Re: [OPE-L] price of production/supply price/value

From: Andrew Brown (A.Brown@LUBS.LEEDS.AC.UK)
Date: Sun Feb 19 2006 - 09:14:28 EST

Hi Ian,
I suspect you are basically doing what Ricardo did and trying to solve a problem that, as Marx puts it, is more difficult than squaring a circle. You are trying to present as actually existing (the two aggregate equalities) what does not actually exist (these equalities do not hold beyond a very abstract level). The root problem is that you are wedded to formal but not also dialectical logic. Marx's discussion of Ricardo is fairly clear on this point. In addition to TSV, see Ilyenkov's chapter on this (a chapter you are not going to like I fear):
You write "(ii) the relations between VCC/TCC/OCC are not
immediately relevant to the modern form of the TP. Rather than
avoiding or criticising the conditions of the modern form of the TP I
want to tackle it head on, and enjoy the contradiction."
I disagree.  You cannot separate the tackling of a problem from the justification of the terms in which the problem has been set. To challenge those terms is *precisely* to tackle it head on, in this case.  You make clear below that we effectively disgree on *how* to taclkle the TP head on, not on the need to do so.
You wrote: " I very much agree with you that the trans-social
necessity to allocate labour implies that prices must have a necessary
relation to labour-time (we have agreed on this point a number of
times in previous OPE-L exchanges). This point is particularly clear
in Rubin. The fact that this relationship is absent in Sraffa's system
is very odd."
I disagree that it is absent. To the contrary I think it is blatant. Prices of production and SNLT are close relative to market prices, and this is clear from the Sraffian calculation. You seem to me to be conflating 'necessary relation' with 'exactly proportional relation'. There may also be a distinction between formal and real necessity that is relevant here. Formally, I can choose any degree of correlation between OCCs on the one hand and wage goods, means of production and 'luxuries', on the other, that I like. In reality, the correlation is not going to be very great because these refer to real structures that have no necessary structural relationship.  
You continue " Unlike Lippi, who therefore concludes that there is
something wrong with Marx's value theory in particular and historical
materiaism in general, I am inclined to think there must be something
wrong with Sraffa's framework. The existence of the modern form of the
TP is an anomaly that requires explanation, rather than a result that
requires the rejection of the LTV. That's the hypothesis I have
applied to the N-R literature."
In our different ways we agree on this.
You continue:

"The narrow question I want to answer is therefore: why does Marx's
value theory appear to break apart under these specific, "modern"
conditions? This question, is, I think, different to the questions you
are asking. This question is also perhaps not so interesting from your
point of view because, I think, you do not believe that the modern
conditions of the TP are crucial for value theory and for
understanding capitalism.  Here we differ, because my intuition is
that the following kinds of assumptions:

- simultaneous determination
- homogeneous labour
- uniform profits
- self-replacing equilibrium
- simple reproduction etc.

are conditions under which the value-price relations should hold,
"ideal" conditions so to speak. Many people over many years have
focussed attention on this contradiction and have formulated various
responses to it. But I am at the stage where I am still struggling
with it. I can very much understand, however, if
those who have moved on from it find the whole thing pretty tiresome
by now. My excuse is that the problem is new to me, so I still find it
I reply:
I do not find it 'tiresome'. Rather I think your intuition is one of someone who is well trained in formal logic (and in maths) but not in dialecical logic. The whole point Marx was making, his essential critique of Ricardo's 'formal abstraction', was that reality does not conform to your intuition.

You wrote:
 "In the modern context of the TP, technical change is abstracted away,
and it is legitimate to do so, just as Marx, say, abstracts away from
unequal exchange in vol I. On the other hand, I do think technical
change and dynamics are essential to a complete understanding of value
I reply:
You have not replied to my point (in my previous posts) that once technical change is introduced then it becomes clear that the deviations between SNLTs and prices of production are small relative to fluctuations in market prices. 
You wrote:
 "But I agree with Saad-Filho when he says that the "aggregrate
equalities are essential for Marx". They are essential because price
is meant to be a form of value and, at a certain level of abstraction,
price must represent labour-time and there must be quantitative
identity (via a MELT, and given the abstractions)."
I reply:
But I agree with this too and have set out clearly the level of abstraction at which they hold. You are essentially disagreeing with me at what the appropriate level of abstraction is. In fact you must be saying more than this, you are saying that the level of abstraction that I propose does not tackle the TP's challenge to the LTV.
You quote me:

> (6) How then can the aggregate equalities hold, even given that the
> inputs are transformed? Once again, the answer must lie in introducing
> assumptions not present in the Sraffian calculation. I would
> speculatively look at as follows. We know that total value = total
> prices = c+v+s (by assumption regarding the value of money). Furthermore
> we know that *if* the average composition of capital of all means of
> production is equal to the social average *then* total value of c =
> total price of c. Likewise for those goods that are in the 'wage bundle'
> [traditionally the subsistence bundle plus moral and historical element,
> but I would think in terms of consumption norms etc... alas this is a
> difficult issue], i.e. *if* the goods that make up the 'consumption
> basket' have the same average OCC as the social average OCC *then* then
> total value of v = total price of v. It would follow automatically that
> total value of s = total price of s. Therefore I speculate that Marx
> makes these requisite assumptions when he claims that the aggregate
> equalities hold together after the transformation of the inputs.

You then comment "I see. Marx does not make these assumptions (granted, they are
unfinished notes)."
I reply:
Yes he does! They are far more justified by the text than the sort of assumptions the 'modern' TP attributes to Marx such as homogenous labour existing apart from its price expression - more to the point they are far more appropriate to reality, to the comprehension of capitalism. The point is that there is no strong real connection between OCCs and relevant categories of goods (wage, means of production, 'luxuries'). Their affects on price will therefore cancel out over time and space. Probabilistic reasoning is clear enough in Marx (reference to 'never attained average' etc.) 

You continue:

"I think it essential that Marx's conservation claims hold generally,
without conditions on the composition of capital. Otherwise Marx's
value-theory cannot form a basis for understanding the laws of motion
of capitalism, for as soon as the system moves away from these ideal
conditions, the necessary relation between labour-time and price is
I reply:
No conditions are externally imposed on the OCC. To the contrary the OCC is examined and found to have no structural relation to the relevant categories of goods hence a random relationship will actually hold. This is obvious from a probablistic point of view. To challenge it theoretcally you would have to argue for a strong structural relation between OCCs and the relevant cataeories of goods (this is an essentially qualitative task foreign to the formal mindset of many economists). To challenge it empirically you would have to demonstrate this relation. Frankly, while one might argue for a relatively 
weak relationship one could not argue for a strong relationship and keep a straight face. (The terms 'strong' and 'weak' are defined relative to real effects on capitalistic reproduction).

You quote me:

> You are arguing that, even considering dynamics, the neo-R calculation shows
> prices and SNLT have no necessary relation. Well, it seems to me that
> this calculation shows the opposite. It shows that the two aggregate
> qualities will not usually precisely hold together but Marx's LTV does
> not propose that they should, at the level of abstraction at which the
> Sraffian calculation is pitched.

And you reply:

"I am not sure about that. As you know, in the Sraffian framework,
Marx's aggregrate equalities hold under conditions of (i) zero
profits, (ii) uniform organic composition of capitals, and (iii) when
the economy is in `standard proportions'. (The latter might be
connected to your reasoning in terms of averages)."
I reply:
Yes, and this means that in the static framework only a *formally* possible, but *really* impossibly strong, correlation of OCCs and categories of goods (wages, means of production and 'luxuries') is going to cause a significant dent in the two agregate inequalities. When one considers that profits calculated through prices of productions require a net surplus, which generally implies surlpus labour, then this reinforces the point that market prices (where capital is allowed to temporarily move away from suplus production altogether) are going to deviate far more from values than do price of production from values (see my previous posts). 
You continue: "Also, compared to the interpretation you are providing, I think the
the simultaneous reading of the TP, in which Marx admits he should
have transformed the cost-prices of inputs, is more likely and
conceptually simpler."
I reply:
You misunderstand my interpretation. My interpretation also strongly supports this reading. It is far, far simpler and far, far more consonant with Marx because it *also* admits a more abstract level, palpably present in Marx, where the inputs aren't transformed. The question is one of levels of abstraction not of whether or not the inputs must be transformed.

You write: 
"I have been discussing the N-R critique -- I haven't added or changed
anything about it, at least not intentionally. I mentioned that the
"no necessary relation" between the dual accounting systems of
labour-value and price, in the Sraffian framework, generate the
redundancy (cannot go from labour-values to prices) and inconsistency
(value rate of profit is not the price rate of profit) critiques. This
accounting mismatch between labour-time and price is the common
element that underlies Steedman's redundancy and inconsistency

I reply:
That is *your* interpretaion. But a read of Steedman's Marx after Sraffa reveal that *he* stresses redundancy and inconsistency (take a look at the summary chapter). I think he is wise to stress this because any adequate dynamic model admitting realistic capital movements is going to show that SNLT and prices of production are very close relative to the boom and bust pattern of market prices. Please refer to my arguments for this above and in the previous posts.
You finish:
"You do not expect Marx's equalities to hold under the Sraffian
conditions. I do expect them to hold. According to the N-R critique
they do not. So I have a theoretical headache, but you are less
troubled. I think that's about the sum of it!"

I reply:
I would say that we have a key disagreenment that leads to our different expectations. I would stress the difference not the similarities between us because I think these differences imply a very different view of the CMP. Which doesn't stop us having, indeed contributes to, what I have found a very fruitful discussion.
Many thanks,

This archive was generated by hypermail 2.1.5 : Tue Feb 21 2006 - 00:00:02 EST