The Marxian Transformation Problem
Gérard Duménil+ and Duncan Foley++
Abstract: The origins of the Marxian Transformation Problem lie in the differences between
two central abstractions of classical political economy, the labour theory of value and the
equalization of the rate of profit through competition among capitals. Marx proposed that
these two principles could be reconciled by distinguishing the production of surplus-value
through the exploitation of labour in the process of production from the realization of surplus-
value through the price system, and claimed that the equalization of the rate of profit could be
viewed as a redistribution of a given surplus-value among sectors. Marx's treatment can be
seen as a generalization of the discussion of this problem in Smith and Ricardo. Modern
criticisms of Marx's discussion claim that the labour theory of value is an unnecessary detour
to the determination of prices and profit rates, and that Marx's claims that total value and
surplus-value are conserved when prices equalize profit rates is unfounded. The Single-
System Labour Theory of Value (SS-LTV) interpretation maintains Marx’s two fundamental
claims in the following formulations: (1) the price of the net product is the expression in
prices (price form) of the total value-creating labour expanded during the period, and (2) total
profits are the price form of surplus-value, determining the value of labour-power as
unallocated purchasing power (UPP) on any set of commodities workers can buy from their
wages. These properties hold for any set of prices, not specifically prices of production. The
various positions in this controversy are illustrated in a mathematical formalization of the
circulating capital model of production.
Marx's framework: value, surplus-value, prices and competition
Marx consistently distinguishes the notions of value and price, in contrast to contemporary
economic language, which uses the term 'value' to refer to prices in a situation of general
equilibrium, though the use of the term is rather flexible; for example 'value added' is actually
the value of net product measured in price terms. For Marx, value is a 'social substance'
manifested in economic relations in the 'form' of prices, though prices are not necessarily
proportional to values as we will see.
Value and surplus-value
We first recall Marx's basic concepts (see also Marx's analysis of capitalist production).
Central to Marx's framework of analysis in Capital is the labour theory of value (LTV), which
defines the value of a commodity as the 'socially necessary' labour time required by its
+ MODEM-CNRS, Université de Paris X-Nanterre, 200 av. de la République, 92000 Nanterre, France. Email:
gerard.dumenil@u-paris10.fr.
++ New School for Social Research, New York. Email, U.S. foleyd@newschool.edu.
production, that is, the labour time required by average available techniques of production for
workers of average skill.
The LTV is central to Marx's theory of exploitation, a term he uses to describe a situation in
which one individual or group lives on the product of the labour of others. According to the
LTV, when commodities are exchanged through sale and purchase, no value is created. But
this principle does not apply to capitalists' purchase of the labour-power of workers. Workers
sell their labour-power, that is, their capability to work, to a firm, owned by a capitalist. The
buyer uses this labour-power in production to add value to the commodity produced. The
value of labour-power is the labour time required by the production of the commodities the
worker buys. But the worker can typically work more hours than are on average required to
produce this bundle of commodities. For example, the goods the worker can buy may require
8 hours of labour per day, when the labour-day lasts 12 hours. The difference, 4 hours, is
unpaid labour time. If an hour of social labour on average produces a value whose price form
is $10, 4 hours of unpaid labour time results in a surplus-value whose price form is $40,
which is appropriated by the capitalist. The rate of surplus-value is the ratio of unpaid to paid
labour time, in this case, 4/8, that is 50%.
Two laws of exchange
Marx situates his discussion in the context of the distinction made by Adam Smith and David
Ricardo between 'market prices' and 'natural prices'. Market prices are the prices at which
commodities actually exchange from day to day in the market. Smith and Ricardo, however,
regarded market prices as fluctuating (or 'gravitating') around centres of attraction they called
'natural prices'. ('Gravitation' means that the economy is in a permanent situation of
disequilibrium, though in a vicinity of equilibrium where natural prices would prevail.)
In the above analysis, Marx assumes that commodities tend to exchange at their values (at
prices proportional to values), that is, in proportion to the labour time embodied in them.
'Tend' means here that deviations are obviously possible, but that such prices will 'regulate'
the market, in the sense that if the prevailing set of prices systematically under-compensates
the labour used in the production of a commodity, labour will move to the production of
better-paid commodities. As a result, the supply of the under-compensated commodity will
decline, and its price will rise. In reality prices would gravitate around values, which would
play the role of natural prices in such an economy. This is the commodity law of exchange.
In a capitalist economy, however, capitalists buy not only the labour-power of workers (which
Marx denotes as variable capital), but also non-labour inputs, such as raw materials, and fixed
capital, such as machinery (which Marx denotes as constant capital). If natural prices were
proportional to labour inputs, as the commodity law of exchange posits, capitalists using more
constant capital per worker than the average would realize smaller profit in comparison to
their total capital advanced, that is, lower profit rates. Marx accepts the idea that competition
tends to equalize profit rates in various industries, despite differences in capital advanced per
worker, which is the capitalist law of exchange. Marx uses the term prices of production to
describe a system of prices which guarantee to the capitalists of various industries a uniform
profit rate. Capitalists will invest more where profit rates are larger, and conversely in the
symmetrical case. They move their capital from one industry to another seeking maximum
profit rates, and this movement result in a gravitation of market prices around prices of
production. Marx regards prices of production as the centres of gravitation of market prices,
and thus the natural prices relevant to a competitive capitalist economy.
Is the theory of surplus-value compatible with the theory of competition?
The problem is posed of the compatibility of the capitalist law of exchange at prices of
production with the theory of exploitation as extraction of surplus-value. Marx's line of
argument is that surplus-value is created in production through the exploitation of labour, that
is, in proportion to labour expended, but realized proportionally to total capital invested.
According to Marx, this separation between the locus of extraction and the locus of realization
does not contradict the theory of exploitation so that capitalist competition is compatible with
his theory of exploitation through the appropriation of surplus-value from unpaid labour time.
To support this argument, Marx presents a pair of tables (1981, chapter 9) showing the
redistribution of surplus-value through deviations of price from values proportional to
embodied labour times. All variables are measured in hours of labour time, and as a result
prices of production are expressed in the same unit. Because Marx's own calculations involve
some extraneous complexity (differential turnover rates among sectors), it is more useful to
consider the simplified case shown in Table 1. Two industries exist, each of which advances
the same capital of 100, but divided in different proportions between the purchase of non-
labour inputs (C) and labour inputs (V). All capital is used up during the period, so that the
rate of profit is the ratio of surplus-value to total capital advanced, r =s/(c+v). The rate of
surplus-value is uniform and equal to 100%. Consequently, surplus-values are equal to
variable capitals. Surplus-values and values are computed in each industry. When prices are
proportional to values, profit rates differ between the two sectors. Prices of production are
determined in Marx's procedure by summing up all surplus-value, a total of 40, and
redistributing it in proportion to total capital, that is 20 in each industry, to equalize profit
rates on the capitals advanced.
Table 1:
Industry
Constant
capitals,
C
Variable
capitals,
V
Total
capitals,
K=C+V
Surplus-
values,
S=V
Values of
commodities
produced,
Λ=K+S
Profits,
Π
'Prices of
production' of
commodities
produced,
P=K+Π
1 70 30 100 30 130 20 120
2 90 10 100 10 110 20 120
Total
economy 160 40 200 40 240 40 240
The procedure illustrates a straightforward 'redistribution' of surplus-value. Clearly, the sum
of prices, 240, is equal to the sum of values, and total surplus-value is, by construction,
conserved in the form of profit. These observations are expressed in two Marxian equations
concerning the entire economy:
Sum of values = Sum of prices of production
Sum of surplus-value = Sum of profits
Note that these compact formulations are not rigorous, since values and surplus-value are
measured in labour time and prices and profits in money. Thus, 'Sum of values' should read
'Sum of prices proportional to values'. A simple way out of the problem of units is to use one
of these equations to define the general level of prices. For example, the sum of prices of
production could be set equal to the number of hours corresponding to the sum of values.
Then, Marx's line of argument implies that the surpluses in both sets of prices are equal, as in
the second equation. This simple calculation illustrates the idea that profits are 'forms' of
surplus-value, that is, unpaid labour.
Approximations
Marx is, however, aware that the type of computation illustrated in Table 1 is not satisfactory,
since the evaluations of constant and variable capital have not been modified despite the fact
that prices have changed:
1) When natural prices are prices of production, non-labour inputs are purchased on the
market at prices of production, not at prices proportional to values. It is, therefore, not correct
to conserve the evaluation of constant capital:
'We had originally assumed that the cost-price of a commodity equalled the
value of the commodities consumed in its production. But for the buyer the
price of production of a specific commodity is its cost-price, and may thus
pass as cost-price into the prices of other commodities. Since the price of
production may differ from the value of a commodity, it follows that the
cost-price of a commodity containing this price of production of another
commodity may also stand above or below that portion of its total value
derived from the value of the means of production consumed by it. It is
necessary to remember this modified significance of the cost-price, and to
bear in mind that there is always the possibility of an error if the cost-price
of a commodity in any particular sphere is identified with the value of the
means of production consumed by it. Our present analysis does not
necessitate a closer examination of this point.' (III,9)
(References for quotations are to the volume and chapter in Capital. Sources can be found on
the internet, in the Marx-Engels Library: http://www.marxists.org/archive/marx/works/, or in
Marx, 1976, 1978, 1981.)
2) There is a similar problem concerning variable capital. When commodities exchange at
prices of production, workers will not be able to buy the same bundle of commodities with a
wage corresponding to a purchasing power expressed, as in Marx's calculation, as a certain
number of hours of labour time, as when prices are proportional to values. Marx is also aware
of this problem:
'[…] the average daily wage is indeed always equal to the value produced in
the number of hours the labourer must work to produce the necessities of
life. But this number of hours is in its turn obscured by the deviation of the
prices of production of the necessities of life from their values. However,
this always resolves itself to one commodity receiving too little of the
surplus-value while another receives too much, so that the deviations from
the values which are embodied in the prices of production compensate one
another. Under capitalist production, the general law acts as the prevailing
tendency only in a very complicated and approximate manner, as a never
ascertainable average of ceaseless fluctuations.' (III,9)
It is not easy to understand Marx's position from these notes (which he never revised for
publication). It does seem that the analysis requires a 'closer analysis', since the revaluation of
constant capital at prices of production will in general make the sum of prices of production
deviate from the sum of values, or make the sum of profits deviate from the sum of surplus-
values. While it is true that a redistribution of surplus-value through a system of prices of
production does not alter the living labour expended in production, so that over the whole
economy the deviations from value 'compensate one another', the value of labour-power will
remain constant only if workers consume commodities in the same proportion as they are
produced in the whole economy, which is implausible. The phrase 'average of ceaseless
fluctuations' suggests the averaging out of market prices to prices of production rather than
the averaging of surplus-value across sectors.
If Marx's use of the term 'approximately' is taken literally, it would appear that the LTV and
the theory of exploitation he introduced in Volume I of Capital are only 'approximately' true!
Although Marx is conscious of the problem, it is impossible to consider his solution as
rigorous. In the formulation of the two equations above, it appears that, when the calculation
is done rigorously as in the formal setting below, the second equation does not hold! Later
critics have judged this a devastating refutation of Marx's theories of value and exploitation,
which in turn has led to ongoing controversy.
Earlier approaches
The foundations of the transformation problem can be found in the first analyses of
competition and prices in capitalism, beginning with Adam Smith and David Ricardo, on
which Marx elaborated. The distinction between values and prices remains somewhat fuzzy in
these authors. Smith fails to establish a clear relationship between value and profit rate
equalization as the principle determining 'natural prices'. Thus, one characteristic feature of
these approaches, from which Marx was unable to depart completely, is that two sets of prices
(the two laws of exchange above) are considered, one proportional to values (embodied
labour times), and the other equalizing profit rates (a dual system), when only one price
system prevails in real-world capitalism (a single system):
1) A system of prices proportional to values (embodied labour times) plays a role in the
analyses of Smith, Ricardo and Marx. Only Marx, however, clearly distinguishes the
two systems from the start.
2) The determination of the 'surplus', when such a concept exists (as in Ricardo and
Marx), is posed in the first system and imported into the second, instead of being
analysed directly within the second system.
This dual system approach lies at the basis of the phrase 'transformation problem', which
refers to the transformation from one system into the other.
Adam Smith
Smith's point of departure is an 'early, rude' state of society, before the establishment of
private property in land and means of production. There, Smith contends, products of human
labour will exchange in proportion to the labour time required to produce them. Smith offers
as an example that if it requires two days on average to kill a beaver, but one day to kill a
deer, a beaver will tend to exchange for two deer. Smith's argument supporting this
conclusion rests on the assumption that any hunter can choose to allocate time to hunting deer
or beaver, so that if the exchange ratio were higher or lower than the labour time ratio, hunters
would shift from the under- to the over-remunerated productive activity, and force the
exchange ratio back toward the labour time ratio. The viewpoint is clearly that of the
commodity law of exchange.
Smith applies the same type of reasoning to argue that once means of production have become
private property (which he calls 'stock', and later economists called 'capital') the ability of
owners to shift their capital from one line of production to another will tend to equalize the
profit rate across different sectors of production. The viewpoint is now that of the capitalist
law of exchange.
David Ricardo
Ricardo critiques and corrects Smith's analysis. Ricardo originally based his theories of prices
and distribution on Smith's first principle that the labour expended in producing a commodity
determines its price in exchange. But Ricardo, elaborating on the dual system approach,
examines the necessary quantitative difference between the two principles that might
determine natural prices more carefully than Smith. Ricardo understood that the proportion
between capitals invested in non-labour inputs and labour is not uniform across industries,
and that this fact implies a discrepancy between the two sets of prices, but he regarded these
deviations as quantitatively limited. Prefiguring Marx's investigation, Ricardo was concerned
to work out the properties of the first system (values) to derive conclusions concerning
distribution, which he supposed were also valid in the second system (prices of production):
1) When natural prices are proportional to values (embodied labour times), it is
obvious that there is a trade-off between the shares of output which respectively go to
workers and capitalists: Workers create all the value added to inputs, and buy a share
of output whose production requires less labour time than they expend. In contrast to
Smith, Ricardo had a clear view of this mechanism. This division of total output
between workers and capitalists was crucial to his analysis, because of its implications
in terms of economic policy. (For example, Ricardo was in favour of a low price of
corn, which, in his opinion, would increase the profits of capitalists by lowering
wages—and encourage capital accumulation).
2) Ricardo would have liked to conserve the straightforward distributional properties
he derived from the assumption of prices proportional to values, even while
acknowledging the quantitative difference between such natural prices proportional to
values and natural prices that would equalize profit rates across industries. But
Ricardo understood that, in the profit rate-equalizing system, the natural prices of
commodities may change with a change in the real wage (due to the distinct
compositions of capital) even if the labour required in production remains unaltered,
contrary to what happens in the first system, where values remain unchanged with a
change in the wage. Thus, with Ricardo's analysis, we are getting closer to Marx's
framework and problems.
The rebellious classical