[OPE-L:7116] [OPE-L:617] Re: Re: New Evidence on Sectoral Prices and Values

Paul Cockshott (clyder@gn.apc.org)
Sun, 7 Mar 1999 20:46:05 GMT

At 23:31 03/03/99 -0500, ope-l@galaxy.csuchico.edu wrote:
>A reply to OPE-L 576.
>Alejandro asked: "Is not possible to use another variable to
>"deflate" the aggregates other than costs?"
>Yes, but costs are the RIGHT choice. Think of it this way. Let P
>be aggregate price, V aggregate value, C aggregate costs, and X
>aggregate output. Note that
> Pj Cj
>Pj = ---*---*Xj
> Cj Xj
> Vj Cj
>Vj = ---*---*Xj
> Cj Xj
>The first terms on the RHS were my regression variables. The middle
>terms are per-unit costs.

I dont understand the point of this procedure, surely all you are
introducing here are sucessive multiplications of values and prices
by unity, since you introduce two new variables C and X
on both the numerator and the denominator.
>You would need to get rid of the last term, Xj (if you knew it),

Why why do you need to get rid of the term Xj as it is already
canceled out in the above equation?

>because it causes spurious correlation. Large sectors, large
>prices, large values; small sectors, small prices, small values.

What is this term x, where have you derived it from and how
is it supposed to cause spurious correlation. I find this
explanation of your idea totally confusing.

Your basic idea
seems to be that if an industry is 'large' both its value and
its price will also be large, and that there is some abstract
largeness which is then multiplied by a value term and a price
term to arrive at the data in the input output tables etc.

But what could this 'largeness' be, and what could the value
and price terms that multiply it be.

One would have to have for each industry a value term that
was in effect
'hours per unit of bigness'
and for each industry a price term that was
'dollars per unit of bigness'

But this is a nonsense, since values and prices are not
defined in terms of abstract units of bigness but in terms
of use values. And such use values, being of different types,
can not be comensurared to form a scale of bigness.

No such X that is dimensionally consistent can exist.

How is industry scale normally measured?

There is no denying that the idea of industries having
different scales makes some sense. If one were asked, outside
of the context of this debate, to rank industries by scale
how would one do it?

Most economists, I suspect, would rank the industries by
the monetary value of their aggregate output. Thus they
would view 'paper and allied products' with a gross output
of $80 billion as smaller than 'petroleaum and natural gas'
with a gross output of some $550 billion.

Some economists might alternatively rank industries by
their employed labour force.

Marxist economists might chose to rank industries by the
value of their output.

The idea that industry costs are the appropriate way to
measure an industry's scale would, I suspect, have little
support. It suffers from the disadvantage of underestimating
the scale of industries with high levels of profit.

Thus if one used costs as a basis, a reduction in wage levels
in an industry, with no change in the number of workers
employed nor any change in the value of its output, would
result in it being ranked as smaller than it previously was -
surely a peverse result.

But if one takes the obvious measure of industry scales, their
monetary output, the whole idea of deflating prices and values
to compensate for spurious scale effects is revealed as absurd.
One would then be attempting to correlate V_i/P_i with
P_i/P_i. Since the latter term is a constant no correlation
can be performed.

The notion of spurious correlation only makes sense if there
is some independent causal factor, X which produces variations
in values and in prices. But scale is not such a factor X, it
is merely a re-labeling of the total price of an industy's

In ranking industries by price of output to measure their scale
one is operating at the sensible or exoteric level. The underlying
cause of an industry's scale however, is determined by the law
of allocation of socially necessary labour time - in other words
the law of value.

Reductions in the amount of labour socially required to produce
corn, have, over the last 150 years, drastically reduced the
relative scale of the arable farming industry in developed
countries. At the same time, the absolute amount of grain
produced has risen.

The law of value predicts that as a consequence of this decline
in socially necessary labour time, both the relative scale
of the industry in employment and its share of monetary national
income will also decline. This has in fact occured. But if we
were to take Kliman's argument seriously, both the decline in
the relative monetary value of arable farming and in its labour
use stem from some independent scale factor which caused both
to fall. His argument says in effect

"No wonder the relative price of arable farming and its employment
have declined, this has happened because the scale of the
industry has contracted."

But this is to confuse causes with effects. It is the labour
necessary to meet society's different material wants that determines
both the scale of the industries devoted to these wants and,
through the law of value, the price of their outputs.

Paul Cockshott (clyder@gn.apc.org)