Re: OPE-L:_Wage_share

From: Paul Cockshott (wpc@DCS.GLA.AC.UK)
Date: Thu Sep 16 2004 - 04:47:03 EDT

If memory serves the 50:50 split in F&M is deduced as a consequence
of Lukacs' Theorem, in the (arguably very special?) case that the
rate of surplus value has a *degenerate* distribution, i.e. has a
uniform value.

Since the theorem concerns a property unique to the gamma (and which
is indeed used as a test for "gamma-ness" of data -- I can dig out
ref.s if needed) I don't readily see how it might apply if the rate
of profit is not a gamma.
Obviously if s/v has a degenerate distribution =1, then the global
rate is necessarily =1.

But if not, offhand I can't think of any necessary relationship
between the global rate and the distribution: no reason, for example,
why all firms bar one could not have s/v=0, and the remaining firm[i]
having s/v = sum(s)/v[i].

Paul C
Farjoun and Machovers proposition that 
a) the distribution of the profit share will be Gamma
b) that the narrow the dispersion the closer the wage share will be to
Is in principle testable. David has come up with some figures
which seem to support the proposition, what is missing is
a test of the extent to which the distribution is actually
a Gamma one. However the interesting thing about F&M's argument
is that it points one to look at something one would not
otherwise have looked at - the dispersion of the profit share
as a determinant of the global average rate of surplus value.

As to all firms but 1 having zero s/v and the other having
all the s/v, well there is no 'necessary' reason why this
will not happen. There is also no necessary reason why my teacup
should not experience significant Brownian motion, it is
just the laws of chance are against it in both cases.

This archive was generated by hypermail 2.1.5 : Fri Sep 17 2004 - 00:00:03 EDT