These are reasonable points Jurriaan, and some of us who are active in this 
area are working on a systematic
answer to these and similar objections. Let me summarise a couple of the themes 
we will be using in the
answer.

1. The use of wages as proxies for labour inputs.
This is a valid objection because the rate of exploitation may differ between 
industries.
However, the point about the Swedish data, is that the raw information in the 
I/O tables is
in the form of person years expended in each industrial sector ( presumably 
obtained from
the Swedish Employment statistics). The Swedish data are thus not subject to 
the objection
that you raise, but they show a similarly strong correlation.

2. You write that we have to be careful that the results are not the result of 
just the law of
averages and the way the statistics are collected. Well if that were the case, 
then one
could take any input into production : coal, forestry, steel etc and get 
similarly strong correlations.
In fact you do not get similarly strong correlations using other value bases. 
This indicates
that the effect must be due to some special role that labour plays in the input 
output tables: basically
that it is the predominant and most universal input to production.

The laws of averages are certainly important, but one could argue that the idea 
of labour value is
itself founded on the concept of averaging. Marx has a discussion of the size 
of work team that had
to be employed in an agricultural project for the differences in strength and 
productivity between
individual worker to even out. As I understand it he defines the labour value 
in terms of the
average amount of person hours required to do something. It is therefore 
something that asserts
itself more and more strongly the larger the group of workers that are involved.
________________________________________
From: [email protected] [[email protected]] On 
Behalf Of Jurriaan Bendien [[email protected]]
Sent: Sunday, March 06, 2011 12:11 PM
To: Progressive Economics
Subject: Re: [Pen-l] Marginalism wrong or not even wrong

In a 2005 paper, Paul Cockshott colloquially explains the input/output
technique for obtaining labour-values:

"If we divide the directly utilised labour by the dollar value of the
industry's output, we get an initial figure for the amount of [direct]
labour in each dollar of the output. For industry A we see that 0.32 units
of labour go directly into each dollar of output. Since we already know the
number of dollars worth of A's output used by every other industry, we can
use this to work out the amount of indirect labour used in each industry
when it spends a dollar on the output of industry A. This gives a second
estimate for the labour used in each industry, which in turn gives us a
better estimate for the number of units of labour per dollar output of all
industries. We can repeat this process many times and as we do so, our
estimates will converge on the true value."
www.dcs.gla.ac.uk/~wpc/reports/rethinking.pdf

As I noted however in 2008
http://ricardo.ecn.wfu.edu/~cottrell/ope/archive/0807/0135.html one problem
of this iteration procedure is that it relies on the methodological
assumption of a fixed ratio between labour time worked, paid labour time,
and the value of gross output produced.

It is assumed, that the magnitude of the indirect labour contained in each
part of the output sold and transferred as an input by each sector {A} to
other sectors {B,C,D...} will be accurately determined by applying the same
labour-output ratio established for sector A's total gross output.

Most likely this assumption is arbitrary (think of joint production, and
qualitatively different outputs transferred by one sector to other,
different sectors) and it introduces a margin of error, but this error is
not corrected by additional iterations, nor can we establish what the
magnitude of error is.

The aim of the whole exercise is to demonstrate a strong correlation between
labour-inputs and output values, but in reality labour-inputs are derived
from output and input magnitudes which are themselves estimated using
numerous statistical assumptions (including the law of averages, categorical
assumptions, valuation adjustments, and imputations for missing data).

Paul Cockshott doesn't deny the methodological problem and the problem of
data accuracy, but he claims "what is interesting is that despite all these
difficulties, the actual correlations between sectoral prices and values
remains so strong."
http://ricardo.ecn.wfu.edu/~cottrell/ope/archive/0807/0139.html  "The bottom
line Jurrian, is that despite all of these possible sources
of error in the data we work with the results are still very good."
http://ricardo.ecn.wfu.edu/~cottrell/ope/archive/0807/0171.html

This is scientifically not really satisfactory however (some would say it's
crap, or propaganda), because what we require specifically is a clear proof
that the strong correlation obtained is not simply attributable to the
chosen methodology itself (an artifact of research design and data
constructs), and that the strong correlation obtained is superior to any
alternative positive or negative correlations which might also be obtained.

For such a proof, it would be useful that all the data assumptions and
methodological assumptions implied in the calculation procedure are spelled
out, and their likely margin of error is estimated, but to my knowledge this
has never been done, since the data sets are simply accepted as given. The
"science" conveniently stops at the point where a result is obtained which
appears to clinch the case being made.

Jurriaan



_______________________________________________
pen-l mailing list
[email protected]
https://lists.csuchico.edu/mailman/listinfo/pen-l

The University of Glasgow, charity number SC004401
_______________________________________________
pen-l mailing list
[email protected]
https://lists.csuchico.edu/mailman/listinfo/pen-l

Reply via email to