productive/unproductive again

Allin Cottrell (
Mon, 2 Feb 1998 22:14:21 -0500 (EST)

I'd like to follow up, with a brief argument and some questions.

Alan has recently attacked the notion of "real" magnitudes, as
in the standard GDP accounts. He points out that the so-called
"real" magnitudes are in fact monetary magnitudes (sums of
prices), but using the prices of a chosen base year rather than
of the current year. That's true -- as any orthodox economist
would recognize -- but it skirts the point. The point is that
if one _compares_ the "real" GDP, so defined, at any temporally
separated observations, one is abstracting from price changes
between the periods and is therefore dealing with a
quasi-physical quantitative difference. The "quasi" is needed:
one can talk of a _definite_ (scalar) physical-quantitative
difference between the outputs of different periods only if the
relative proportions of physical outputs of different
commodities have remained unchanged, -- if, in other words,
there has simply occurred a uniform scaling of a given vector of
physical outputs -- in which case the use of base-year prices as
a tool of "real" measurement is redundant. But we never see
such uniform scaling, and the use of base-year prices is the
best the economics profession has come up with for assessing
quasi-physical growth. It seems to me this is not useless.

Furthermore I suspect that if one rejects this notion, one can
make little _practical_ sense of Marx's distinction between
productive and unproductive wage labour.

Various contributors have proposed thought experiments involving
competing open economies and variation in the terms of trade.
That is very complex, and I'd rather stick with a conceptually
simpler thought experiment: parallel closed economies, identical
in all respects other than those specified. So, let
(capitalist) economy A devote a large proportion of social
labour to advertising, while B devotes minimal labour to this
use, instead allocating the labour to straightforward current
production of use-values, or research and development pertaining
to new technologies and products.

I want to claim, first of all, that Marx would see a significant
difference here -- the proportion of productive labour is higher
in B than in A. I also have the gut feeling that this is right
and important; that there really will be a significant
difference between A and B.

I want to ask: From what perspective(s) is this putative
difference "visible"? I suspect that from a purely "monetary"
perspective no difference will be visible, and likewise from a
pure "labour-time" perspective. [The only way one can
distinguish the (really subordinated, waged) labour-time devoted
to advertising rather than to production is by helping oneself
to the distinction between productive and unproductive wage
labour, but with what justification, if not that some
labour-time actually produces "real" stuff, and some does not?]
Second: If the difference between A and B is a significant one,
how exactly will it manifest itself?

Allin Cottrell
Department of Economics
Wake Forest University, NC