‘Labour and Value: Rethinking Marx’s Theory of Exploitation’ by Ernesto Screpanti reviewed by Bill Jefferies

Labour and Value: Rethinking Marx’s Theory of Exploitation

Cambridge, Open Book Publishers. 2019. 144 pp. £25.95 hb.
ISBN 9781783747801

Reviewed by Bill Jefferies

About the reviewer

Bill Jefferies’ book Measuring National Income in the Centrally Planned Economies; Why the …


Screpanti’s Labour and Value is an examination of what he deems to be the manifest and insurmountable flaws of Marx’s value theory.

Screpanti opens with a discussion of whether Marx had an ethical opposition to capitalism. His view leans strongly on Althusser’s separation between the young-Hegelian, naive, ethical opponent of capitalism, influenced by Feuerbach (14), and the older, wiser and more cynical scientist who was “not interested in a moral condemnation of the abuses of capitalism” (16). Screpanti points out that Marx “argues that capitalists extort surplus labour in the production process” (16). Extort means to obtain from a person by force, intimidation, or undue or illegal power, while the extraction of surplus value is exploitation. Notwithstanding, Screpanti counterposes normative and descriptive (moral and amoral) connotations and makes an “unequivocal choice” in favour of the latter (17).

Screpanti provides an explanation of the relationship between value and abstract labour in Marx, and considers that for Marx “all surplus value is a result of exploitation” (20). But, “Marx the scientist does not aim to demonstrate the existence of exploitation. He endeavours to explain it” (Screpanti’s emphasis, 20). If the existence of exploitation cannot be demonstrated, how can it be explained? Sidestepping this dilemma, Screpanti explains that in his view “labour values are variables of a purely technological nature” (21) that ““do not play any role in determining the production of surplus value”   theory only serves as an instrument of measurement” (22). The theory of value does not determine the social relations which explain the production and distribution of value and “in fact, labour values are determined independently of profits. They hold in a non-capitalist economy and are therefore unsuitable for measuring surplus value” (22). Not only do labour values not determine the social laws that determine the nature of capitalist production, they are not even unique to capitalism but exist generally. They are both nowhere and everywhere.

Marx asserts that in the exchange of twenty kilos of coffee for ten meters of fabric labour is the common factor to the exchange. Screpanti insists: “the fact that 20 kilos of coffee exchange for 10 meters of fabric does not imply that the two commodities have some substance in common” (32). Certainly, taken in isolation, this relationship is merely incidental, but Marx’s example was of capitalist manufacturing production, with repeated, innumerable, essentially unlimited exchanges, wherein competition forces commodities to sell for their “natural price”.

Following Smith and Ricardo, Marx made the axiomatic, indeed tautological observation (tautologies are true) that assuming the production of these commodities was multiplied without any assignable limit, to paraphrase Ricardo, and that these commodities were not scarce or finite, then what determined this natural price was the substance that they all shared as products of human labour. This is basic set theory. If there is one set of natural products, and a subset of human products, then what differentiates human products from natural products is humanity. Humanity alone is the property that they all share. Therefore, it was the quantity of this socially necessary (i.e. not unnecessary as in wasted or superfluous) human labour embodied in the product, that determined its value. Screpanti nonetheless claims that “the identification of the value magnitude as a quantity of embodied labour is a result of the restrictive hypothesis of zero profits. Therefore, the proposition that abstract labour is, in general, the substance of value is not proved. It has to be assumed axiomatically” (32). Citing Chris Arthur, Screpanti claims that “Marx assumes it in the first pages of Capital, in which the zero-profits hypothesis is implied by the model of simple commodity production” (32). This is explicitly not assumed by Marx, who famously explains that wealth within capitalism consists of a mass of commodities, that is useful products, produced for sale. Profits, let alone zero profits, are not mentioned. Nonetheless, Screpanti sums up “the meaning of ‘creates’ in the metaphor of value creation by abstract labour is obscure and devoid of any scientific merit!” (34). Following Sraffa, Screpanti asserts that “knowledge of the technical coefficients is sufficient to determine labour values, while knowledge of the rate of exploitation is not necessary” (34).

Screpanti then describes what he understands by Marx’s theory of surplus value. The wage buys a worker’s labour power, or the use of their labour for a given period and activity and not the labourer themselves. This is what distinguishes the free labour of wage-slavery from slavery. According to Screpanti “in exchange for the wage paid to the worker, the capitalist obtains the establishment of a relationship, not a thing”. This relationship prompts the “utilization and appropriation of labour”. This process is “qualitatively different from the exchange of commodities and is its direct opposite” (42), but of course it is not. Labour power is a commodity. Marx’s theory of surplus value explains how, on the basis of the laws that govern the exchange of all commodities, surplus value is extracted. Commodities cost the amount of necessary labour to produce them, but some of that labour is unpaid. As only humans can own property, so only humans can be unpaid for their property. Thus, only human unpaid property, the amount of labour extorted from the producers in production, can be the source of surplus value. This explanation rests on the distinction between the reproduction cost of labour, the wage, and labour power, the thing bought. Screpanti refers to some of these points, but his discussion is confusing. He explains that “A commodity is a thing, an object produced by a subject using concrete labour” but of course a commodity does not have to be a thing; it may be a service, and so a relationship, or an object. He continues “Value, contrarily, is not a thing, and cannot be supposed to be created by a subject”. But if value is not created by a subject, how does the labourer embed value into the product during production? Screpanti maintains value “is an economic relationship among commodities, and a result of the social relations prevailing in productive activity” (54), but of course value is not a relationship between commodities, but people. The exchange of commodities is merely a vehicle for the alienated social relationship of people, not something separate from it.

Screpanti observes that “exploitation occurs when the value added for commodities is higher than wages” (54). Yet this does not account for “the production of surplus value” (54). According to Screpanti “surplus value is explained as the result of the capitalist’s ability to compel workers to attain a labour productivity higher than the wage” (54).  Yet wages do not have productivity while labour does. As earlier noted, assuming exchange is a one off, then only relative prices exist, so “the labour unit only serves as an instrument of measurement. Thus, one of the implications of the analysis […] is that the labour theory of value is not necessary to explain exploitation” (54).

Screpanti illustrates his argument in a perfectly competitive economy, with constant returns to scale, no scarce resources or fixed capital or joint production or luxuries or complex labour or growth. Screpanti’s model is a non-capitalist, utterly stagnant, slave society (80). Screpanti continues, “two equations may refer to an economy that produces a single good, in which case all symbols represent scalars and I=1” (80). Assuming away production (a labour process which physically transforms inputs into different outputs and so conceptually requires at least two goods, where the output is incommensurate and physically different to the input); Screpanti notes that this model can “also be interpreted as referring to an economy producing n commodities” (80), not one, but a definite and finite number of physically identical inputs and outputs. Assuming away free labour, capitalism and production, it is possible to demonstrate that labour values are superfluous. Who would deny it?

Screpanti notes that “Marx uses labour values to measure exploitation” and claims that he “often provides examples based on a single commodity” (80). This is quite wrong and indeed absurd. Marx never refers to a single commodity example, not once anywhere in Capital or anywhere else. Sraffians widely criticise Marx for exactly this mistake. But of course, it is no mistake, there are no physical constants in a real economy. When Marx refers to means of production or means of consumption, he is well aware that these are constantly changing, physically incommensurate quantities of physical inputs and outputs. Screpanti later explains that to simplify the model “assume that each sector produces a single commodity, a consumer good (sector 1) and a capital good (sector 2)” then the “equations represent two conditions of equality between demand and supply” (111). Marx apparently is damned if he does and damned if he doesn’t. He doesn’t. For Marx (and Leontief) the consumer and means of production sectors produced use value types, not single use values, with labour as the only scarce commodity.

Screpanti necessarily concludes that that “there is no need” for an “essentialist philosophy” to “determine value”, to “explain capitalist social relations”, or “the rate of surplus value” (95). He embraces the refutation of labour values attributed to Nobuo Okishio (95) and considers that the transformation of values into prices of production, which contradicts the requirements of linear algebra, means that “reproduction conditions in a capitalist economy cannot be determined in labour values” (111), when all it in fact means is that the transition itself was disproportionate.

Screpanti’s discussion of Marx’s theories of labour and value is their total rejection. Screpanti is hardly alone in this endeavour and there is not much original in this presentation. The problem Screpanti faces (as all Sraffians do) is the other worldly, paradoxically mystical nature of the assumptions necessary to achieve this mathematically correct (if not logically consistent) objective. As the Sraffians reject an external standard of value (to physical production), so they must find an internal one. This requires physical constants (or that inputs be physically identical with outputs) to allow commensurability and so measurement. As production is a process which physically changes inputs into physically different incommensurate outputs, this assumption precludes all actual production. As physical constancy requires one, or a finite quantity of products, so relative prices must be fixed, so that one commodity may be substituted for another, such that relative prices become absolute prices. As capitalist production is a process that constantly revolutionizes physical prices, this assumption precludes capitalist production. As the cost of production is the quantity of physical inputs destroyed (although somehow not destroyed) in producing the output, and as surplus is the difference between costs and prices, this identity precludes all actual profits. As class society requires exploitation, this precludes class society. As there must nonetheless be a surplus, so surplus appears without equivalent from nothing. This assumption precludes material reality, the conservation of energy, and the physical universe.

17 November 2020


  1. When Screpanti says that “the fact that 20 kilos of coffee exchange for 10 meters of fabric does not imply that the two commodities have some substance in common”, I think he is correct.

    The only thing required, is that people want to exchange them, on terms that are agreeable. Commodities can exchange on all sorts of terms, under all sorts of conditions. And we might also say, that they normally can be exchanged, simply because they are priced goods (although they need not even necessarily to be priced, in order to be exchanged). In that case, the common factor would be simply that the traded commodities are priced, which they normally are.

    The question is, however, what *regulates* the terms of exchange for great masses of commodities which are all traded at the same time – in particular, because the transactors must be both willing to cooperate with the exchange, and because they compete to get the best deal. For economic science, that issue of regulation is not primarily a logical problem, but an empirical one.

    When Marx argues that ultimately the regulating value of commodities is regulated by the current replacement costs in labour time, he cannot provide a conclusive *logical* proof of this. This has been known for more than a hundred years. In fact, Marx himself already explicitly rejected the very idea of a “logical” proof of the concept of value (in a famous letter to Kugelmann, 11 July 1868 – unfortunately the MIA archive did not publish the entire text online). Marx said that “all that palaver about proving the concept of value” was ridiculous, and that, anyway, his analysis as a whole showed the validity and coherence of his concept, even if he had never written a specific chapter on the value of commodities.

    Instead, Marx considered that the problem was to explain how the regulation of exchange value of commodities by comparative expenditures of labour time actually operated, when great masses of commodities are traded all at the same time, forming a universal market. In his own era, the econometric data to test all this out was very sparse, and a lot of the evidence that Marx used was anecdotal.

    What you can do today, is to provide proofs of the logical coherence and consistency of the theory, you can test it out against the available facts of experience, and you can compare the theory with rival theories for their relative explanatory and predictive power. Marx’s theory of the regulation of commodity values by labour is in fact *not* the only possible labour theory there is, there are many such theories.

    The main problem you strike with theories of economic value is, that they are often tautological, i.e. they are purported to be true, in virtue of the definition of the terms used. If, for example, it is argued that prices are determined by other prices, then you strike the problem of an infinite regress of prices which determine other prices… which determine other prices. It leads to the conclusion, that prices are simply determined by “the market”, that is the totality of priced transactions across an interval of time.

    Marx regarded such theories as pretty vulgar and superficial, that is, from the point of view of the explanatory and theoretical task of economic science. Such theories might be able to explain particular instances of trade in commodities, but not the trade in commodities in total. The obsessive focus on the trade in particular commodities either omitted the “big picture”, or ran into major contradictions with it.

    Already around the time that Adam Smith wrote “The Wealth of Nations”, utility theories of value existed, and Joseph Schumpeter documents this in some detail. Marx and Engels knew that already, at the time they wrote The German Ideology manuscript (where they refer to it). But they regarded the utility theories as implausible and vulgar, insofar as those theories made no sense of the overall movements of markets across longer intervals of time.

    If those movements are to be explained simply by individual preferences, which can change at a moment’s notice, you don’t obtain a theory with a lot of predictive power. If demand rose more than supply, or supply rose more than demand, all you can say is that individual preferences changed. Why did preferences change? It may be hard to say, there could be all sorts of different causes, in all kinds of different cases. It yields a bunch of eclectic theories, and it becomes difficult to know which theory might actually explain any particular case, or how this was consistent with other theories. That does not provide much orientation or guidance, or predictive power. The least that we might expect of a theory is, that it provides a useful general orientation.

    This does not mean that preferences are not important in the political economy of consumption, but that those preferences themselves are also shaped by objective circumstances that people confront. Among other things, the consumer is usually *simultaneously* a worker. If we place workers and consumers in different theoretical baskets, we disregard very important realities about their connection.

    Any plausible theory of value must refer to both objective and subjective conditions. It is merely that in economic theory, the worker is often portrayed as an abstract object (an input or cost, among other inputs and costs), rather than as the living subject of production. There seems to be an abstract “labour force”, but it has no role in markets, except as wage costs. Labour as a living subject appears only in management theory and organization theory.

    A lot of the trade in assets does not occur on terms which are in proportion to total labour costs. It never did. Marx knew this very well, judging by his comments on banking and credit. That is probably a reason, why Marx never called his own theory a “labour theory of value”. He was not talking about *all* value, but the value of commodities, labour-power and physical capital.

    The label “labour theory of value” only emerged sporadically after David Ricardo published his Principles of Political Economy, and became common when people began to contrast it with the newer marginalist theories of value from the late 1870s onward, that is, after Marx’s Capital Vol. 1 was published.

    Probably, the growing popularity of marginalist theories of value owed much not only to the growth of markets with relatively stable prices, but also to the search for a reply to the growing power of the labour movement in the political sphere, which sought the right to vote, a reduction of working hours, fair wages, and a say in the organization of production. The labour movement protested and campaigned against the exploitation of workers. The global impact of the Russian revolution decisively changed the developmental paths of economic science.

    It is quite possible in global markets for a significant dissynchrony to emerge between trading prices for products and their true total labour costs of production. But Marx argues that no escape from the regulation of product prices by the law of value is possible in the long term; at a certain stage, when the dissynchrony between market prices and labour costs for products has become too great, the market and the value-chains cave in, “like a house of cards”, and price-levels are realigned with real costs through competition.

    Does a labour theory of value still have a lot of explanatory and predictive power nowadays, with such large trade in assets, services and information, that do not directly involve significant tangible or material products, and when enormous global disparities occur in the organic compositions of capital, wages, and terms of trade?

    I think the evidence tells us it does, but (1) it is probably more difficult to observe and test, insofar as the people collectively producing (part of) the products and the people consuming them are often far removed from each other in space and time; (2) the exact ways in which services and information are priced in markets, and what regulates those prices, is still often not very well understood; (3) the total economic reproduction process is changed greatly by the enormous growth of assets which are neither an input nor an output of current production (durables, real estate and financial assets).

    The traditional Marxist economic reproduction models assumed, that the earnings from the output value of production are either reinvested or consumed, but this is in reality not true. It is an error, to equate the “mode of production” in a society with “society as a whole”, or “production” with “the whole economy”. That is easily verifiable from the national accounting systems we now have.

    Perhaps the more important point is that labour theories of value, and Marx’s theory of value in particular, cannot be the basis for the political unity of labour movements and Leftwing parties. They never were either. Yes, there is still exploitation, there always was. But its critics are better off showing what particular cases of it tell us about the state that society is in.

  2. One thing that struck me in Screpanti’s eclectic text (which I think also contains some big mistakes) is his propensity to conflate clear conceptualization and theorizing with algebraic formalization.

    Mathematical equations cannot substitute (1) for conceptual precision in the definition of measuring units and concepts and (2) for the defence of theoretical assumptions with empirical evidence; they can only reveal logical and quantitative implications of concepts, theories and measurement units.

    Screpanti makes it seem that the results of formalization can simply and straightforwardly “prove or disprove” the plausibility of the concepts, theory or measurement units. One crucial mathematical equation seemingly suffices to rule out or enthrone a concept. Prima facie, it seems to be a very erudite, rigorous and “scientific” approach.

    In reality, that is not the case, since the conclusion of the formalization holds, only if a particular definition of the concept (and its assumptions) is already accepted. Only if you accept the definitions, the logical implication follows. But that can be, and often is, a big “if”.

    The crucial question for modelling and theorizing is then, “where we get our concepts, assumptions and measurement units from”, and how they can be combined in a coherent theory.

    For Marx, the main flaw of Hegelian thought was, that it imported assumptions and premises into the inferential system *speculatively*, while subsequently suggesting that one definition derives *logically* from another.

    In this way, Marx says, “Hegel fell into the illusion of conceiving the real as the product of thought concentrating itself, probing its own depths, and unfolding itself out of itself, by itself, whereas the method of rising from the abstract to the concrete is only the way in which thought appropriates the concrete, reproduces it as the concrete in the mind. But this is by no means the process by which the concrete itself comes into being.”

    It is certainly true that to achieve something, we need an idea, but it is a mistake to attribute its achievement only to an idea. The apparent rigour of Hegel’s dialectical reasoning turns out to be spurious, because it repeatedly assumes what it intends to prove. And it is illusory, because ultimately the object is considered to arise out of its concept, rather than the other way around (“the real” is “the product of thought”).

    Already at the beginning of The German Ideology manuscript, Marx and Engels recognize this view as a grotesque error. Mockingly, they tell the anecdote of a “valiant fellow” who believed that men were drowned in water, only because they were possessed with the idea of gravity. If they could only get rid of this idea of gravity, they would be safe from any danger of water.

    By contrast, Marx does not try to import his categories and establish their interconnections speculatively, but develops them out of (i) a critical inspection of the empirical facts, (ii) from the actual historical development and practices which give rise to them, in various formulations, and (iii) through the criticism of the theories which try to combine them in a coherent whole, in order to explain them rationally.

    In this way, the inferential process in Marx’s own story often mirrors the actual way in which the relevant concepts were formed and developed in the real world, from very simple beginnings to quite complex structures.

    Thus, Das Kapital begins with analyzing the meaning of the simplest observable object of commercial trade, the commodity, which already existed for thousands of years prior to industrial capitalism, and builds a theory out of that, which step by step incorporates more complex objects of trade – without needing to abandon the original definition of a commodity at all. It is merely that at the end of the analysis, the same traded object appears as the product of capitalist production, and not just as its most basic precondition.

    Screpanti is by no means a Hegelian, but his procedure suffers a similar defect, mainly due to a reification of mathematization, as a seemingly infallible rational authority to adjudicate the veracity of concepts. First an idea is imported in the inferential system, and then the idea is presented as the product (or a waste product) of the inferential system itself.

    Formalization (just like speculative thought) certainly has its place, but it should not be made to do more than it can validly do. Hayek and Keynes admitted that. Michael Polanyi among others noted, that to formalize a concept, we need to be able to identify it in the first instance without mathematization.

    The idea that mathematization can prove the valid meaning of a concept, and rule out all other meanings, is flawed. Algebra can only be an aid to understand the semantic and objective relations of phenomena, but these phenomena also exist beyond mathematics.

  3. I recently spotted a book „Reconstructing Marxian Economics: Marx Based upon a Sraffian Commodity Theory of Value” (New York, NY: Praeger) and read some chapters. Author is Spencer J. Pack and book is from 1985. Maybe you could consider it for some future review.

    By the way, I read your comment in Brazilian Journal of Political Economy on Araujo’s attempt to show that Sraffa’s price system is compatible with Marx’s theory and Araujo’s response to your article.

  4. I think neither Araujo nor Jefferies show clearly why exactly Sraffa needs a “standard commodity” in the first place.

    Strictly speaking, it isn’t true that Sraffa’s theory is “compatible” with Marx’s theory – simply because it is a different theory of the same phenomena.

    Nevertheless you can say, that Sraffa’s theory is at least compatible with “a labour theory of product values”. That is also what Ron Meek (and others) meant, back in the 1970s. There is no evidence, that Sraffa himself ever denied or rejected that possibility.

    Via Sraffa’s concept of direct and indirect labour (and, implicitly, prices), Pasinetti’s concept of vertical integration, and Leontief’s input-output tables, you can construct econometric indicators for a *measurable* labour theory of value, using price data and labour data.

    This work began to be done from the 1980s onward by scholars like Anwar Shaikh, Eduardo Ochoa, Ed Chilcote, Alan Freeman, Lefteris Tsoulfidis etc.

    In that sense, Sraffa’s conceptual work proved helpful, by contributing to the empirical verification of a labour theory of product values and product prices.

    Marx and Sraffa agreed on the classical principle, that the levels of market prices for products are regulated by production prices constituted by cost-prices and average profit rates. Sraffa shows, how you can model that classical principle, without any necessary reference to a controversial labour theory of value.

    Where Marx and Sraffa obviously differed, is in exactly how the levels of production prices themselves are determined.

    Marx didn’t actually have a “dated labour” theory, but rather a theory of a “current replacement cost (or reproduction cost) in average labour-time” (and this doesn’t have anything directly to do with the Ricardian “embodied labour” or the Smithian “labour commanded”, as I was among the first to point out). Marx’s theory is, in some respects, more difficult to test than the modified Sraffian labour theory.

    Nevertheless, the econometric theory of labour-value is already a huge step forward in understanding and exploring the empirical validity of a labour theory of product values.

    There are Marxist inquisitions which would like to fry Anwar Shaikh at the stake, for failing to explicitly swear r-r-revolutionary fealty to Marx’s Labour Theory of Value in his new book. Fred Moseley for example complains that Shaik’s analysis “doesn’t have anything to do with Marx”.

    But the critics haven’t understood the book, and in fact Shaikh’s economics is fully compatible with a labour theory of product values. For some economic analyses, it has little relevance what particular theory of value you uphold, while for other analyses it does.

    It is just that economics is about so much more, than academics jumping up and down, and huffing ‘n’ puffing to emphasize the “labour substance of value”, the “exploitation of labour”, and the “falling rate of profit”.

    Ultimately, the ultra-leftist “Marxism” that is so popular today is a very barren science, that can tell you almost nothing about how to understand, predict and solve any real problem of resource allocation – never mind organizing a planned economy that really works.

  5. I think myself that you have to distinguish carefully between “Piero Sraffa” and “the Sraffians”.

    Sraffa published his first academic articles when Benito Mussolini came to power. It was a time, when people really had to watch out with what they said. You could get prosecuted, jailed or killed.

    If Sraffa had not been there, Antonio Gramsci probably wouldn’t have been able to write his Prison Notebooks, and the Notebooks probably would never have been published either. Sraffa went into exile, after he refused to pledge loyalty to Mussolini’s regime (Mussolini himself was originally a socialist).

    Sraffa was always a man of the Left, a socialist. He didn’t attack Marx, to the contrary he valued Marx’s work highly, and personally owned a lot of Marx memorabilia. Sraffa attacked the theoretical foundations of the marginalist theory, and he criticized the economic theories of Alfred Marshall and Friedrich Hayek. He wasn’t fighting against his own side, but doing battle with the opposition.

    At first, Sraffa’s PCMC book was generally interpreted as a “modernized” defence of Marx, Ricardo and classical political economy. In the 1960s, Maurice Dobb and Ronald Meek often presented it that way. The difference between Marx’s and Ricardo’s theory of value was regarded as rather slight. Sraffa had provided a new way to get a respectable hearing for a classical economics perspective, in what was known as the “cold war” era.

    However, in the 1970s and 1980s, Sraffa’s neo-Ricardian alternative also became an instrument to attack Marx’s theory of value (and its Stalinist dogmatization). This was mainly a new development, and a shift in emphasis. Sraffa had supposedly “proved” that Marx’s theory was wrong, and Marx’s theory was supposedly fully superseded by Sraffa’s own approach. The focus shifted from the “production of surplus value” to the “distribution of value-added”. Exponents of this trend were e.g. Ian Steedman, Mark Blaug, Geoffrey Hodgson, and Steve Keen. There were numerous replies from Marxist academics.

    After the Wende from 1989, the Sraffian offensive lost much of its steam, since Marx, Marxism and socialism went out of fashion anyway. Having peaked in 1980, Marx was supposedly a dead dog after the Berlin Wall fell. In addition, more and more attention went to the phenomena of “financialization” and financial bubbles.

    One scholarly challenge today is to reclaim the real legacy of Piero Sraffa (contrary to the unthinking vilifications of the “super radical” academics in the past), and create a comprehensive scientific alternative for Ricardian, Institutionalist, and Post-Keynesian economics – without fanatical dogmatism and empty rhetoric. That is what I understand Anwar Shaikh (and other theorists inspired by classical economics) to be doing.

    It is obviously difficult to dialogue or debate with people with whom you just don’t have anything in common, or whom you have already written off totally; it becomes a ritualistic, feigned debate. Yet many Sraffian-inspired economists are egalitarian liberals, or socialists of one stripe or another. That is, many of them are actually on the Left, even if they are not Marxian economists. That makes a “constructive controversy” possible, from which you can learn something new.

  6. What did you think of Araujo’s reply Nelson? I think all the attempts to reconcile Sraffa and Marx are basically misconceived. It is a physical alternative price system to abstract labour and Sraffa is absolutely explicit about that. Sraffa insists wages enter into his system on the same footing as fuel for machines or feed for cattle. For Sraffa there is nothing distinct about human labour per se. Therefore, he has no explanation for prices or surplus. In order to enable commensurability he has to assume that inputs and outputs are physically identical. How else to measure them? This is an impossibility that contradicts the purpose of human production, to change physical inputs into different, more useful, outputs. For Sraffa surplus has to appear from nowhere without equivalent, as the property form, is irrelevant. Whereas for Marx, as surplus extraction is a form of property redistribution, the cost of labour input is extracted from the workers in production, it is a real cost, but to the workers not the capitalists. It is unpaid labour. Finally, notwithstanding Araujo’s complaints relative prices are fixed, otherwise the standard commodity cannot be a standard. This in effect means the economy is a one commodity economy, as the only commodity that is relatively fixed against itself is in fact the same commodity, so again it is not a model of production.
    The reason for the theories popularity among Marxists is not because Sraffa was a Marxist. He wasn’t. He was a mate of Gramsci. You won’t find much mention of Marx in his writings, he’s a pretty orthodox Marshallian. His system is closer to Walras than anything else. It is because it was championed by Maurice Dobb, the Stalinist professor and Cambridge, who thought it was a method for reinserting value into the Material Product System of the Soviet plan. Everything else is history.

  7. I agree with you completely, and Araujo’s reply does not seem to me to be an reply at all. He still sticks to Srafa’s assumptions.

  8. Yes it seems to me that he doesn’t address any of the key points at all. For example he quibbles about my assertion that production is to make things more useful and asserts that this is “neo-classical”, not at all. Marx comments, “It is as clear as noon­day, that man, by his industry, changes the forms of the materials furnished by Nature, in such a way as to make them useful to him”. Clear as noonday is evidently not clear enough!
    He then simply ignores the question of commensurability and surplus, so I agree its not an answer at all.

  9. For the record, I am not concerned with “reconciling” Marx with Sraffa, but I am still interested in both thinkers, each in their own right. If you are seriously interested in science or politics, you have to willing to understand different points of view. I believe both thinkers had valuable insights to offer, in different contexts; but to really understand what they are saying, you have to know about the intellectual milieu, and who they were arguing with. I also have some criticisms of both, but that is another story.

    I have never claimed that Sraffa was a “Marxist”, and Sraffa never claimed to be a Marxist himself either, as far as I know. At best you might say, that Sraffa tried to follow the spirit or intention, rather than the letter of Marx’s endeavour. It is certainly a mistake to call Sraffa an “orthodox Marshallian” (see on this e.g. Gary Mongiovi, Sraffa’s critique of Marshall: a reassessment, Cambridge Journal of Economics, Volume 20, Issue 2, March 1996, Pages 207–224). Sraffa as Walrasian? I never heard of that one before.

    In 1981, Geoffrey Hodgson remarked that “The Sraffa system, like many stationary-state general equilibrium models, contains no good which, uniquely, possesses all the important features of money.” (Geoff Hodgson, “Money and the Sraffa System”. Australian Economic Papers, June 1981, p. 83). Which is true. This obviously affects the ability of Sraffa’s own models to explain or predict price fluctuations.

    Generally Sraffa assumes in his economic models, that all commodities (inputs and outputs) are valued in a uniform way, according to a standard valuation. This is like Marx assuming in Capital Vol. 1, in Vol. 2 and in the first part of Vol. 3, that there is no divergence between commodity prices and commodity values. To explain price fluctuations consistently, presupposes that credible conceptual foundations have already been shaped up to do that. Otherwise, we don’t much further than the trivia of supply and demand curves.

    So this is a type of “highly abstract modelling” which deliberately restricts the possible causes and modalities of price fluctuations (basically, prices changes must be due to changes in the production system or remuneration systems). It is also analogous to the SNA-type national account for gross product, which measures and adjusts prices according to a standard (uniform) valuation.

    In the Sraffa system, the operation of the market cannot predetermine how exactly the “net value added” (the developed form of “surplus”, i.e. the value of production less the costs of production) will be shared out between economic agents, but it does set limits for how it can be shared out. The Sraffian surplus is essentially determined by productivity, and in turn productivity is determined by the development of technology.

    In one sense, it is quite true, that there is for Sraffa “nothing distinct about labour as such”. As the title of his book already suggests, labour is treated in his economics as a “commodity” with a certain price tag, just like machines, cattle, cattlefeed or any other sort of input or output of production.

    Yet of course the title of Sraffa’s book is also *deliberately* ironic, and an oblique reference to Marx’s theory. Sraffa himself never believed, morally speaking, that labour “ought” to be treated as a commodity. He is merely saying that, scientifically speaking, human labour is as *a matter of fact* treated as a commodity in the market economy, and on that basis, he distinguishes between different magnitudes and dimensions of the “labour commodity”.

    Sraffa implies, that the market system itself is largely amoral: it does not provide any intrinsic morality, except for basic requirements and agreements among economic agents to settle transactions according to accepted rules. If they don’t honour their obligations, they risk reputation damage or legal sanctions, but they will accept tolerance for differences in opinion, when interacting in the marketplace to make deals.

    The amoral market system can in principle function with all kinds of different belief systems, so long as the competing economic agents do not transgress the rule of law enforcing civil behaviour (think here of the official and media response to the recent storming of the Capitol in Washington DC).

    Sraffa’s suggestion is not simply that economic science is subject to “distortions by commercial reifications”. The deeper problem is, that conventional economics tries to provide apparently “scientific” rationalizations justifying those reifications – as proofs of the ultimate rationality of the market system, and of the market as the optimal way to allocate resources. This leads to plenty misunderstandings and theoretical errors.

    At the very core of the “scientific” striving of conventional economics, is a theory of capital and value which Sraffa shows to be irresolvably flawed, basically because it assumes what it has to explain. Sraffa then tries to provide a logically rigorous alternative conceptualization, which is free from arbitrary definitions and tautological arguments.
    Yet, we should also not regard Sraffa’s theory as a complete theory of the economy. Sraffa was quite clear about this himself, and also extremely modest: he subtitled his book “a Prelude to the Critique of Economic Theory.” He had, as it were, just cleared up some basic concepts, and had broomed away a lot of deadwood, theoretical nonsense etc., as the *precondition* for a serious scientific criticism of mainstream economics.

    Prof. Anwar Shaikh followed that path, but he took the scientific endeavour much further. He started with a paper on “the “Humbug production function” (1974), a powerful logical critique of the concept of production functions. His rejoinder to Robert Solow’s response to that paper was actually suppressed, and much later, the New Palgrave article on the Humbug Production Function (1990) was also deleted from the second (Durlauf) edition of the New Palgrave.

    Nevertheless, Prof. Shaikh has crowned his career with a very large treatise on capitalism covering vastly more topics than Sraffa ever did, but with a similar intention. I think both Karl Marx and Piero Sraffa would have been very pleased. As Marx put it in a preface to Capital Vol. 1: “Every opinion based on scientific criticism I welcome. As to prejudices of so-called public opinion, to which I have never made concessions, now as aforetime the maxim of the great Florentine [Dante] is mine: “Segui il tuo corso, e lascia dir le genti.”

    PS – About Maurice Dobb, another day, I just ran out of time for that.

  10. If Anwar Shaikh had extended his humbug function to Sraffa maybe we would get somewhere.
    It’s really very simple. Humans produce their world. As physical inputs and outputs are incommensurate so they must be measured by an external social standard – abstract labour time. The only factor common to all human production.
    Sraffa does not understand that value is a form of property. As only humans can own value, so only humans can produce value, which is a measure of the distribution of output between humans.
    NO! says Sraffa. This is mystical nonsense. Prices must be derived from material means of production themselves. There is no external standard of value! Surplus is a crop! But of course simply denouncing the alternative doesn’t help the argument.
    How do you measure output if their is not external standard of value? There must be an internal standard of value. What is it? Only the material existence of commodities themselves. But there are no physical constants in production. Inputs are different from outputs. How to measure different physical things against each other? Assume they are the same. Assume inputs are physically identical to outputs. Assume that there relative prices do not change. Assume that profits are yielded in identical proporions always.
    Hence the fixity of relative prices, the identity of inputs and outputs and the creation of surplus without equivalent. Hence the absence of production.
    Sraffa’s model is mathematically correct, Shaikh proves that (among many others) but only by paradoxically assuming away production. It’s about as unscientific as humbug.

  11. When Sraffa proposes that price relativities for commodities sold in markets are ultimately determined by “physical real costs” (the costs of material inputs used up to produce new products), I don’t think there exists any “commensurability problem”, if all such costs are priced.

    After all, it is not *the physical things themselves* which, according to Sraffa, determine value, but rather what they (the physical things) happen to “cost” on average, and in his theory, that cost is expressed in price terms. So as long as inputs and outputs are both expressed in prices, then no “commensurability problem” arises between them, and there is no problem of trading them for each other, or for any other goods (a “valuation problem” could arise, if the product chains involve significant production time).

    Admittedly, David Ricardo at one time toyed with the idea, that because in corn production corn was simultaneously an input and an output, that value might ultimately be determined by corn production, but he dropped that idea again; at the end of his life, he concluded that in reality no numeraire or constant existed to express the value of all commodities, the chosen numeraire remained only a hypothetical assumption.

    The more substantive point is, that Sraffa’s highly abstract modelling still doesn’t get anywhere near to what happens in the real world, i.e., he has (as far as I can see) no plausible explanation of the connection between Ricardian-type cost-prices and real market prices for products.

    That said, econometric techniques do exist nowadays to test out the price theories of Ricardo, Marx and Sraffa statistically, using national accounts data, labour data, price data and input-output tables. You can compute labour, capital and material inputs implicated in output values. In this way, you can create estimates for each theory, to see which one is the best predictor of observed price movements.

    Of course, the theorists often like to think that there is only “one principle” that determines all market trade, or “one principle” that determines all product prices etc. In the real world, that is probably not the case, or very rare. It is in principle quite possible, that a Marxian theory will score better in statistical tests for some product markets, and a Sraffian theory scores better for some other product markets. It is also quite possible that, even if price-levels in product markets are regulated by labor quantities, this regulation process operates in a somewhat different way in different sectors.

  12. What is a price? It is a ratio of physical things. For two physical constants, that’s fine, but production is a process which changes physical constants, it uses them up. It transforms inputs into a product, an output that is different to the inputs. So how much is it worth? As it is physically different so it is incommesurate. It is unknowable.
    OK says Sraffa, the output is the cost of the inputs destroyed in producing it.
    A watch is the amount of brass, iron, glass, leather and so on required to produce it. Fair enough. In which case as there is a smaller physical quantity of brass, iron, glass and leather in the watch than before it was manfactured, necessarily as production destroys the physical form of inputs, then manufacturing is a process of value destruction. What is this nonsense?

  13. In the Sraffa system, both the inputs and outputs of capitalist production are all commodities, and consequently, capitalist production is the production of commodities “by means of commodities”.

    Each commodity is created as a composite of other commodities, and the production system can reproduce itself via a circular flow of commodity transactions. All commodities are priced. The prices express exchange value in units of money. All commodities are exchangeable using money, and can be combined or processed in innumerable different ways.

    Labour as an input to production is also a priced commodity, and wage costs are regarded as part of the “physical real costs” (this is broadly consistent with the SNA conceptualization of “production” and “value-added”).

    On this basis, Sraffa examines, for example, what happens to prices, profits and rates of profit, if the cost composition of inputs and outputs changes (in abstraction from price inflation, or changes in the money supply), and he invents a method for decomposing the cost structure of commodities.

    I am not necessarily defending Sraffa’s theory here, but this is what it is. It is true that Sraffa is not particularly concerned with what specifically happens in the production process itself (as Marx was), but that was because Sraffa’s aim is different from Marx’s.

    Sraffa intended to provide a consistent classical model as an alternative to a neoclassical system of long-term competitive equilibrium prices under constant returns to scale, given a self-reproducing production system which can constantly renew the conditions of its own perpetuation.

    To a considerable extent Sraffa was successful and changed the academic discussion. However, conventional economics (just like Marxism) is a mixture of science and religion, fact and ideology, fiction and non-fiction. Just because economic theories are in truth demolished by logical and empirical evidence, does not necessarily mean at all, that economics abandons those theories (although some theorems are quietly dumped and/or replaced with other theorems).

    So the marginalist theory of value and the neoclassical equilibrium model still persist. Marginalism survives also because it contains a “kernel of truth” about the subjective dimensions of economic value, which is ignored or destroyed by objectivist theories of value. Economic value has both objective and subjective dimensions – this is obvious to a phenomenologist – but the objectivists and the subjectivists cannot accept that, because it would undermine their own theories, which reduce value either to subjects or to objects.

  14. Its absolutely untrue that Sraffa expresses prices in units of money. There is no money in Sraffa’s system, as money is an external numeraire. A system of values outside of physical production itself.
    The contradictions in Sraffa’s system arise precisely from the requirment that prices can be measured only internally. Sraffa points out that if relative prices are fixed, or more accurately if there is only one commodity, there is no need for money. Although of course if there is only one commodity, then there is no exchange, no price, and no production either. The standard commodity, the eternal composite commodity, is required precisely as there is no money, not external standard of value.

  15. Price is exchange value of commodity in terms of money. Ok. But, as Jefferies asked, what does money measure?

    It must be that, as Marx said, exchange-value is the mode of expression, the ‘form of appearance,’ of a content distinguishable from it, and that content is value.

  16. There does exist money in the Sraffa system, but the circulation of money has no independent influence on prices, and there is no variability in the value of money. It is simply assumed that goods are and can be supplied, and that they are and can be sold.

    That is also why, as I mentioned, Geoff Hodgson noted that “there is no good in Sraffa’s system which has all the characteristics of money”. Money does implicitly enter the story, because otherwise the reference to prices would be meaningless. But money functions merely as a medium of exchange, and it has no other role.

    In chapter 2 of PCMC , Sraffa explains that he will *deliberately* avoid a lot of classical and neoclassical “terminology” to talk about prices, costs and capital assets, because he considers that such terminology is to a large extent embedded in theories which assume types of causal relationships (“what determines what”) which he rejects in his own theory.

  17. OK so there is money in Sraffa’s system, it’s just he never mentions it and it plays no role. Gotcha.

  18. Well if you want to exchange and circulate commodities, money is required, even if, let us say, it is only a referent (as in countertrade, which Sraffa did not deal with). Profit is a sum of money. The wage is a sum of money. The capital stock is a sum of money. Incomes and expenditures are sums of money. This is all presupposed.

    But I agree Sraffa’s model of the production and distribution is developed at an astronomically high level of abstraction. That’s a limitation. It also raises the question of what the fuss is all about, its importance, really. That is not easy to summarize in a quick way.

    In his (now probably forgotten) book “An introduction to Post-Keynesian and Marxian theories of value and price” (1983), Peter Lichtenstein commented quite correctly: “In Sraffa’s theory of price – a theory which we have called Ricardian because of its emphasis – the labor embodied in production actually becomes incidental to the whole process of price formation. Although costs of production can be reduced to dated labor, this is not essential to the analysis. For this reason the word ‘value’ is frequently indistinguishable and indeed often synonymous with the word ‘price’ in Sraffa’s Ricardian theory” (p. 152-153).

    Lichtenstein stated Sraffa’s core thesis simply as follows: “The real wage and the conditions of production, when specified in terms of actual physical quantities, are sufficient to determine prices of production and the rate of profit. No demand theory is necessary”. (ibid. p. 114)

    In the classical Ricardian model, changes in relative prices of production can be caused by (1) a change in the technology of production, where the distribution of the surplus between wages and profits remains constant, or (2) a change in the distribution of the surplus, given a certain technology of production. (cf. ibid. p. 123). But there is a theoretical problem:

    “It is the payment of profit which causes prices to diverge from their labor values. Because capital-labor ratios differ among industries, changes in the wage-profit distribution will also change prices. And as prices change, the size of the pie (social wealth) changes too! Thus, the size of the aggregate output would reflect not only the quantity of commodities actually produced, but the distribution of income as well”. (p. ibid. 122) “This would be like dividing an apple pie between you and a friend. Every time you give yourself less, and your friend more, the size of the pie changes! When you want to discuss distribution, you want the quantity to be distributed to remain the same irrespective of who gets what share” (ibid. p. 119-120).

    To solve this problem, David Ricardo believed “a commodity would have to be found whose price would not change as the profit rate (and hence distribution) changed. All other commodities could then be measured in terms of this special standard commodity”. (ibid. p. 123) This is the concept of a standard commodity. Ricardo never found that commodity, and concluded in the end there wasn’t any. Sraffa however computes a “theoretical” standard commodity, a composite, which solves that problem, so that all the equations add up, and all the variables in the model relate in a logically sound way.

    That’s an achievement in itself (Sraffa had mathematical help from other scholars) but what now is the theoretical (or ontological) status of the production prices thought to be the determinants of market prices for products? Point is, if the production prices exist only “in theory”, how can they truly shape market prices in the real world? Probably the most popular way out, is to say that the production prices are “centres of gravity”, i.e. price levels which market prices will tend to approximate, or oscillate around. But where is the proof that this is really the case? How does the “gravitation process” actually operate?

    There are no “logical proofs” available to answer these questions, all you can do is empirical tests of the processes and results of production, using macro-economic and other data, to build a theory informed by empirical findings. That is also what Anwar Shaikh and others have done, and the classical theory of value and price stands up quite well to the evidence. It’s a vastly bigger achievement than Sraffa’s, but Sraffa’s “Prelude to a critique of economic theory” nevertheless made an important contribution to “thinking through the problem”. A theory of demand is necessary (Keynes is not simply “reformist rubbish”), but it can be integrated in an empirically based, systematic theory of competition and cooperation in the capitalist economy, that incorporates all the important research since Sraffa published his book.

  19. Is it acceptable to fill in the gaps? To add to the system where you find it lacking? To ignore or discount its incohence and absurdity? To add money or abstract labour or embodied labour or any of the other categories necessarily excluded by it?
    Nelson is quite right. The moment you accept the existence of money, you refute the premise for Sraffa’s system. For money is an expression of social value external to physical production. The very existence of which Sraffa’s system is designed to refute.
    Is Sraffa’s system highly abstract? Not really, in fact the premise of it is that physical material relationships are all that is necessary to determine price. This is how it was used by Samuleson, Roemer, Steedman and co. It enabled them to “mathematically prove” that Marx’s values were obsolete, illogical and wrong. An unnecessary redundant step. That prices could be derived directly from physical production. So did they prove it? What are the assumptions of Sraffa’s sytem necessary to “mathematically prove” such a result?
    As if maths means much. Didn’t philosophers mathematically prove the number of angels on the head of a pin?
    Sraffa’s assumptions are only slightly less metaphysical, for they assume away production as such. To enable commensurability, there can be no physical change between inputs and outputs, or relative prices and surplus has to appear from nowhere without equivalent. Is it possible to have important research based on such a premise? Well, assuming of course, the existence of angels…

  20. Is it acceptable to fill in the gaps? Sure, if you have to see your dentist, haha. But it is always a good idea to maintain a good relationship with your dentist, if you want to get the best results, clean your teeth and wash before you go, and so on (think of the dentist as a worker, who has to work on people’s teeth every day of the week, it takes stamina, endurance, care and patience).

    There are plenty Sraffians or Post-Keynesian economists who have tried to develop and concretize Sraffa’s ideas to obtain a more realistic theory of the economy, including the monetary regime as an additional variable. The contemporary theories are just vastly more sophisticated than they were in 1960, also because there is vastly more data about economic life, produced by a burgeoning stratum of economic and commercial researchers.

    The core idea of the classical theory of capitalist production, as Anwar Shaikh would say, centres on the idea, that the business is driven along by an incessant competition to cut costs, increase sales, and increase profits, like it or not. http://www.anwarshaikhecon.org/sortable/images/docs/publications/political_economy/1989/The_current_economic_crisis.pdf

    For many classical and modern economists that was in fact also the “progressive” thing about capitalism: businessmen were impelled by competition to economize resources and improve efficiency, and if they did that, then society as a whole was better off (in contrast, supposedly, to the much maligned Soviet Union, where a lot resources were wasted, due to bureaucratic shenanigans).

    The neo-Ricardians do want to keep that central idea about costs, sales and profits in some form, but they are often not much interested in Marxist ideology, a labour theory of value, not to mention revolution etc. Sraffa’s legacy is somewhat ambiguous, scholars draw different political conclusions from it. And this is clearly shown by the reception of his ideas across six decades.

    As I mentioned, originally Sraffa was mainly interpreted to offer a “modernized” interpretation of the ideas of Ricardo and Marx, but subsequently Sraffa’s ideas were also deployed specifically to attack Marx’s theory of value (e.g. Gerald Cohen, Geoffrey Hodgson, Ian Steedman, etc.) and even more specifically Marx’s theory of the exploitation of labour.
    It is rather obvious, that if you want a good job higher up, harping about “labour exploitation” is more likely to be a hindrance, than a help. But existential problems or victimology aside, the theory about labour exploitation seemed questionable, if not incoherent, according to a new crop of scholars, and they felt it was legitimate to question its validity, or ditch it. Perhaps historical materialism had a validity, but then without the “awkward” exploitation idea. That would have existed in the past, but not now, or it existed only in faraway countries.

    Social scientific ideas are obviously not developed in a void, even if you happen to live in an ivory tower. There are times and epochs when exploitation is a topic of great controversy, people can see very clearly, that there is great exploitation, and that it is unjust, and there are also times when it isn’t. That has a big effect on the academic interest in the issue of exploitation.

    During an economic boom, there are plenty opportunities to better your lot, the vast majority can make gains, even if the gains are unequal. In that situation, exploitation is more likely not to be a dominant or primary theme in social protests, except among people who are really exploited terribly. When the economic crash arrives, and a recession or depression sets in, that changes, competition for jobs and positions intensifies, and people can clearly see the increase in social inequality and misery, and they are more receptive to ideas about exploitation.

    Of course, I do not want to vent a crude “economism” here, the social contradictions that people experience can, at times, explode with outbursts of all sorts of grievances that previously nobody had ever expected, even if conditions of life are actually relatively good. Just before the upheavals of May ’68 happened, the French sociologist Raymond Aron declared confidently that the French Republic had “never been so stable” before. He just wasn’t very much in touch with the youth, and missed the real meaning of the youth radicalization developing in the 1960s.

    The patterns of radicalization of public opinion can take all sorts of forms. It is nevertheless true, that in every era or epoch there are certain dominant themes and “metaphysical” ideas, which very strongly influence the academic “discourses”. That also explains why social science never progresses in a “linear” way. The previous intellectual preoccupations and research programs are, at a certain point, simply “pushed aside” by new events that grab attention and capture the imagination (and the finance for research, cynically you might say that “the talk follows the money”). Usually only a minority of thinkers steeped in the literature grasp the real evolution of ideas across long intervals of time. Most people are only interested in what is happening now because that is what they practically have to deal with.

    The same thing happened to Sraffa’s ideas. They were originally voiced in a certain intellectual milieu, they tackled certain problems that arose in that milieu, and they had a definite impact on the academic discourse about economics. But certainly after the Wende, three decades later, they were increasingly “pushed aside” and overtaken by new preoccupations and concerns, such as “globalization”, the “new world (dis-)order”, “financialization” and all sorts of new developments in economic life. In retrospect, I can understand it, if people think now that Sraffa’s ideas are rather “absurd”. But at the time he published, he was revered by many as the “real MacCoy” who struck a blow at the neoclassical paradigm. That paradigm has persisted, even through deep crises. Why?

    Partly because it is a faith in the ultimately rationality of the market economy, even if we have yet to discover why it is rational, and because of the lack of any comprehensive alternative. But partly also because it is an “eclectic kaleidoscope” theory, an amalgam of a cluster of partial theories dealing with different aspects of economic life. When big economic events happen, largely unforeseen, therefore, there is no one clear conclusion about why all that happened. It could be explained in many different kinds of ways, all kinds of conclusions and lessons could be drawn which might well (and do) contradict each other.

    It is not true though that Sraffa thought “that physical material relationships are all that is necessary to determine price.” That is because the distribution of the “surplus” (the net value-added) is not exclusively economically, technologically or commercially determined. Workers can fight for better wages and better jobs, and employers can cut wages and jobs, flexibilize contracts etc. that is a dimension of competition and the outcome of that is not predetermined. That was very clear for example in the 1970s.

    When I said that Sraffa’s theory is pitched at an “astronomically high level of abstraction”, what I meant is, that his models involved a highly idealized, stylized, and simplified picture of economic life, which disregarded all influences which he deemed non-essential to the issues he was dealing with. This is common in economic discourse and model building, which operates constantly with “ceteris paribus clauses” (assuming “other things being equal”). Sraffa regarded it as a prologue or a foundation for a much better economic theory, he aimed only to tackle some core issues in classical economics.

    Personally I wouldn’t disparage mathematics, I realize it is enormously important to understand the proportions and implications of the claims staked out about economic and social life. Although I worked as research statistician for some years, I am not a great mathematician myself, but I regret losing my notebooks in which I dutifully wrote down the principles of algebra, trigonometry, calculus, analysis of variance etc. Mathematics and statistics are necessary, to see things in true proportion.

    When we are confronted with a problem, we want in the first instance to understand properly what the problem is. But we also want to know how big or small the problem is, why it matters, and what the implications are. For this quantitative insight is often very important. About half of all adults in advanced capitalist societies like Britain only have numeracy skills at the level of an 11-year-old child. If the population had much better grasp of math, I think we would likely have a better and more prosperous society.

    There is of course also such a thing as the fetish of numbers, numbers can be used to bamboozle people, particularly people who have no real grasp of what it means, sometimes the reverence for mathematics is misplaced. Nevertheless I think it is an important tool to help solve problems.

    As I have already said, I disagree with your commensurability thesis and your thesis that the Sraffian surplus appears out of nowhere. Sraffa never believed that the surplus appears out of nowhere, to the contrary he identifies determinants of its size, principally the productivity achieved with given technology.

    I can understand it, if you think that Sraffa disregards “at a high level of abstraction” all sorts of aspects of human labour and production which are essential to study from a Marxian perspective. But why aim your barbs at Sraffa specifically, after he has receded into relative obscurity? Most people today, including economists, wouldn’t even know who he was, or what he argued, or what the background was for what he said. You would be better off taking aim at Greg Mankiw, or somebody like that. The world has moved on, and the disputes we had decades ago no longer speak much to the controversies that there are today, or at any rate, they take quite a different form.

  21. I don’t disparage maths, linear algebra is always fun, although not before breakfast. I think most of your explanation for the Marxian fascination with Sraffa is correct. They wanted a stick to beat Marx with, and there were subjective and objective reasons for that.
    I also agree the debate is largely archane. Although if Marxism is ever to revive (who knows) its important to settle scores with its attackers.
    Sraffa’s model is basically a one commodity model. These are widely used by economists to model – god knows what – as of course they exclude production and so economics. But since when did that ever stop anyone?

  22. Well actually Sraffa starts his book PCMC with not one, but *two* commodities produced by two industries (wheat and iron), and then he adds the production of pigs, with perhaps a whiff of satire. He then investigates briefly the conditions for balanced reproduction of exchanges in this simple triangular production system.

    I think myself, that one problem with the discussion about “production prices” is that Marx’s theory has been interpreted as being quite close to Ricardo’s theory of production prices or “natural prices”. I think that is not the case, and I think a careful examination of Capital Vol. 3 can show this. But I haven’t worked this out in fine detail yet, I am putting that aside, until I have worked out some issues in my own life.

    One of the advantages of an econometric approach to values and prices is, that you have to get very clear about the concepts for which you try to find measures. So I regard the econometric approach as beneficial in that sense (also, in showing which magnitudes really matter empirically, for the overall results, or what difference particular choices of measures can make to the results).

    Generally, the research appears to show mostly a close statistical correlation between movements in labour requirements and movements in price levels, just as we would expect, and that result doesn’t even change significantly, when we adopt different kinds of relevant measures. The dismissal of a “labour theory of value”, or great controversy about the transformation problem is therefore rather odd, because the data show that there just isn’t a good reason for that.

    But I am somewhat heretical, because I think that Marx did not actually have a labour theory of “value”, he had a labour theory of “product value”, a labour theory of the values and prices of commodities. He actually never referred to “the labour theory of value”, only to “the theory of value” which across several centuries sought to discover the forces objectively regulating market production, distribution and consumption, beyond moral notions of fair trade. Marx’s theory was a theory of the capitalist mode of production, but not of the whole capitalist economy.

    Judging by his comments on banking and the credit system, Marx was quite aware that the value of many assets is *not* regulated by labour requirements, at least not directly. This is somewhat confusing to many people, because how then do you define the “value” of, say, financial assets? There are different approaches to this problem, but I think it can be solved. Basically, it means that there are dimensions and modalities of values and prices that Marx never analyzed himself, but which assume increasing significance in the developed late capitalist economy. There is already a burgeoning literature about that.

    The classical reproduction model was clearly a model of production, of how the production system can generate the conditions for its own reproduction. This is also the perspective that Sraffa had, although admittedly he does not analyze the production process itself, only the circulation of capitals via production, at a high level of abstraction. But the forward march of the accumulation of capital has as a consequence, that a greater and greater stock of assets is accumulated, which is neither an input nor an output of production, and which alters the whole economic reproduction process, and indeed the structure of the whole world economy.

    Next to production capital, that is, there is an ever increasing stock of capital assets external to production, which is also traded for profit in various forms. So you get more and more trade in already existing assets, i.e. additional circuits of trade. Thus, whereas there is an observed long-run tendency for the rate of profit on production capitals to fall, this does not by itself cause a collapse of the system, because of the existence of a quantitatively very important sector of capital assets external to production.

    More likely, the biggest financial “system shock” for this type of capitalist economy will be huge debts which cannot be repaid at all. How that will be resolved, is as yet unclear, but it is likely that it would have to involve a drastic overhaul of the whole financial system and very strong state intervention. Non-financial “system shocks” are of course also possible.

    And there is the political outfall of all that, so that what we call the “neo-liberal” regime could possibly turn into its opposite, a very authoritarian or even a military regime which adjusts human behaviour by force, to keep things going. There are many historical examples already to illustrate what that involves, but they could become much more pervasive, simply because of the malfunctioning of market trade almost everywhere.

    However, nobody really knows for sure what will happen in the longer term. There are all sorts of possibilities, variants and scenarios being mooted. It is rather futile to engage in somber predictions of the distant future, what matters is what you can achieve in the here and now. All you can do, is the best you can for your own side. Most of the elements of a better solution to the predicaments of modern society already exist, it is more a question of whether they will become generalized in their application, or not. That is ultimately a political issue, a matter of human volitions. History shows us that sometimes things have to get pretty bad, before people are driven to do something better.

    There is obviously a huge amount to be said about all these issues, I can only give a quick “thumbnail sketch” here. I don’t pretend to have all the answers, I am more often preoccupied with the questions.

  23. Well he has a number of commodities, in fact you could have any number of commodities, provided its a finite set that never changes.
    These commodities have to be unchanged by the production process, identical as inputs and outputs, to enable direct physical commensurability. Further they must always be in relatively the same proportions, such that the standard commodity (a composite of all commodities or any commodity in fact) could serve as a standard of value for any commodity.
    This is a one commodity model as any commodity can substitute for any other commodity and the relations between them are fixed. As the only commodity that can be fixed relative to itself, is itself, it means that every commodity is in fact the same commodity. Such that it is de facto, if not de jure, a one commodity economy.
    Surplus arises without equivalent as, if the price of a commodity is its cost of production (the amount of commodities destroyed in producing it) then there can be no surplus by definition.
    The “surplus” has to be additional to the cost of production and so cannot have been produced in production, so it has come from nothing. It cannot have an equivalent for if it does then it is not a surplus. As the surplus is made from nothing so it is nothing. If its value/price (same thing) can be measured in the original goods (it must be or it has no value as its value cannot be measured) so they too are nothing. It is a model of nothing.

  24. Hey Bill, I don’t know what is going on in your world, or why you feel the need to repeat your nihilistic gripe about Piero Sraffa to the nth degree, but this comment of yours is totally over the top.

    I don’t get any indication that you understand what Sraffa’s argument actually is, or that you understand the economic meaning of the concepts of gross output, net output, value-added and surplus, and it appears to me that you are willfully misrepresenting Sraffa’s theory for some reason that I can only guess at, given your political background and stance.

    In order to have any constructive scientific dialogue, there has to be at least some common ground and respect for the evidence. If we cannot even agree at the most basic level about what a carefully written text means and intends, no fruitful discussion is possible. You make it very clear, that our views about this matter are “incommensurable” – there is no common ground.

    For that reason, I think there is no point in pursuing our conversation any further. All I can do, is wish you luck in your career as British academic – hoping that the light of Reason will reach you some day – and leave it at that.

  25. I think Jeffеries is absolutely right and that he is one of the few who questioned Srafa’s absurd assumptions. These comments are similar to Araujo’s reply, which doesn’t address Jeffries’ main point about Srafa’s model. The only thing that remains unclear to me is why so many marxists have accepted these assumptions.

  26. TBH I think its all the other way around. Having said that, the result’s the same. Best.

  27. @Nelson: I don’t actually think it is true that “many Marxists accepted the Sraffian assumptions”. As soon as critics attacked the labour theory of value with a Sraffian perspective in the 1970s, there were a multitude of vigorous and sharp replies and rebuttals from Marxist scholars. The contest was never firmly decided in favour of Sraffa, by any means. The rifts in opinion went very deep indeed.

    What is probably true, is that after the Wende, a lot of supporters of radical economics dropped their Marxist affiliations, and opted for heterodox economics. There just weren’t so many jobs anymore for Marxian scholars, as a matter of fact.

    In turn, that left a rather smallish group of dedicated Marxist scholars – some of whom nevertheless became or were quite well-known – who specifically promoted Marx’s value theory and value-form theory as a theoretical foundation of the Marxist endeavour.

    Allow me to briefly cite two highly respected Marxian scholars, Ben Fine and Alfredo Saad-Filho, as a piece of evidence. They commented on contemporary Marxist economics, and said in 2018 that:

    “The key to the continuing relevance and analytical strengths of Marxist political economy lies in its capacity to provide a framework of analysis for unifying disparate insights into and critiques of the contradictions of capitalism across the social sciences. The instrument for forging that unity is Marx’s theory of value.”

    Whether or not you agree with that, that reflects pretty much the state of the discourse. The idea here is, that if you strip away Marx’s theory of value, then there is nothing much left of Marx’s critique of capitalism. I think myself that there is a whole lot more to the life and work of Marx and Engels (and many of their more intelligent followers). Okay, they lived in the 19th century and we live in the 21st… even so, we are still grappling with many of the same sorts of questions as they did, in whatever languages we use to express it.

    The predicament of “dissident” or “system-critical” economists – of whom there are still quite a lot – is, that on the one side, economics has to be concerned with the optimal allocation of scarce resources. If you aren’t concerned with that as economist, you aren’t yet much of an economist, you have to be able to provide some real answers to that, even if they are relatively modest.

    On the other side, such economists also have a critique of how resources are actually allocated in the capitalist system, since that is often far from “optimal”. Point is, you don’t actually need to be a certified Marxist, to be convinced of that. Reality cries out for better answers to the problems there are.

    As an economist or theoretician, you have to be able to live, study and do your scientific work, if you have this combination of views, as your position. Not always an easy ride, particularly in a highly reactive, internetized world, where everything you do is under inspection, and invites instant responses to what you say and do, even to the smallest details.

    Marxian economists will accentuate the critique of capitalism, and super-radical people will cheer them on. But real economists also have to be concerned with real and workable alternatives. If all you have to say, is that the solution is, for the working class to abolish capitalism and the wages system, and that all the rest is just putting band-aids on the capitalist order, you are just not going to get much traction these days, as economist, or in just about any other job. Such a stance might have been fine in 1975, but it is not so today.

    Yep, that suggests contradictions in the system-critical endeavour, and leftwing economists have to find their own solutions to dealing with those contradictions. It can be a real struggle. A very painful struggle at times. The question then is, do you want to be supportive to them, or would you rather have that they are not there at all, because they are a “reformist scourge” detracting from the revolutionary path of the world proletariat?

    Personally, I think the most sensible, savvy and sober attitude is, that you ought to be very mindful of who your real friends and enemies actually are, in this world as it really is. Believe me, leftwing economists aren’t really the problem. And, I think that it is a good idea, if people on the Left would focus more on getting more friends, academic or non-academic.

    To make new friends, we have to accept, that real people have to live with and struggle with the contradictions of life, and that they are entitled to their own beliefs and strivings. We don’t really help people along, if we deny the validity of their beliefs, because they don’t happen to conform to our own standard of what is the correct ideology.

    We need a real, constructive and honest ongoing dialogue, so that we can together achieve better understandings and better actions. Would you talk to anybody? Probably not. But there are places where you can exchange ideas.

    The root meaning of dialectics is “dialogos”, i.e. a dialogue, a living dialogue between different viewpoints. Democracy is effectively dead, without a real and meaningful dialogue. All real democrats affirm this. To have that dialogue, we must allow people to say what they want to say, within a framework of civilized norms.

    Okay you have to take responsibility for what you say, but there has to be a space for people to say it, even if they are embroiled in highly contradictory situations. Those spaces are there. We can use them wisely, but also allow for experiments in dialogue of which we do not know in advance whether they will succeed.

    I am not trying to be a manic preacher here, or sell a moral panacea, but I do believe this is the way to go. That is also why I thought this site and its formats were a good idea. I can learn all sorts of things from it, from thoughtful, talented and pretty rational people from across the world who think differently than I do, and unexpectedly find that I have much in common with them too.

  28. I think you would be hard put to find a single Marxist economist who does not agree with Sraffa, at least in the sense of the justness of the implicit critique of Marx.
    If you want to say why was Sraffa’s critique so successful, when it’s basically nonsense, then it comes back to the influence of Dobb and Meek and linear algebra.
    Anwar Shaikh for example praises Dobb’s role in particular. Dobb promoted Sraffa as he saw his physical price system as analogous to the material product system of the Soviet plan, which had physical balances of a similar kind to Sraffa. Although of course, the USSR was a centrally planned economy where the issue of commensurability was irrelevant, as nothing was bought or sold. Stalin had famously announced there was “value” in the plan and Dobb thought Sraffa was the way to find it.
    The other key influence is linear algebra. Leontief’s input-output analysis of the US economy built on the Soviet Balance of 1923/24 to show how the interrelationships between sectors of the economy could be modelled with linear algebra.
    Famously linear algebra also implied that Marx’s solution to the transformation question was “logically mistaken” and so value theory, it was implied, was wrong. Marxist mathematicians worked long and hard to find an alternative mathematical solution, but basically failed. There isn’t one at the level of maths.
    In fact, all that was actually required was to embrace the contradiction. Marx’s model showed that there was a disproportion in the transition between values and prices of production. The “solution” to the transformation question is simply to check whether there was. There was.
    But this claimed logical mistake, combined with a mathematically correct (albeit absurd) Sraffian alternative, was enough for Marxians to dump value theory in favour of Sraffa.

  29. I regard all your claims about Sraffa and Dobb as false, slanderous and nonsensical, and I am not going to discuss it any further, it serves no good purpose and is a waste of time. I hope that is clear now.

  30. Thank you for your answers. However, I think that Jefferies convincingly showed, especially in his articles, but also in these comments, that Sraffa’s model is not a model of production and that his assumptions are absurd. Where did he go wrong? I read all the comments and his critique of Sraffa seems completely correct to me.

Make a comment

Your email address will not be published.