9

The Negation of
'Negative Values'
Sungur Savran
Within the multi-faceted critique directed at Marx's theory of value
by those economists who base themselves on the work of Piero SrafTa,
there is an area which has intrigued many if only because of the
singularity of its very terms. I am referring to the debate on so-called
'negative values' and 'negative surplus-value', a debate initiated by
Ian Steedman in 1975/ and since then a central feature of his critique
of labour values. 2 To many, it seemed that since the extension of the
Marxist theory of value to various fields (i.e. joint production and
fixed capital) yielded such absurd results, there certainly had to be
something irredeemably wrong about it. Given the impeccable nature
of Steedman's mathematical argumentation, there seemed to be no
escape route. The contradictions simply had to be admitted.
The truth is that the mathematical argumentation that leads to
these results is based on a fallacious reconstruction of the theoretical
structure of Marx's concept of value. It can be shown that in both
cases, the demonstration rests on the inacceptable replacement of
Marx's concept with another which is totally alien to it. The
anomalies reached in this manner are then falsely presented as
resulting from Marx's own theory. My purpose in this article is to
argue that these anomalous results have no relationship to Marx's
theory of value. 3

1. Social Values

What I will do is to show that, contrary to his claims, Steedman

reaches negative values and negative surplus-value through the
application not of Marx's concept of value but of a totally erroneous
and caricatured version of it.
In order to do this, one has to reduce the situation depicted
by Steedman to its true dimensions. This situation is not joint

?11
1·&"
212 Savran 213

production pure and simple, as would have been the case if two I.n [Link] ofth.i~ general approach, Marx investigates the effect
commodities were produced by a single process. It is a situation where which dlfferen~ conditions of production have in determining social
each of the two commodities in question is produced using two value. A detailed analysi~ of this investigation is unnecessary for
different processes, this is to say, there are prevalent in society two our purposes. What IS Important is the fact that in different
different production methods for each commodity. If, therefore, circum~tances? different conditions of productivity ma~ influence or
Marx's theory of value is to be applied to the situation in question, as determIne social value. That is to say, social value can be determined
Steedman claims to do, one has to follow Marx's own method of by thos~ producers who work at the lowest or highest levels of
dealing with such questions. This is briefly what we shall first have to productiVity, as well as those who work at average levels. It is,
review. therefore, absurd to contend, as Steedman does in his debate with
In the first volume of Capital, where he investigates commodity and Morishima, that 'in his mature works, Marx repeatedly asserted that
'capital in general', that is capital in its sole relationship to wage- ... he would .[Link] the value of [a] commodity by reference to the
labour, Marx abstracts from the relations among different producers average conditions of production and not by reference to the most (or
of the same commodity. Hence, the effects of these relations are least) favourable conditions.'6
excluded by the nature of the level of abstraction. In the investigation That Steedman distorts Marx's position on this question should
of the 'isolated commodity' of the first volume, the labour embodied not obscure the even more important fact that for Marx the same
in the commodity has to be socially necessary, no more, no less. 4 That cOI?m.o~ity, when produced with different methods, has a mUltiplicity
which determines socially necessary labour is defined as 'the of IndiVidual values, none of which is in general equal to the social
conditions of production normal for a given society'.5 This first value of the commodity. The latter comes about only as a result of a
definition of socially necessary labour is, therefore, abstract: in other process of equalization.
words, it is not explained how the category 'normal' is determined in We are now ready to evaluate Steedman's claim that what he
the concrete. This category is indifferent to the various conditions of applies to the .case of joint production is Marx's theory of value.
production that may exist in real life, be they conditions of average, ~[Link] embarkIn~ upon a detailed critique of Steedman's procedure,
high or low productivity with respect to the situation prevalent in It I~ useful t? POInt out that the result of negative surplus-value is
society. entirely contIngent upon the existence of negative values, so that, once
In the third volume, on the other hand, where the relations between the concept of negative values is done away with, there remains no
various capitals are included in the investigation, the competition problem to be solved with respect to negative surplus-value.
among producers within a single branch of production can no longer ~ second point of considerable importance is that, in the example
be abstracted from. This, of course, results in the analysis of situations which Steedman constructs, 7 which analyses a situation where two
where different capitals produce the same commodity with methods commodities are produced jointly by two different processes, one of
of differing productivity. the processes represents a higher productivity of labour. That is to
The concept of socially necessary labour is now subject to new say, one process is more productive than the other.
determinations, beyond that generality appropriate to the level of Using Stee?man's example, if one compares the two processes, the
abstraction made in the first volume. It is, in other words, made net product (In Sraffa's sense) of the first process contains one unit of
concrete. It now turns out that in a branch of social production where each commodity, while the net product of the second contains three
various capitals produce under different conditions of productivity, units of the first commodity and two of the second. This is no mere
the social (or market) value of a unit of the commodity produced in coincidence. It ha.s .been proved that negative values can only arise
this branch is determined by the division of the whole quantity of under such conditIOns. In other words, an absolute difference of
labour expended in the branch to the total mass of commodities productivity between the two processes is a necessary condition for
produced in the branch. Commodities under different conditions of the appearance of such negative values. 8 Hence, if it can be shown that
production have distinct individual values which are equalized, Steed~an's method is inappropriate in this specific case, then the
through a process, in the social value of the commodity. questIOn of negative values will totally disappear.
 ~..

214 Savran 215

Figure 1 second process than in the first. Therefore, to determine the value of
the second commodity, a smaller amount of labour has to be divided
Inputs Outputs Net product
into a larger number of units in the second process. Hence the
Com- Com- Com- Com- Com- Com- individual value of the second commodity is necessarily smaller in the
modity modity modity modity second process than in the first. Equality between individual values in
modity modity the two processes is not possible for commodities simultaneously.
This result is very important, for it implies that in the context of an
2 Lab. 2 2
absolute difference in productivity between the two processes, the
Process 1 5 0 6 1 1 1 individual values in the two processes of at least one of the
Process 2 0 10 3 12 3 2 commodities are unequal. Hence a method consistent with Marx's
has to investigate the method of equalization of these two distinct
individual values. Steedman claims that he is applying Marx's
2. The Equalization Process method to his own special case. Yet what he does is to ignore this
difference between the individual values (which has conclusively been
Once this has been established, we can now proceed to investigate proved) and to declare, through the use of simultaneous equations,
Steedman's example in the light of Marx's framework. In a system of their equality. Thus he totally sets aside Marx's own analysis
one-product processes, given the values of all other commodities, the regarding the determination of social value. Whereas Marx speaks of
individual values of a certain commodity produced under several a process of equalization between distinct individual values, with
different conditions can be determined by recourse to the individual Steedman these lose their quality of distinctness only to be assumed
processes. With joint production, on the other hand, it is not possible equal from the outset. Hence the existence of negative values. These
to 'read off' individual values from individual processes, since at least arise as a result of the forcible equalization of distinct and unequal
some processes produce more than one commodity. Nevertheless, magnitudes.
where an absolute difference exists between the two processes with It has, therefore, been shown that in the unique case where negative
respect to productivity, one fact can be established unequivocally: values can arise, Steedman has reached them by applying a theory of
that the individual values of both commodities cannot be value different from Marx's. Hence the futility of his allegation that
simultaneously equal for both processes. For instance, if the Marx's theory of value leads to self-contradictory results in the
individual values of the first commodity are equal in the two context of joint production. It is a caricature ofthe Marxist concept of
processes, those of the second are necessarily unequal, and vice versa. value that leads to these absurdities. Of course, as has already been
Of course, in all probability, the individual values of both noted, once the existence of negative values is disproved, so is the
commodities will be different for the two processes, but as a special existence of negative surplus-value, so that there is no contradiction
case one ofthe two commodities may have identical individual values in either the theory of value or the theory of surplus-value developed
in the two processes. However, even in such a case, this cannot be true by Marx.
for both commodities at the same time. Although it has already been proved that Steedman's method of
The proof is simple. Pick a commodity at random and assume that simultaneous equations is patently inappropriate for representing
its individual values in the two processes are equal. Once due Marx's method, it may still be necessary to dwell upon Steedman's
deduction is made for this commodity, the quantity of labour that own claim that the former method is in fact due to Marx. Since the
remains for the second commodity in the first process is necessarily method of simultaneous equations does not occur in Marx's own
smaller than that remaining in the second process. This is because, in voluminous work, Steedman bases his claim on Marx's calculation of
the second process, a greater amount has been produced of the first value through the addition of that part of constant capital used,
commodity, so that whatever [Link] is allocated to this commodity in variable capital and surplus-value. In either of the following cases, it
the first process, more has to be allocated in the second. On the other can be regarded as appropriate to express this method of calculation
hand, more has been produced of the second commodity in the through the use of simultaneous equations: firstly, if each commodity
216 Savran 217
is produced by a single method or, secondly, if individual values have the contention that joint production is the only method of
already been reduced to social values. However, if and when there satisfactorily dealing with fixed capital. In what follows, both
exist more than one method of production for a commodity, it is contentions will be seen to be wrong. Limitations of space do not
impossible to determine the social values of this commodity permit the reproduction of Steedman's arithmetic examples. The
simultaneously with the others by the use of a system of equations. interested reader is referred to the relevant chapter of his book. 9
There are two reasons for this, one theoretical, the other formal. Steedman's point of departure is the observation that value
The theoretical reason is that this method assumes equal individual depreciation with respect to time may not be linear, for example a
values where there are only unequal magnitudes. The formal reason machine may tum out to be less efficient in the first year of its life and
is, very simply, that if two production equations are included in the more efficient in the following years. He admits that Marx is aware of
system but in both cases the same identical value is attributed to the this fact but assumes linear depreciation as a first abstraction. He
commodity, the number of equations will exceed the number of claims, however, that once this first level of abstraction is abandoned,
unknowns and the system will be indeterminate. Hence the method of Marx's method of treating fixed capital, namely the transfer of value
simultaneous equations is inappropriate for the reduction of distinct to the final product, can lead to self-contradictory results. To show
individual values into one social value. Note that this is true even in that this is the case, he constructs an example where the same machine
the general case, before joint production is introduced. Not being of different ages is used by different capitals. His first step is to reach
valid in the general case, naturally it cannot be extended to two different values for the same commodity, produced by the use of
Steedman's case. Therefore Steedman's claim that he is applying these machines of differing ages, through the application of Marx's
Marx's method to joint production falls to pieces. The 'difficult' method. What is more, both of these values are different from the
problems of negative values and the coexistence of positive profits 'correct' value, calculated by the use of the net product method.
with negative surplus-value are no more than pseudo-problems that Therefore, concludes Steedman, the use of this method within Marx's
arise from mistaking mathematical relations for real relations. theory of value leads to internally inconsistent results. 10
One final point before we leave the matter. Not only is the method According to him the correct method of calculation is to attribute
applied by Steedman to the analysis of joint production not Marx's different values to machines of different ages and solve the problem by
method, but, equally, this method is itself internally inconsistent. means of a system of simultaneous equations. In other words, it is to
According to Steedman, value is determined by the labour embodied treat fixed capital as a joint product. This does lead to consistent
in commodities, which in its tum is totally dependent upon the results but also to other sorts of bizarre consequences. Depending on
method of production used. In his own example, two different the case, either the depreciation quota, that is the value transferred
methods of production are used for the production of each from fixed capital to the final product, or the value of fixed capital at
commodity and, consequently, the quantity of labour embodied in the end of the period may tum out to be negative. 11 Hence even in
the commodity is different in the two processes. The value of the those cases where it does not lead to unacceptable results, the theory
commodity when produced in one process should therefore be of value ends up in a state of contradiction with its bases.
different from that of the same commodity when produced in the The problem seems serious indeed. Negative value transfer or a
other. Steedman contradicts himself by assuming these different, negative value for the used fixed capital is contradictory with the
individual, values of the commodities to be equal. fundamentals of Marx's treatment. As Marx says, 'the means of
production can never add more value to the product than they
themselves possess independently of the process which they assist.'12
3. Fixed Capital This is contradicted by a negative value for fixed capital, for what this
implies is, in effect, that fixed capital imparts to the final product more
A third area where Steedman claims to have shown an inconsistency value than it possesses and thereby acquires a negative value. On the
is in Marx's treatment offixed capital. This question is also important other hand, the value of the means of production 'is determined not by
because the significance of joint production for the theory of value has the labour process into which it enters as a means of production, but
been defended not on the grounds of 'pure' joint production but on by that out of which it has issued as a product.'13 This is in blatant
218 Savran 219

contrast with negative value transfer. What happens in such a case is a by applying Marx's method but by abandoning it. Therefore, these
flow of value from the product to fixed capital, or, in other words, the results have nothing to do with Marx's theory of value.
determination of the value of fixed capital by that process into which Once this is understood, it is easy to see that value calculation
it has gone as a means of production. according to Marx's own method, that is value transfer to the final
All these important results are obtained, however, on the basis of product proportional to use-value depreciation, is totally consistent
unsound arguments. To show that this is the case, it is better to start and adequate. It was already noted that Steedman obtains three
with the second of Steedman's calculation methods. As in the case of different values for the same commodity through the application of
joint production, this allegedly correct method relies on the forcible this method. Once it is grasped that two of these three values are the
equalization of the unequal individual values of different species of a individual values related to two different production processes it is
commodity produced under different conditions. It is of course very very simple to explain firstly, why these values are not equal to 'eaCh
easy to show that individual values are not equal: different degrees of other, a~d seco~dl~, wh~ both of them are distinct from and not equal
efficiency having been assumed for the different ages of the machines, to the. thud, ~hlch IS s~cJaI val~e. In fact, far from being the symptom
equal numbers of living labour will produce unequal quantities of the of an mconslstency, this result IS a perfect manifestation of the logical
commodity in the different cases. This of course implies unequal conherence of Marx's work in its totality. The analysis offixed capital
individual unit values. merges here into the analysis of the formation of social value in the
What should have been done, on the contrary, is to admit the context of the existence of different methods of production for a
inequality of individual values and to investigate the process of the comm?dity. Steedman's allegations, both of inconsistency and of the
formation of social value. The answer to this correct question is not necessity of the method of joint production in the treatment of fixed
difficult to provide, since as 'the means of production transfer their capital, fall to pieces on careful examination.
value to the product only in so far as they lose their exchange-value
along with their independent use-value'/4 they impart to each unit
that can be produced during their lifespan the same amount of value. Conclusion
Which means that the basis of Marx's conception is linear
depreciation not with respect to time but with respect to use-value. In The argument presented in this article shows that the anomalous
this case, the value of each such unit is equal in magnitude. During results reached by Steedman in his treatment of joint production and
those years in which the means of production are less or more efficient fixed capital have no bearing on Marx's theory of value, for they are
than average, individual value rises above or falls below social value. derived on the basis of a total misrepresentation of the relevant
In the first case, the capital in question receives a lower rate of profit aspects of this theory. Once this is seen, the mystique that surrounds
than average, in the second, it is the opposite that holds. But the concepts of 'negative values' and 'negative surplus-value' van-
calculated over the whole lifespan of the instrument, the rate of profit ishes. Non-existent in theory, 'negative values' are revealed to
of both the branch and the individual capital that uses this instrument embody wasted intellectual labour in practice.
is equal to the general rate of profit. The rebuttal of the allegations concerning joint production and
The application of Steedman's allegedly 'correct' method is not fixed capital removes yet one more foundation of the post-Sraffian
only inconsistent with the bases of value theory, it also produces critique of the Marxist theory of value. Of modest importance on its
absurd results. To take but a single instance, ifthis method is adopted own, this result is significant insofar as it contributes to the all-sided
but the example is modified, the value of the commodity will fluctuate demise of this supposed critique.
from one year to another. 1 5 That this result is absurd in the context of
the theory of value is obvious, for the products of the various years
will have different social values even though all are produced under
what Marx calls 'normal' conditions prevalent in society. But this is
secondary. What is of primary importance is that Steedman has
arrived at negative-value transfer and negative fixed-capital value not