[OPE-L:8178] Re: Marx and labour theory of value

From: Michael Eldred (artefact@t-online.de)
Date: Sun Dec 15 2002 - 12:23:59 EST

Cologne 15-Dec-2002


Paul Cockshott clyder@gn.apc.org schrieb Sat, 14 Dec 2002 00:21:36 +0000:

> Quoting Michael Eldred <artefact@t-online.de>:
> > If there is any confusion here, it is your own, since, without correcting
> > you, I merely adopted your mode of expression when you wrote: "differential
> > value creating powers of different labours" (sic) and claimed the effect is
> > "quite modest".
> I agree that I am to blame for expressing myself loosely. Conceptually I think
> that one should not really speak of the value creating powers of different
> labours. What I should have written was that the two techniques for
> computing the labour coefficients place an upper bound on the effect
> of using wages as a surrogate for labour time. The use of wages as such
> a surrogate is usually justified by saying that the higher wage rates
> correspond to higher value creating powers, for my part I consider this
> argument rather confused for the reasons I set out in my last letter,
> but I just used the conventional terminology used to discuss the issue
> in my previous letter - I should not really have done that.
> > The question stills holds substituting 'labour-power' for
> > 'labour'. According to Marx's formulation, complicated labour-power creates
> > more value than simple labour-power in equal periods of time. So we agree
> > that there is this difference is value-creating potentials between
> > different, concrete kinds of labour-power (but see further below, where you
> > apparently deny this).
> No this is where I disagree, both because of the reservation about
> value creating potential, and because I consider that the difference
> between simple and compound labour applies only within the context
> of a given concrete task. I dont think that qua value, different kinds
> of labour count differently.

I think your ontology compels you to do this, and that this is quite consistent
within your social ontology (see below).

> > The question is how to gain any empirical access to
> > this difference _without_ proceeding through the value-form, since the
> > labour that forms the purported substance of value must be independent of
> > the value-form if it is to be its causal determinant.
> My view is that because it is only labours of the same concrete sort
> that can be compared for productivity, then it is only in this
> context that one can speak of simple and compound labour.
> > ME That is the claim made by the LTV. You are now obviously understanding
> > "socially necessary" as "socially average". What is the justification for
> > this? Why not "socially minimum" or "socially median", e.g.? How is the
> > qualification "socially" to be empirically or even theoretically specified?
> It is justified within the general problematic of classical political
> economy from Smith to Marx, which is that society has a certain quantity
> of labour that has to be distributed between different activities.

Actually, in truth, it doesn't. The workforce available to a given society is not
just a quantity, but a highly differentiated social 'structure' or 'complex' (for
want of a better word, a 'together' of many 'folds'). Not labour, but labour-power
is distributed to various labouring activities, and this distribution itself is a
complex phenomenon depending on many factors such as training, willingness to move
location, adaptability vs. social inertia, various traditional bonds, etc.

> In
> the absence of special conditions relating to agriculture - dealt with
> under the theory of differential rent - a 10% increase in the output
> of a given industry will require a 10% increase in labour allocated.
> In computing this increase in labour one is automatically performing
> an averaging operation over the labour of all the workers currently
> employed. If 100,000 are currently employed, we know that a 10% increase
> in output will require 10,000 new workers of average ability.

This only holds under the ceteris paribus condition that labour productivity
remains the same, i.e. that new know-how enabling more productive ways of doing
things is left to one side. But knowing how to get more out of a production
process is a continual, on-going process accompanied by never-ending
transformation. That is one of the key features of capitalism -- that, mediated by
competitive markets, it tolerates no standstill. It is rather the state that
stands for inertia, for the protection of conservative inclinations and entrenched
interests. This is one of the main issues for politics -- the tendency to inertia
and habitual ways in social living.

> The value of a commodity is just the change in the division of labour
> required to increase its production, and here the norm is to assume
> an average. The assumption of constant returns here consists of the
> minimal assumption. If one has explicit reason to think that constant
> returns will not apply then one does not take average labour to
> be regulating - hence Ricardos rent theory.

Yes, the phenomenon of natural conditions of production and subsequent rent is one
reason why equal amounts of labour do not produce equal amounts of product. But
another, equally fundamental reason is that human know-how (Greek: _technae_) is
in a continual state of flux. The use of things and how to bring them forth
(literally: pro-duce them, 'guide them forth') are continually showing themselves
in other ways that seem more advantageous and increase productivity. This takes
place in capitalism not just under the constant striving to decrease costs (which
can also just lead to cheap and nasty products), but also in the effort to produce
better products or qualitatively different products with new use-values.

This is a continual social process of truth in the sense that use-values and
production processes constantly disclose themselves to human understanding or
know-how in new ways. Truth is a way of disclosure to human understanding, and
this way of disclosure is always historical and always social in the sense that
new ways of disclosure, such as more efficient methods of production, are
inevitably _shared_ by people.

> > You are apparently understanding "social" here as a simple aggregate, as a
> > set of factual data. Such a mass-aggregation of the social dimension to a
> > set of data represents a crude ontology of the social dimension.
> It may strike you as crude, but it does not strike me that way.
> It is an approximation, a simplification, but unless one first
> examines the simple case, one does not know how significant
> second order perturbations are going to be.
> Can you give some reason other than personal preference why this
> approximation is to be rejected?

The "reason" can only be an alternative ontology of the social dimension. Such an
ontology would not be along the lines of being intent on grasping, i.e.
comprehending, social phenomena such as commodity exchange and production from the
quantitative side.

The reason for preferring a richer social ontology over a predominantly
quantitative social ontology can only be that the former is more adequate to
showing up the social phenomena in their many aspects than the latter, which is
always dependent on finding a quantitative reduction of the phenomena in order to
work. An explicit ontology of social relations would make these issues clearer.
Why? Because social ontology decides on how social phenomena show up and define
themselves in presenting themselves to human understanding.

With regard to capitalist society, I think that the commodity exchange relation
has an exemplary, foundational role in comprehending the kind of society this is.
The predominant proclivity to quantity and to claiming that this is the only
'rational' possibility for social science in the modern age (since the fifteenth
century) prejudices how social relations can appear in their truth. Strictly
speaking, the oft-used term 'social relations' is pleonastic, since the very term
'social' already implies a relation.

I think one has to be very clear about how our thinking is held under the sway of
simple casts or paradigms of thinking exemplified quite astoundingly by texts such
as Descartes' _Regulae_, _even though_ there is hardly a natural or social
scientist living today who has studied this eye-opening text. Certain
philosophical texts unfold their history-shaping effects in subterranean ways,
unbeknowns to those caught up in an age.

> > ME If "socially necessary" is defined as "of average productivity", then by
> > definition, differences in skill are reduced to the average.
> And of course within the framework of i/o analysis that is just
> what socially necessary implies. I agree that for long term projections
> linear i/o analysis breaks down, but over a perspective of a year or
> so - time covered by the i/o tables it is reasonably accurate.
> The representation of intractable non-linear systems by locally
> linear approximations is a common technique.

Then you get a curve composed of linear segments.

> > ME Now you are shifting from different workers working on the same task to
> > whole industries which comprise not only "hundreds of thousands of workers"
> > but also tens of thousands of occupations and at least tens of thousands of
> > concretely different labour-powers. How do you compare the concrete labours
> > of, say, an electrician and a welder? There is no simple measure of
> > productivity just in terms of time. Time to do what?
> >
> >ME There are enormous differences among houses built (ranging, say, from luxury
> > mansions to cheap little huts) and dresses made (ranging, say, from haut
> > couture to simple, basic smocks).
> PC Obviously any serious work study has to control for the task done,
> if you want to compare two people's productivity you set them both
> to do a standardised task.

That is a possibility, but it doesn't help when dealing with the factually given
diversity of use-values of a given kind such as houses or dresses or hi-fi
systems. The empirical test of the LTV depends on quantitative aggregation.

> > If the average builder builds h average houses in a week, and the average
> > dressmaker makes d average dresses in a week then what does it mean to
> > equate "the two averages"? Since a house is the final result of the labour
> > of at least hundreds of different individual trades, and a dress likewise,
> > there is no way to simply equate h houses to d dresses (unless it be through
> > the value-form).
> No you don't have to resort to the value form. As Marx, a poor Darwinist,
> remarked the key to the anatomy of the ape is the anatomy of man - that
> may be crap, but applied to history and society it has some sense.
> If you consider the problem that a socialist society has in allocating
> labour then you can understand what the labour theory of value is all
> about. Suppose a socialist society wants to rate people according
> to the productivity of their labour. How to compare the labour of
> a particular dressmaker with a particular typesetter?
> Let dressmaker d be 150% as productive as the average dressmaker.
> Let typesetter t be 75% as productive as the average typesetter
> Equate the average productivities as simple labour, so we get
> that d is twice as productive as t. The whole thing makes perfect
> sense in terms of the allocation of social labour.

I agree that for the purposes of socialist planning and allocation of labour, this
way of measuring -- for better or for worse -- would be an appropriate way
allocating or assessing social labour. What I am questioning is whether commodity
value has much to do with such an orderly, quantitative allocation or assessment.

What I mean is the following: For a socialist society in which there is a
structure of organs empowered to decide on an overall plan of production for
society, the problem of the sociation of labour has been solved precisely in this
way, in the medium of political power (including deliberation, decision-making,
prioritization, factional fighting, etc. etc.). But in capitalist society, the
sociation of labours is done first and foremost through commodity exchange.
Adaptations of the total social production process take place only through a kind
of feedback loop which regulates the dissociated production processes of myriad
individual small, medium and large capitals.

> > ME It seems to me that to be consistent, you would have to negate all
> > differences whatsoever among concrete labour-powers and concrete labours by
> > reducing uniformly to durational time t. But this would contradict your
> > previous admission that there are "differential value creating powers of
> > different labours".
> yes
> > > PC The point is that differences in skill apply within a trade, between
> > > trades what applies is the abstraction from concrete labour to
> > > abstract social labour. There is not, in Marx's scheme, any prejudice
> > > with regard to some trades being more noble or worthy than others.
> > > There is no surplus value proper to, or specific to, a profession.
> > > Builders are not more worthy than dressmakers, nor are professors of
> > > agriculture more worthy than plumbers.
> >
> > ME I presume that you mean by "more worthy" "more value-creating".
> yes
> > ME That would
> > mean that different kinds of concrete labour-power of average quality, no
> > matter whether one kind of labour is complicated and the other is simple,
> > create an identical amount of labour value in the same mathematically
> > measurable time t. But Marx claims just the opposite in Kapital Chapter 1:
> >
> > "Complicated labour counts only as _potentiated_ or rather _multiplied_
> > simple labour, so that a smaller quantum of complicated labour is equal to a
> > larger quantum of simple labour. The fact that this reduction is constantly
> > taking place is shown by experience. A commodity may be the product of the
> > most complicated labour, but its _value_ sets it equal to the product of
> > simple labour and therefore itself represents only a certain quantum of
> > simple labour." (MEW23:59 emphases in the original)
> I think that there is some confusion on Marx's part here.

Good, that is now clear.

> > You now apparently dispense with this multiplier or coefficient of
> > value-creation in your version of the LTV, and so depart from Marx's
> > formulation of the so-called law of value. Such a departure is in line with
> > the Cartesian _Rules_ (_Regulae_), according to which phenomena are to be
> > grasped purely quantitatively in order to gain certain and reliable
> > knowledge of them. You reduce all the different concrete kinds of
> > labour-power to a homogenous, abstract labour-power whose exercise can be
> > measured purely by a mathematical time variable t (where t should not be
> > confused with the phenomenon of time as such). The only requirement is that
> > the labour-power has to be averaged over all the factically available
> > labour-power of a certain kind. The factical total performed labour of a
> > certain kind  is then defined to be socially necessary labour-time, which is
> > then spread as an average over individual labour-powers.
> >
> > This reduction of performed labour (a highly complex and differentiated
> > phenomenon) to an abstract homogeneous labour measurable by a mathematical
> > time variable t is akin to Newton's first law of motion of physical bodies
> > (and Galileo's earlier preliminary formulations of the same law) which
> > cannot tolerate any differences in space. A total homogenous space is
> > "imagined" (mente concipio, Galileo 1638). Why? Because otherwise there
> > would be no _mathematical_ access to the phenomenon of motion formulable in
> > a simple law amenable to a mathematical interpretation and calculation.
> I broadly agree with this.
> > ME A highly trained engineer earns much, much more that a simple manual
> > labourer, regardless of whether education costs are footed by the state or
> > not. Your argument here is highly counter-intuitive. If surplus-value
> > production were greater through employing simple labour-power, then why
> > should there be a tendency within the development of capitalist economies to
> > shift more and more to the higher end of the so-called 'value-creation'
> > chain (value being understood now in the sense of 'realizable in money'),
> > requiring more and more highly trained labour-power? Why is the market value
> > of the total product of highly developed economies such as Switzerland with
> > highly trained workers per capita immensely higher than economies based
> > largely on simple labour? Why are certain developed industries (such as the
> > chemicals industry in Germany) dependent upon a highly trained, but very
> > expensive workforce (say, of engineers and chemists) to maintain their
> > profitability?
> This relates to the constraints imposed by the necessities of concrete
> labour. If one is to produce certain products one has, until
> automation advances sufficiently, no option but to employ expensively
> trained labour. As soon as automation or improved work organisation
> allows the employment of less skilled labour - that is what
> capital does.

Yes, "automation or improved work organisation allows the employment of less
skilled labour", but such a complex production process still requires its top
level of highly trained engineers, etc. who know and understand the process
holistically and are able to continually modify it and fix it when there are
hitches (which also happen continually). What I'm saying is that there is no
static state which could dispense with highly trained and educated workers, and
that the know-how which such workers embody is acknowledged by both considerably
higher wages and higher sales revenues (either per unit or overall).

> > ME I agree that, according to the LTV,  the varying value-creating potencies
> of
> > labour-powers of various kinds cannot be read off the differences in pay
> > scales. If one professional worker is paid twice as much on average than a
> > worker in another occupation, this does not imply that twice as much value
> > is created by the former worker compared to the latter. But you are here
> > denying (contrary to previously -- see above), it seems, that there are
> > differences in value-creating potencies of labour-powers at all. And that is
> > quite consistent of you. A mathematical treatment cannot tolerate such
> > differences. Differences are reduced to the average with greater or lesser
> > variance around these averages.
> >
> > An alternative way of thinking to the LTV would be to allow the qualitative
> > differences in labour-powers, labours, use-values, commodities, etc. to have
> > their full weight and to dispense with the attempt to provide an explanation
> > of commodity prices in terms of a simple, abstract, mathematically amenable
> > variable.
> >
> > E.g. there are hundreds of different types of steel suitable for various
> > different uses. To make highly specialized (say, super strong, super
> > lightweigt) steel requires specialized know-how that is not generally
> > available among steel workers en masse. The high quality of this steel is
> > acknowledged on the market in higher selling prices by those consumers
> > (other capitalist enterprises) which similarly make high-quality,
> > specialized steel products (say, aircraft). It is not the amount of
> > additional labour that is reflected in the higher prices of high quality
> > steel (relative to the prices for run-of-the-mill steel churned out by heavy
> > industry with little know-how), but it is the especially high quality of the
> > labour-power, i.e. its special _know-how_, which enables a higher price to
> > be had for high quality steel on the market and which enables higher wages
> > to be paid for highly trained labour-power.
> I doubt this. I suspect that the higher costs of specialised steels
> directly reflect the shorter production runs and higher labour intensity
> of production.

I suspect that the issue of "higher labour intensity" represents a similarly
intractable problem for you as the difference between complicated and simple
labour between qualitatively different labours. The national accounts don't
provide any foothold for taking labour intensity into account. It too has to be
averaged and reduced to mere duration time, t, as if all labour were of average

> This is not to exclude the existence of  deviations of prices
> from values induced by shortages of certain skilled labours.

The admission of "deviations of prices from values" is what the LTV is
theoretically compelled to do in order to account for any phenomenon that presents
an exception to the quantitative averaging procedure. This amounts to defending
the core of the quantitative social ontology upon which the LTV is based and
dealing with aberrant phenomena by means of auxiliary theoretical constructions.
But, in my view, this amounts to theoretical violence.

> > ME The phenomena of branding and advertising and product design also show that
> > not only labour is recognized socially in product selling prices, but also
> > perceived use-value. The perceived use-value (and all use-value is
> > perceived) on the part of consumers, their desire to have and use a product
> > can be influenced by making the product appear to be a cut above the rest
> > (successful branding), or by inventing a new use-value through product
> > design (games are a good example). There is no fixed set of use-values.
> > Rather, use-value itself is a social phenomenon constantly in a state of
> > flux depending on how things _show themselves_ in their usefulness to
> > members of society. In capitalist society, this self-display of products is
> > mediated by the value-form of money, and is also honoured in the money-form
> > in the shape of enhanced prices (which, say, justify the enormous pay
> > package for a successful product designer). (In your paper with A. Cottrell
> > you yourself refer to "the development and application of new technologies;
> > the development of new products..." as factors enhancing capital
> > valorization.)
> Yes but these factors relate to price not value.

I say there is no difference between the two. There is no substance which
constitutes a value that then expresses itself in money. The fundamental concept
of Marx's theory of capitalist society, value, is not and cannot be a concept of
substance. Value is a relational concept that conceives the elementary 'cell'
social relation of capitalist society. Value 'is' only in the relation of one
commodity to others, and ultimately in relation to the peculiary thing, money.
This relation can be either 'potential', _dynamei_, or it can be 'at work',
_energeiai_, actual when commodity exchange takes place.

> > What I am proposing as an alternative to the LTV amounts to enriching the
> > concepts to suit the richness of the social phenomena instead of going in
> > the opposite direction of abstracting from the phenomena to make them
> > mathematically manipulable (in your case through statistical methods).
> >
> > > If the working week is reduced to 35 hours, then over time the
> > > petroleum industry will still have to meet the cost of
> > > reproducing the labour of its engineers which will still
> > > be euro 5000pa, but now has to meet this out of 35hours a
> > > week value added. The point is that the amortised cost of training does
> > > not vary with the working week wheras the value added does.
> > > A sufficient reduction in the working week in an economy with
> > > high education costs will entirely eliminate surplus value.
> > > Skill aquired through education appears on the debit rather
> > > than the credit side of the national accounts.
> >
> > This apparent dilemma seems to be a consequence of your postulation of an
> > abstract, homogenous labour as the substance of value. Then, a highly
> > trained production engineer with a lot of know-how will 'create' an equal
> > amount of value as a simple rig worker.
> No it has nothing to do with this. It is simply a matter of doing
> the national accounts in hours and showing that a rise in education
> cost can exhaust the social surplus product.

I don't know my way around the national accounts. In the end, it is all a matter
of how the magnitudes represented there, which are in the main monetary
quantities, are interpreted through an adequately founded concept of value.

> > ME My problem is the conceptual problem of how to empirically identify that
> > labour which counts as "socially necessary" and whose measurement in
> > labour-time therefore is supposed to regulate, in some fashion, monetary
> > prices. Your procedure plainly relies on taking the brute aggregates,
> > levelling off all differences by way of averages and abstract positing, and
> > then claiming that the result is value substance measured in time.
> > ME A "close and systematic relationship" between national labour accounts and
> > monetary accounts established by means of statistical analysis does not yet
> > say anything at all about causes. Causal relationships have to be
> > theoretically explained in a persuasive was and used to interpret empirical
> > data.
> I agree. But I think the F&M explanation is well argued and plausible.
> > > PC The proposition of the Law of Value, is that VALUE regulates Exchange
> > > Value.
> >
> > I agree, with the proviso that you homogenize the substance of labour-value
> > more than Marx does (see above).
> > ME How is the average hourly wage empirically established, i.e. without
> knowing
> > the aggregate labour hours worked in an industry?
> The hourly wage rates in each industry are collated by the
> department of trade on the basis of reports filed by firms.
> > > As a communist my principle concern in economic investigation is to
> > > come up with better planning mechanisms for communist economies.
> > > I only get into discussing value under capitalism to establish that
> > > capitalist economies - behind the scenes - are doing calculations in
> > > terms of labour times.
> > >
> > > One of the points Allin and I make in writing on communist economies is
> > > that a deficiency of the gosplan accounting system was that it did not
> > > use labour time calculations sufficiently in deciding between
> > > alternative
> > > possible projects. We hold that the integration of macro and micro
> > > planning would be better if labour accounts were used explicitly instead
> > > of Roubles.
> >
> > That is at least consistent communism. A total social cybernetic plan for
> > production and distribution can only fulfil this total control by abandoning
> > the value-forms altogether. The issues then become those of the politics of
> > production and executing the total social production plan efficiently and
> > adequately. The sociating medium of money is replaced by the sociating
> > medium of power and power struggle.
> Yes that is the specific relation between levels in the communist
> mode of production.

Communism is then socially cybernetic in the sense that it envisages a structured
social _archae_ which governs what, how, how much, of what quality is to be
produced and distributed throughout society. This contrasts with capitalism with
its multi- or rather myriad-_archai_ system of individuals and collectives
composed of individuals each pursuing their self-interest, mediated by market
exchange. The principle or _archae_ which ultimately rules in such a dissociated
system sociated through the medium of money and markets is the simple principle of
capital, viz. that money advanced as capital not dwindle. This principle imposes
discipline on the self-interested actors in a capitalist economy, whereas in a
communist form of sociation, such discipline would have to be conscious, political
discipline. The contradiction between self-interest and collective or universal
interest assumes other forms.

_-_-_-_-_-_-_-_-_-_-  artefact text and translation _-_-_-_-_-_-_-_-_-_
_-_-_-_-_-_-_-_-_-_-_-_-_-_-_- made by art  _-_-_-_-_-_-_-_-_-_-_-_-_-_
http://www.webcom.com/artefact/ _-_-_-_-_-_- artefact@webcom.com
_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_ Dr Michael Eldred -_-_-

This archive was generated by hypermail 2.1.5 : Mon Dec 16 2002 - 00:00:00 EST