John Kilcullen
Copyright (c) 1996, R.J. Kilcullen.
In this lecture I will try to explain and assess some of the criticisms made of Marx's analysis of capitalism by later economists.
Marx was an exponent of the 'Labour theory' of value, held also by Adam Smith, Ricardo, James Mill, John Stuart Mill and some early 19th century political economists. This is the theory that human labour is the sole source of value (of exchange value), and that the value of a commodity is the quantity of socially necessary labour embodied in it. When we exchange commodities they exchange in the ratio of the quantities of labour required to produce them and bring them into the market.
The chief rival theory at the time was the 'supply and demand' theory: that the exchange value of a commodity in a competitive market depends on the quantity brought into the market and the quantity required by buyers: when there is a glut the price falls, when there is a shortage the price rises. These movements take place because of competition - there are many potential buyers and sellers, who look round for the best deal.
Marx gave supply and demand and competition the minor job of explaining short-term fluctuations in prices. A temporary shortage causes a temporary price rise. But what still needs explaining are the longer-term average exchange ratios, the levels at about which the short-term fluctuations occur. Why is it that through all the fluctuations houses are regularly worth many times the value of a pair of shoes? To explain this Marx pointed to the different quantities of human labour necessary by normal methods to produce these two kinds of commodities. As long as the same productive methods remain normal, houses and shoes will have the same underlying or average relative value, through all the short-term fluctuations of price.
After Marx, starting in the 1870s, political economists worked out a more sophisticated version of the supply and demand theory, which seemed to them capable of explaining whatever the labour theory of value could explain, and of explaining other economic phenomena as well, by means of a few simple assumptions which seemed to correspond realistically with actual human behaviour. (Criteria of theory choice).
The revised supply-and-demand theory was not worked out in response to Marx, whose work was not known to its pioneers. The labour theory of value Marx got from Ricardo, and it was Ricardo and J.S. Mill who were the chief targets of the new school - the 'marginalist' school as it is usually called. But when Marx's Capital later came to the attention of political economists its theory of value seemed to them to be obsolete. (M. Blaug, Economic Theory in Retrospect (2nd edn) (HB/75/.B664), p.307).
The main point about the marginalist theory is that they switched the focus of attention from the average or normal price, value or cost to the price, value or cost of the marginal item, that is the item just inside or just beyond the edge of what is actually produced or sold. I suppose the term came from agriculture: marginal land is either land at the edge of settlement as it spreads, or (sometimes the same thing) the land which it is just profitable to cultivate - the land that will go out of cultivation first if prices fall.
Let us consider first the consumer's evaluation of a commodity. It is very seldom, if ever, the case that a consumer will want as many items of a given type as he or she can get, and with equal intensity. Some people want peanuts like that, but no-one would buy a second appendix operation no matter how low the price fell. At any given price, if you buy several items there will be one or a few just worth having at that price - if the price had been a little higher you would have done without that item. It is the marginal item at that price. If the price were higher, and you did buy fewer, then at that higher price there would again be one or more marginal items - ones you would not buy if the price rose higher still. Even if you buy only one, at some price that one would be marginal.
The same idea applies if we think of things you make or do for yourself, rather than buying or selling. Some of the things you do are just barely worth doing. The contribution they make to your total satisfaction or wellbeing just barely justifies the trouble. These are your marginal activities: if life became a little harder you'd leave them out. Consider this notion of satisfaction or well being for a moment: this should be taken in a sense that is open to whatever interpretation your value system calls for. 'Satisfaction' for a hedonist means pleasure. For a religious person it might mean peace of conscience, or something like that.
No matter what are your values or priorities there are activities which just barely earn their place on your agenda and commodities which just get onto your shopping list.
Let us look at this idea now from the viewpoint of the producer of commodities. At a given selling price there will probably be some items just barely profitable to produce. The cost of production per item almost certainly varies with the volume of production. To produce just one may be very expensive because of the need for an initial investment in equipment and for learning how to do it. If you produce another thousand you may not have to spend much more on equipment or learning, so cost per item falls. But eventually it rises again - you get to the point, as production increases, where buildings need extensions, new machinery is needed, new workers have to be hired and taught, and so on. In other words, as production expands, you may come to the point when all resources are strained, and the last increase in production is barely profitable. The items barely worth producing are at the margin.
Now another idea basic to this theory, the idea of substitution. Look at it first from the consumer's point of view. If margarine becomes more expensive you think of using butter instead. If butter becomes too expensive to buy you think of making your own. If movies become too expensive you stay home and make your own drama, and so on.
Look at substitution also from the producer's point of view. Suppose the farmer of a mixed farm finds that some produce fetches such a low price that he makes no profit: he can divert his effort and resources into some other line. It is not so easy to reallocate highly specialized equipment to another line; but over time the firm can reorient its activities - it can cut back on maintenance, or sell off some of its present equipment, to make it possible to switch to some other line of production.
Note that the reorganization of production takes time: it will be necessary to accept a lower level of production of one commodity, and therefore less profit, for the present, to achieve a higher level of production of another commodity, and a higher profit, at a later date. The producer has to choose between more now and more later.
A consumer will maximise his or her overall satisfaction, and a producer will maximise net income, by substituting an item or items of one commodity for an item or items of another: if an item of one type not at present on your agenda or shopping list would contribute more than an item of another type just inside the margin, then the latter should be deleted and the former added. You delete some of the items you up till now thought barely worth having, to make room for other things you would now rather have instead. You can't just add the other things, if you are fully extended, with no spare time or money; so you give up some activities or purchases to make room for something else. And the things you give up are things at the margin. When all such substitutions have been made - when it is no longer possible to do better by deleting and substituting - then satisfaction or income is at as high a level as you can reach.
On this account the 'cost' of an item (its cost not in money, which after all you rationally want only for what it can buy, but cost in real terms - its cost in terms of what money can buy) is any one of the range of items that could be had instead if this one were deleted. To get this I have to go without that. Goods which are not commodities such as leisure, or freedom from bother, can be considered along with commodities. So the cost of a commodity may be the leisure I have to forego to earn the money to by it, or it may be some other commodity I could buy if I deleted this one.
The viewpoints of consumer and of producer can be combined: a producer is also a consumer, and a consumer may be a producer or potential producer. A person decides whether and what and how much to produce, and what to buy with income from products he sells (or from other sources), by making substitutions at the margins between various items including leisure. And if you can't be bothered making all these substitutions, then apparently a certain freedom from bother is a good you prefer to others you could have if you calculated more carefully. So bother, lack of leisure, hard work, as well as commodities, are possible equivalents to express the real cost of any commodity. And this of course is the plausibility of the labour theory of value: it treats labour, the foregoing of leisure, as the cost of anything, in the last analysis. But this is not the only cost, as we will see; and insofar as it is one of the costs, it finds a place in the marginalist theory. (This illustrates one of the points of theory choice: the marginalist theory finds a place for labour as a cost, but also for other costs the labour theory does not acknowledge.)
The facts that reorganization of production takes time, and that meanwhile income/satisfaction is reduced, can also be incorporated into this scheme. The items I survey when deciding what substitutions to make must be dated: I must consider whether so many items of X in a year's time will contribute more to my 'overall satisfaction' (over time) than however many items of Y I have to do without in the meantime. I must of course take into consideration the uncertainty of the future - I may be dead before the investment pays off. And perhaps it is rational (though many deny it) to give some preference to present good over future. So taking into account these points, I must decide whether so many items of X in a year's time are worth the Ys I have to do without meanwhile to have them.
Now a technicality: a 'demand' or 'supply' schedule for a given consumer or supplier is a statement of how many items of each commodity he or she would buy or supply at various hypothetical prices. This will depend of course partly on the prices of other things. If margarine is on special this week buy so many extra - unless there is an especially good special on cheese: then spend the money on that. No-one has a fully-worked-out set of demand or supply schedules: we work out only enough to cover likely price movements - not even that much, if we can't be bothered. Each person has his or her own set of demand and supply schedules - their 'shape' differs from one person to another, and for a given person from one time to another. How much of what will be demanded and supplied in a market on a given date at a given price depends on the 'shapes' of everyone's demand and supply schedules.
In a competitive market prices fluctuate with demand and supply: the price of an item moves until demand at that price and supply at that price are equal. Where that will be depends upon the shapes of the many demand and supply schedules of the various individuals who participate, or could participate, in the market. It does not depend only on quantities of labour required, as if leisure were the only thing people go without to have this item; they also go without other commodities. The equilibrium point is never actually attained, and anyway it keeps shifting: people's demand and supply schedule's change all the time, partly for subjective reasons (changes of taste, of mood, of fashion etc.), partly because of the fluctuations in the availability of natural resources, partly because of technological changes: but not only because of technological change, i.e. change in the amount of socially necessary labour. It may be true that for years at a time houses will be worth on average so-many times a pair of shoes, because people's demand and supply schedules may change only slowly in respects which affect the relative prices of houses and shoes, or because there is not much change in technology or in the availability of relevant natural resources.
Now for some more specific comments on the relation between this theory and the labour theory of value.
First, it is a theory not of value but of prices. Value in Marx's theory is the main determinant of long run price. Exponents of the marginalist theory have no use for a notion of 'value' as of something such as congealed labour that commodities have, the more they have the higher their long run price. What determines price is not what is in the commodity, but the complex continually shifting interrelations of everyone's demand and supply schedules. Marginalists have no theory of value as distinct from price: exchange value and price are synonymous, as far as they are concerned.
Second, there is in the marginalist theory no reason to dismiss use value in discussing exchange value. Exchange values are a function of the uses people find for the commodity, and of the uses they would find for the goods they have to do without to get the marginal item of that commodity. Which goods those are depends on availability of natural resources and technology. If air has no exchange value, although it is very useful, that is because it is so abundant that we do not have to go without other things to get it, or go without it for other purposes if we use it for this one.
Third, the marginalist theory attaches no special importance to human labour as the sole source of 'value', or as a cost. Leisure, or freedom from toil, is a good which has to be considered in deciding what to substitute for what, but it has no special place among the possible alternatives.
You may be inclined to say that by rights no income should accrue to the mere possession of a natural resource, whereas it is right that income should compensate labour.
Ethically this is not self-evident: why should those able to labour insist on a reward? In a family do the members who contribute most expect the biggest share, as an ethically appropriate reward? But the ethical point is really irrelevant. Suppose natural resources were all state-owned, so that no individual was giving up anything that was his in allowing these things to be used: still, the use of a scarce natural resource would have to count as a cost. If we use oil, for example, as if it had no value (apart from the labour put into bringing it to the user), then eventually we will have to do without things for which oil is needed: these things, at future dates, are the cost, in addition to labour, of present use of this oil.
It may be that no natural resource is absolutely non-renewable and indispensable: the cost of using it wantonly will then be what has to be invested to make some alternative available, or what has to be done without if we fail to make that investment.
Fourth, Marx's references to 'normal' methods seem unsophisticated to exponents of the marginalist theory. It may be (and usually is) the case that what is supplied is produced by a mixture of methods, some obsolescent, some inefficient. Some of what is actually supplied is produced by the marginal supplier (the one who would go out of production if prices fell slightly), or by the marginal component in some producer's plant (e.g. an old machine on the point of replacement or closure). If the price is not enough to keep the marginal producer or machine in operation the supply will fall, the price may then rise, demand may fall (as marginal purchases are deleted), and so on. It is these adjustments at the margin, against the background of everyone's demand and supply schedules, that determine price, not the 'normal' or 'average' method of production.
Fifth, there is part of Marx's theory that takes into account the fact that industries are more or less labour-intensive. Marx appealed to value and surplus value to explain long-run prices and profit, but in his theory there is another important factor behind price, namely the ratio of constant to variable capital, i.e. the proportion of total capital that can be used to buy the right to use labour power. If a type of commodity requires a lot of machinery to produce, then in Marx's theory its price will be above value (even in the long run), and the capitalists' profits will be above the corresponding amount of surplus value that capitalists in that industry extract. He has to say this, because rates of profit tend to be equal across all industries - otherwise capital would move out or unprofitable and into profitable industries. If a less labour-intensive industry could get profit only from surplus value produced by its own workers, then its rate of profit would be low. So Marx says that competition in the market for capital, labour and commodities transfers surplus value from more to less labour-intensive industries. The commodities of capital intensive industries sell at above the value that corresponds to the quantity of labour they actually embody. Competition between industries for capital and for workers tends to equalize rates of profit and wages. If the product of some capital-intensive industry is demanded strongly enough to keep such businesses going, then its price must be enough to yield about as high a rate of profit as is got in labour-intensive industries: therefore the price must be above value, and exchanges at this price must transfer surplus value from other capitalists. (These exchanges are either directly with firms in labour-intensive industries, if the product exchanged is a means of production, or indirectly if it is a consumer good: the workers in other industries, in their role as consumers, buy it at an over-value price, and their wages must therefore be higher: so value is transferred from capitalist in labour intensive industries by way of higher wages and higher prices of consumer goods to the capitalist in the less labour-intensive industries.) Notice that all this falls back on the mechanism of competition and supply and demand. So to deal with the fact of unequal ratios of constant to variable capital Marx must fall back on the 'competition-supply-and-demand' theory (of which Marginalism is a sophisticated version): whereas the marginalists never need to resort to the labour theory. (This illustrates again that the marginalist theory can explain whatever the labour theory can, whereas the labour theory at some points has to be supplemented by the supply-and-demand theory.)
Sixth, there is a disagreement about profit. In one sense profit is just the difference between the firm's income and expenses. But economists sharpen up the notion. Suppose one person saves the capital, manages the firm, and does some work at the bench. Then part of what he calls his profit is interest on his savings, part of it is the equivalent of a salary paid to a manager, part equivalent to a wage paid to a tradesman. The rest is profit in the strict and proper sense. And in fact there may not be any profit in that sense. Now what is this profit in the narrow sense? Marx says it is the charge the owner of the means of production levies on workers for access to the means of production and thereby to the market. The marginalists may acknowledge that that is a component, but typically they regard that as a kind of rent, and in talking about profit emphasize something else, corresponding to Marx's notion of exceptional profits made by selling a product at a value set by the normal methods of production when your actual method of production requires less than the normal labor. In other words, profit in part is the return got by the innovator - the entrepreneur. And this 'entrepreneurial' profit may in fact be got by anyone who makes any sort of innovation, not only by an owner or employer, and not only by technological innovation. It might be a new method of selling, that increases the commission earned by a salesman. Now in comparison with Marx's theory there is here not only a difference in terminology, but also a conceptual difference. The modern economist groups things together differently: what Marx calls profit he puts together with other kinds of rent, and he calls 'profit', and gives great salience to, something that comes under a different part of Marx's theory, namely his account of the dynamics of capitalism.
Seventh, there is the problem of explaining the level of wages. The wage is the price of the right to use labour power, and this price in Marx's theory is in the long run a function of value, of the labour required to produce and maintain a labourer. Marx expects the long-term wage to be below value for various reasons, e.g. because the gale of innovation creates and recreates a reserve army of unemployed. But marginalists say there is no reason why there should be any particular relation between wages and the quantity of labour needed to produce the labour, unless it is assumed that labourers are produced like other commodities are, to make a profit: and then the relation will be explained best by the interaction of supply and demand as the marginalist theory explains it. It is to some extent true that labourers are produced as commodities for profit (supply of their training, recreation, and so on may be influenced by commercial considerations), but not true enough to make the labour theory of value realistic as an account of wages.
Further, the labour theory cannot explain why more talented workers get a higher wage. The additional value of a trained worker's labour power may be due to the extra labour required to produce it, i.e. in training; but talent is not deliberately produced in this way, as a commodity, yet fetches a higher return. This requires no special explanation in the supply-and-demand theory, and labour theorists can deal with it only, once again, by falling back on supply-and-demand.
The criticism of a theory by refuting some assumption (or set of assumptions) on which it rests should be tested by seeing if you can think of some way of restating the theory in which that assumption does not figure. If there is some set of premises which do not include this assumption, which holders of the theory might accept, from which its essential conclusions follow logically, then the theory is logically independent of those assumptions, and criticism of them leaves the theory intact.
Now the essential conclusions of Marx's theory seem to be those concerned with the dynamics of capitalist society, with how such a society is likely to develop over time. These are the predictions about the falling rate of profit, the centralization of capital, the immiseration of the working class, the likelihood of frequent and intense crises. It seems to me that Marx's argument to these conclusions could start at the point where he explains why capitalists introduce machinery and seek foreign markets. He explains that the capitalist who has a new method which requires much less labour than the normal method will introduce it because the exchange value of the commodity produced will still be set by the amount of labour necessary by the normal method. Why not just say that the new method will be introduced if it reduces costs of production and thereby increases profit - without using terminology or concepts peculiar to the labour theory?
But at this point modern economists will make an objection not especially connected with the marginalist theory, but relevant on either theory. That is that there is no reason to suppose that every innovation must increase the ratio of constant to variable capital, or for any other reason lead to a fall in the rate of profit. Some innovations (e.g. replacement of steam engines with electric power, and innovations in communications) make it possible to produce or do more with less investment in machinery. It seems impossible to predict any overall trend in one direction or another. We can't say in advance whether in future more, or less, capital will be tied up in the production of the commodities we know about, or in others yet to be invented. (What is invented depends not just on what ideas people happen to have, but also on what needs are perceived, which depends partly on perceived possibilities of making profit.)
If there is no trend to a higher ratio between constant and variable capital, then there is no reason even on Marx's theory to predict a falling rate of profit. The modern economist will add that in any case what he calls profit does not depend on the ratio of labour to other factors of production. There will be profit as long as there is the possibility of innovation, and as far as we know there is no fixed quantum of possible innovation. Profit in this sense will still do what profit does in Marx's system, namely motivate new investment; so the difference in meaning here makes no difference: assuming that there will always be possibilities of innovation there is no reason to expect the capitalist economy to stagnate.
Neither is there any reason to expect increasing centralization. This expectation rested on the assumption that innovation would make industry more and more capital intensive, and that the falling rate of profit would make larger and larger investment necessary to yield the same absolute return: small people would increasingly not be able to make a living as owners of businesses and employers. In every crisis some of them would be swallowed up by larger rivals. But if there is no trend to greater capital intensiveness, no trend to lower profit rates, there is no reason why smaller business should disappear. It may be a true generalization that in a given line of business small businesses amalgamate or are taken over, but there is not just a finite number of lines of business. Innovation may at any time open a new kind of possibility, which small capitalists may move into. (E.g. the beginnings of the computer industry: no chance these days of starting up your own coal mine, cotton mill, iron works, the growth industries of the early 19th century; but there are new possibilities not anticipated then.)
Sociologists Economists and Democracy (JC/423/.B2625)
Gide and Rist (HB/75/.G5/1948)
Barber (HB/75/.B147)
Routh (HB/75/.R673)
Schumpeter (HB/75/.R64)
Deane (HB/75/.D29)
Dobb (HB/75/.D446)
See lecture:
Historical Materialism
Return to Teaching
Materials on Political Thought
Return to Home Page