I think that Duncan's idea of a web site with access to
different data sets is a very good one. But someone
would have to take responsibility for this. Any

4. Alan mentioned that one of the main remaining
disagreements among us is the estimation of the stock of
constant capital and hence of the rate of profit. Alan,
what disagreement(s) are you referring to? Whether
constant capital should be estimated in terms of current
or historical costs? Although there has of course been
much recent theoretical controversy over this issue (on
OPEL and elsewhere), all the estimates of constant capital
that I have seen are in terms of current costs (including
Shaikh and Tonak, and Dunemil and Levy, and even
Alan's own estimates for the UK economy, as I
understand them). Alan, are you suggesting that we
estimate constant capital in historical costs? Are your
estimates for the UK economy in historical costs? Or do
you have other disagreements in mind, that do not come
to my mind at the moment?

5. Jurriaan said:
"It isn't possible to measure Marx's pure concept of
productive labor."

I agree that it is difficult to distinguish between
productive labor and unproductive labor, especially with
a high degree of precision. But I think that it can be done
in a rough way which yields interesting results, especially
with respect to trends in the distribution of income,
(as I have argued recently on OPEL), and that these results
would probably not be altered significantly with more precise estimates.

6. Duncan said:
"For many purposes, can't existing official data sets
[e.g. which do not distinguish between productive
labor and unproductive labor? FM]], appropriately
interpreted, be used to approach the questions
Marxists are interested in?"

For the purpose of estimating trends in the distribution of
income from a Marxian perspective, I would say that one
has to distinguish between productive and unproductive
labor. Perhaps for other purposes, the official estimates
will do. Duncan, what other specific questions do you
have in mind?

7. Jurriaan also said:
"Shaikh and Tonak don't really say how to allocate
the wages of unproductive workers in the Marxian
account. Clearly though, they are a component of the
circulating constant capital.

This is a minority view. The only authors I have read
who have this interpretation are Shane Mage and OPEL
listmember Murray Smith (Murray, are you still there?).
But it seems clear to me at least that, for Marx,
unproductive costs are paid out of the surplus-value
produced by productive labor. Constant capital is the
capital invested in the means of PRODUCTION (i.e.
material inputs to the process of PRODUCTION). Constant
capital does not include any capital invested in the
functions of circulation, including the wages of
unproductive labor. This is also clearly Shaikh and
Tonak's interpretation (pp. 45-51).

I look forward to further discussion.