[OPE-L:3046] Marxian empirical research revisited

Andrew Trig (A.B.Trigg@open.ac.uk)
Wed, 18 Sep 1996 08:32:14 -0700 (PDT)

[ show plain text ]

From: Andrew Trigg
In reply to: Jerry's OPE-L:3013

Thanks very much for your welcoming show of interest in my work on
Marxian emprical issues. You asked four questions:

1)How do I use micro data? Some of you will be familiar with this work,
but the basic idea is that a micro data set such as the UK Family
Expenditure Survey contains information whichi is potentially useful
for Marxian purposes. In particular the hours of work information for
each worker in a micro sample can be used to estimate quantity of hours;
and the individual wage data, together with other individual characteristics
such as education, age, sex, occupation etc. can be used to estimate a
microeconometric wage equation. The proposal is that this can be used
to estimate the quality of labour time - at least it might improve on
previous studies which assume that all the wage instead of part of it
is correlated with labour quality.

2) The approach may improve slightly on Cockshott, Shaikh etc. in so
far as labour quality is treated in a more sophisticated manner. N.B.
Other people such as Greg Bongen at the New School are working on a
different way in to this.

3) How do I evaluate the recent debate in Capital and Class on the
Cockshott et al paper? I think it makes a very open and honest
contribution to the empirical literature. One confusion, which I
am trying to grapple with is how this debate relates to the input-output
approach. Cockshott et al argue that since labour values and prices
are closely correlated then it follows that price data can be used to
for Marxian empirical work. Shaikh and Tonak, on the other hand, argue
that the best way to proceed is to interpolate between input-ouput and
money national accounts. The critique of Cockshott et al in the Summer
issue of Capital and Class (by the person from Greece who's name
escapes me) argues in favour of Shaikh and Tonak but does not address
this issue - unless I've missed something. This leaves me not knowing
how important it is whether prices and values should be closely

4) On the relative merits of input-output and national accounts, this
is still to be resolved in my own thinking, depending on whether labour
values (or measures of labour embodied if you prefer) are still close
to prices once labour quality is dealt with. I am also not clear whether
the method of interpolation from input-output to a time series of
national accounts in Shaikh and Tonak makes too many assumptions to be

Thanks again Jerry for your kind welcome.

In Solidarity.