> * Why can't the empirical research proceed as follows --
>
> -- develop *two* (or more) sets of defining criteria which
> can be used with the same data?
> If for no
> other reason, it might be useful to develop a "range" and then
> discuss how important (or unimportant) the different definitions
> are to the empirical "results".
No you are not "off-base" Jerry, this is a good idea, and in fact a number
of authors, including Shaikh & Tonak, follow this approach. In the New
Zealand case, we also tested for the effects of different definitions.
Similarly Alan Freeman has suggested that we could "apply whatever
criterion you find convivial to decide which part of [labour] is
productive, and which is unproductive". The question at stake is really
what is the best measure of the new value added, and which activities
contribute new value instead of preserving or transferring it. The
argument is that the distinction between new value (new income) and
preserved value in official acounts is rather capricious, in which case we
need a more rigorous concept. But it can often be shown that a number of
different measures all follow the same sort of pattern over time, the only
real difference being the exact level of the quantities involved. I do
agree with you that the productive/unproductive distinction is important,
since it is argued that up to half of the gainfully employed workforce
engages in capitalistically unproductive labour these days. That's a lot
and it does make a significant difference to the figures.
Regards
Jurriaan Bendien