Re: [OPE-L] workers' consumption and capitalists' consumption

From: Ian Wright (wrighti@ACM.ORG)
Date: Sun Jun 11 2006 - 18:37:34 EDT

Hi Ajit

You have very rapidly focussed your attention on some of the key points.

> For Sraffa, this makes no difference
> to the labor-values of corn. But for you this
> increases the labor-value of corn from 500 to 1000.
> How does this happen? You argue (implicitly or
> explicitlt) that capitalists consumption of the 4 tons
> of corn should be put back as input in the system.
> Thus your method of production changes to:
> 2 t. of corn as raw mat.+4t. of corn as cap. con.
> +4000 hrs of labor--> 10 t. of corn
> This obviously makes the value of corn equal to 1000
> hrs of labor per unit of corn. Now the absurdity of
> this change in the method of production should be
> obvious. But I'll not belabor on that.

The method of production:

(A) 2 t. of corn as raw mat. + 4 t. of corn as cap. con. + 4000 hrs of
labor --> 10 t. of corn

which seems absurd to you is the correct description of self-replacing
equilibrium -- because the real-costs balance on each side of the

Of the 10 tons of output, 2 tons replaces the corn used-up as raw
material, 4 tons replaces the corn consumed by capitalists, and 4 tons
is the real wage that replaces the direct labour expended.

Contrast with the representation that you think is not absurd:

(B) 2 t. of corn as raw mat. + 4000 hrs of labor --> 10 t. of corn

Of the 10 tons of output, 2 tons replaces the corn used-up as raw
material, 4 tons is the real wage that replaces the direct labour
expended, leaving 4 tons undistributed or unaccounted for ...

Where did the 4 tons of capitalist consumption go? It is on the RHS
because it is an income. But it is missing on the LHS -- even though
capitalists consumed 4 tons of corn during the period of production.
The real-cost accounts do not balance.

The two representations (A) and (B) constitute a very simple example
of the difference between working with a distributed rather than an
undistributed surplus. In (A) the surplus is fully distributed, in (B)
it is not. Capitalists continually and repeatedly consume corn in
simple reproduction. It therefore must appear both on the LHS, as a
cost, and on the RHS, as an income. It is not just an income.

Consider this: the adoption of (B) implies that capitalists never
consume their 4 tons of corn, because labour is never expended to
replace it, despite the fact that capitalists bought their 4 tons of
corn, due to the balanced price accounts in Sraffa's price equation.
If you grasp this point then you will be free of the Sraffian
real-cost accounting error.

I have no gripe with Sraffa's price equation because it does balance
and it is conservative in nominal costs. But the "half an equilibrium
system" that Joan Robinson alluded to in her review of PCMC (1961) is
Sraffa's asymmetrical treatment of the distribution of the surplus:
he's happy to slide up and down on the nominal wage-profit trade-off
while ignoring what that means for the real wage and capitalist
consumption. That's ok -- until he turns to labour-cost accounting
(and real-cost accounting in general, which he drops after Ch. 1).

> I'll just underline a few more obvious points: (1) when you
> strip the system of all unnecessary complications,
> then you can see the whole idea of 'money capital'
> which is produced in capitalist household and price of
> money capital, etc. melts away. The firms have corn
> and corn is what they need as raw material as well as
> for wages and the devidends to the capitalists are
> also paid in corn. Nobody in the system needs anything
> else than corn. Thus my point that the firms always
> have their needed commodity capital should be clear
> now.

Yes you can strip away the explicit representation of money-capital
because we can map from the circular flow, which explicitly represents
money-capital, to Sraffa's surplus equations, which do not. I swap
between the two throughout the numerical example.

But without the concept of money-capital and the mapping to the
circular flow then we miss the principle of real-cost. The Sraffian
labour-cost accounting error is not caught. The circular flow
representation yields the j-inverse measure of real-cost, and it helps
us to understand why the Leontief inverse, which is employed in
Sraffa's formula for labour-values, fails to vertically integrate over
the whole real-cost structure.

> (2)Your procedure amounts to throwing back
> physical capitalists consumption on the side of the
> inputs in the methods of production.

Basically -- yes. Although a better description would be "balancing
the real-cost books".

Method of production (A) augments the (1-dimensional) technique with
capitalist consumption. But this does not imply capitalists input corn
to production, just as the technique augmented by workers consumption
does not imply workers input corn to production. Capitalists input
money-capital to production, and workers input labour. The method of
augmenting the technique "short-cuts" these inputs. In linear
production theory there are often multiple equivalent representations
of the same economic state-of-affairs. The circular flow
representation is another example of this.

> (3)In a multiple
> commodity model you cannot properly allocate
> capitalists consumptions as inputs for various sectors
> unless you solve for rate of profits from Sraffa's
> surplus equations.

I agreed with this point in an earlier message. In the numerical
example, I do solve Sraffa's surplus equation in order to construct
the circular flow representation, just as you say (bottom of p. 77).

But remember that the circular flow still has Sraffa's surplus
embedded within it -- except now it is also physically distributed,
not just nominally distributed. The circular flow price solution is
identical to Sraffa's surplus equation price solution.

> Thus your attempt to simultaneously
> determine all prices and the rate of profits along
> with all capitalists consumption treated as inputs in
> the system is not possible--the system lacks one
> equation.

The system lacks a numeraire equation. So the circular flow matrix has
a degree of freedom (if we construct it from Sraffa's starting point).
You can see that in the matrix on p.78. In the example, both the
capitalist consumption coefficients and the money-capital coefficients
are dependent on the price of corn. I discussed this in an earlier
message. However, it does not have the import that you think it has.
We would not be discussing a numerical example If the system really
lacked an equation and was not possible to solve.

> The whole idea of a separate money capital
> sector, which produces a commodity called money
> capital--a commodity which does not appear in the
> input output structure but somehow gives you an
> independent equation is not making any common sense.

The money-capital commodity does appear in the circular flow
input-output structure. That's the output of the capitalist household
sector, the last row of the circular flow matrix. We get rid of it
when we map back to Sraffa's surplus representation, or construct the
technique augmented by capitalist consumption, such as in
representation (A) above.

Best wishes,

This archive was generated by hypermail 2.1.5 : Fri Jun 30 2006 - 00:00:03 EDT