The Labour Theory of Value (LTV) is one of probably only a handful of economic theories, along with Francois Quesnay’s Tableau Economique, which have actually been completely abandoned over the past couple of centuries. So, in the interests of combating blind Whiggery, allow me to revive it (maybe I’ll do Quesnay another day, though here’s a sort-of modern version). I’m not going to argue the LTV is necessarily correct; I am merely interested in clearing up some common misconceptions.
My initial reaction to the LTV was the same as almost everyone’s: hostility. Why is there even a need for such a thing? Has it not been discredited on many fronts: logically, empirically, ethically? Are there not ways to preserve the important aspects of Marx, while at the same time ditching his dated and irrelevant theory of value? And yet, after a while, the hostility fades as you realise that:
- You should never be hostile to a theory based only on its name and what other people have said about it, and
- The theory, properly understood, is valid and highly illuminating, and explains many real world phenomena.
One common source of confusion with the LTV is the lack of appreciation that it only applies under capitalism, when goods are produced with wage labour, for the purpose of sale (this is what makes them ‘commodities’). For this reason it doesn’t apply to, say, artifacts; this blog post; or house work. This historical specificity can be a problem for economists, even heterodox ones, as they are generally wont to find principles which extend across different times and societies. This, for example, was the chief problem with Arun Bose’s critique of the LTV, which argued that no matter how far back you go, you will always have a commodity residue embedded in the value of a current commodity, and so labour is not the only source of value. Bose failed to consider that if you go back far enough, you will not have ‘commodities’ but simply naturally occurring objects, or objects not produced for sale. Only when labour was applied to these for the purpose of sale was ‘value’ created in the Marxist sense of the word.
In a nutshell, Marx’s theory goes as follows: under capitalism, the value of commodities is determined by the “socially necessary” amount of labour required to produce them (‘variable capital’), plus the current necessary cost of the capital used up in production (‘constant capital’). Fixed capital, such as machines, adds value at the same rate it depreciates, while raw commodities are used up completely and so add all of their value. Labour is generally paid less than the value it adds, and therefore is the sole source of profit.
Here’s a brief mathematical example: say an hour of labour adds £1 of value, and a certain type of chair requires 8 hours of labour (‘labour-time’), uses £2 worth of wood and depreciates a saw worth £10 by 1/10th (i.e. after the saw is used 10 times it will break). It follows that, according to the LTV, the value of this chair is:
(1/10)*£10 + (8*£1) + £2 = £11
The only way the capitalist can make a profit is to pay the labourer less than the value he creates (for the most part, Marx suggested wages were determined by a social subsistence level). So if the wage is, say, £0.50, the capitalist will have £4 worth of profit. Contrary to what many think, this does not imply that capital-intensive industries will have lower rates of profit, as the rate of profit will tend to equalise between industries, ‘sharing out’ the total surplus value produced in the economy.
The qualifiers of “necessary” costs and “socially necessary” labour are also important. It’s logically possible that a madman could purchase a saw for £100 for some reason, or that a lazy labourer could take 10 hours to make the chair, but this would do nothing to alter the resultant value of the chair. Marx was concerned with general rules, not specific cases, which could obviously fluctuate wildly as they are based on human behaviour.
The main implication of this theory is that, since capitalists tend to use labour saving technology to increase productivity, over time they use relatively more constant capital – which cannot be a source of surplus – and this drives down the rate of profit: the Tendency of the Rate of Profit to Fall (TRPF). Though the first capitalist who uses the technology will be able to sell at the market price, and thus gain, once the technology is widely adopted, the value of the commodity will decrease – a ‘fallacy of composition’. Again, this may not be true in particular industries at any one time, but it holds true across the economy as a whole. The result will be intermittent crises as capitalists face lower profits and try to increase them by pushing down wages, devaluing their constant capital, or through technological progress. Marx never predicted capitalism would collapse in on itself, though he did suggest that the working class would revolt as their wages were pushed down.
Does the subjective theory of value (STV) ‘refute’ the LTV?
The basic point that ‘value is in the eye of the beholder’ is often thought to be where the debate ends. This is surprising, because it has little to do with Marx’s main theoretical implications, which as I have noted, concerned economy-wide trends. Marx was well aware that price and value may differ wildly based on monopoly, demand and other fluctuations, but considered it irrelevant to his theory of value. He never claimed to be able to predict the day-to-day movements of prices, and purported attempts to use the LTV to do this are erroneous.
So while it may well be true that the tastes of rich people propel the price of diamonds upwards (whether this is due to the fact that they ‘subjectively value’ diamonds higher than poor people or the fact that they’re rich is up for debate, but I digress), but according to the LTV, this imbalance between price and value must be offset somewhere else, by some other commodity selling below its value. The important thing for Marx was the aggregate equality of price and value*.
Neither were Marx or the classical economists unaware of the utility of commodities and the array of use they might be put to: they called this the ‘use-value‘ of a commodity, though they did think this was socially determined rather than subjective. In classical economics, use-value is not quantifiable: a commodity either is a use-value or it isn’t, and there’s no definitive way to gauge the ‘value’ of sitting on a chair. This means it is not possible for use-value to translate into prices, as use-values are incommensurable between commodities.
The inherently intangible nature of use-value caused Marx and other classical economists to ask: what is it that makes commodities expressible by the same yardstick (money) under capitalism? The answer was that commodities had a twin expression of value: exchange-value. A commodity needed a use-value to have an exchange-value (i.e. it needed to be useful to be worth anything), but the two types of value were not equivalent or even on the same plane. The classical economists decided that exchange-value was determined not by utility, but by the labour required to make a commodity. This is because labour was the common element between all commodities: even capital ultimately reduced to labour, making it the prime candidate for the determination of exchange-value**.
In fact, the subjective theory of value is in many ways a fairly crude attempt to combine use-value and exchange-value, and doesn’t really offer anything new compared to the classical theory. The only way the quantitative expression of subjective valuation can be determined is either circular, ‘revealed’ by purchasing decisions ex post, or in the case of neoclassical economics, completely deterministic based on how a model is constructed. Therefore, as far as market prices are concerned, the theory doesn’t have any more predictive power than simply saying ‘use-value cannot be formalised’. Furthermore, though mutually beneficial trade is achieved through exchange of use-values, as far as exchange-values go trade is a zero-sum game: money exchanged must sum to zero. Neoclassical models of utility obscure this fact.
The ‘transformation problem’ and all that
Hopefully, I have cleared up some of the qualitative misconceptions surrounding the LTV. However, the critique which has probably done the most to ‘discredit’ Marx in academic circles is the idea that the maths simply doesn’t add up, usually based on ‘physicalist’ or Sraffian critiques (in fact, Paul Samuelson relied on this to steer his students away from Marx, despite rejecting Sraffian economics elsewhere). I will avoid the maths here and just try to get to the crux of the misconceptions, which is actually quite simple to do: once the basic methodological misinterpretations are highlighted, the purported complications simply disappear.***
In physicalist/Sraffian models, key variables such as prices and the rate of profit are all determined physically (generally by technology & distribution). The result is that the rates of value are superfluous and completely different to prices: when you try to ‘transform’ them, you run into problems. Yet this only really renders Marx logically inconsistent by interpreting him in a manner that…renders him logically inconsistent. For while physicalist models determine output and input prices simultaneously, Marx actually modeled them temporally, so they could easily differ. As the two prices depend on different transactions at different points in time, this is a fairly reasonable modelling tool by anyone’s standards, but it seems to have been lost in the wilderness of economic debate.
The root of this fundamental incompatibility is that the physicalist notion that key variables are determined simultaneously completely contradicts one of Marx’s premises: that value is determined by labour-time. Intuitively, this makes sense: in a simultaniest world that doesn’t really have time, how could the time spent labouring have any relevance? To show the inconsistency more rigorously, determination of value by labour-time implies that value will fall as productivity rises, which implies that the rate of profit will fall relatively in value terms compared to physical terms (more will be produced, but it will be worth less). Hence, the physical rate of profit – determined simultaneously by the parameters of the model (technology, distribution) – and the value rate of profit, determined by labour-time, differ, and physicalism is incompatible with Marxism.
Once we allow value to be determined by labour-time, and therefore allow output and input prices to differ, a myriad of supposed refutations of the LTV fail: the Okishio theorem; Ian Steedman’s Marx after Sraffa; V. K. Dmitriev’s labourless theory of value (unavailable online). The temporal method can also be combined with the ‘single system’ interpretation of the LTV, which suggests that instead of having two separate systems of price and value, as so many critics do, the two are linked: the necessary cost of inputs at the time of purchase is equal to the sum of the value transferred in production.
Thus, the most plausible mathematical interpretation of Marx’s LTV is the Temporal Single-System Interpretation, which I find a valid and illuminating way of modelling production. The basic elucidation of the theory, and the relationship between values and prices, simply becomes a lot less complicated, and the alleged ‘transformation problem‘ loses its venom as prices and values interact and adjust temporally.
There are a myriad of ways one can object to the LTV, but the idea that is is nonsensical and incoherent is simply based on misunderstandings. One may well disagree with the premise that labour is the source of value (I do, simply because I have no positive reason to believe it). One may also endorse alternative theories over the LTV. But, based on a clear understanding, there is no a priori reason not to develop a comprehensive understanding of Marx’s theory, and treat it in the same way one would treat any other theory in economics.
*It was these appeals to aggregates and totals, instead of the immediate behaviour of the system, that led Bohm-Bawerk to term Marx’s theory ‘tautological’. Yet this rests on Bohm-Bawerk’s own premise that money is neutral, which is controversial to say the least. In any case, Marx’s theory has clear, non-tautological implications, such as the TRPF.
**I find this explanation unsatisfying as labour as well as capital is heterogeneous. Marxists reply to this by arguing that labour under capitalism shares a common element: abstract labour, performed specifically to create commodities. This is partially convincing, but it doesn’t alter the main fact that different types of labour are highly disparate in nature.
***In this section I am drawing heavily on Andrew Kliman’s ‘Reclaiming Marx’s Capital: A Refutation of the Myth of Inconsistency‘ . It’s a great book: clear and concise, and a great example of a ‘remorseless logician’. I am, however, undecided on whether he started with a mistake and ended up in bedlam.