Journal of Economic Behavior and Organization 212 (2023) 887–896

Contents lists available at ScienceDirect

Journal of Economic Behavior and Organization

journal homepage: www.elsevier.com/locate/jebo

Marx’s concept of “exploitation” and the problem of

measurement ✩
Heinz D. Kurz a, Neri Salvadori b,∗
a
Emeritus Professor at the University of Graz, Austria
b
Dipartimento di Economia e Management, Emeritus Professor at the University of Pisa, Via Ridolfi 10, Pisa 56123, Italy

a r t i c l e i n f o a b s t r a c t

Article history: The paper scrutinizes Marx’s proposal to measure the “degree of exploitation” in terms
Received 2 October 2022 of the “rate of surplus value”. This amounts to defining an equivalent to the device of the
Revised 16 April 2023
thermometer in measuring temperature. The condition to be met is that a concept must be
Accepted 12 June 2023
suggested that allows one to express the degree of exploitation in an independent manner
and to show that the two move in parallel with one another. It is then shown that the
Keywords: rate of surplus value does not satisfy the sought criterion, because it may move in opposite
Choice of technique direction of the degree of exploitation. This is demonstrated in terms of a simple numerical
Degree of exploitation example with a choice of technique concerning the lengthening of the working day.
Karl Marx © 2023 Elsevier B.V. All rights reserved.
Measurement in empirics and in theory
Rate of surplus value
Piero Sraffa

1. Introduction

Peter Flaschel was deeply fascinated by Karl Marx’s contribution to political economy and throughout his academic career
scrutinized it and elaborated on it. As he and his co-authors wrote in Peter’s last book, Value, Competition and Exploitation,
which contains a kind of legacy of his research program, Marxian value theory provides “an arguably deeper analytical
framework than current mainstream theory” (Cogliano et al., 2018: xvi). A central concept in Marx’s analysis is that of
“exploitation”, and this concept assumes centre stage in Part II of the book entitled “Value and Exploitation: Marx’s Legacy”.
In defining the concept, the labour theory of value has always played an important part. It has been under attack ever
since Marx put forward what he called the “law of value”; more recently, employing the analytical tools provided by Sraffa
(1960), it was criticised anew and with it also the concept of “exploitation” (see especially Steedman 1977). Flaschel et al.
reject the view that the concept of exploitation stands or falls with the labour theory of value. They do so by defining
labour values differently from the conventional definition, which interprets them as employment multipliers in standard
input output models, and they also insist that there is no “transformation problem” to be solved, starting from Marx’s
labour theory of value. Since, as they maintain, “there is no obvious definition of labour values in economies with joint
production, fixed capital, and so on “ (ibid: 379), they use the degree of analytical freedom by interpreting labour values

✩
We have written this paper in memory of Peter Flaschel, a good friend and highly esteemed colleague, with whom we spent many hours of fruitful
discussions. Roberto Veneziani has kindly read an earlier version of this paper and given us useful information about the literature that is pertinent to
our theme, and Ian Steedman has sent us valuable comments on the paper. We are most grateful to an anonymous referee of JEBO for several useful
observations on the paper and suggestions how to improve it.
∗
Corresponding author.
E-mail address: neri.salvadori@unipi.it (N. Salvadori).

https://doi.org/10.1016/j.jebo.2023.06.016
0167-2681/© 2023 Elsevier B.V. All rights reserved.
H.D. Kurz and N. Salvadori Journal of Economic Behavior and Organization 212 (2023) 887–896

“as indices of the real direct and indirect cost of producing commodities measured in labour units: labour content is the
average labour time ‘embodied’ in a good, in the sense of full-cost accounting in terms of labour time spent on average in
the production of commodities” (ibid). They state that their interpretation is based on a “rigorous reading of Marx’s texts”
and has labour values that are “well defined, logically consistent, nonnegative … and satisfy several theoretically desirable
properties. Labour values are not a logically incoherent construct” (ibid).
In this short paper we will bypass the issue of the relationship between labour values and the concept of exploitation and
focus attention on Marx’s quantitative expression of the concept of exploitation and his recurrent equalization/identification
of the “degree of exploitation” with the “rate of surplus-value”.1 The thrust of the argument consists in the following: Up
until now, important issues in Capital such as the struggle over the length of the working day have almost exclusively been
discussed by assuming implicitly that there is no choice of technique available to capitalists and that therefore the technique
employed does not change if, for example, the length of the working day is increased. However, in case there is a choice of
technique available to cost-minimizing capitalists, they may be led to adopt a technique different from the one employed
before the increase in the length of the working day. (This is analogous to the case of a change in the real wage rate, which,
as Marx made very clear, may result in the adoption of a different technique.) Taking into account this obvious possibility
affects the results and, as will be shown, questions the identification of the concept of exploitation with the rate of surplus-
value. One ought to add in this context that Marx is keen to root the analytical concepts he elaborates in socioeconomic
history and stresses the correspondence between his theory and actual facts. In the historical chapter on “The Struggle for
the Normal Working-Day”, which covers many decades, he for obvious reasons never presents the issue on the premise
that capitalists cannot choose the technique they consider cost-minimizing. In fact, his respective observations are peppered
with remarks about how a change in the length of the working-day, for example, may impact on the social organization of
the labour process and the process of production operated by workers. Marx was clearly no advocate of a narrow partial
analysis and his concern with the long-term “law of motion” of capitalism defied the use of such an analysis. In capitalism,
he insisted, the dominant economic motive is the profit motive, and it would make little sense to assume that this motive
is prevented from developing its full potential by disallowing the adoption of cost-minimizing methods from the set of all
alternatives at a given point in time or from the changing set as a consequence of inventions over time. Marx famously
wrote:
No capitalist ever voluntarily introduces a new method of production, no matter how much more productive it may
be, and how much it may increase the rate of surplus-value, so long as it reduces the rate of profit. (Marx 1959: 264)

Therefore, our reasoning in this paper may be said to be fully in accordance with Marx’s general approach to the problem
under consideration and not the partial approach, which artificially constrains the space for manoeuvre of capitalists. While
in given conditions a partial analysis may give rise to the impression that the rate of exploitation is bound to increase with
an increase of the working-day, a more general analysis that takes into account the said space for manoeuvre may show
the opposite, even if the real daily wage is held constant. Hence, in ordner not to be misled by the narrowness of the
perspective assumed, one has to take into account the choice of technique options available to capitalists at any moment of
time and over time.2
This paper consists of an elaboration of a paper written in Italian by one of us in 1977, when he was a visiting student
in Manchester and submitted it to his supervisor, Ian Steedman, at that time. Ian Steedman liked the issue and asked
permission to insert what he considered the analytical result of that paper in a chapter entitled “Miscellanea” of his book
(see Steedman 1977: 105–109, and especially footnote 17 on page 109). Despite the fact that a large part of the analytical
issue dealt with here has been available to scholars since then, the implication of this analysis for the validity or otherwise
of the concept of exploitation appears to have been overlooked in many studies on exploitation published since then (see,
for instance, Desai 1979 and Roemer 1981). A reason for this might be that Steedman’s treatment of the issue did not include
any historical consideration concerning the role of the rate of exploitation as a measure. This problem was, however, very
important to Marx, as will be shown in this note.
The composition of the note is the following. In Section 2 we provide a selection of passages from Marx’s Capital, which
document his conviction that the rate of surplus-value is to be considered an independent measure of the degree of ex-
ploitation. In Section 3 we briefly recall Piero Sraffa’s concern with the problem of measurement in economic theory in
his intervention at the famous Corfu conference dedicated to the theory of capital and income distribution and then in his
preparatory notes that led up to the publication of his 1960 book. Section 4 turns to the meaning of Marx’s concept of
“exploitation”: it draws the attention to the problem of measuring exploitation, which amounts to defining an equivalent to
the device of the thermometer in measuring temperature. The condition to be met is that a concept must be proposed that
allows one to express the “degree of exploitation” in an independent manner. Is the concept of the “rate of surplus-value”
the sought autonomous concept that satisfies the requirement? Marx appears to have thought so and many of his followers
still share this view. Section 5 shows that the rate of surplus-value does not meet the requirement, because the degree of

1
For a thorough analysis concerning the measurement of the quantity of labour bestowed upon a commodity vis-à-vis the heterogeneity of labour, see
the paper by Naoki Yoshihara and Roberto Veneziani (2022) in this Gedenkschrift for Peter Flaschel.
2
Marx’s view that the organic composition of capital will rise as a consequence of the particular form of technical progress that according to him is
congenial to the capitalist mode of production was, of course, based, as the passage cited in the above alludes to, on an analysis of the choice of technique
problem. (As Piero Sraffa, Nobuo Okishio and others have shown, his respective analysis cannot be sustained.)

888
H.D. Kurz and N. Salvadori Journal of Economic Behavior and Organization 212 (2023) 887–896

exploitation and the rate of surplus value may move in opposite directions. This is illustrated in terms of a simple numerical
example relating to an economic system that produces a single commodity only (but see also the presentation by Steed-
man1977: 105–109). The case discussed is that of a lengthening of the working day, ceteris paribus. It is shown that under
certain conditions, which are by no means extraordinary or special, what may be called “perverse” behaviour of the two
magnitudes may occur. The phenomenon is then analysed more deeply and the conditions of its occurrence are specified.
Next, the investigation is extended to the case with n commodities. Section 6 offers some concluding remarks and draws
the implications for Marx’s concept of “exploitation”.

2. Marx on exploitation and its measurement

The concept of “exploitation” in Marx is well known. He speaks of exploitation whenever a part of the working time of
an individual worker materializes directly or indirectly in products that cannot be used or consumed by him- or herself or
other persons with whom he or she is deliberately prepared to share them, but are used or consumed by a third party:
If the labourer wants all his time to produce the necessary means of subsistence for himself and his race, he has no
time left in which to work gratis for others. Without a certain degree of productiveness in his labour, he has no such
superfluous time at his disposal; without such superfluous time, no surplus labour, and therefore no capitalists, no
slave-owners, no feudal lords, in one word, no class of large proprietors (Marx 1967: 511).

Exploitation means that only a fraction of the working population produces what is consumed by the working population
as a whole, with the rest of it producing commodities for the remaining part of society that does not contribute itself to
production:
Just as the individual labourer can do more surplus labour in proportion as his necessary labour-time is less, so with
regard to the working population. The smaller the part of it which is required for the production of the necessary
means of subsistence, so much the greater is the part that can be set to do other work (Marx 1967: 513).

Here, Marx may be said to refer to what was later called the “integrated wage-goods industry”, a concept to which we
will come back below.
It is also well known that in Volume I of Capital Marx had set himself the task to clarify, how capitalist society, just
like previous societies, was also based on the exploitation of humans by humans. However, whereas in previous societies
the exploitative relation was explicitly enacted in law, in capitalism, Marx insisted, it was not, but resulted in spite of, and
indeed through, the exchange of equivalents that characterises the world of commodities.
What interests us here is that Marx gives the concept of exploitation a quantitative expression and also introduces the
concept of the degree of exploitation which he considers to be reflected in, and actually equal to, the rate of surplus labour.3
It suffices to draw the readers’ attention to three passages in Volume I of Capital, which express clearly Marx’s concern with
the quantitative aspect of exploitation and, hence, the concept of the “degree of exploitation”.

1. Marx wrote:

Since, on the one hand, the values of the variable capital and of the labour-power purchased by that capital are
equal, and the value of this labour-power determines the necessary portion of the working day; and since, on the
other hand, the surplus-value is determined by the surplus portion of the working day, it follows that surplus-value
bears the same ratio to variable capital, that surplus labour does to necessary labour, or in other words, the rate
of surplus-value, s/v = (surplus labour)/(necessary labour). Both ratios, s/v and (surplus labour)/(necessary labour),
express the same thing in different ways; in the one case by reference to materialised, incorporated labour, in the
other by reference to living, fluent labour.

The rate of surplus-value is therefore an exact expression for the degree of exploitation of labour-power by capital, or of
the labourer by the capitalist. ([1867] Marx, 1967: 217–8; emphasis added)4

3
This quantitative aspect is frequently referred to in the secondary literature on Marxian economics. It is, however, surprising that the problem of
measurement raised in this note is hardly ever mentioned, let alone analysed. It is typically taken for granted that the rate of exploitation is equal to the
rate of surplus value and that the two magnitudes move in step with one another, whatever the changes the economic system experiences. Here it suffices
to refer to Wolfstetter (1973), Nuti (1977), Desai (1979) and Roemer (1981). In the case of the so-called “Single-System Labour Theory of Value” (see, for
example, Duménil and Foley 2008), the requirement that magnitudes must be measured by some other, independently ascertained magnitude of the same
nature, is removed by assumption.
4
Desai (1979: chap. VII) distinguishes between several measures of the rate of surplus-value in Marx, which in equilibrium are said to be all equivalent.
Desai himself appears to share Marx’s above view, which becomes clear when he uses the formulation “the rate of surplus value (or the rate of exploitation)
…” (ibid: 47).

889
H.D. Kurz and N. Salvadori Journal of Economic Behavior and Organization 212 (2023) 887–896

And in a note he added:

Although the rate of surplus-value is an exact expression for the degree of exploitation of labour-power, it is, in no sense,
an expression for the absolute amount of exploitation. For example, if the necessary labour = 5 hours and the surplus
labour = 5 hours, the degree of exploitation is 100%. The amount of exploitation is here measured by 5 hours. If, on
the other hand, the necessary labour = 6 hours and the surplus labour = 6 hours, the degree of exploitation remains,
as before, 100%, while the actual amount of exploitation has increased 20%, namely from five hours to six. (Ibid:
218–19; emphasis added)

2. In the chapter on “Different formulae of the surplus-value ratio”, after having recalled some “classical political economy”
formulae of the s/(v + s) type, which differ from his s/v formula, Marx added:

In all of these formulae (II.), the actual degree of exploitation of labour, or the rate of surplus-value, is falsely ex-
pressed. Let the working day be 12 hours. Then, making the same assumptions as in former instances, the real degree
of exploitation of labour will be represented in the following proportions.

6 hours surplus − labour Surplus − value of 3 sh.

= = 100%
6 hours necessary labour Variable Capital of 3 sh.
From formulae II. we get very differently,
6 hours surplus − labour Surplus − value of 3 sh.
= = 50%
Working day of 12 hours Value created of 6 sh.
These derivative formulae express, in reality, only the proportion in which the working day, or the value produced
by it, is divided between capitalist and labourer. If they are to be treated as direct expressions of the degree of self-
expansion of capital, the following erroneous law would hold good: Surplus-labour or surplus-value can never reach
100%. (Ibid: 531–2)

This confirms that Marx was concerned with the problem of the measurement of exploitation.

3. Marx stated:

It will be remembered that the rate of surplus-value depends, in the first place, on the degree of exploitation of
labour-power. Political Economy values this fact so highly, that it occasionally identifies the acceleration of accumu-
lation due to increased productiveness of labour, with its acceleration due to increased exploitation of the labourer.
(Ibid: 599)5

We now turn to Sraffa’s view of the problem of measurement in economic theory, on the one hand, and in empirical
studies, on the other.

3. Sraffa on measurement in empirics and in theory

In one of his interventions at the Corfu conference on “The theory of capital” in September 1958, organized by the Inter-
national Economic Association (Lutz/Hague 1961: 305–306), Sraffa is recorded of having insisted that “one should emphasize
the distinction between two types of measurement. First, there was the one in which the statisticians were mainly inter-
ested. Second, there was measurement in theory. The statisticians’ measures were only approximate and provided a suitable
field for work in solving index number problems. The theoretical measures required absolute precision. Any imperfections
in these theoretical measures were not merely upsetting, but knocked down the whole theoretical basis.”
He added with explicit reference to the theory of capital:
The definition in this case must be absolutely watertight, for with a given quantity of capital one had a certain rate
of interest so that the quantity of capital was an essential part of the mechanism. One therefore had to keep the
definition of capital separate from the needs of statistical measurement, which were quite different. The work of J.
B. Clark, Böhm-Bawerk and others was intended to produce pure definitions of capital, as required by their theories,
not as a guide to actual measurement. If we found contradictions, then these pointed to defects in the theory, and an

5
The referee perceptively observed in his comments that in this passage Marx does not say that the rate of surplus-value is ”an exact expression for
the degree of exploitation”, but that “the rate of surplus–value depends, in the first place [in German: ‘in erster Instanz’], on the degree of exploitation”.
Does this mean that Marx now distances himself from his earlier statement? We do not think so. The passage is to be found in the context of a criticism
of certain views of political economists (he refers explicitly to David Ricardo and John Stuart Mill), who are said to have confounded the acceleration
of accumulation and the increase in labour productivity. They have overlooked, Marx surmises, the fact that an increase in the pace of accumulation is
repeatedly brought about by a reduction of the wages of labour below the value of labour power, which implies, of course, an increase in the rate of
exploitation. Since Marx also does not withdraw his former formulation as misleading or incomplete, we retain our interpretation. (Incidentally, Desai
(1979: 48) maintains that Marx assumed “that wages never fall below the exchange value of labour power and occasionally rise above it”. However, in the
passage in Capital referred to here, Marx takes the reduction of wages below it to be not infrequent.)

890
H.D. Kurz and N. Salvadori Journal of Economic Behavior and Organization 212 (2023) 887–896

inability to define measures of capital accurately. It was on this – the chief failing of capital theory – that we should
concentrate rather than on problems of measurement.

Students of Sraffa’s papers kept at Trinity College, Cambridge (U.K.), will know that the problem of measurement in
theory as opposed to that in statistics permeates much of his notes and manuscripts. Here it must suffice to provide just
two examples. They concern, first, his re-constructive work of the theory of value and distribution of the classical economists
and, secondly, his criticism of the marginal theory of capital and interest (profits). Important insights into both matters he
had put forward as early as the late 1920s.
In late 1927 and in 1928, Sraffa reformulated the classical theory of value and distribution, first in terms of a system
without a surplus product in what he called his “First Equations”, and then in terms of a system with a surplus product in
his “Second Equations”. In the latter he assumed given real wages paid ante factum and a uniform rate of interest (profits)
on the values of capital employed in the various industries of the economy. He then began to investigate the mathematical
properties of the system more closely by reformulating problems Ricardo had tried to analyse within his own analytical
framework more than a century before him, such as the impact of a change in wages on the rate of profits and relative
prices in what he called his “Third equations”. In this context he at one point took up an idea echoing an example in
Ricardo’s Principles (Works I: 50 and 64–6), in which gross outputs of the various commodities are proportional to the
aggregate amounts of these commodities consumed productively as inputs in the various industries. The upshot of Sraffa’s
respective argument is summarised in a document dated February 1931, in which he stated: “it may be said that the value
of total capital in terms of total goods produced cannot vary [consequent upon a change in the rate of interest (profits)],
since the goods are composed exactly in the same proportions as the capitals which have produced them”.
That is, the capital-to-output ratio is taken to be constant by assumption. He observed immediately that actual economies
will not exhibit this property and that the proposition therefore is “false”, but “may contain an element of truth” (D3/12/7:
157(3)).6
Twelve years later, in November 1943, he interpreted this statement as referring to the “statistical compensation of large
numbers” (D3/12/35: 28) in systems in which production is conceived as a circular flow. The idea recurred variously in his
papers under the name “My Hypothesis” or simply “Hypothesis”, before Sraffa succeeded in elaborating the concept of the
Standard commodity in early 1944.7 The corresponding Standard system, Sraffa emphasised, provides “tangible evidence of
the rate of profits as a non-price phenomenon” (D3/12/43: 4). This confirms Ricardo’s conjecture that “the great questions
of Rent, Wages, and Profits … are not essentially connected with the doctrine of value” (Ricardo, 1951, Works VIII: 194).
Measuring wages and prices in terms of the Standard commodity reveals properties of the economic system, which any
conventional measurement of the magnitudes belonging to the actual system would not. Sraffa thus arrived at the “ideal”
measure – ideal for his purpose – by starting from a numerical example in Ricardo, which he then interpreted as reflecting
a fact, which in actual economies would at best only be roughly true, and then translated this fact into an analytical device
to study the actual system, which does not have the properties of Ricardo’s example, from the point of view of the Standard
system. The latter provides a point of observation of the actual system by hypothetically changing its proportions without,
however, changing any of its mathematical properties. Assuming this point of observation reveals at once what otherwise
would be hidden and could be seen only by following a more complicated research strategy. The measure Sraffa elaborated
met the criterion of “absolute precision” he referred to in Corfu.
The second example concerns his criticism of the marginal theory of capital and interest (profits). The attempt to explain
the competitive rate of profits in terms of the relative scarcity of a factor, called “capital”, reflected in its “marginal pro-
ductivity”, requires a measure of the quantity of capital that is independent of the variable that is to be determined – the
general rate of profits. This Sraffa also pointed out in the late 1920s with regard to the contributions of major marginalist
authors, such as Eugen von Böhm-Bawerk, Knut Wicksell or John Bates Clark. Böhm-Bawerk had elaborated a concept of
the quantity of capital in terms of the “average period of production” which, he surmised, could be ascertained without
pre-empting the issue of the determination of the rate of profits. However, as Sraffa stressed in a note dated 8 July 1928,
this was generally not possible:
The length of the period of production is not a purely physical (objective) fact, which can be measured by a clock;
and which is independent from the way in which, after it is completed, the product is going to be divided between
workers and capitalists.

The length of the period is strictly dependent upon the way in which the product (of “last year”) is divided between
wages and profits. (D3/12/7: 90)

In fact, the measure implied a given distribution of the product and could therefore not be used to determine this
division. It contradicted a major requirement of measurement, which the French engineer and socialist Robert Louzon (1882–
1976), to whom Sraffa referred in the given context, had expressed in the following way: “Une grandeur ne peut en effet
être mesuré que par une grandeur de même nature. (A magnitude cannot be measured except by a magnitude of the same

6
Sraffa’s papers are cited according to the catalogue prepared by Jonathan Smith, archivist, Wren Library, Trinity College, Cambridge.
7
On the concept of the Standard commodity in single-product systems, see Sraffa (1960: chap. IV) and Kurz and Salvadori (1995: 116-118).

891
H.D. Kurz and N. Salvadori Journal of Economic Behavior and Organization 212 (2023) 887–896

nature.)” (D3/12/9: 58) As an example, Louzon mentioned mercury to measure temperature. And in the summer of 1929
Sraffa stressed:
In order to have a marginal theory of distribution (or, in fact, any such theory) we must have a physical measure of
the quantity of each factor, independent (i.e. that can be carried out without knowing …) of its share in distribution.
[…] This physical measure implies that each factor is perfectly homogeneous with itself, i.e. all its composing particles
are identical to one another; this requirement does not extend to all their possible qualities (if it did, there would be
absolute identity between the particles, also as to position in space and time, i.e. there would be only one particle)
but only: a) to the quality that is taken as the basis of the physical measurement, and b) to the quality-ies [sic] which
compose its productiveness (these are unknown but must always be proportional to those of the phys. measurement:
i.e. they must be the same!) (D3/12/13: 17(5))

These considerations are echoed in Sraffa (1960: 38), when he insists that the “period of production” does not provide
“an independent measure of the quantity of capital which could be used, without arguing in a circle, for the determination
of prices and of the shares in distribution.”
After these prolegomena, we now turn to the problem of measurement in the case of Marx’s concept of degree of ex-
ploitation in terms of the rate of surplus-value.

4. The problem of measurement in the Marxian case

In order to be able to say that the height of a mercury column is a measure of temperature, we would need to consider
the quantitative effects of variations of temperature on the height of the mercury column. Similarly, in order to be able to
say that the rate of surplus-value is a measure of the degree of exploitation, we would have to consider the quantitative
effects of changes in the degree of exploitation on the rate of surplus-value. Alas, this requirement is neither explicitly
mentioned in Capital nor in the recent literature on Marxian economics. At first glance it may even seem that the involved
procedure is ruled out by definition, since we would have no other way of saying that the degree of exploitation in one
society is greater than in another one than by comparing the respective rates of surplus-value.
If this happened to be the case, the concept of the degree of exploitation would have no meaning that is independent of
that of the rate of surplus-value, and taking one as a measure of the other would be meaningless: it would be a metaphysical
operation, not susceptible to falsification or verification. If this were the case, Marx’s identification of the rate of surplus-
value with the degree of exploitation would only have an ideological, but no scientific meaning: it may arouse certain ideas
in the readers’ minds and excite their fantasy, but it would not express the phenomenon in another independent way.
However, Marx suggests how to deal with the issue at hand, at least in some particular cases. The concept of the rate of
surplus-value is in fact a scientific entity, whereas the degree of exploitation is meant to describe an aspect of reality and
as such must find a counterpart or correspondence in reality, and it is here that Marx goes to look for it.8 He seeks to show
that its measure, the rate of surplus-value, is a powerful tool by means of which one is able to analyse and explain specific
historical realities that have characterised capitalism since its beginnings. These include the continuous struggle to limit the
length of the working day against the capitalists’ attempts to lengthen it (viz. the “creation of absolute surplus value”) and
increase the intensity of labour by revolutionizing the means of production (viz. the “creation of relative surplus value”).9
It certainly corresponds to the heuristic meaning of the concept of the degree of exploitation that if exploitation exists,
the increase of the working day, given the real wage expressed in commodity terms, implies an increase in the degree of
exploitation. Furthermore, if we insist that exploitation must be quantifiable, then struggles between those exploited and
exploiters are typically struggles to decrease or increase exploitation. Marx realizes this and makes it clear in one of the
finest historical chapters of Capital, Chapter X, in which the reader is taught that “capital is dead labour, that, vampire-like,
only lives by sucking living labour, and lives the more, the more labour it sucks” (Marx [1867] 1967: 233) and how the
lengthening of the working day implies an increase in the rate of surplus-value.10
More to the point is the consideration that the degree of exploitation increases when there is technical progress, every-
thing else being equal, that is, new techniques previously unknown are introduced. Marx, however, had to include in his
theoretical framework the struggles between workers and machines that had characterized the entire period of the estab-
lishment of capitalism, or to use Marx’s words, the transition from formal to real subsumption of the worker to capital.
In another of the historical chapters on capitalism, Chapter XV, Marx shows how technical progress increases the power
of those who manage the process of production, that is, the class of capitalists. Far from undermining the concept of the
degree of exploitation, this fact is seen to be in perfect harmony with it. In Chapter XV, Marx shows how technical progress

8
As such there is also an intuitive, heuristic concept attached to it, and those who use it in a scientific context assume, consciously or unconsciously, to
formalise and specify this heuristic concept.
9
Marx wrote: “The establishment of a normal working day is the result of centuries of struggle between capitalist and labourer” (Marx [1867] 1967:
270). And: “The creation of a normal working day is, therefore, the product of a protracted civil war, more or less dissembled, between the capitalist class
and the working-class. As the contest takes place in the arena of modern industry, it first breaks out in the home of that industry – England” (Marx [1867]
1967: 299).
10
“But in the capitalist the greed for surplus labour appears in the straining after an unlimited extension of the working day, in the Boyard more simply
in a direct hunting after days of corvée.“ (Marx [1867] 1967: 237)

892
H.D. Kurz and N. Salvadori Journal of Economic Behavior and Organization 212 (2023) 887–896

and any form of reorganization of production (co-operation, manufacturing, machinery and large industries) leads to an in-
crease in the rate of surplus-value and thus in the degree of exploitation of labour power, provided it affects industries that
produce commodities that enter directly or indirectly into real wages.11
Following Marx, we can therefore say that there are at least two cases – the lengthening of the working day (the pro-
duction of absolute surplus value) and innovations (the production of relative surplus value) – in which it is possible to say
a priori, without the need to calculate the rate of surplus-value, that the degree of exploitation increases.
In terms of the above metaphor of the thermometer, we do not need such a thermometer to say that boiling water is
warmer than ice. Indeed, if the height of a mercury column, passing from boiling water to ice would not behave in an
unequivocal manner, we would be forced to consider the statement that the height of a mercury column can be taken as
an index of temperature as false.
Similarly, if we compare two economies that employ the same technology in terms of hours worked and the same real
wage rate per day, except that in one of them the working day is longer than in the other (and therefore also the wage rate
per hour is different). In this case we can say that in the economy with the longer working day, the degree of exploitation
is larger, without the need to measure surplus value. If we compare two economies that are equal to one another in every
respect except that in one of them the working day is longer than in the other, we can say that in the economy with a
longer working day the degree of exploitation is larger, without the need to measure the surplus value.12 Indeed, if the
surplus value were higher in the economy with a shorter working day, we would be forced to consider Marx’s idea that the
degree of exploitation is measured by the magnitude (or rate) of the surplus value as false. In short, there is a criterion for
falsifying, or verifying, Marx’s respective idea.
In the next section we will show in terms of a numerical example that whenever there is more than one technique to
produce one or several commodities, it is possible that an increase in the working day may be associated with a decrease
in the rate of surplus-value, even if the commodities are exchanged at their (labour) values.13 We will then ascertain, when
such “perverse” movements occur. Then we will analyse the effects of technical progress on the rate of surplus-value and
draw the attention once again to the possibility of “perverse” movements.
The reader might be of the opinion that the comparison should not only be in terms of the same set of available tech-
niques, but also in terms of the same actually chosen technique. In this case Marx’s position would, of course, be confirmed.
But exploring Chapter X of Capital, we see that the struggles concerning the length of the working day were fought with
no regard to the power of capitalists to choose the best technique amongst all those available to them, that is the cost-
minimizing one: the choice-of-technique problem was no integral part of the struggle over the length of the working day.

5. An example

Let us consider an economy in which there is only one commodity (or any number of commodities all produced with the
same “organic composition of capital”, so that their exchange ratios are proportional to the quantities of labour bestowed
upon them). Call the commodity “corn”. There are two different techniques to produce corn by means of (homogeneous)
labour and corn. Wages are paid at the end of the production period.

5.1. The example

Assume that workers’ daily wages equal 1/14 units of corn, the working day is 8 hours long, and the two techniques are
tabulated in the following Table 1, where labour is measured in days of work.

11
“The object of all development of the productiveness of labour, within the limits of capitalist production, is to shorten that part of the working day,
during which the workman must labour for his own benefit, and by that very shortening, to lengthen the other part of the day, during which he is at
liberty to work gratis for the capitalist.” (Marx [1867] 1967: 321) “Like every other increase in the productiveness of labour, machinery is intended to
cheapen commodities, and, by shortening that portion of the working day, in which the labourer works for himself, to lengthen the other portion that he
gives, without an equivalent, to the capitalist. In short, it is a means for producing surplus-value” (Marx [1867] 1967: 371).
12
Marx is actually more precise than this and in fact stresses: “When the working day is prolonged, the price of labour-power may fall below its value,
although that price be nominally unchanged or even rise. The value of a day’s labour-power is, as will be remembered, estimated from its normal average
duration, or from the normal duration of life among the labourers, and from corresponding normal transformations of organised bodily matter into motion,
in conformity with the nature of man” (Marx [1867] 1967: 527). These considerations by Marx strengthen, rather than weaken, our argument. Readers are
well advised to read carefully a footnote Marx appended in the given place, in which he quotes Grove (1846), who had stated: “The amount of labour which
a man had undergone in the course of 24 hours might be approximately arrived at by an examination of the chemical changes which had taken place in
his body, changed forms in matter indicating the anterior exercise of dynamic force.” William Robert Grove (1811-1896), a Welshman, was, amongst other
things, a physicist, who gained considerable recognition in the field. His On the Correlation of Physical Forces, which Marx cited, contains an early exposition
of the principle of the conservation of energy. The book appeared one year prior to Über die Erhaltung der Kraft (On the Conservation of Force) of the
German physicist Hermann von Helmholtz (1821-1894), who had independently discovered the principle. It is interesting to note that both Marx and then
Sraffa at the beginning of the 20th century were keen to stay abreast with what was happening in physics and chemistry. However, they drew different
conclusions from their respective studies, as is explained in Kurz and Salvadori (2005).
13
Hence the kind of criticism put forward here is entirely independent of the (in)famous problem of the so-called “transformation” of labour values in
prices of production. We take this opportunity to point out that the transformation problem in Marx resulted from his identification of the “substance of
value” and “abstract labour” and his attempt to establish the validity of the “law of value” in aggregate terms, that is, the conservation of the mass of
value in the transformation process. For a demonstration that, starting from a physical real costs approach as the classical economists had proposed it, and
elaborating on it coherently as Sraffa (1960) did, there simply is no transformation problem. Marx may be said to have struggled in vain with a problem
that he could well have avoided. See Gehrke and Kurz (2018).

893
H.D. Kurz and N. Salvadori Journal of Economic Behavior and Organization 212 (2023) 887–896

Table 1
Technical alternatives from which cost-minimizing
producers can choose.

corn labour corn

technique α 4/5 1 → 1
technique β 20/27 2 → 1

Table 2
Technical alternatives when the working day is longer.

corn labour corn

technique α  4/5 1 × 8/10=8/10 1

technique β  20/27 2 × 8/10=8/5 1

It is easy to see that capitalists will choose technique α , which yields a profit rate of about 16.17%, while technique β
yields a profit rate of about 15.7% only. The rate of surplus-value corresponding to technique α turns out to be
 
1− 4
5
1
− 14 λ
= 1.8
1
14
λ
where λ is the amount of labour embodied in one bushel of corn.
Consider now another economy that is identical to the previous one in all respects, except that workers work for 10 hours
a day. Then the two techniques are tabulated in Table 2.
It is easy to see that capitalists will now choose technique β  that yields a profit rate of about 19.5%, while technique α 
yields a profit rate of about 17.8% only. The rate of surplus value will now be:
 
1− 20 8 1
27 − 5 14 λ 274
= = 1.26851851... < 1.8.
5 14 λ
8 1 216

5.2. Discussion

The reader might think that the above example is very special and indeed peculiar. But this would be a misconception.
In order to see this, it suffices to show that in an economy with only a single commodity (corn), if there is an increase in
the working day that is sufficiently small relatively to the status quo, then for each pair of techniques there is a range of
levels of the real wage such that in the economy with the longer working day the value of corn is smaller.
Consider Fig. 1, where w is the wage rate and r is the profit rate. The two straight lines Rα Wα and Rβ Wβ are the w-r
relationships corresponding to two techniques, which we will call α and β . Suppose there is an increase in the working day
and call z the ratio of the longer to the shorter of the two working days. Each w-r relationship will move in such a way that
for each profit rate the corresponding wage rate is z times the wage rate prior to the increase in the working day. In Fig. 1,
these relationships rotate from Rα Wα and Rβ Wβ to Rα Wα  and Rβ Wβ  . Accordingly, for a given rate of profit the real wage
rate will be higher or, put alternatively, for a given real wage rate the rate of profit will be higher. The switch-point from

Fig. 1. The w-r relationships.

894
H.D. Kurz and N. Salvadori Journal of Economic Behavior and Organization 212 (2023) 887–896

one technique to the other shifts from (w1 , r1 ) to (w2 , r2 ), where w2 = zw1 and r2 = r1 . We do not yet have an equation
to express the rate of surplus-value, but it can easily be seen14 that it is equal to: y−w
w , where y is the per capita income in
terms of corn which in this case is equal, for each technique, to the maximum wage rate (W). Focus the attention now on
Wβ −w
the open range (w1 , w2 ). In this range the rate of surplus-value equals w in the economy with the shorter working day
Wα  −w
and w in the economy with the longer working day. Now it is easy to see that if z is such that Wα  = zWα < Wβ , that
Wβ
is, if 1 < z < Wα , then the rate of surplus-value is lower in the economy with the longer working day.

5.3. Technical progress

The same example can be used to deal with the other case considered by Marx, that is technical progress. Indeed, let
us assume that only technique β  exists and that the workers’ daily wages equal 1/14 units of corn. It is easily checked
that when technique α  is introduced as a consequence of technical progress, the rate of profits goes up, so that the cost-
minimising technique is technique α  , but the rate of surplus value goes down. This analysis is also useful, since it is not a
static comparison between two different circumstances. Now, the choice of technique is a part of the argument and cannot
be eluded or questioned.

5.4. Beyond the one commodity world

Before extending the argument in a more general model, consider a system of production and therefore a technique with
n industries. Assume that the n industries exist in such proportions relative to one another that the economy produces a net
physical product that consists precisely of that basket of commodities that workers actually consume in the real economy,
neither more nor less. The reference is obviously to a particular sub-system, which in the respective economics literature is
known as the “integrated wage commodity industry”15 . Its defining feature is that its net physical product consists of wage
commodities exactly as the actual wage itself, so that if y is the net physical product per capita measured in terms of the
number of baskets of the “wage (composite) commodity” and w is the wage rate also expressed in terms of this measure,
then the rate of surplus-value is equal to y−w
w .
16

Consider now a pair of techniques with n and m commodities, respectively, which enter directly or indirectly into the
production of the wage commodity. Let s be the number of commodities common to the two techniques (so s – 1 processes
will be common to the two techniques). It is known (see Bharadwaj 1970) that in this case there are up to m + n – s
switch-points between the two techniques. If h is one half of m + n – s, in case m + n – s is even (or h is one half of
m + n – s + 1, in case it is odd), then there are up to h switch-points, such that on the right-hand side of each of these
switch-points there is a technique that has a higher maximum wage rate (see again Bharadwaj 1970).
Therefore, for each pair of techniques and for a sufficiently small increase in the length of the working day there are up
to h intervals of levels of w, such that for every w belonging to one of these intervals, it is true that the economy with a
longer working day exhibits a lower rate of surplus-value. The demonstration is analogous to the one given at the beginning
of the section, provided we use the “wage commodity” as the standard of value.
In the concluding section we summarize the argument and draw the implications for the Marxian concept of the degree
of exploitation.

6. Concluding remarks

This is not a paper on exploitation nor even a paper on the rate of exploitation. It is a paper on the rate of surplus value
as a measure of the rate of exploitation. In one of his preparatory notes leading up to Production of Commodities by Means
of Commodities (1960), Piero Sraffa referred to the “Man from the Moon” as the ideal unprejudiced and impartial spectator,
whose judgement ought to be trusted. Paraphrasing Sraffa, the message of this note may be put in the following way.
Suppose the man from the moon encountered on earth two economies which had basically everything in common except
that the working day was longer in one economy than in the other. The man from the moon does not know why this is
so, but he has come across earthlings complaining about “exploitation” and explaining the concept to him. He concludes
that the economy with the longer working day exhibits a larger exploitation than the other one. At the same time, the man
from the moon realizes that the surplus value, a concept with which he has also been made familiar, may be higher in the
economy with a shorter working day. From this he concludes that the rate of surplus value cannot serve as a measure of
the rate of exploitation.
In Section 2 we recalled that the concept of the “degree of exploitation” and that of the “rate of surplus-value” have
typically been considered, in Marxian analysis, as two autonomous concepts. Proving or disproving that either one of them
may be taken as a measure of the other, as has widely been assumed in the Marxian literature, is therefore of genuine
significance for Marxian economic theory: Does the Marxian proposition that the rate of surplus-value is a measure of

14
This can be seen with reference to the concept of “sub-system”; cp. Sraffa (1960: 89).
15
See Garegnani (1970: 419).
16
This formula is only true if the rate of surplus value is uniform across industries. This uniformity can easily be demonstrated.

895
H.D. Kurz and N. Salvadori Journal of Economic Behavior and Organization 212 (2023) 887–896

the “degree of exploitation” stand up to close scrutiny or not? In Section 5 we have shown within the framework of a
little model that allows for a choice of technique of cost-minimizing capitalists that this is not generally the case, since
examples can easily be constructed in which the relationship between the “degree of exploitation” and the “rate of surplus-
value” is not unambiguous: one of the two may rise, while the other one falls. Taking the technique actually employed as
unalterable would, of course, avoid this result, but the implicit assumption backing it up that capitalists cannot choose the
cost-minimizing technique in given circumstances makes no sense: it would certainly contradict the general analysis of the
problems at hand Marx elaborated and the sought congruence with the historical facts he reported.
What follows from this, retaining the assumption that there is exploitation? Then at least one of the following proposi-
tions is true:
(a) Exploitation cannot unambiguously be quantified: it makes no sense to speak of a “degree of exploitation”.
(b) Exploitation cannot be defined solely with reference to the existence of “surplus value” (or “surplus labour” or “unpaid
labour”).
For if (b) is false, then, on the basis of the results set out in this note, (a) is true.
This conclusion may be related to the following observations. Marx was apparently convinced that showing that in cap-
italist society there is “surplus labour” and “surplus value” eo ipso proves that there is exploitation. This was perhaps un-
derstandable at a time before the elaboration especially of the marginalist explanation of profits in terms of the marginal
productivity of “capital”. However, with the elaboration of this theory things changed fundamentally. Advocates of this the-
ory such as William Stanley Jevons or Eugen von Böhm-Bawerk did not dispute that workers did not receive the entire
net social product and that there was what Marx called “surplus labour”. But, they insisted, workers had no right to the
entire product, because their labour productivity would have been much smaller without the capital goods at their dispo-
sition: these supported workers in production and were responsible for a significantly larger product per worker. Therefore
the proprietors of capital, who were taken to have accumulated it by saving parts of their incomes, that is, by “abstaining
from consumption”, deserved to participate in the sharing out of the product and getting compensated for the productive
contribution of their capital.
As Sraffa’s unpublished papers show, he was very clear that ascertaining the existence of surplus labour was a fairly triv-
ial task compared to showing that the explanation of profits in terms of the marginal product of capital could not generally
be sustained. Viewed against this background, pointing out a problem of measurement with regard to the concept of “ex-
ploitation” may be considered as providing an extra reason in support of Sraffa’s concern with the tenability, or otherwise,
of the marginal theory of capital and distribution.

Declaration of Competing Interest

No potential conflict of interest was reported by the authors.

Data Availability

No data was used for the research described in the article.

References

Bharadwaj, K., 1970. On the maximum number of switches between two production systems. Schweiz. Z. für Volkswirtsch. Stat. 106, 409–429.
Cogliano, J.F., Flaschel, P., Franke, R., Fröhlich, N., Veneziani, R., 2018. Value, Competition and Exploitation. Marx’s Legacy Revisited. Edward Elgar, Chel-
tenham, U.K., and Northampton, MA, USA.
Desai, M., 1979. Marxian Economics. Basil Blackwell, Oxford.
Dumenil, G., Foley, D.K., 2008. The marxian transformation problem. The New Palgrave Dictionary of Economics, 2nd ed. S. Durlauf and L.E. Blume. Palgrave
Macmillan, London.
Garegnani, P., 1970. Heterogeneous capital, the production function and the theory of distribution. Rev. Econ. Stud. 37, 407–436.
Gehrke, C., Kurz, H.D., 2018. Sraffa’s constructive and interpretive work, and Marx. Rev. Polit. Econ. 30 (3), 428–442.
Grove, W.R., 1846. On the Correlation of Physical Forces. Samuel Highley, London.
Kurz, H.D., Salvadori, N., 1995. Theory of Production. A Long-Period Analysis. Cambridge University Press, Cambridge.
Kurz, H.D., Salvadori, N., 2005. Representing the production and circulation of commodities in material terms: on Sraffa’s objectivism. Rev. Polit. Econ. 17
(3), 413–441.
Lutz, F.A., Hague, D.C., 1961. The theory of capital. In: Proceedings of the Conference held by the International Economic Association. New York. Stockton
Press.
Marx, K., 1967. Capital, Volume I. The Process of Capitalist Production. International Publishers, New York Translated from the third German edition by S.
Moore and E. Aveling..
Marx, K., 1959. Capital, Volume III. The Process of Capitalist Production As a Whole. London: Lawrence & Wishart.
Nuti, D.M., 1977. The transformation of labour values into production prices and the Marxian theory of exploitation. The Subtle Anatomy of Capitalism J.
Schwartz. Goodyear Publishing Co, Santa Monica, CA.
Ricardo, D., 1951. The Works and Correspondence of David Ricardo. Edited by P. Sraffa with the Collaboration of M. H. Dobb, Volume I. Cambridge University
Press, Cambridge, cited as Works, followed by the volume number in Roman.
Roemer, J.E., 1981. Analytical Foundations of Marxian Economic Theory. Harvard University Press, Cambridge MA.
Steedman, I., 1977. Marx After Sraffa. New Left Books, London.
Wolfstetter, E., 1973. Surplus labour, synchronized labour costs and Marx’s labour theory of value. Econ. J. 83 (331), 787–809.
Yoshihara, N., and Veneziani, R. (2022). The measurement of labour content: an axiomatic approach, in this volume.

896