Chris Harman

Zombie Capitalism

Part One: Understanding the System: Marx and Beyond

* * *

Marx and His Critics

The neoclassical critique of his theory of value

Marx’s theory of value has been under attack ever since Capital was first published. The most common form this attack takes is to claim that capital as well as labour creates value. After all, it is said, a worker using a machine produces much more than a worker without one, and all the time workers are being replaced by machines that do the same job. It is even possible to conceive of an economy in which all work is done by machines. So neoclassical economists argue that not only labour but also capital is involved in producing things which satisfy human need. And just as labour gets paid according to what it contributes to wealth creation, so does capital. Each “factor of production” gets a “reward” equal to its “marginal output”.

There is a central fallacy in this argument against Marx. It rests of a static picture of the economy, in which capital and labour simply exist alongside each other. It ignores the palpable reality that the means and materials of production themselves have been produced. Machines and factory buildings are not things that exist in their own right. They are the product of previous human labour. The wheelbarrow which aids the toil of the labourer is itself the product of the toil of the metal worker. That was why Marx called the means of production “dead labour” (as opposed to present work, which is “living labour”). They are the products of labour that has taken place previously – and can, if necessary, be replicated by the application of labour today. The amount of socially necessary labour needed to reproduce them determines their current worth.

The failure of neoclassical theory to take into account the creation of the means of production by labour is no accidental failing. Its founders in the late 19th century – the Austrians Carl Menger and Eugen von Böhm-Bawerk, the Englishmen W. Stanley Jevons and Alfred Marshall, the Frenchman Leon Walras, the Italian Vilfredo Pareto, and the American John Bates Clark – built the assumption of a static system into their theory. They viewed the whole economy as like a street market where the buyers calculate what combination of goods gives them the best value for the money they have got in their pockets, while the stallholders calculate the best price they can get for each of their goods. The mutual adjustment of the price each seller is willing to accept and each buyer is willing to pay leads to all the goods being sold. And, since each seller is in turn the buyer from someone who has bought from someone else, a whole network of prices is set up which ensures that what is produced is exactly what people want. Walras claimed to show how this works for a national economy with hundreds of pages of equations and graphs.

Inbuilt into the whole approach was a very unreal view of capitalism. For, whatever else capitalism is, it is not static. In a real street market, people do not agree instantaneously on the prices for buying and selling. But neoclassical theory assumed that through the mediation of a central auctioneer they could arrive at agreed prices instantaneously. Real life haggling often takes quite a long time, with prices across the market as a whole being arrived at through a process of successive adjustments. Once that is taken into account divergences open up between the actual prices of goods and those presupposed in the theory. The actual production of goods that are to be sold is always a process taking place in time. “Price signals” do not tell you what will be wanted when production is finished, but what was wanted before it began. The simultaneity of the theory is a myth, and the simultaneous equations developed on the basis of its assumptions bear little relation to really existing capitalism.

Faced with the reality that production occurs over time, how did the founders of neoclassical economics react? They did not allow it to affect their theory one iota. Walras, for instance, recognised that “production requires a certain lapse of time”. But he then wrote he would deal with the “difficulty purely and simply by ignoring the time element at this point” [1]; and when he returned to the issue it was to assume that “data” remained “constant for a given period of time” [2], as if the transformation of the whole productive apparatus with economic growth would not mean continual transformation of the structure of supply and demand. Marshall went as far as to admit that “time is the source of many of the greatest difficulties in economics” [3], since “changes in the volume of production, in its methods, and its costs are ever mutually modifying one another.” This did not, however, stop him teaching the theory and a whole generation of mainstream economists taking it as proof of the efficacy of market capitalism. An updated version of Walras’s mathematical model was produced in the early 1950s by Kenneth Arrow and Gerard Debreu which attempted to take time into account. But Arrow himself recognised that the model only works “if you assume no technological progress, no growth in population and lots of other things”. [4]

The refusal of the neoclassical school to grasp the basic point that capitalism is a system undergoing continual transformation that disrupts the old price structure and prevents any settled equilibrium means that it provides at best an apologetic description of things as they exist at any moment in time, not an account of economic developments and dynamics.

The neoclassical school’s own theory of value, which it counterposes to the theories of Smith, Ricardo and Marx, is in terms of the utility which a commodity gives – that is, how individuals evaluate the commodity compared to other commodities. But this leaves completely unresolved the basis for measuring the utility for one person compared to another. How do you measure the “utility” of a glass of water to someone in the desert with the “utility” of a diamond tiara to a princess? The most you can do is list the preferences of individuals. But to explain why the preferences of some individuals matter more than the preferences of others you have to explain why some are wealthier than others – and that depends on factors to do with the structure and dynamic of capitalist society which “utility” theory ignores.

Pareto replaced the term “utility” by “ophelimity” [5] because, as his American contemporary Irving Fischer put it, “the great untutored and naive public ... find it hard to call an overcoat no more truly useful than a necklace, or a grindstone than a roulette wheel”. [6] Some later neoclassical theorists dropped any notion of value altogether – although “marginal utility” continues to be taught in school and college textbooks to this day as the “modern” answer to the labour theory of value.

Neoclassical economists have not succeeded in giving an objective basis to their theory of value, despite more than a century of effort. Of course, ultimately, someone must want to use something (or, at least be able to sell it to someone else who will use it) if they are going to pay for it. But it is not use that determines price.

Nor can “marginal output” as defined by the neoclassical school provide an answer. This is measured, they argue, by the value of the capital used up in producing it; but when they define the value of that capital they do so in terms of the marginal output. They end up saying, in effect, that “the marginal value of capital equals the marginal value of capital”, or “profit equals profit”. Statements of this sort cannot explain anything. All they do is to state that if something exists, it exists.

Orthodox economics in fact does no more than state that certain things are bought and certain things are sold at present, without explaining why these things are produced and not others, why some people are rich and some poor, and why some goods pile up unsold while people who desperately need them go without, or why sometimes there are booms and at other times slumps.

These points were made against marginal economics more than 80 years ago by the Austrian Marxist Rudolf Hilferding and the Russian revolutionary Nicolai Bukharin. They have been put across more recently in a rigorously logical form by dissident academic economists known as the “Cambridge School”. [7] But the capacity of dissident economists to point out the absurdities in neoclassical theory has not weakened its hold on academic economics. It has simply led to ever more obtuse mathematical models being used to provide an appearance of scientific rigour. As Joan Robinson pointed out half a century ago:

Quantitative utility has long since evaporated but it is still common to set up a model in which quantities of “capital” appear, without any indication of what it is supposed to be a quantity of. Just as the problem of giving an operational meaning to utility used to be avoided by putting it into a diagram, so the problem of giving a meaning to the quantity of “capital” is evaded by putting it into algebra. [8]

Recognition of the difficulties with their own theory has, on occasions, forced those who otherwise accept the neoclassical system to try to reinforce it with elements from the labour theory of value. So Marshall suggested there might occasionally be merit in using a labour theory of value: “the real value of money is better measured for some purposes in labour rather than in commodities”, although he hastened to add, “This difficulty will not affect our work in the present volume ...” [9] John Maynard Keynes also half grasped the limitations of the very neoclassical system whose postulates he took for granted. At one point in his most famous work, the General Theory of Employment, Money and Interest, he recognised that you cannot simply add together different sets of physical commodities at one point in time and compare them with a different set at a later point. [10] To make such comparisons involves “covertly introducing changes in value”. [11] To deal with this problem, he dropped the usual assumptions of neoclassical theory and made half a turn to a labour theory of value [12], suggesting that output could be measured by “the amount of employment associated with a given capital equipment”. [13] He explained later:

I sympathise with the pre-classical [sic] doctrine that everything is produced by labour, aided by what used to be called art and is now called technique, by natural resources ... and by the results of past labour, embodied in assets ...” [14]

Neither Marshall nor Keynes was prepared to go further and jettison the neoclassical system as a whole. But if they had taken their own observations of these points seriously, they would have been compelled to do so.

The failings of the neoclassical system provide at least a partial, negative, proof of Marx’s approach. For his theory of value avoids such a subjective and static approach. Marx’s theory is objective because it is not based on individual evaluations of a commodity, but on the necessary amount of labour needed to produce it given the level of technology existing in the system as a whole at a particular point in time – both the direct living labour of the worker and “dead labour” embodied in the equipment and materials of production used up in the production process. For Marx, it is the pressure different capitals exert on each other, not the evaluations of individuals, that matters, since any capitalist who prices a commodity at a level higher than the amount of socially necessary labour needed to produce it will soon be driven out of business. The law of value is therefore an external force operating on every capitalist through the interaction of all capitalists once there is the general production of commodities for exchange mediated by money. Since the “individual capitalists”, writes Marx, “confront one another only as commodity-owners, the ‘inner law’ enforces itself only through their competition, their mutual pressure upon each other, whereby the deviations are mutually cancelled”. [15]

The relation between capitals cannot be understood as something fixed and unchanging. It is a dynamic process, based on the interaction through time of different capitals, so that the average “socially necessary labour” at any point is the result of individual processes of production organised independently of each other with different, often changing, amounts of concrete labour. The first capitalist to introduce a new technique in any section of industry will be able to produce goods with less than the amounts of labour prevailing in the system as a whole, and will be able to capture markets from others. But once other capitalists adopt the technique, this advantage is lost. Only a capitalist who controls a very large part of the market for particular commodities, or who can exert political pressure to impede others accessing his markets, will be able to get away for longer or shorter periods of time with charging prices which reflect amounts of labour higher than those that are socially necessary. The law of value only operates as the result of the pressures these different capitals exert on each other through time. Any still photograph from the moving film of capitalist development will always show discrepancies – and sometimes large ones – from the law of value. But the film itself will show the discrepancies eventually disappearing under the pressure of inter-capitalist competition even as other discrepancies arise.

Value and prices

It is this dynamic aspect to Marx’s theory that enables it to deal with a problem that beset the attempts by Smith and Ricardo to base value on labour. This is that the ratio of labour to investment varies from industry to industry. Yet in practice the rate of profit (the ratio of the surplus value to investment) does not vary in the same way, even when wages are more or less the same level and the rate of exploitation must be about the same. The prices of goods seem to depend not on the amount of socially necessary labour needed to produce them, but on a mark up on the cost of capital investment. The bigger the capital investment the bigger, it seems, is the mark up. A capitalist selling something produced by one person working on an expensive machine will expect a bigger mark up than for something produced by one person working on a cheap machine. The fact that some industries are more “capital intensive” than others implies that prices have to diverge from values in terms of labour if profitability is not to be much lower in some cases than others.

This is what led Adam Smith to dilute his labour theory of value with another, contradictory approach. The sale of goods produces a payment that is divided up into different “revenues” – wages for the workers, profit for the industrialist, interest for the banker who lent the industrialist money, and rent for the landlord. Smith contradicts his own initial labour-based theory of value by arguing that each of these revenues adds to value. David Ricardo was more consistent than Smith and tried to stick to the pure labour theory. But this left a gap in his theory that economists who came after him could not solve – one which eventually opened the way to the neoclassical abandonment of the labour theory of value.

Marx could, however, deal with the problem – usually called the “transformation problem” – precisely because his model is a dynamic one that operates through time. His solution depends on looking at how firms will react to the emergence of different profit rates. Those with lower profit rates will begin to move their capital elsewhere. This will cause a potential shortage in what they have been producing, leading to a rise in prices above their value in labour terms. Other firms who use those products as inputs to their own production (either directly or through paying their workers to buy them to replenish their labour power) [16] are forced to pay the higher prices, in the process effectively handing over some of the surplus value in their own hands. The equalisation of the rate of profit takes place through the redistribution of surplus value within the capitalist class.

This does not in any degree alter the fact that the surplus value came from the exploitation of workers in the first place, and that every change in the socially necessary labour time needed to produce a commodity has an effect on its price. It is the flow of already produced surplus value from one capitalist to another through time that equalises the rate of profit [17] – which is also why there can be big differences between the rates of profit in different parts of the system when there are impediments to the flow of value through the system (for instance, when firms have very large amounts of investment tied down immovably in certain sorts of fixed capital or when states prevent investment moving out of what they see as priority industries).

Marx’s solution to the problem posed in Smith and Ricardo was attacked within two years of it appearing in Volume Three of Capital by the marginalist Böhm-Bawerk. The same arguments he used have been employed repeatedly every since. They have often thrown Marxists onto the defensive, with many accepting the core of the criticism and retreating from the attempt to understand the dynamics of capitalism using Marx’s concepts. This happened, for instance, soon after the revival of interest in Marxism after the events of 1968. Figures on the left such as Ian Steedman and Geoff Hodgson took up arguments essentially the same as those used against Marx by Böhm-Bawerk (although they did not accept the marginalist theory of value) and his successors like Samuelson. [18 ]Marxist scholarship, already on the defensive for political reasons inside university economics faculties, often retreated into scholastic debates over texts or into obtuse mathematical calculations as remote from the real world as those of their mainstream colleagues. The result overall was, as Ben Fine has put it, “an increasingly and exclusively academicised Marxism” [19] and “limited engagement with the world of capital as opposed to that of Capital”. [20]

The criticism of Marx’s approach centres around the contention that simply looking at the movement of value between capitals after production has taken place cannot explain final prices, since it does not explain the prices of the inputs into production (the means of production and labour power). For the inputs themselves are commodities with prices different to their values. So Marx’s method, it is claimed, explains prices in terms of prices, not in terms of labour values. [21]

The Ricardian, Ladislavs von Bortkiewicz, attempted in 1907 to solve the problem of deducing prices from labour values mathematically, using simultaneous equations. He used a model in which there is no change in the amount of capital investment from one cycle of production to the next (what is called “simple reproduction”). His equations supposedly showed that any attempt to provide a generally applicable way of transforming labour values into prices led to one of the “equalities” taken for granted by Marx not working. Either total price did not equal total value, or total profit did not equal total surplus value.

Every attempt to deduce prices from values for most of the 20th century ran into the same problem. The response of Marxists was either to abandon the central feature of the labour theory of value or to conclude, as for instance Paul Sweezy did in 1942, that “the Marxian method of transformation is logically unsatisfactory” but that the “patterns of development” of value and price “will differ only in minor details”. [22] A somewhat similar conclusion was arrived at by Miguel Angel Garcia and Anwar Shaikh among others in the late 1970s using models that were much less mathematical and easier to follow than von Bortkiewicz’s. [23] Shaikh showed that total price could equal total value, but total profit would not always equal total surplus value. Garcia claimed to prove that both equalities could hold. But he could only do so by allowing a change in the rate of exploitation from one production cycle to the next, since the shift in prices caused by the movements of value between sectors caused changes in the relative prices of wage goods and capital goods. [24]

Since then, however, a number of Marxists have been able completely to rescue Marx’s position by challenging the fundamental assumption made by von Bortkiewicz, Sweezy, Shaikh and many others – the reliance on simultaneity. [25] The method of simultaneous equations assumes that the prices of the inputs to production have to equal the prices of the outputs. But they do not. The outputs are produced after the inputs have gone into production. Or, to put it another way, the value of the inputs for a process A will differ from that of the same inputs for a later process B – even if in terms of their physical composition as use values they are identical. The value of a ton of steel used to make a machine today will not be the same as the value used to make an identical machine next week. [26]

But, argue the critics of Marx, this still leaves the inputs into production as prices, not as values, and to reduce them to labour values involves an infinite regress. The investment need to produce the inputs needs to be broken down into labour values, but that is not possible without breaking down in turn the investment needed to produce it, ad infinitum.

There is a simple response to those who pose the problem like this: Why? Why do the investments needed to produce the inputs have to be broken down in terms of their labour value when they themselves were produced? [27]

The starting point for looking at any cycle of production is the money price of the inputs needed to undertake it. The exercise of labour in the production process then adds a certain amount of new value, which forms the basis of the new commodity, the price of which in turn is formed through the movement of surplus value from capitalists who would otherwise get a higher than average rate of profit to those who would get a lower one.

There is no need to go back in history to decompose into labour values the prices of things which were paid for at the beginning of the production round, in order to understand the impact of creating new value and surplus value on the dynamics of the system. It is no more necessary than it is in physical dynamics to decompose the momentum of an object that strikes another into all the forces that have previously acted on it to create that momentum, going right back to the foundation of the universe with the big bang; or than it is necessary in biology to know the whole history of the evolution of an organism, going right back to the first formation of organic life forms, in order to see what the effect of a genetic change will be in the present.

As Guglielmo Carchedi has pointed out, “If this critique were sound, it would mean the bankruptcy not only of Marx’s transformation procedure but also of social science in all its versions” including those that criticise Marx:

This critique, in fact, would have to apply to any social phenomenon inasmuch as it is determined by other phenomena, both present and past. Social sciences, then, would become an endless quest for the starting point of the inquiry. [28]

It would never be possible to analyse how some current actions related to the cumulative products of past actions.

Skilled and unskilled labour

The same dynamic character of Marx’s model also dispels what has been presented from Böhm-Bawerk onwards as another problem for the labour theory of value. This is how the contribution of skilled labour to value creation is to be measured. Marx seems to see this as easily solved. He writes that:

skilled labour counts only as simple labour intensified, or rather, as multiplied simple labour, a given quantity of skilled being considered equal to a greater quantity of simple labour ... The different proportions in which the different sorts of labour are reduced to unskilled labour as their standard, are established by a social process that goes on behind the back of the producers and, consequently, appear to be fixed by custom. [29]

This explanation is fully adequate when the same job is done by a skilled worker and unskilled worker, with the skilled worker doing it much more quickly. An hour of the skilled labour will be worth more than one hour of the average “socially necessary” labour in the system as a whole, while the unskilled labour will be worth less than that.

There is a problem, however, when it comes to skilled labour that cannot be replaced by a greater quantity of unskilled labour. It does not matter how many unskilled labourers a capitalist employs, they will never be able to do the same task as a skilled toolmaker or a systems analyst. How then can the value produced by the second group be measured in terms of hours of labour of the first group? It seems that any attempt to do so must involve an arbitrariness that undermines the basic theory. Böhm-Bawerk argued that when Marx writes that “a social process” explains the measurement, he is taking for granted that which he is trying to explain. For Böhm-Bawerk this proved that it is not the amount of labour in goods which determines their prices, but the way people evaluate them in relation to other goods (their “utility”) and that this deals a death blow to the labour theory of value.

However, the problem for the theory evaporates once the law of value is seen as something working through time. The development of technology again and again leads to jobs emerging that can only be carried out by those with particular skills. At first there is no objective measure of the amount of socially necessary labour time needed to produce them, and those in possession of such skills or the goods produced by them can receive payments which bear no obvious relation to labour time. In effect, value flows to those controlling a monopoly of these skills from the rest of the system. But this is only a transitory phase, even if sometimes a long one, as capitalists elsewhere in the system will do their utmost to try to gain control of some of the benefits of the new skills for themselves.

There are two ways they can do this. They can train new groups of workers to acquire the skills. This effectively amounts to using one sort of labour to create new labour power capable of doing the skilled work, so that the final labour is in fact composite labour, made up of the living labour of the immediate workers and a form of dead labour embodied in their labour power as skills. The capitalists can get this extra element in labour power directly by on the job training for workers (as with apprenticeship systems), they can leave its provision to the workers themselves (when workers pay to go through courses to get skills qualifications) or they can rely in part on the state providing it through its training courses. But in each case, dead labour is embodied in the enhanced labour power and then transferred into the products of the labour process, as with the dead labour embodied in means and materials of production. [30]

But this still leaves a question unresolved. Who trains the trainers? Skilled trainers cannot themselves get their skills from unskilled workers. If their skills are monopoly skills and they produce goods that cannot be produced by unskilled workers, however many work together at the job, then those who own those goods will be able to charge monopoly prices that do not reflect labour values, but simply how much people are prepared to pay.

This will be true of certain skills and certain goods at any particular point in time. But over time this labour too will be reduced to some objective ratio of other labour. Capitalists elsewhere in the system will actively seek out new technologies that undermine such skill monopolies by enabling the tasks to be done by much less skilled labour. In this way, the reduction of skilled labour to unskilled labour over time is a never ending feature of capitalist accumulation. If enough unskilled labour is trained up to the level of skilled labour needed to produce particular commodities, those commodities will cease to be scarce and their value will fall to the level that reflects the combination of the labour needed to reproduce average labour power and the extra cost of the training.

As Carchedi puts it:

Due to the introduction of new techniques in the labour process, the level of skills required of an agent is lowered. The value of his or her labour power is then devalued. We can refer to this process as devaluation (of labour power) through dequalification (of skills). It is this process which reduces skilled to unskilled labour and thus (at least as far as the value of labour power is concerned) alters the exchange relations between the commodities of which those different types of labour power are an input. It is this real process which justifies the theoretical reduction of skilled to unskilled labour, or the expression of the former as a multiple of the latter ...

The process of devaluation through dequalification is a constant tendency in capitalist production, due to the constant need capitalists have to reduce the level of wages. On the other hand, the same techniques create new and qualified positions (the counter-tendency) which, in their turn, are soon subjected to dequalification ... At any moment in time we can observe both the tendency (the dequalification of certain positions and thus the devaluation of the agents’ labour power) and the counter-tendency (the creation of new, qualified positions for which agents with a high value of labour power are needed). [31]

Labour may not be reducible to socially necessary labour time instantaneously. But it is so reduced over time through the blind interaction of different capitals with each other. Again the law of value has to be understood as pressurising the individual components to operate in a certain way, not as a formula establishing fixed, fast frozen relations between them.

Marx’s basic concepts survive all the criticisms once they are not interpreted through the static framework, ignoring the process of change through time that characterises the neoclassical system.

* * *


1. Leon Walras, Elements of Pure Economics (London, George Allen, 1954 [1889]), p. 242.

2. As above, p. 372.

3. Alfred Marshall, The Principles of Economics, 8th edition (London, 1936), p. 109.

4. Interview with Kenneth J. Arrow, in G.R. Feiwel (ed.), Joan Robinson and Modern Economic Theory (London, Macmillan, 1989), pp. 147–148.

5. Meaning “desirability”.

6. Irving Fischer, Is ‘Utility’ the Most Suitable Term for the Concept it is Used to Denote?, American Economic Review, Vol. 8 (1918), pp. 335–7.

7. For a series of articles discussing this problem and the failure of different marginalist economists to deal with it, see J. Eatwell, M. Milgate and P. Newman (eds.), Capital Theory (London, Palgrave, 1990). The article by L.L. Pasinetti and R. Scazzieri, Capital Theory: Paradoxes, pp. 136–147, provides a useful and relatively accessible summary of the arguments. See also Joan Robinson, Economic Philosophy (London, Watts, 1962), p. 60.

8. Joan Robinson, Economic Philosophy, p. 68.

9. A. Marshall, The Principles of Economics, p. 62.

10. “Two sets of incommensurable collections of miscellaneous objects cannot in themselves provide the material for a quantitative analysis”, John Maynard Keynes, General Theory of Employment, Interest and Money (London, Macmillan, 1960), p. 39.

11. J.M. Keynes, General Theory of Employment, Interest and Money, p. 39.

12. Although it is a version in which the wage is the measure of value, rather than Ricardo’s version, where it is labour performed.

13. J.M. Keynes, General Theory of Employment, Interest and Money, p. 41.

14. As above, pp. 213–214.

15. Marx, Capital, Volume One, p. 858.

16. A rise in the price of basic consumer goods will cause workers to pay more, so raising the price capitalists have to pay for their labour power in the form of wages. Of course, in reality, none of this happens without struggles between different capitalists and between capitalists and workers.

17. This whole argument for Marx’s position is put simply and at length in Andrew Kliman, Reclaiming Marx’s Capital (Lexington Books, 2007), pp. 149–175.

18. See Ian Steedman, Marx after Sraffa (London, New Left Books, 1977); for an influential debate on the issue see Ian Steedman and others, The Value Controversy (London, Verso, 1981). See in the same collection, Geoff Hodgson, Critique of Wright: Labour and Profits, pp. 75–99. For the mainstream account of Marx, see P.A. Samuelson, Understanding the Marxian Notion of Exploitation: A Summary of the So-Called Transformation Problem between Marxian Values and Competitive Prices, Journal of Economic Literature, 9:2 (1971), pp. 399–431.

19. Ben Fine, Debating the ‘New’ Imperialism, Historical Materialism, 14:4 (2006), p. 135.

20. As above, p. 154.

21. For a fuller and very influential account of this critique, with mathematical presentations, see Paul Sweezy, The Theory of Capitalist Development (London, Dennis Dobson, 1946), pp. 115–123; see also G. Carchedi, Frontiers of Political Economy, (London, Verso, 1991), pp. 90–92; A. Kliman, Reclaiming Marx’s Capital, pp. 45–46.

22. P. Sweezy, The Theory of Capitalist Development, pp. 115 and 128.

23. Miguel Angel Garcia, Karl Marx and the Formation of the Average Rate of Profit, International Socialism, 2:5 (1979); Anwar Shaikh, Marx’s Theory of Value and the ‘Transformation Problem’, in Jesse Schwartz (ed.), The Subtle Anatomy of Capitalism (Santa Monica, Goodyear, 1977).

24. This is an argument I accepted and spelt out in my book Explaining the Crisis (London, Bookmarks, 1984), pp. 160–162.

25. They have formulated what are known as “temporal” interpretations of Marx.

26. For various presentations of this argument, using mathematical examples, see G. Carchedi, Frontiers of Political Economy, pp. 92–96; A. Kliman, Reclaiming Marx’s Capital, pp. 151–152; Alan Freeman, Marx without Equilibrium, MPRA Paper No. 1207 (2007) available at

27. This interpretation of Marx on this point is known as the “single system interpretation”. Some of those who hold to it also accept, like me, the temporal interpretation, with the combination of both being known by the cumbersome (and somewhat off-putting) phrase “The Temporal Single System Interpretation” or TSSI.

28. G. Carchedi, Frontiers of Political Economy, pp. 96–97.

29. Marx, Capital, Volume One, p. 44.

30. This is essentially the account provided by Bob Rowthorn in Capitalism, Conflict and Inflation (London, Lawrence and Wishart, 1980), pp. 231–249. Except that Rowthorn sees training as taking place mainly via the state and therefore, for him, outside of capitalism. Carchedi sees training as adding to the value of labour power, but does not explain how this transfers into the final product (Carchedi, Frontiers of Political Economy, pp. 130–133). Alfredo Saad Filho and P. Harvey both strongly reject the interpretation I have put here because “it conflates education and training with the storing of up of labour in machinery and other elements of constant capital” – see Alfredo Saad Filho, The Value of Marx (London, Routledge, 2002), p. 58. For Harvey, “Skilled labour creates more value in equal periods of time than does unskilled labour because it is physically more productive, and there is no reason to suppose that any determinate relationship exists between this increased physical productiveness and the physical productivity of the extra labour needed to produce the skill” – P. Harvey, The Value Creating Capacity of Skilled Labor in Marxian Economics, Review of Radical Political Economy, 17:1–2, quoted by A. Saad Filho, in The Value of Marx). But what Harvey is assuming is that there is some way of measuring the comparative physical productivities of labour that might be producing quite different products, i.e. that you can equate the value of commodities by comparing their use values. This leaves him (and A. Saad Filho) open to the very objection made by Böhm-Bawerk.

31. G. Carchedi, Frontiers of Political Economy, p. 133.

Last updated on 7 May 2021