ONTOCAST interview on the Theory of the General Law of Value [Transcript]

The nice people at the ONTOCAST podcast interviewed me about my theory of the general law of value. I was asked some great questions, including (i) what differentiates the general theory from Marx’s theory, (ii) the meaning of Marx’s volume 3 transformation procedure, (iii) the relations between Marx’s values and super-integrated values, (iv) the dynamics of the law of value, including the causal relations between values and prices, and finally (v) how the more general theory solves, or more accurately dissolves, one of the main problems of Marx’s theory of value. We also had time to reflect on the differences between Marxist and neoclassical theories of value and price formation. So a lot of ground was covered, and I hope others find it useful.

Prefer audio? You can listen to the interview here.


Q. What differentiates your general theory of value from Marx’s theory of labour-value?

My proposal of a “more general labour theory of value” is a contribution to the Marxist theory of capitalism. I view this more general theory as very much in the tradition of classical, orthodox Marxism.

The general theory is, in a very precise way, a strict generalisation of the theory of value that Marx presents in Capital. It includes what Marx established, but adds something new. And so, in this sense, my general theory is not very differentiated from Marx’s theory at all. It is almost identically the same theory.

For example, I don’t propose a new or unusual interpretation of Marx. I don’t propose to drop or remove any of his core claims. I don’t attempt to give a creative re-reading of any of his core terms or concepts, or anything like that.

Marx uses a single definition of objective value throughout Capital. I think the phenomenon of capitalist production demands that we add a second, complementary definition. And when we do that we get a more general theory of value, which yields an improved understanding of capitalism, and therefore contributes to its critique.

But before I can explain what I add, I think I need to make some basic remarks about Marx’s theory of value.

When Marx says “value” he means the “socially necessary labour time that, given the prevailing techniques of production, is necessary to produce a commodity”. But, of course, this “value” also has a nominal representation in terms of the unit of account, or money, which Marx calls “exchange value”. So, for complete clarity, I prefer to use the term “labour-value” instead of “value” just to make it clear when we’re talking about a real cost of production, rather than its monetary representation.

Anyhow, Marx, as I’m sure you know, defines the labour-value of a commodity as fundamentally composed of 2 elements: First, the labour-value of the constant capital that gets used-up to produce it, where constant capital is a catch-all term for the means of production, such as any inputs to the production process, including machinery and tools etc. Sometimes we call this dead labour, or past labour, or indirect labour.

The second element is the labour directly supplied to produce the commodity. We sometimes call this new labour, or living labour.

During production, workers add their own labour to the output, and transfer the labour-value of constant capital to the output. So the labour-value of a commodity is the sum of dead plus living labour.

The labour-value of a commodity represents how much labour time, given the prevailing techniques of production, are supplied to produce it. It’s the labour-time needed to produce the constant capital, plus the work needed to take that constant capital and transform it into a new output ready for sale.

All this is very clearly specified by Marx in Volume 1 of Capital.

We have to point out here, as Marx did, that a labour-value isn’t simply the sum of concrete labour-times, or simple clock-times. If workers decide to slack off and take double the amount of time than what’s typical, they don’t therefore add double the labour-value to their product. And workers don’t add more labour-value if they use outdated techniques, and therefore take more time to produce the same commodity compared to workers in a competing firm. Marx points out that not all concrete labour-time actually counts, that not all concrete labour time is “socially necessary”.

For simplicity let’s assume that all firms producing the same commodity use exactly the same techniques. And let’s also assume that workers producing the same commodity take the same amount of time to do it. This isn’t what happens in reality. But making these assumptions will help us to focus on some essential aspects of Marx’s theory, without getting distracted by more advanced issues.

Given these assumptions then Marx’s labour-values are a relatively simple function, or property, of the prevailing techniques of production.

Marx’s concept of labour-value is incredibly important for a proper understanding of the nature of capitalist production.

Let me give a quick example of this. Marx’s definition gives us a way of measuring the technical difficulty of producing things over time. Take a relatively stable commodity, say shoes. Perhaps 20 years ago the labour-value of a typical shoe was 4 hours. And perhaps today it is 1 hour, because our techniques of production have improved. In terms of technical difficulty of production then shoes have got cheaper to manufacture. Or, putting this in a different way, the productivity of labour has increased. We have to work less now to make them.

This is much clearer than looking at the market prices for shoes. Because, market prices obscure the underlying changes in the technical productivity of labour, because they are affected by temporary supply/demand imbalances, inflation and so on.

Another quick example. Marx’s labour-values can quantify by how much the capitalist class is ripping off the working class; that is how much labour time the working class supplies to the capitalist class for free, as economic tribute. It tells us the quantitative degree of economic exploitation.

So the labour-value of a commodity is the sum of dead labour plus living labour. But workers supply more labour to the economy than they take out in the form of their real wages. Say that I supply 10 hours of labour a day to the firm I work for. They pay me a money wage. I spend that on a real wage, which is a collection of goods and services, which themselves have a labour-value. The labour-value of my real wage is typically less than the labour I supplied to the firm. In other words, I supply an excess of labour-time.

Marx calls this surplus-value. Some of the surplus-value supplied by workers may produce new means of production, new constant capital, to increase the scale of production, and grow the economy. Some of the surplus-value may produce consumption goods for capitalists, who receive luxuries, and enjoy freedom from work, purely in virtue of their ownership of the firm, and not in virtue of supplying any labour.

So Marx’s concept of labour-value very quickly tells us that, if workers didn’t have to produce luxury goods for capitalists, then the overall length of the working day could be reduced by so many hours without reducing the real wage of the working class. In other words, as Marx so eloquently described, the working day splits into two components, that which is needed to sustain workers, and that which simply is not, which is an unnecessary and exploitative extension of the working day.

So that’s just two examples which illustrate how Marx’s definition of labour-value is incredibly important for a proper and deep understanding of capitalist production.

Now, getting back to your question: what differentiates the general theory from Marx’s theory? The general theory states that Marx’s concept of labour-value is necessary, but not sufficient, to understand capitalist production. In other words, what we’ve just said isn’t enough. The phenomena of capitalist production demands more than this.

A more complete understanding requires another definition of labour-value, which is complementary but different to Marx’s definition, and which we use for a different purpose.

So this is the main difference between Marx and I. If Marx were here — sadly he is not — I’d like to say to him, “Karl, it’s not enough. There’s something you missed. We need a bit more, there’s a problem in our theoretical understanding of capitalist production, we’re not fully reflecting what’s actually happening. Here let me show you how”. And then I’d explain why we need to generalize his concept of a labour-value.

I hope this gives a preliminary answer to your question. Why we need to generalise, and how we go about doing that, I guess we’ll get into the details shortly.

Q. What are the main problems of classical labour-value theory?

I think the main problem with the classical labour theory of value, and Marx’s theory of value, is that they fail to fully demonstrate the lawful connection between money and labour-time.

Adam Smith, quite famously, restricted the applicability of the labour theory of value to an “early and rude state” of society before capitalists earned profit. Ricardo very much wanted to remove this restriction. He wanted to show that a labour theory of value applied to a mature capitalist economy. But he couldn’t quite close a theoretical gap between measuring economic costs ultimately in terms of labour-time and relating those real costs of production to monetary magnitudes, to the prices in the market.

Now Marx, quite correctly in my opinion, claims that money represents labour time. This is the essential content of his theory of value. The value-form, these numbers that we throw around to organise our economic behaviour, actually refer to a material content, which is fractions of society’s total-labour time. According to Marx, the dynamics of capitalist competition instantiate lawful regularities between flows of money and changes in the division of labour. So the value-form represents labour-time not because we say so, or we think that it should, but in virtue of objective economic laws that bind the value-form to its content — which is labour-time. In other words, it’s the “law of value” that makes the labour theory of value true.

Now this claim implies a bunch of things. One implication is that in certain special situations, we expect to see a 1:1 correspondence between monetary magnitudes and labour-values.

And why should there be a 1:1 correspondence? Think about a thermometer for a moment. A necessary condition for a thermometer to represent temperature is that the height of its mercury column varies 1:1 with the ambient heat. If the height did not vary 1:1 then the thermometer’s gauge wouldn’t measure temperature. It would measure something else, or perhaps nothing at all, or perhaps a mixture of temperature and something else.

The same reasoning roughly applies to the theory of value. If we’re saying that prices represent labour time — and not something else, or some mixture of things — then we need to demonstrate a quantitative 1:1 relationship between prices and labour-values. Prices shouldn’t vary independently of changes in the labour-time necessary to produce commodities.

Now, Marx, and the classical economists before him, were quite aware that the labour theory of value directly contradicts empirical reality. If the demand for a commodity increases, perhaps due to a change in consumer tastes, then it’s price will increase, quite independently of its labour-value. This is obvious.

The classical economists viewed these mismatches between market prices and labour-values as completely necessary, and an essential part of the mechanism by which a capitalist economy, in a distributed manner, communicates price signals that reallocate the division of labour in society to different productive activities in order to meet social demand. We can get into the details of this dynamic process later. The only point I want to emphasise here is that the existence of such mismatches, the fact that prices do vary independently of the underlying labour-values, is very much part of the labour theory of value.

Nonetheless, in certain special situations, we do expect prices to vary 1:1 with labour-values. And the main problem with the classical labour theory of value, and Marx’s theory, is that prices don’t vary 1:1 with labour-values, even in some of the special situations.

So to explain the root of the problem, I need to explain the special situation, and why it’s important.

Let’s assume, as Marx does, that market prices have completely gravitated towards stable prices, which Marx calls prices of production. This happens when supply equals demand everywhere: so there are no mismatches between what’s getting made, and what’s getting consumed. In order for this equilibrium state to manifest empirically then, during the process of gravitation, the techniques of production must remain relatively stable. And we also need to assume a fully competitive economy, so no firm or sector enjoys monopoly profits. In consequence, the profit-rates, in all the different sectors of the economy, are uniform in this steady-state. And that means, capitalists have no incentive to re-allocate their money-capital from one sector to another to earn higher returns. Marx analyses this steady-state in Volume 3 of Capital.

Before continuing, we need to be clear about the empirical status of this steady-state. Some people think that, because this steady-state is a type of equilibrium state, and because capitalism is never in equilibrium, but forever turbulent and changing, then the steady-state is irrelevant to reality, and we can ignore it. Profit-rates, either in a nation-state, or across the globe, are never uniform. And therefore Marx was wasting his time even thinking about these special circumstances.

I think Marx was correct to analyse this steady-state for the simple reason that it’s real. Not real in the sense of actually empirically manifesting at any moment of time, but real in the sense that it’s an attractor-state for the dynamics of capitalist competition. In other words, at all times the economy is being continually pushed towards this steady-state by the law of value. Its effects are real, always.

But, of course, there are many other mechanisms at play in economic reality, not just the law of value. Innovation changes the techniques of production. External shocks can disrupt production. The state can intervene to boost the economy, or change the price structure via taxation. And on and on. And therefore the actual trajectory of capitalist economies, although always permanently and partially controlled by the law of value, the actual trajectory never reaches the steady-state. But the steady-state is nonetheless an attractor for the dynamics, and continually exerts a real influence.

So I interpret Marx, in Volume 3 of Capital, as performing a counterfactual exercise to reveal the hidden mechanism of the law of value, to reveal what direction the law of value is pushing the economy. He’s saying that if all commodities are reproducible, and techniques are relatively stable, and there’s perfect competition, and the total labour of society can be shifted around in the division of labour, then the dynamics of the law of value, if allowed to work without interference, would empirically manifest this steady-state. And, further, this steady-state, although never fully manifested, nonetheless partially explains the actual empirical trajectory of capitalist economies.

I’ve taken a bit of time to explain this, because I think unfortunately there’s a tendency to reject Marx’s equilibrium methodology, but — and this is a dialectical point — you cannot explain disequilibrium trajectories without understanding what equilibrium attractors are in play.

OK, so let’s get back to this steady-state that Marx discussed in the early parts of Volume 3. In this state, supply equals demand, and therefore the equilibrium price structure should perfectly reflect labour-values. The truth of the labour theory of value should become particularly clear and obvious. We expect that, in these circumstances, planes cost more than pens because planes require more of society’s labour-time to produce. So it’s here, in this steady state, that the 1:1, lawful correspondence between prices and labour-values should manifest.

But it doesn’t. We don’t see a 1:1 correspondence. The structure of prices is different from the structure of labour-values. There’s still a mismatch, even in this special situation, even when all the accidental mismatches, due to supply and demand in the market, have melted away.

Marx completely accepts that there must be a mismatch, and he also understands why there must be a mismatch.

In the steady-state profit-rates are uniform. So all industries generate the same profit-rates. Let’s say the uniform profit-rate is 2%. The price of each commodity, therefore, is a 2% mark-up on top of the money cost of producing it. And the money cost of producing a commodity is the sum of the cost of the constant capital, plus the wages of the living labour employed to make it. So Marx’s prices of production, the steady-state prices, are proportional to the total money cost of producing each commodity, where the constant of proportionality is the profit-rate.

But the labour-value of each commodity is independent of the profit-rate. Labour-values depend on the technical conditions of production. They don’t depend on money prices, or capitalist profit-rates, at all. In fact, Marx proposes that labour-values ultimately explain profits.

So, on the one hand, Marx’s prices of production vary with the profit-rate. But Marx’s labour-values do not. And therefore, in general, the prices of commodities — even in this special steady-state — are not in 1:1 correspondence with labour-values, and can vary independently of labour-values. The value-form seems to be radically disconnected from its content.

This is a problem for any labour theory of value. And this mismatch between prices of production and labour-values is the reason why Smith restricted the labour theory to pre-capitalist societies, and why Ricardo failed to resolve his theoretical difficulties.

This problem wasn’t hidden. Everyone knew it. In fact, the problem was so famous that Engels, in his introduction to the 2nd volume of Capital, boasted in advance that Marx had already solved it. Engels challenged other economists to propose their solution before the big reveal when Volume 3 was published.

So what was Marx’s solution?

According to Marx, money profits represent the surplus-value supplied by workers. And the surplus-value supplied by workers in different industries depends on how much living labour is employed. Because the source of surplus-value is living labour.

So Marx, in Volume 3, considers a situation where money profits are non-uniform, and proportional to the living labour employed in each industry. In this situation, profits are clearly and transparently just surplus-value, and also prices and labour-values have a 1:1 relationship. In this situation, labour-intensive industries yield more profit than capital-intensive industries, because labour-intensive production yields more surplus-value.

Marx doesn’t give us a detailed causal explanation of the economic dynamics. But he knows that, in a situation of non-uniform profits, capitalists have an incentive to re-allocate their capital away from low-profit industries and towards high-profit industries.

So Marx assumes that, when capitalist competition kicks in, the economy gravitates to a steady-state where uniform profits prevail. So the relatively high profits in industries that are labour-intensive, and the relatively low profits in industries are capital-intensive, are evened-out and become uniform. But in this new situation, in the steady-state, prices don’t have a 1:1 relationship with labour-values.

So Marx has given us a before and after.

Before we had non-uniform profit-rates and a 1:1 relationship between prices and labour-values. After we have uniform profit-rates and the absence of a 1:1 relationship.

But Marx says that the process of profit-rate equalisation is identically a process by which individual capitals grab a share of the surplus-value produced from all industries according to the distributional rule that equal amounts of money-capital invested in production earn an equal return, regardless of whether that investment occurs in capital or labour-intensive industries.

This is Marx’s transformation. Surplus-value is produced unequally in different sectors, but by the magic of profit-rate equalisation, all capitalists, in whatever sector they invest, receive equal shares of that surplus-value proportional to the size of their investments.

And because this transformation occurs purely in the realm of exchange-value, because the transformation is purely a matter of changes in the price structure of the economy, and not due to any changes in production, then the transformation must be conservative. No new surplus-value is created or destroyed. Labour-values don’t change. So the original surplus-value is merely redistributed.

And so Marx states that, although prices of production don’t appear to be related 1:1 to labour-values, in fact they still are.

Marx claims that we can recover the 1:1 relationship by considering larger aggregates of commodities. He says that: the uniform profit-rate in the economy is equal to the ratio of the total surplus-value divided by the total labour-value of means of production and labour-power. And he says that the total profit in the economy is proportional to the total surplus-value, and the total price of all commodities sold is proportional to their total labour-value. So he states these conservation constraints, which must hold.

So the transformation scrambles the 1:1 relationship between prices and labour-values, but it’s a conservative transformation, and so the 1:1 relationship still holds in the aggregate.

So Marx’s theory of the transformation upholds the core claim of the labour theory of value, that human labour is the source of profit, and that monetary magnitudes represent labour-time.

Marx’s answer to the famous problem of the classical labour theory of value is a very good answer, full of important and true insights about the dynamics of capitalist competition and the origin of profit in human labour.

But — and here we come to a great controversy, which started as soon as Marx’s transformation was published — there is a problem with Marx’s transformation.

Actually, these issues are so controversial, even claiming there is a transformation problem has itself become controversial.

But first let me say what the problem is. Marx himself first identified a potential problem with his transformation, but he didn’t pursue it, and just left us hanging. We should always remind ourselves that Volume 3 of Capital was assembled by Engels from unfinished manuscripts.

The problem, quite simply, is that Marx’s transformation cannot be a conservative transformation. The 1:1 relationship between prices and labour-values is actually lost during the transformation.

The reason why Marx’s transformation cannot be conservative is actually very simple. Labour-values are a property of the techniques of production, about the actual human cost of making things. But the profit-rate is a property of the class struggle, about how much workers get and how much capitalists get, about how the working day divides into necessary and surplus parts. And so prices fluctuate according to a distributional conflict between workers and capitalists, but labour-values do not. And that means labour-values cannot fully account for the structure of prices, even in the special circumstances of this steady-state.

Many people have demonstrated this problem, in lots of different ways, and many people have denied the problem exists, also in lots of different ways. The transformation problem is incredibly controversial.

Some Marxists, for example, say that we don’t need Marx’s transformation at all because the steady-state never occurs in empirical reality. I think that’s wrong because the steady-state is an attractor that affects empirical reality all the time. And we can see, in the empirical data, a tendency for profit-rates to equalise, even if that tendency is never fully realised. And so, it would be odd to hold a labour theory of value whose truth fluctuates depending on the distribution of profit-rates at any particular time. But anyway, a transformation problem, except in a small number of cases, still arises even when profit-rates aren’t uniform.

Some Marxists say that Marx’s theory of value was never really about explaining prices anyway, and therefore the 1:1 relationship just isn’t important, and the whole transformation problem is a waste of time. I think this attitude reflects an understandable impatience with the lack of progress on this issue. Marxists and non-Marxists have been arguing about the transformation problem for over 120 years! The problem seems intractable, and getting us nowhere, and distracting energy from more important things.

However, we can’t simply wish scientific problems away, and so this attitude ultimately drops some of the essential content of Marx’s theory.

Some Marxists propose new interpretations of Marx’s work that make him consistent, and so the transformation problem doesn’t arise. There are lots of these interpretations, which differ in subtle ways. The new interpretations typically redefine some of Marx’s core concepts by conceptually mixing-up monetary magnitudes with labour-time. Typically, when you look closely into the details, the new interpretations fail to theoretically sustain an ontological separation between labour-value and exchange-value. Labour-values become defined, ultimately, in terms of exchange-value. And so the causal mechanism of the law of value is ultimately lost because there can’t be quantitative mismatches between labour-values and prices. And that’s a bad mess to get into. By re-interpreting Marx to make him consistent, we end up with new inconsistencies.

Some Marxists accept that there is a transformation problem in Marx’s theory. They will say, yes OK, there isn’t a 1:1 relationship between labour-values and prices of production, but nevertheless, most of the variation, something like 93% of the variation, in prices of production is in fact explained by labour-values. So Marx’s theory is approximately right, and that’s good enough. The remaining, unexplained 7% we can ignore.

That’s a reasonable thing to say, but it’s not what Marx wanted to say, and more importantly it’s no longer a labour theory of value. We can no longer say monetary magnitudes are really labour-time magnitudes, and that profit really is, despite appearances, the surplus-value produced by workers. So the core claims of Marxist theory start to fall apart.

And finally, non-Marxists, at least the sophisticated ones, point out that the transformation problem completely undermines Marx’s theory of value. And therefore Marx’s value theory, and by extension, his whole critique of political economy is in error.

I could go on. The transformation problem is like a huge island in an ocean. Wave after wave of Marxists and non-Marxists come up against it, and are dashed upon it, splitting into lots of different interpretations and arguments about the true nature of economic value. It’s a mess, and it’s still a mess even to this day. In the future, I think the transformation problem will be considered an incredibly fascinating episode in the history of science, when science was operating under the rule of capital.

Anyhow, to get back to your question: I think the biggest problem with the classical labour theory of value is what ultimately became known as Marx’s transformation problem. It’s been responsible for decades and decades of controversy. And it’s an incredibly important scientific problem because it’s essentially concerned with what these numbers we hold in our pockets, that we throw around the entire world, actually mean, what they actually represent. What is money, really? What is the meaning of this social symbol that dominates our lives? This is why the transformation problem was, and remains, an essential and important scientific problem.

Luigi Pasinetti

Q. Could you elaborate on the differences between classic, generalised and super-integrated subsystems?

Classical labour-values are what we’ve just been discussing. They measure the total labour-time necessary to produce a commodity, including producing any used-up constant capital.

But in the 20th-Century some economists, including some self-identifying Marxist economists, realised that classical labour-values don’t measure the total labour-time necessary to produce commodities in all circumstances.

The work of Luigi Pasinetti is especially important here. Pasinetti is an Italian economist, often associated with Cambridge University in the UK, which is where he completed his PhD. Cambridge was then the home of a vibrant and left-leaning post Keynesian school of thought, which partially engaged with Marx’s theory of value.

Pasinetti pointed out that, in a growing economy, some of the output takes the form of investment goods, which augment the existing means of production. Some of the output isn’t consumed, but invested. This additional constant capital, when combined with additional living labour, enables the economy to grow, producing outputs on an ever-increasing scale. The additional labour might be drawn from the pool of unemployed labour, or from population growth.

Pasinetti realised that, in this situation of growth, classical labour-values do not measure the total labour-time necessary to produce commodities. They systematically undercount that time because they don’t count the additional labour-time necessary to grow the economy, that is to produce the additional means of production.

In these circumstances of growth additional labour has to be supplied to produce a commodity. Consumption goods don’t get produced without simultaneously producing additional means of production. And so this growth imperative translates into additional real costs of production.

Now the techniques of production haven’t changed. So classical labour-values, or Marx’s labour-values, continue to tell us how the productivity of labour changes over time, or the degree of exploitation and so on. But the institutional circumstances of production have changed. There is a growth imperative. And that means classical labour-values don’t measure the total labour-time necessary to produce commodities in these circumstances. They measure the total labour-time if, counterfactually, the economy was not growing.

This is very interesting. And Pasinetti immediately proposes a generalisation of classical labour-values, which he calls hyper-integrated labour coefficients, which in fact measure the total labour-time necessary to produce commodities in these growth conditions. The details of how this is done don’t really matter.

So Pasinetti generalises the classical concept of a labour-value. And he’s been forced to do that because he wants to measure the total labour-time necessary to produce commodities in more general conditions.

And we now have classical labour-values, and Pasinetti’s hyper-integrated labour coefficients, and they both tell us different things about the very same economic system. They go together in the same theory. They just happen to measure different things about the same economy.

So that’s Pasinetti. He goes on to provide a generalisation of the transformation problem in the circumstances of a growing economy. And Pasinetti, therefore, doesn’t think that Marx’s labour theory of value applies to real economies. I think Pasinetti is very wrong about that, but that’s what he thinks.

Now this distinction between the technical conditions and the institutional conditions of production turns out to be very important. And so does the distinction between classical labour-values, which measure the labour-time that’s technically necessary to produce things, and total labour-values, which measure the labour-time that’s actually necessary to produce things, because of the presence of non-technical factors at play in the economy.

Now, from this more general point-of-view, it becomes very clear that classical labour-values in general don’t measure the total labour-time necessary to produce commodities. And, in fact, they don’t even measure the total labour-time in the steady-state situation that Marx analysed in Volume 3. They are not total labour-values, but partial labour-values.

And the reason for this is quite simple.

In the institutional conditions of a capitalist economy nothing gets produced without simultaneously producing consumption goods for the capitalist class. This imperative to extract a tribute from the working class translates into additional real costs of production. Every time a commodity is produced, for consumption by the working class, then some additional labour is supplied to produce goods and services for the capitalist class. Nothing can be made without the supply of tributary labour. And this is precisely why workers can’t clock-off their job after the necessary part of the working day has completed. They have to keep going, to produce surplus-value for capitalists.

Now, classical labour-values, because they focus on just the technical conditions of production, don’t count this additional labour. That’s why they don’t function as total labour-values.

But we can count this additional labour. And when we do that, we get what I call the super-integrated labour coefficients, or super-integrated labour-values. The super-integrated labour-values are, by construction, total labour-values. They measure the total labour-time that’s actually necessary to produce commodities, in the institutional circumstances of a capitalist economy, whether in a steady-state, or growing.

In contrast, classical labour-values measure the labour time that, counterfactually, would be necessary to produce commodities if capitalists weren’t extracting their tribute, if they didn’t exist.

So I hope that answers your question. As I hope you can see, the phenomenon of a capitalist economy demands that we consider multiple definitions of labour-values, where each is a different viewpoint on the same economy, and where each kind of labour-value gives us different kinds of insights about its operation.

Computing labour-values by vertical integration

Q. How are the super-integrated value coefficients formed?

Let me answer this question by first thinking about Marx’s labour-values.

So when we say that Marx’s labour-values are a property of the technical conditions of production, this is simply shorthand for saying that it’s the totality of ways of making all the goods and services in an economy — the actual methods by which concrete labour activities combine with machinery, materials, tools to materially produce things — that determine labour-values. Labour-values, as Marx says clearly and unambiguously, are determined in production, not exchange.

Now, of course, if the demand for goods and services change, then labour-values, in general, also change. But we have to be very careful here, especially as this can cause confusion. Let’s say the demand for mobile phones increases. This will have lots of consequences, including an increase in demand for copper. Now the copper-producing industries will increase production to meet this demand. And in consequence it’s very likely that their techniques of production will also change. For example, copper may have to be mined from more difficult spots, and so more time is needed to get the stuff out of the ground, compared to before.

So changes in demand, as transmitted by exchange-values in the market, can cause changes in the techniques of production. But this fact doesn’t imply that labour-values are a property of exchange-values, or the market. Labour-values remain a property of the current techniques of production, however those current techniques have been arrived at.

Marx’s use of the modifier “socially necessary”, when he first introduces labour-values, can sometimes confuse people. Marx, by using this phrase, is simply pointing out that not all concrete labour-time, individual wall-clock times, automatically determine the techniques of production. For example, if a small group of workers in a copper-producing firm decide to work-to-rule, or slack-off, and therefore take more time to produce copper, it doesn’t follow that their copper has more labour-value. No, the labour-value of copper is a social property that depends on the techniques of production in the copper-producing industry as a whole, across all the firms in that industry. Labour-values aren’t noticeably affected by individual acts of wayward production or non-production.

So many factors can affect how classical labour-values are formed. But classical labour-values are always a property of the prevailing techniques in production.

So how are the super-integrated labour-values formed? Well the super-integrated values include classical labour-values as a component part. So they’re also subject to the same factors that form classical labour-values. Any changes in techniques also affect them in the same way.

But the super-integrated labour-values include the labour-time necessary to produce the goods and services that capitalists consume. They include the additional tributary labour that must be supplied when producing commodities for consumption by the working class.

And so, super-integrated labour-values are also affected by the class struggle. If the capitalists consume more or less of the surplus product then the super-integrated labour-values will, all other things being equal, either increase or decrease. And so, they are not just a property of the techniques of production, but they are also a property of the distribution of the surplus product, how it gets divided between workers and capitalists. They depend on the level and composition of the basket of goods and services that capitalists consume. And in consequence the factors that form the super-integrated labour-values are correspondingly more varied.

In fact, to fully elucidate the factors that form them would require a theory of the distribution of income in capitalist societies. Now this is a big topic in itself, and I’m not going to get into it. The only thing I’d like to say is that the super-integrated labour-values are unambiguously defined given a particular distribution of the net product, however that specific distribution has been arrived at. They’re essentially agnostic, and independent of, any particular theory of the causes of the division of the surplus between workers and capitalists.

So, to sum up, the super-integrated labour-values, just like classical labour-values, may change due to changing market conditions, but they nonetheless are completely ontologically distinct from exchange-values, and the price system.

Q. How are prices formed as labour-values?

Let’s follow Marx and assume, initially, that profit-rates are different across sectors. Some sectors are very profitable, others less so, or even making losses.

There are many reasons why profit-rates vary. But, for simplicity, let’s consider just one reason, which is that, in general, the supply of commodities, the quantity of different things being produced, doesn’t equal the demand, where demand consists of demand from firms for their inputs, and final demand from consumers, both workers and capitalists. So let’s assume supply and demand are completely out of whack.

Market prices, as Marx and the classical authors of course understood, are partially determined by supply and demand. If a commodity is in under-supply relative to its demand, then firms tend to raise their prices because buyers are willing to outbid each other to obtain the scarce product. And, in the opposite way, if a commodity is in over-supply then firms tend to lower their prices in order to sell to scarce buyers.

So scarce commodities obtain a higher price in the market. And therefore firms that produce these commodities get additional revenue, over-and-above their costs. And this translates into a higher profit-rate. Conversely, firms that produce commodities in over-supply will tend to have lower profit-rates, or even negative profit-rates. When they’re selling prices are lower then their input costs, then they’re making a loss.

So market prices fluctuate according to mismatches between supply and demand, which in turn affect profit-rates.

Now, at the commanding heights of the economy, we have a large collection of privately-owned, competing capitals. And there’s a scramble for profit. Capitalists inject and withdraw their money-capital to and from different sectors of the economy according to profit-rate signals. High-profit industries get additional injections of money-capital, and low-profit or loss-making industries experience withdrawals of money-capital. So competing capitals, at the top of the tree so to speak, continually allocate and re-allocate their money-capital in a ceaseless search for high returns.

The inward flows of money-capital to profitable sectors fund the purchase of additional means of production and labour, which means firms in these sectors can increase their output; and withdrawals of money-capital from low-profit sectors reduce the funding, and therefore reduce the level of output.

So the profit-rate signals, which here we’re assuming reflect underlying supply/demand imbalances, attract and repel capital investment, which ultimately cause changes in the quantities produced in each sector of production.

And so the competitive scramble for profit has the consequence of increasing production for commodities in under-supply, and reducing production for commodities in oversupply. The imbalance between supply and demand begins to reduce. And, at the same time, the total labour of society, which was misallocated, becomes re-allocated to different sectors of production to meet the social demand.

Marx is most associated with explaining crises of capitalism. But he was just as much concerned in explaining how capitalism persists and reproduces itself over time. The coordination of millions of independent production activities in a large-scale market economy isn’t perfect and it isn’t equitable but nonetheless we should be far more surprised by coordination than by disorder. Marx understood that a capitalist economy has intrinsic dynamics that are both stabilising and destabilising. The law of value, which is basically what we’re talking about here, is a major stabilising mechanism.

Marx, in volume 3, outlines some of the conditions that need to hold for the law of value to operate to completion. One condition is that there are no natural or artificial monopolies that prevent it. So we’re assuming a competitive market here, and the ability for all kinds of resources, including labour resources, to be shunted around.

As I’ve mentioned, we don’t expect the law of value to operate to completion in empirical reality. But it does define an attractor state, which affects the trajectory of the economy at all times, even without empirically manifesting.

Now, the attractor state of the law of value, in a capitalist economy, is essentially the steady-state we discussed earlier, which Marx looked at in Volume 3. As money-capital is reallocated, and supply and demand start to balance, then profit-rates in the different sectors also start to equalise. In the equilibrium limit, we have a steady-state economy where market prices have stabilised to Marx’s prices of production, and profit-rates are uniform.

OK, that’s a brief sketch of the dynamics of the law of value.

So what’s the relationship between prices and labour-values during this process? The first point is that market prices, which we can think of as scarcity prices, have at any point in time no relationship at all to the underlying labour-values in the economy.

This empirical fact is often raised by people, even trained economists, to reject the labour theory of value. But, of course Marx, and all the classical authors, including Adam Smith and David Ricardo, understood that market prices, at any time, are determined by supply and demand, and therefore necessarily differ from labour-values.

So I think the classical economists wouldn’t be particularly interested in the results of, say, Walrasian-style general equilibrium theory, which raises this empirical fact to a theoretical principle, and views the economy as always being in an equilibrium characterised by scarcity prices, which are supposed to form outside of time due to the assumption of instantaneous market clearing. This theoretical approach, the neoclassical approach, tends to obscure the relationship between prices and labour-time.

In contrast, Marx’s law of value is essentially about how an economy allocates the total labour of society over historical time in a distributed and unplanned manner. So it’s essentially a dynamic theory of how an out-of-equilibrium economy adjusts its prices, the scale of production, and the employment of labour, in order to meet social demand.

So, most of the time, market prices bear no relationship to the underlying labour-values. But as the economy adjusts, and supply begins to equal demand, then scarcity prices dissipate and start to converge toward equilibrium prices that, if they fully manifested, should transparently represent the underlying labour-values in the economy, if the labour theory of value is right.

And why is that? Simply because, in the steady-state, the price structure is no longer determined by supply and demand, but is primarily determined by the objective difficulty of making things, which is determined by the prevailing techniques of production. In equilibrium, planes cost more than pencils because manufacturing a plane uses-up more of society’s labour-time compared to making a pencil.

So this, briefly, is how prices are formed by labour-values. Prices are signals to reallocate labour. And when the total labour of society is properly allocated then prices will reflect the objective costs of production.

Q. How does your general theory solve the classic problems of labour-value theory?

OK, so let’s return to Marx’s transformation problem, which is that prices don’t reflect objective costs of production in this steady-state.

And that’s because equilibrium prices include a profit component that is completely unrelated to classical labour-values. In the steady-state, capitalists still earn interest on their money-capital tied-up in production. They’re still extracting profits from every production process. And so equilibrium prices are not entirely determined by the techniques of production. They are also determined by the class struggle. A change in the equilibrium profit-rate can change the entire price structure of the economy, even when the techniques of production remain constant. So, as we mentioned earlier, prices of production vary independently of classical labour-values, and so the 1:1 relationship that we’d expect to see doesn’t show up, neither at the level of individual prices, nor at the aggregate level as Marx proposed.

And the reason for this is ultimately very simple. Equilibrium prices include a profit-rate component, which is the monetary expression of the fact that capitalists grab a share of the net product. So if the profit-rate increases, and capitalists grab more of the net product, then prices will change. But classical labour-values, because they are purely a function of the techniques of production, don’t change at all.

We can state this another way: prices depend on the institutional situation, where production cannot take place without workers supplying tributary labour to capitalists; but classical labour-values are independent of the institutional context, and only measure technical costs of production. They don’t include the tributary labour that must be supplied in the institutional circumstances of a capitalist economy.

Prices are total monetary costs of production, but classical labour-values are partial objective costs of production. There cannot be a 1:1 correspondence between total costs and partial costs, even in this steady-state. It’s like comparing apples with oranges.

To think there could be a 1:1 correspondence between prices determined by the institutional conditions of production, and labour-values determined by solely natural or technical conditions of production, is in fact a very subtle and pervasive conceptual error.

The conceptual error is present in Smith, it’s present in Ricardo, and it’s also present in Marx. It’s the specifically classical conceptual error. And once you’ve seen it, you can’t unsee it. And it’s very fruitful to re-examine the writings of the classical authors, including Marx, from this point-of-view.

Deep conceptual errors like this are difficult to spot. And they tend to generate intractable theoretical problems, precisely because anyone who attempts to solve the problems has already accepted, and therefore been captured by, the conceptual error. And this explains the longevity of the transformation problem. Because, the whole debate around it, the whole history of it, is essentially a history of a subtle conceptual error, which hasn’t been noticed.

On the one hand, some Marxists either deny the importance of the transformation problem, or solve it by redefining Marx’s core concepts; and on the other hand, we have some Marxists and non-Marxists accepting that there is a transformation problem, and therefore dropping either some essential aspects of the labour theory of value, such as the proposition that money represents labour-time, or dropping Marx’s theory of value altogether.

But in my view both sides of this debate, at least on this specific issue, are wrong. There is a transformation problem. But the problem only arises because classical labour-values are being misused, they’re being asked to do a job they cannot do. They don’t measure the total labour-time necessary to produce commodities in the institutional conditions of a capitalist economy. The super-integrated labour labour-values do this job.

So, in this more general theory, we have prices of production, which represent total monetary costs of production, and super-integrated labour labour-values, which represent total labour costs of production. And so now we’re comparing total monetary costs with total labour-values. We’re comparing apples with apples. And when we do that, we start to prove theorems that demonstrate that prices of production are 1:1 proportional to the super-integrated labour values.

Let me try to put this as simply as I can. The labour-time that prices of production represent is captured by the super-integrated labour labour-values. It is not captured by classical labour-values.

Profit, in this steady-state, is surplus labour, and is clearly and unambiguously the unpaid labour-time of the working class. But money profit does not represent the excess of the classical labour-value of the net product, over-and-above the classical labour-value of the real wage, as Marx supposed. Money profit doesn’t refer to that difference. Instead, money profit represents the excess of the super-integrated labour-value of the net product minus the super-integrated labour-value of labour-power. And so on, and so on. So we get all Marx’s conservation claims, as long as we use the super-integrated labour-values, the values that reflect the actual time it takes to make things under the rule of capital.

So, in this more general theory the transformation problem doesn’t arise because we avoid committing the conceptual error of comparing institutional cost structures with technical cost structures. And therefore the relationship between monetary magnitudes and the labour-time supplied to produce things becomes completely transparent and clear.

But we have to be clear about the purpose of the different measures of labour-value. We have more theoretical tools now. And some tools are right for some jobs, but wrong for others.

If we’re asking questions about labour productivity, or the degree of exploitation, then we use classical labour-values as before. Because they get right down to techniques. They tell us what work would need to be done if the capitalist class was abolished and workers didn’t need to supply tribute. But if we’re asking questions about what prices represent, what labour-time they express, within a capitalist system, then the answer is the super-integrated labour-values.

So we have the beginnings of a richer, more general theory of value.


Q. What elements do you believe show the advantage of your theory of price formation in relation to the neoclassical theory?

First I should point out that I don’t believe I’m proposing a new theory of price formation. I see myself contributing to the classical and Marxist theories of price formation. I think some of my mathematical models formalise the classical theory in new ways, and are therefore especially illuminating, but I’m not proposing a new theory of price formation, although I guess I am proposing a new theory of what prices mean.

If we zoom out to 1000 feet and look down then I think the main difference between the classical-Marxist theory of price formation, and the neoclassical or marginalist theory, is that the classical-Marxist theory is mainly concerned to explain prices in terms of objective forces, in terms of what happens “in the hidden abode of production”, whereas the neoclassical theory is mainly concerned with subjective forces, or what happens “in the higgling and haggling of the marketplace”.

Let me try to add some detail to this very abstract statement.

The classical approach begins with the production of reproducible commodities as its object of analysis. A commodity is reproducible if the size of the workforce is the only enduring constraint on its level of supply. For example, Ricardo, writing in 1817, noted that commodities, the value of which is determined by scarcity alone (things like rare painting and so on) “form a tiny part of the commodities exchanged daily in the market”.

So the classical approach tends to postpone the analysis of non-reproducible commodities, and what determines their prices, to a later stage. For example, Smith, Ricardo and Marx both analyse land rent much later in their works.

And it’s Ricardo I think who first introduced marginal concepts in his theory of differential rent. He wanted to understand how the scarcity of land modified his labour theory of value.

The classical approach also tends to focus on longer-term equilibrium states of the economy where the supply equals effective demand. They are less interested in short-term states where mismatches of supply and demand, which may be due to all kinds of accidental reasons, cause market prices to fluctuate.

And, in the classical approach, the principal method of coordination is capitalist competition, which is a mechanism that allocates productive capacity, which is considered fungible, to meet demand.

So classical economics tends to focus on the objective causes of relatively stable prices of reproducible commodities due to the effects of capitalist competition occurring in historical time. And when we do that, we can begin to see the underlying objective values that explain the trajectory of market prices. We can begin to see the law of value.

Neoclassical economics basically takes the opposite point-of-view on all these issues.

The starting point, which unfortunately is the starting point for almost every university trained economist, is ubiquitous and permanent scarcity. So the property of economic
scarcity, rather than reproducibility, dominates.

In the traditional neoclassical vision of an economy, market participants arrive at the market endowed with different endowments, let’s say commodity bundles, which are scarce because their quantity is given, is a fixed and exogenous, variable.

The market participants have different preferences that define the subjective utility they obtain from consumption. And the principal method of coordination is then market exchange, which is a mechanism that (we are told) efficiently allocates the given scarce resources to meet demand.

So the neoclassical approach typically analyses short run “market clearing” states of the economy, and the market exchanges that maximise the utility of each participant given their budget constraints. And so prices are scarcity prices that clear the market, everyone goes home having swapped their commodities for other things, and feeling happier.

The neoclassical approach starts with market exchange and scarcity prices, and then applies this vision later to the analysis of production, and the re-allocation of productive capacity over time. So production is talked about last.

A small example of this is that neoclassical theorists, in the 50s, were very surprised by proofs of the “non-substitution theorem” that states that, under certain technical conditions, market-clearing prices are independent of consumer demand.

Now I know that all the conceptual machinery of the general theory — such as Marx’s labour-values and the super-integrated labour-values — and its main theorems — like the theorem that production prices are 1:1 with labour time — can all be recast in the language of Walrasian general equilibrium theory. That’s not such a difficult thing to do, and it might even be useful.

But ultimately we’re talking about competing political and philosophical visions about the meaning of economic phenomena, and the status of capitalism as a social system.

Marxism is fundamentally a critique of capitalism. And therefore it’s theory of value, and its theory of price formation, is a conflict-based, class struggle theory, where prices obscure the exploitation of the working class. Mainstream economics is fundamentally an apology for capitalism. And so it’s theory of value, if it has one at all, eradicates the distinction between classes, and the formation of prices through market competition is believed to fairly and justly distribute rewards according to each factor’s contribution to production. So labour gets it just reward, capital does, and so do the landlords. Everything is fair, and as it should be.

Q. In short, what does your proposal for a general law of value contribute to the Marxist economic debate in relation to the theory of socialist transition?

I think that, first and foremost, my proposal of a more general law of value is a scientific contribution. I am absolutely concerned with understanding the society I live in. And, to me at least, the most important social phenomenon that demands a clear and deep understanding is the fact that we live under the rule of capital, because that is the most dominant force, the thing that controls every aspect of our lives. It’s control reduces right down to the simple constraint of the amount of money we hold in our pockets. What is this social substance? What do these numbers that we throw around between us, that fly around the whole world, actually mean? This is why I’m interested in the theory of value, and also why I think that — ultimately — it is the most important question in contemporary social science.

With regards to socialism, I hope that the more general theory will eventually contribute to raising class consciousness. I think that, even amongst self-identifying socialists, there’s a certain lack of confidence, or nervousness, about the scientific validity of Marx’s economics and his theory of value. And that’s understandable, because mainstream economics, what’s perceived to be scientific economics, is virulently anti-Marxist. And so if you’re a worker who has, let’s say naive, views about scientific production under the rule of capital, then you’re unavoidably going to be influenced by the dominant views and ideology. Let’s add into the mix that Marx’s critique of political economy is incomplete, and even incorrect in places. Marxists must be the first to recognise and acknowledge any theoretical flaws, because only true ideas can be effective, and we want our politics to be as effective as possible. So some of the mainstream criticisms of Marx’s theory have real substance, actually point to real problems. Life would be very easy if the class struggle in the realm of ideology was simply a matter of one side automatically being entirely right, and the other side automatically being entirely wrong. History and science don’t work like that. And so I see my contribution as deepening and strengthening Marx’s critique of political economy.

In terms of the socialist transition. I mean, again a huge topic. But we can say something quite precise here. I think Luigi Pasinetti’s work is very important for Marxism and socialism more generally. Unfortunately, it is not very well known outside of the academy, and there’s a barrier to entry because he uses lots of linear algebra. There are many re-interpretations of Marx’s Capital, of varying quality, which simply require the ability to read and understand abstract concepts. So the barrier to entry is low. But the bad news is that natural language isn’t sufficient for a deep scientific understanding of the social reality we live in. So for Pasinetti, the barrier for entry is higher, because the reader needs to understand some mathematics, but so are the rewards.

Pasinetti’s mathematics really sings, and once you wrap your head around it, really simplifies things. A good starting point is his book, Structural Economic Dynamics, where he deliberately models a (seemingly simple) pure labour economy, without capital and without land. That seems like an extreme simplification, but it’s incredibly illuminating, and cuts to the heart of the matter, and identifies important objective constraints on production. And these hard, natural, material constraints will be present in both the socialist transition, and afterwards. I think it’s a very dangerous utopian attitude to think otherwise, to think we can simply wave a magic wand and believe that the abolition of the value-form automatically abolishes the necessity for any society to materially reproduce itself. In Pasinetti’s work we clearly see the objective consequences of the hard limit of the available workforce (i.e., Marx’s total working day) and the unavoidable challenge of reallocating that workforce to different activities in the context of ceaseless technical change and innovation, whether that is achieved by the market, by democratic planning, or some mixture of the two, or even by methods we haven’t thought of yet.

So Pasinetti’s work I think is very relevant for transitioning to a dynamic socialism, which democratically controls the length of the working day, and yet maintains a very high quality of life for everyone. Pasinetti makes extensive use of generalisations of Marx’s definition of labour-values to get at the root of phenomena. And he has to, because the phenomena demands it. So, I think my further generalisations, and my work on the dynamics of the law of value, could play a role here too.

Supplementary material:

Wright, I. (2014). A category-mistake in the classical labour theory of value. Erasmus Journal for Philosophy and Economics, 7(1), 27-55. https://doi.org/10.23941/ejpe.v7i1.155

Wright, I. (2018) Marx’s transformation problem and Pasinetti’s vertically integrated subsystems. Cambridge Journal of Economics, Volume 43, Issue 1, January 2019, Pages 169–186 https://academic.oup.com/cje/advance-article/doi/10.1093/cje/bex068/5057684?guestAccessKey=a51b4b1a-eb0d-49ce-a1a2-eb93a228e899

Wright, I. (2017) The general theory of labour value. Workshop on Input-Output and Multisectoral Analysis: Theory and Applications OU, Milton Keynes https://ianwrightsite.files.wordpress.com/2017/04/general-theory-labour-value2.pdf (see accompanying video https://youtu.be/jROxFYv1bks)

Wright, I. (2016). The law of value : a contribution to the classical approach to economic analysis. PhD thesis The Open University. https://eastsidemarxism.files.wordpress.com/2017/04/wright-thesis-deposited.pdf

Wright. I. (2009) On nonstandard labour values, Marx’s transformation problem and Ricardo’s problem of an invariable measure of value. BOLETIM DE CIÊNCIAS ECONÓMICAS, VOLUME LII https://digitalis-dsp.uc.pt/bitstream/10316.2/24730/1/BoletimLII_Artigo4.pdf?ln=pt-pt

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s