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

From: Ian Wright (wrighti@ACM.ORG)
Date: Fri Jun 23 2006 - 02:21:25 EDT

Hi Ajit

I think if we can agree on given assumptions then we will make progress.

> The 2t. of corn as raw material is a historical
> datum for Sraffa, it will not change whatever you do
> with the 10t; of the output. However, if I give less
> or more to the workers or the capitalists from the
> output, in your case that must show up in the input
> otherwise your accounting system breaks down.

Apply this same reasoning to technical change that occurs in the
production period. Of the 10t of output, say that only 1t now need be
allocated to raw materials, due to corn-saving technical change.
Sraffian labour-values are a function of the technical matrix. Hence,
according to your argument we must conclude that in this case Sraffa's
labour-value accounting breaks down (and his price accounting to

My point is this: the Sraffian system in particular, and linear
production theory in general, is not a dynamic theory. It abstracts
from change that occurs in a production period.

For example, the TSS school has complained that Sraffa's theoretical
framework breaks down when we introduce technical change in the
production period. Your criticism of real-cost accounting is a
mirror-image: you state that real-cost accounting breaks down when we
introduce a change in the distribution of real income in the
production period. Both critiques are external, in the sense that they
reject the assumptions of the model class. Sraffa abstracts from
technical change and changes in the distribution of real income (he
considers only nominal changes, and then, arguably, only in a
counterfactual manner). Real-cost accounting also makes these
assumptions because it derives from an immanent critique of Sraffa.

However, I sympathise with both these external critiques: any
realistic model of an economy must handle dynamic change. In contrast,
Sraffa, and linear production theory in general, can only deliver
comparative statics. For example, real-cost accounting allows
comparison of two configurations of an economy with different real
income distributions and different techniques.

Implicit in your critique, however, is the claim that Sraffa's labour
values, because they are independent of the real distribution of
income, do not break down when we consider changes in the distribution
of real income in the production period. But as I mentioned -- and you
have yet to respond to this -- the claim begs the question.

The real-cost crtique of Sraffa's labour values denies they are the
correct measure of replacement costs even if we assume no changes in
the distribution of real income in the production period. The validity
of Sraffian labour-values in the special case of self-replacing
equilibrium is precisely what is now in question. Once it is shown
that Sraffian labour-cost accounting does not hold in this special
case, and under-estimates replacement costs, then what rational
justification is there for claiming general validity?

In summary, the terms of this debate, stated at the beginning of the
thread, is that we are considering a state of self-replacing
equilibrium in which the surplus is distributed in real terms,
abstracting from technical change and changes in the real distribution
of income in the production period. This is a natural set of
assumptions when reasoning about equilibrium linear production models.

> What I have been trying to tell you is
> that you can either have a 'closed' model, which is
> represented by subsistence economy, or have a
> 'surplus' model but not both. You are trying to have
> both, which entails logical contradiction.

You think there must be a logical contradiction because you fail to
distinguish between an undistributed surplus and a distributed
surplus. More on this below.

> It is you who had introduced the time-scripts to
> counter my point that outputs cannot be inputs to
> itself.

I assure you that you first introduced time-scripts. That's because,
in my view, you began to shift the terms of the debate from
self-replacing equilibrium to a non-equilibrium situation.

> In your above statement you have again
> introduced time-script.

To illustrate your use of time scripts.

> Your 2 tons of corn as raw material is a historical datum--logically, i.e. there
> is no logical posibility of changing it after you have got your 10 tons of output out of it.
> Can you say the same for your 4 tons of corn as capitalist consumption?


> If you can, then it is also a historical
> datum and you cannot logically change it after getting
> your 10 tons of corn.

Yes. (Although I hesitate to use the term "historical datum").

> Now you cannot say that
> logically I cannot give more to the capitalists from
> the 10 tons of corn, can you?

No I cannot.

> Since you cannot, then I am free to give 6 tons of corn to the capitalists and
> 2 tons to the workers out of the 10 tons.

Yes you are.

> This cannot be logically denied. But when I do something which is
> logically possible, your principle of accounting breaks down.

Yes; but only in the sense that all linear production models break
down if we introduce certain kinds of changes within the production
period. The correct thing to do, given the absence of a dynamic
theory, is to formulate a new closed model that reflects the new
distribution of real income and perform a comparative static analysis;
there are some examples in the paper.

Allow me to repeat an earlier point in a slightly different form: This
external critique that you have introduced also applies to Sraffa's
labour-cost accounting. For example, if you are free to change the
distribution of real income then I hope you will grant me the freedom
to change the technique and introduce technical change in the
production period. In which case, Sraffa's labour-cost accounting
"breaks down". Do you accept the legitimacy of this external critique
and do you therefore agree that Sraffa's labour-cost accounting must
be rejected?

But if this kind of external critique is granted then we have nothing
to debate, because we have not agreed to share the same set of

For example, although I am free to introduce technical change, I don't
choose to, because I have accepted the terms of Sraffa's theoretical
system. I am more interested in an immanent critique and precise
reasoning about well-defined theoretical objects. The merit of this
approach is that reasonable people can iterate towards agreement on
the properties of such theoretical objects, given that they adopt the
same starting points, assumptions and rules of mathematical logic. I
invite you to join me in thinking about a state of self-replacing
equilibrium in which there are no changes in
- technique, and
- distribution of real income
in the production period. We will then be able to resolve a relatively
straightforward matter: whether Sraffa's labour values correctly
measure replacement costs in this case. Your talk of "logical errors"
and accounting systems "breaking down" is a kind of displacement
activity that gets in the way of really thinking about this important

> How do you get 4000 hours of labor when your both
> moments are identical?

By "moment" I meant "production period", which I think was clear from
the context. The 4000 hours are applied in the production period.

> The capitalist consumption is not part
> of the 'methods of production' neither in the standard
> or non-standard interpretation of methods of
> production, except yours.

My reference to "standard interpretation" was not that capitalist
consumption is part of Sraffa's "method of production". I know that it
is not. This is why Sraffa's labour-cost accounting is
non-conservative. My reference to "standard interpretation" is that it
is standard to consider that commodities are used-up and replaced
within a production period.

> Of course you cannot replace within the same
> production period, unless you do away with the notion
> of production period altogether and say that
> production is instantaneous (even then I think it will
> be illogical to replace input in the same
> period--actualy I don't know if word 'same period'
> will have any meaning). In that case, what meaning can
> one attach to your 4000 hrs. of labor and that 1 ton
> of corn having 1000 hrs. of labor-value.

Means of production are used-up and replaced within the production
period. This is a standard interpretation of a "method of production"
in the single production case. I can provide quotations, but it is an
elementary point. The 4t of corn of capitalist consumption on the LHS
is intepreted in this standard manner: it is consumed and replaced
within the production period.

> And so per unit of value of corn itself is 1000 hrs.
> of labor. But then you don't tell us what this 4t. of
> cc does in the production process?

The 4t is consumed by capitalists in the period of production. It is
the real-cost of the money-capital supplied. The key point, missing in
Sraffa's surplus representation of this economy, is that under
capitalist conditions commodities are also produced by means of the
commodity money-capital. Speaking plainly: Sraffa's PCMC excludes from
consideration the very commodity that distinguishes simple commodity
production from capitalist production. Sraffa's objectivism does not
extend to counting the money-capital inputs in a capitalist economy.
Hence, his labour-cost accounting under-estimates replacement costs.
Some labour is missing.

To avoid any confusion, but at the risk of repetition, the simple
example of a corn economy is represented by an equation in which the
technique is augmented by capitalist consumption, so the explicit
representation of money-capital is absent. However, it is equivalent
to a circular flow representation, in which the money-capital is
present, as the full numerical example in my paper demonstrates.

> If it is needed in
> the production process as the 2t. of raw materials
> then in economics for centuries we have been calling
> it necessity of production. Then your system does not
> have any surplus and it is equivalent to subsistence
> economy.

An economy with a distributed surplus is not a subsistence economy.
Leontief, prior to the 1951 second edition of his opus, employed a
closed model to study the American economy. I don't think he described
the US economy as a "subsistence" economy. Sraffa did however
interpret a set of homogenous equations (a closed model) as
representing a subsistence economy, rather than a surplus economy in
which the surplus is fully distributed. But that is a choice of
interpretation, not a necessary property of the equations.

Consider this: the surplus is produced. What happens to it? It gets
distributed. To be evocative, workers and capitalists don't forever
argue over its distribution. We can represent the distribution of the
surplus in terms of a closed model. According to your terminology we
can then no longer speak of this economy producing a surplus, even
though the very same quantity of net product is produced. Instead of
conflating an open model with undistributed surplus and a closed model
with a distributed surplus you should distringuish the different ways
we can think about a "surplus"; for example:

> A necessity by definition is not a surplus and a surplus by definition not a necessity.

In my view, this sentence neatly encapsulates a root metaphorical
problem in Sraffa's system: the Sraffian surplus is forever
undistributed, frozen just before the moment of its physical
distribution: it can never become a necessity.

But consider that today's surplus becomes tomorrow's necessity.
Incomes are never only incomes -- they also become costs (cf. Marx's
"moral and historical" element that enters the determination of the
real wage, or Rakesh's quote of Grossman's views on capitalist
consumption). Consider also, that your sentence is purely analytical.
I don't think it churlish therefore to mention the "magical" term
dialectics here as a contrast.

I would wholeheartedly agree with your sentence if you replaced
"surplus" with "undistributed surplus" or "new surplus" or "outputs
that are not yet inputs". You are then talking about a process of
change and an open future. How this open-ness is resolved requires a
dynamic theory of the determinants of income distribution (not
Sraffa's exogenously imposed wage-profit trade-off, but this is an
aside). The circular flow simply assumes that the distribution of real
income is a given datum. It is not a dynamic theory.

I think you follow Sraffa and have one foot in and one foot out of the
circular flow. Either jump out, and embrace non-equilibrium, or jump
in and embrace equilibrium. But the indecision is a little agonising.
Or, alternatively, propose some differential or difference equations
that explain how change occurs over time. At this point the time
susbscripts will make an appearance, and we can talk about changes of
technique and the real distribution of income within a production

Best wishes,

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