>under Marx's law of value, what's wrong with the following?
>
>"Can't machines produce surplus-Value? After all, capitalists hire or
>rent machine-power (the use of a machinery during a specific period)
>and can then use it to create machine services in production, more
>than paying for the cost of hiring the machine-power. This creates
>surplus-Value, right?"


There are layers here:

1.  With respect to the circuit of capital M--C--M', Marx defines "surplus 
value" as the positive difference between M' and M, subject to the 
conditions that (a) this difference corresponds to the production of new 
value financed by the initial outlay of M, rather than merely a 
redistribution of existing value, and (b) this difference is appropriated 
by someone other than the direct producer of that value--that is, by the 
capitalist(s) who advanced M.

2.  According to Marx's law of value, this new value must *by definition* 
correspond to the expenditure of productive labor, and furthermore labor 
embodied in the newly produced commodities must exceed the labor embodied 
in the wage goods paid to the workers.  But then it can't be that machines 
produce surplus value, since  "machine services" aren't labor, or to put it 
another way, the value that machines contribute to production is already 
embodied in the value that capitalists purchase with the initial outlay of 
M. This connection between surplus value and surplus labor, *within* a 
framework that defines commodity values in terms of labor, is reflected in 
the "Fundamental Marxian Theorem" (FMT), which asserts that when commodity 
prices correspond to their respective "prices of production" (prices 
determined by average production costs plus a markup determined by the 
economy-wide rate of profit), the rate of profit is positive if and only if 
there is surplus labor, understood in the above sense.  In other words, the 
rate of profit is positive iff the rate of labor exploitation is positive.

3.  But there seems to be a catch.  Suppose that instead of defining 
commodity values in units of homogeneous labor time, we define them in 
terms of one of the commodities--say, steel, assuming machines were made 
out of steel.  Then one can establish an alternative to the FMT asserting 
that the rate of profit is positive if and only if the steel value of 
steel--that is, the difference between the steel embodied in final output 
and the steel embodied in the payment to sellers of steel--is 
positive.   In other words, the rate of profit is positive iff the rate of 
steel exploitation is positive.  One can establish a similar result using 
any other commodity as the basis for defining and measuring commodity 
values.  This result has been established by several different authors; 
John Roemer labels his version the "generalized commodity exploitation 
theorem" (GCET).

What to make of this result?  Two conditions need to be met for something 
like the FMT or the GCET to obtain in a given market economy:  the system 
must be capable of producing a surplus product, net of the payments to all 
the suppliers of tangible inputs (i.e., of labor and intermediate goods 
expended in production), and capitalists must somehow appropriate this 
surplus in the form of profit. Note that these two conditions correspond to 
conditions (a) and (b) in Marx's definition of surplus value summarized in 
point 1.  However, neither the FMT nor its generalization (understood in an 
accounting rather than an explanatory sense) in the GCET provides any 
insight into how these two conditions, and particularly the second one, 
come about.  One might ask whether the needed explanation distinguishes 
labor power and its use value, labor, in some *qualitative* way from other 
potential inputs (say, "machine-power") and their respective use-values 
(say, "machine services")

4.  It does.  The mathematical theorems used to establish the FMT or GCET 
are based on assumed properties of given input-output matrices (and 
corresponding wage vectors) for given multi-commodity economies, but don't 
investigate what systemic conditions give rise to these conditions.  Such 
an investigation would reveal this qualitative distinction between labor 
power and, say, machines:  to purchase a machine is pretty much equivalent 
to securing its productive services, so long as one knows how to run (and 
maintain, etc.) the machine.  There's no prospect that the machine might 
say "no, I'm not going to provide these services to you."  In contrast, to 
purchase labor power is not tantamount to securing labor.  [So, for 
example, the "direct labor" coefficients in the above-named input-output 
matrices *presume* some translation of labor power into effort per labor 
hour, labor hours per worker, and workers per production process, and thus 
treat as exogenous what is, at the level we're now discussing, necessarily 
endogenous.]  Marx addresses the question of how capital has translated 
labor power into labor via the circuit of capital in what I've called his 
"historical" (as opposed to "value-theoretic") account of surplus value, 
which includes his analysis of the "subsumption of labor under capital."

5.  But you don't need a labor *theory of value* to investigate how labor 
power is translated into labor (and into surplus value and profit) in 
capitalist economies, and it's a judgment call whether such a theory is 
even particularly useful in telling the required story. I tend to think it 
isn't, for a number of reasons, but don't think that point is worth (re-) 
arguing over here. 

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

Reply via email to