Marx himself considered his theory of surplus-value his most important contribution to the progress of economic analysis (Marx, letter to Engels of 24 August 1867). It is through this theory that the wide scope of his sociological and historical thought enables him simultaneously to place the capitalist mode of production in his historical context, and to find the root of its inner economic contradictions and its laws of motion in the specific relations of production on which it is based.

As said before, Marx’s theory of classes is based on the recognition that in each class society, part of society (the ruling class) appropriates the social surplus product. But that surplus product can take three essentially different forms (or a combination of them). It can take the form of straightforward unpaid surplus labour, as in the slave mode of production, early feudalism or some sectors of the Asiatic mode of production (unpaid corvée labour for the Empire). It can take the form of goods appropriated by the ruling class in the form of use-values pure and simple (the products of surplus labour), as under feudalism when feudal rent is paid in a certain amount of produce (produce rent) or in its more modern remnants, such as sharecropping. And it can take a money form, like money-rent in the final phases of feudalism, and capitalist profits. Surplus-value is essentially just that: the money form of the social surplus product or, what amounts to the same, the money product of surplus labour. It has therefore a common root with all other forms of surplus product: unpaid labour.

This means that Marx’s theory of surplus-value is basically a deduction (or residual) theory of the ruling classes’ income. The whole social product (the net national income) is produced in the course of the process of production, exactly as the whole crop is harvested by the peasants. What happens on the market (or through appropriation of the produce) is a distribution (or redistribution) of what already has been created. The surplus product, and therefore also its money form, surplus-value, is the residual of that new (net) social product (income) which remains after the producing classes have received their compensation (under capitalism: their wages). This ‘deduction’ theory of the ruling classes’ income is thus ipso factor an exploitation theory. Not in the ethical sense of the word – although Marx and Engels obviously manifested a lot of understandable moral indignation at the fate of all the exploited throughout history, and especially at the fate of the modern proletariat – but in the economic one. The income of the ruling classes can always be reduced in the final analysis to the product of unpaid labour: that is the heart of Marx’s theory of exploitation.

That is also the reason why Marx attached so much importance to treating surplus-value as a general category, over and above profits (themselves subdivided into industrial profits, bank profits, commercial profits etc.), interest and rent, which are all part of the total surplus product produced by wage labour. It is this general category which explains both the existence (the common interest) of the ruling class (all those who live off surplus value), and the origins of the class struggle under capitalism.

Marx likewise laid bare the economic mechanism through which surplus-value originates. At the basis of that economic mechanism is a huge social upheaval which started in Western Europe in the 15th century and slowly spread over the rest of the continent and all other continents (in many so-called underdeveloped countries, it is still going on to this day).

Through many concomitant economic (including technical), social, political and cultural transformations, the mass of the direct producers, essentially peasants and handicraftsmen, are separated from their means of production and cut off from free access to the land. They are therefore unable to produce their livelihood on their own account. In order to keep themselves and their families alive, they have to hire out their arms, their muscles and their brains, to the owners of the means of production (including land). If and when these owners have enough money capital at their disposal to buy raw materials and pay wages, they can start to organise production on a capitalist basis, using wage labour to transform the raw materials which they buy, with the tools they own, into finished products which they then automatically own too.

The capitalist mode of production thus presupposes that the producers’ labour power has become a commodity. Like all other commodities, the commodity labour power has an exchange value and a use value. The exchange value of labour power, like the exchange value of all other commodities, is the amount of socially necessary labour embodied in it, i.e. its reproduction costs. This means concretely the value of all the consumer goods and services necessary for a labourer to work day after day, week after week, month after month, at approximately the same level of intensity, and for the members of the labouring classes to remain approximately stable in number and skill (i.e. for a certain number of working-class children to be fed, kept and schooled, so as to replace their parents when they are unable to work any more, or die). But the use value of the commodity labour power is precisely its capacity to create new value, including its potential to create more value than its own reproduction costs. Surplus-value is but that difference between the total new value created by the commodity labour power, and its own value, its own reproduction costs. The whole marxian theory of surplus-value is therefore based upon that subtle distinction between ‘labour power’ and ‘labour’ (or value). But there is nothing ‘metaphysical’ about this distinction. It is simply an explanation (demystification) of a process which occurs daily in millions of cases.

The capitalist does not buy the worker’s ‘labour’. If he did that there would be obvious theft, for the worker’s wage is obviously smaller than the total value he adds to that of the raw materials in the course of the process of production. No: the capitalist buys ‘labour power’, and often (not always of course) he buys it at its justum pretium, at its real value. So he feels unjustly accused when he is said to have caused a ‘dishonest’ operation. The worker is victim not of vulgar theft but of a social set-up which condemns him first to transform his productive capacity into a commodity, then to sell that labour power on a specific market (the labour market) characterised by institutional inequality, and finally to content himself with the market price he can get for that commodity, irrespective of whether the new value he creates during the process of production exceeds that market price (his wage) by a small amount, a large amount, or an enormous amount.

The labour power the capitalist has bought ‘adds value’ to that of the used-up raw materials and tools (machinery, buildings etc.). If, and until that point of time, this added value is inferior or equal to the workers’ wages, surplus-value cannot originate. But in that case, the capitalist has obviously no interest in hiring wage labour. He only hires it because that wage labour has the quality (the use value) to add to the raw materials’ value more than its own value (i.e. its own wages). This ‘additional added value’ (the difference between total ‘value added’ and wages) is precisely surplus-value. Its emergence from the process of production is the precondition for the capitalists’ hiring workers, for the existence of the capitalist mode of production.

The institutional inequality existing on the labour market (masked for liberal economists, sociologists and moral philosophers alike by juridical equality) arises from the very fact that the capitalist mode of production is based upon generalised commodity production, generalised market economy. This implies that a propertyless labourer, who owns no capital, who has no reserves of larger sums of money but who has to buy his food and clothes, pay his rent and even elementary public transportation for journeying between home and workplace, in a continuous way in exchange of money, is under the economic compulsion to sell the only commodity he possesses, to wit his labour power, also on a continuous basis. He cannot withdraw from the labour market until the wages go up. He cannot wait.

But the capitalist, who has money reserves, can temporarily withdraw from the labour market. He can lay his workers off, can even close or sell his enterprise and wait a couple of years before starting again in business. The institutional differences makes price determination of the labour market a game with loaded dice, heavily biased against the working class. One just has to imagine a social set-up in which each citizen would be guaranteed an annual minimum income by the community, irrespective of whether he is employed or not, to understand that ‘wage determination’ under these circumstances would be quite different from what it is under capitalism. In such a set-up the individual would really have the economic choice whether to sell his labour power to another person (or a firm) or not. Under capitalism, he has no choice. His is forced by economic compulsion to go through that sale, practically at any price.

The economic function and importance of trade unions for the wage-earners also clearly arises from that elementary analysis. For it is precisely the workers’ ‘combination’ and their assembling a collective resistance fund (what was called by the first French unions caisses de résistance, ‘reserve deposits’) which enables them, for example through a strike, to withdraw the supply of labour power temporarily from the market so as to stop a downward trend of wages or induce a wage increase. There is nothing ‘unjust’ in such a temporary withdrawal of the supply of labour power, as there are constant withdrawals of demand for labour power by the capitalists, sometimes on a huge scale never equalled by strikes. Through the functioning of strong labour unions, the working class tries to correct, albeit partially and modestly, the institutional inequality on the labour market of which it is a victim, without ever being able to neutralise it durably or completely.

It cannot neutralise it durably because in the very way in which capitalism functions there is a powerful built-in corrective in favour of capital: the inevitable emergence of an industrial reserve army of labour. There are three key sources for that reserve army: the mass of precapitalist producers and self-employed (independent peasants, handicraftsmen, trades-people, professional people, small and medium-sized capitalists); the mass of housewives (and to a lesser extent, children); the mass of the wage-earners themselves, who potentially can be thrown out of employment.

The first two sources have to be visualised not only in each capitalist country seen separately but on a world scale, through the operations of international migration. They are still unlimited to a great extent, although the number of wage-earners the world over (including agricultural wage labourers) has already passed the one billion mark. As the third source, while it is obviously not unlimited (if wage labour would disappear altogether, if all wage labourers would be fired, surplus-value production would disappear too; that is why ‘total robotism’ is impossible under capitalism), its reserves are enormous, precisely in tandem with the enormous growth of the absolute number of wage earners.

The fluctuations of the industrial reserve army are determined both by the business cycle and by long-term trends of capital accumulation. Rapidly increasing capital accumulation attracts wage labour on a massive scale, including through international migration. Likewise, deceleration, stagnation or even decline of capital accumulation inflates the reserve army of labour. There is thus an upper limit to wage increases, when profits (realised profits and expected profits) are ‘excessively’ reduced in the eyes of the capitalists, which triggers off such decelerated, stagnating or declining capital accumulation, thereby decreasing employment and wages, till a ‘reasonable’ level of profits is restored. This process does not correspond to any ‘natural economic law’ (or necessity), nor does it correspond to any ‘immanent justice’. It just expresses the inner logic of the capitalist mode of production, which is geared to profit. Other forms of economic organisation could function, have functioned and are functioning on the basis of other logics, which do not lead to periodic massive unemployment. On the contrary, a socialist would say – and Marx certainly thought so – that the capitalist system is an ‘unjust’, or better stated ‘alienating’, ‘inhuman’ social system, precisely because it cannot function without periodically reducing employment and the satisfaction of elementary needs for tens of millions of human beings.

Marx’s theory of surplus-value is therefore closely intertwined with a theory of wages which is far away from Malthus’s, Ricardo’s or the early socialists’ (like Ferdinand Lassalle’s) ‘iron law of wages’, in which wages tend to fluctuate around the physiological minimum. That crude theory of ‘absolute pauperisation’ of the working class under capitalism, attributed to Marx by many authors (Popper, 1945, et al.), is not Marx’s at all, as many contemporary authors have convincingly demonstrated (see among others Rosdolsky, 1968). Such an ‘iron law of wages’ is essentially a demographic one, in which birth rates and the frequency of marriages determine the fluctuation of employment and unemployment and thereby the level of wages.

The logical and empirical inconsistencies of such a theory are obvious. Let it be sufficient to point out that while fluctuations in the supply of wage-labourers are considered essential, fluctuations in the demand for labour power are left out of the analysis. It is certainly a paradox that the staunch opponent of capitalism, Karl Marx, pointed out as early as in the middle of the 19th century the potential for wage increases under capitalism, even though not unlimited in time and space. Marx also stressed the fact that for each capitalist, wage increases of other capitalists’ workers are considered increases of potential purchasing power, not increases in costs.

Marx distinguishes two parts in the workers’ wage, two elements of reproduction costs of the commodity labour power. One is purely physiological, and can be expressed in calories and energy quanta; this is the bottom below which the wage cannot fall without destroying slowly rapidly the workers’ labour capacity. The second one is historical-moral, as Marx calls it, and consists of those additional goods and services which a shift in the class relationship of forces, such as a victorious class struggle, enables the working class to incorporate into the average wage, the socially necessary (recognised) reproduction costs of the commodity labour power (e.g. holidays after the French general strike of June 1936). This part of the wage is essentially flexible. It will differ from country to country, continent to continent and from epoch to epoch, according to many variables. But it has the upper limit indicated above: the ceiling from which profits threaten to disappear, or to become insufficient in the eyes of the capitalists, who then go on an ‘investment strike’.

So Marx’s theory of wages is essentially an accumulation-of-capital theory of wages which sends us back to what Marx considered the first ‘law of motion’ of the capitalist mode of production: the compulsion for the capitalists to step up constantly the rate of capital accumulation.