As an economist, Marx is generally situated in the continuity of the great classical school of Adam Smith and Ricardo. He obviously owes a lot to Ricardo, and conducts a running dialogue with that master in most of his mature economic writings.

Marx inherited the labour theory of value from the classical school. Here the continuity is even more pronounced; but there is also a radical break, For Ricardo, labour is essentially a numeraire, which enables a common computation of labour and capital as basic elements of production costs. For Marx, labour is value. Value is nothing but that fragment of the total labour potential existing in a given society in a certain period (e.g. a year or a month) which is used for the output of a given commodity, at the average social productivity of labour existing then and there, divided by the total number of these commodities produced. and expressed in hours (or minutes), days, weeks, months of labour.

Value is therefore essentially a social, objective and historically relative category, It is social because it is determined by the overall result of the fluctuating efforts of each individual producer (under capitalism: of each individual firm or factory). It is objective because it is given, once the production of a given commodity is finished, and is thus independent from personal (or collective) valuations of customers on the market place; and it is historically relative because it changes with each important change (progress or regression) of the average productivity of labour in a given branch of output, including in agriculture and transportation.

This does not imply that Marx’s concept of value is in any way completely detached from consumption. It only means that the feedback of consumers’ behaviour and wishes upon value is always mediated through changes in the allocation of labour inputs in production, labour being seen as subdivided into living labour and dead (dated) labour, i.e. tools and raw materials. The market emits signals to which the producing units react. Value changes after these reactions, not before them. Market price changes can of course occur prior to changes in value. In fact, changes in market prices are among the key signals which can lead to changes in labour allocation between different branches of production, i.e. to changes in labour quantities necessary to produce given commodities. But then, for Marx, values determine prices only basically and in the medium-term sense of the word. This determination only appears clearly as an explication of medium and long-term price movements. In the shorter run, prices fluctuate around values as axes. Marx never intended to negate the operation of market laws, of the law of supply and demand, in determining these short-term fluctuations.

The ‘law of value’ is but Marx’s version of Adam Smith’s ‘invisible hand’. In a society dominated by private labour, private producers and private ownership of productive inputs, it is this ‘law of value’, an objective economic law operating behind the backs of all people, all ‘agents’ involved in production and consumption, which, in the final analysis, regulates the economy, determines what is produced and how it is produced (and therefore also what can be consumed). The ‘law of value’ regulates the exchange between commodities, according to the quantities of socially necessary abstract labour they embody (the quantity of such labour spent in their production). Through regulating the exchange between commodities, the ‘law of value’ also regulates, after some interval, the distribution of society’s labour potential and of society’s non-living productive resources between different branches of production. Again, the analogy with Smith’s ‘invisible hand’ is striking.

Marx’s critique of the ‘invisible hand’ concept does not dwell essentially on the analysis of how a market economy actually operates. It would above all insist that this operation is not eternal, not immanent in ‘human nature’, but created by specific historical circumstances, a product of a special way of social organisation, and due to disappear at some stage of historical evolution as it appeared during a previous stage. And it would also stress that this ‘invisible hand’ leads neither to the maximum of economic growth nor to the optimum of human wellbeing for the greatest number of individuals, i.e. it would stress the heavy economic and social price humankind had to pay, and is still currently paying, for the undeniable progress the market economy produced at a given stage of historical evolution.

The formula ‘quantities of abstract human labour’ refers to labour seen strictly as a fraction of the total labour potential of a given society at a given time, say a labour potential of 2 billion hours a year (1 million potential producers, each supposedly capable of working 2000 hours a year). It therefore implies making an abstraction of the specific trade or occupation of a given male or female producer, the product of a day’s work of a weaver not being worth less or more than that of a peasant, a miner, a house-builder, a milliner or a seamstress. At the basis of that concept of ‘abstract human labour’ lies a social condition, a specific set of social relations of production, in which small independent producers are essentially equal. Without that equality, social division of labour, and therefore satisfaction of basic consumers’ needs, would be seriously endangered under that specific organisational set-up of the economy. Such an equality between small commodity owners and producers is later transformed into an equality between owners of capital under the capitalist mode of production.

But the concept of the homogeneity of productive human labour, underlying that of ‘abstract human labour’ as the essence of value, does not imply a negation of the difference between skilled and unskilled labour. Again: a negation of that difference would lead to the breakdown of the necessary division of labour, as would any basic heterogeneity of labour inputs in different branches of output. It would then not pay to acquire skills: most of them would disappear. So Marx’s labour theory of value, in an internally coherent way, leads to the conclusion that one hour of skilled labour represents more value than one hour of unskilled labour, say represents the equivalent of 1.5 hours of unskilled labour. The difference would result from the imputation of the labour it costs to acquire the given skill, While an unskilled labourer would have a labour potential of 120,000 hours during his adult life, a skilled labourer would only have a labour potential of 80,000 hours, 40,000 being used for acquiring, maintaining and developing his skill. Only if one hour of skilled labour embodies the same value of 1.5 hours of unskilled labour, will the equality of all ‘economic agents’ be maintained under these circumstances, i.e. will it ‘pay’ economically to acquire a skill.

Marx himself never extensively dwelled on this solution of the so-called reduction problem. This remains indeed one of the most obscure parts of his general economic theory. It has led to some, generally rather mild, controversy. Much more heat has been generated by another facet of Marx’s labour theory of value, the so-called transformation problem. Indeed, from Böhm-Bawerk writing a century ago till the recent contributions of Sraffa (1960) and Steedman (1977), the way Marx dealt with the transformation of values into ‘prices of production’ in Capital Vol. III has been considered by many of his critics as the main problem of his ‘system’, as well as being a reason to reject the labour theory of value out of hand.

The problem arises out of the obvious modification in the functioning of a market economy when capitalist commodity production substitutes itself for simple commodity production. In simple commodity production, with generally stable technology and stable (or easily reproducible) tools, living labour is the only variable of the quantity and subdivision of social production. The mobility of labour is the only dynamic factor in the economy. As Engels pointed out in his Addendum to Capital Vol. III (Marx, g, pp, 1034-7), in such an economy, commodities would be exchanged at prices which would be immediately proportional to values, to the labour inputs they embody.

But under the capitalist mode of production, this is no longer the case. Economic decision-taking is not in the hands of the direct producers. It is in the hands of the capitalist entrepreneurs in the wider sense of the word (bankers – distributors of credit – playing a key role in that decision-taking, besides entrepreneurs in the productive sector properly speaking). Investment decisions, i.e. decisions for creating, expanding, reducing or closing enterprises, determine economic life. It is the mobility of capital and not the mobility of labour which becomes the motive force of the economy. Mobility of labour becomes essentially an epiphenomenon of the mobility of capital.

Capitalist production is production for profit. Mobility of capital is determined by existing or expected profit differentials. Capital leaves branches (countries, regions) with lower profits (or profit expectations) and flows towards branches (countries, regions) with higher ones. These movements lead to an equalisation of the rate of profit between different branches of production. But approximately equal returns on all invested capital (at least under conditions of prevailing ‘free competition’) coexist with unequal proportions of inputs of labour in these different branches. So there is a disparity between the direct value of a commodity and its ‘price of production’, that ‘price of production’ being defined by Marx as the sum of production costs (costs of fixed capital and raw materials plus wages) and the average rate of profit multiplied with the capital spent in the given production.

The so-called ‘transformation problem’ relates to the question of whether a relation can nevertheless be established between value and these ‘prices of production’, what is the degree of coherence (or incoherence) of the relation with the ‘law of value’ (the labour theory of value in general), and what is the correct quantitative way to express that relation, if it exists.

We shall leave aside here the last aspect of the problem, to which extensive analysis has recently been devoted (Mandel and Freeman, 1984). From Marx’s point of view, there is no incoherence between the formation of ‘prices of production’ and the labour theory of value. Nor is it true that he came upon that alleged difficulty when he started to prepare Capital Vol.III, i.e. to deal with capitalist competition, as several critics have argued (see e.g. Joan Robinson, 1942). In fact, his solution of the transformation problem is already present in the Grundrisse, before he even started to draft Capital Vol. I.

The sum total of value produced in a given country during a given span of time (e.g. one year) is determined by the sum total of labour-inputs. Competition and movements of capital cannot change that quantity, The sum total of values equals the sum total of ‘prices of production’. The only effect of capital competition and capital mobility is to redistribute that given sum – and this through a redistribution of surplus value (see below) – between different capitals, to the benefit of some and at the expense of others.

Now the redistribution does not occur in a haphazard or arbitrary way. Essentially value (surplus-value) is transferred from technically less advanced branches to technologically more advanced branches. And here the concept of ‘quantities of socially necessary labour’ comes into its own, under the conditions of constant revolutions of productive technology that characterise the capital mode of production. Branches with lower than average technology (organic composition of capital, see below) can be considered as wasting socially necessary labour. Part of the labour spent in production in their realm is therefore not compensated by society. Branches with higher than average technology (organic composition of capital) can be considered to be economising social labour; their labour inputs can therefore be considered as more intensive than average, embodying more value. In this way, the transfer of value (surplus-value) between different branches, far from being in contradiction with the law of value, is precisely the way it operates and should operate under conditions of ‘capitalist equality’, given the pressure of rapid technological change.

As to the logical inconsistency often supposedly to be found in Marx’s method of solving the ‘transformation problem’ – first advanced by von Bortkiewicz (1907) – it is based upon a misunderstanding in our opinion. It is alleged that in his ‘transformation schemas’ (or tables) Marx calculates inputs in ‘values’ and outputs in ‘prices of production’, thereby omitting the feedback effect of the latter on the former. But that feedback eject is unrealistic and unnecessary, once one recognises that inputs are essentially data. Movements of capital posterior to the purchase of machinery or raw materials, including the ups and dawns of prices of finished products produced with these raw materials, cannot lead to a change in prices and therefore of profits of the said machinery and raw materials, on sales which have already occurred. What critics present as an inconsistency between ‘values’ and ‘prices of production’ is simply a recognition of two different time-frameworks (cycles) in which the equalisation of the rate of profit has been achieved, a first one for inputs, and a second, later one for outputs.