=================== HES POSTING ================== Surplus and classical distribution theory Anthony Brewer Department of Economics University of Bristol It is often said that the classical economists had a 'surplus' theory of distribution. I will argue that although there is some truth to this claim, it is one that must be rather carefully qualified. I have become increasingly convinced that the vague and imprecise way the word 'surplus' is used is a hindrance to clear thinking. What is a surplus theory? If x + y = z, and if x and z are fixed independently of each other and of y, then one can say that y is the surplus of z over x. For example, if my brother and I get a cake and if he first takes what he wants, regardless of the size of the cake, then my share is a residual or surplus. What he takes might depend, say, on how he feels, provided it does not depend on the total to be divided. But if we split the cake equally, or by any other rule that makes his share depend on mine or on the size of the cake, then my share is not a pure surplus. In the surplus interpretation, classical distribution theory looks like this: (a) wages are fixed at a given level, (b) wage plus non-wage income equals (the value of) total output, so (c) non-wage incomes are determined by the surplus of output over wages. Valuation of output poses well-known problems, but my concern here is with the problems which would remain even if output were homogeneous. Note in particular that proposition (a) is the key to the surplus theory, since (b) is really quite trivial. A surplus theory with any substance to it must do more than assert proposition (a) - it must offer some coherent mechanism to bring it about. The classical wage theory of wages was directly based on population theory - (a) if wages are above subsistence, population and hence labour supply grows, and (b) wages are governed by demand and supply, in that rising labour supply, ceteris paribus, lowers wages (e.g. Ricardo, p. 94). In a static model it would follow that wages tend to subsistence. There are other ways in which wages could be fixed prior to everything else, but they cannot plausibly be attributed to the classics. Wages could be fixed by law or binding custom, but the classics did not think so since they discussed variations in wages in response to changing economic conditions. Employers could conspire to fix wages at subsistence. Smith did mention conspiracy, but his texts as a whole cannot plausibly be interpreted in those terms. Employers might voluntarily pay subsistence (but no more) because workers cannot work efficiently if they are paid less. This would be a sensible story, but I cannot find it in the the classics. Bargaining theories of wages are sometimes mentioned in this context, but any sensible bargaining theory makes the outcome depend on the total to be divided. Smith did discuss bargaining power, implying that employers could force wages down to subsistence, but immediately added that it did not apply to a growing economy. What of the claim that wages are determined by social or historical factors, over and above physical subsistence? A legal or customary wage could include elements other than physical subsistence, but that is not relevant to classical economics. If there is a social element in classical wage theory it must be built into the population theory. If people will starve (or allow their children to starve) rather than be 'indecently' dressed, then clothes to a socially determined standard of decency are necessary. If, as Cantillon suggested, people will not marry unless they can be sure of an adequate standard of life for the resulting children, then the equilibrium wage will adjust accordingly. It is not enough to utter the words 'historical and moral element' like a mantra. The mechanism has to be explained, and demonstrated in the texts. Did the classics have a surplus theory as defined? For reasons of space, I shall consider only Smith and Ricardo. First Smith. His main theory of wages is very clear from the bulk of the Wealth of Nations, despite the rather confusing alternatives considered at the start of the chapter on wages. He was concerned with growing economies, where wages must be above subsistence to allow the labour force to grow in line with the demand for labour, so wages depend on the growth rate (Smith I.viii.16-27). That could still be counted as a surplus theory in the sense defined above if growth, and hence wages, were independent of output and profits, but they are not. In colonies, for example, high per-capita output on the best land and low rents (or land prices) lead to rapid accumulation and high wages. In countries which have reached their 'full complement of riches', investment opportunities are low, profits are low, and so are wages. Smith had not worked out all the details of these interactions, but it is clear that he had advanced substantially beyond a crude surplus theory. Labour is not homogeneous. The structure of wages clearly affects profits and other variables. To assert, as some do, that the structure of relative wages can simply be treated as given prior to everything else is an evasion, not a solution to the problem. Smith at least tackled the question seriously. Could it be said that although Smith did not hold a subsistence wage theory there is an asymmetry between his treatment of wages and profits which justifies treating his theory as in some way surplus-like? I have some sympathy with this view, but it is difficult to sustain. Labour supply depends on population (inter alia), which grows because of natural human desires. But the accumulation of stock (capital) also depends on a natural human desire, the desire to better oneself. Population growth is checked when the wage falls to subsistence, and accumulation is also checked when profits fall to a minimum set by the risk of loss. This does not amount to perfect symmetry between wages and profits (why should it?), but it is hard to see either as prior to the other. Ricardo, by contrast, did state a crude surplus theory, with wages reduced to subsistence by the demographic mechanism and non-wage incomes as the residual. The problem is that he also understood and accepted Smith's more sophisticated wage theory. The crude subsistence wage theory, with or without social elements in subsistence, is valid in a non-stationary state if and only if demographic changes work through much faster (strictly, infinitely faster) than changes in the demand for labour caused by capital accumulation. It would be absurd to believe that they do (as Marx, for example, pointed out). The simple subsistence wage, and hence the simple surplus element in Ricardo's theory, is its weak point, not its strength. One can reasonably debate whether he thought of it simply as a heuristic simplification or whether there is a real inconsistency (c.f. Hollander, 195-6, Peach, 130-1). In any case, to attribute an unqualified surplus theory of distribution to either Smith or Ricardo is to undervalue and oversimplify their writings. Finally, a very brief word on Marx (who I do not count as a classical economist). He explicitly rejected the Malthusian theory of wages, and could not therefore take over the classical version of surplus theory as it stood. His main statement of the theory of surplus value in Vol. I of Capital depends on a definition of the 'value of labour power' in terms of subsistence requirements (with a 'historical and moral element'). It offers no reason why wages should be at this level and therefore has no substantive content. However, Chapter 25 contains a different story, in which accumulation raises the demand for labour power and drives up wages but rising wages in turn check accumulation, while labour-saving technical change tends to bring wages down again. This incomplete sketch of dynamic interaction between wages, profits, and growth clearly does not meet the conditions for a surplus theory as defined above. I conclude: (1) the substance of a surplus theory of distribution has to derive from the wage theory on which it is based. (2) To the extent that there was a classical surplus theory, it was based on 'Malthusian' population theory and stands or falls with it. Any claim that such a theory can be used in modern conditions is clearly untenable. (3) Even a Malthusian population theory only yields a pure surplus theory of distribution in a static economy. As Smith and Ricardo both knew, in a growing economy distribution and growth interact to produce wages above subsistence, invalidating a simple surplus theory. Hollander, S. Classical Economics, 1987, Blackwell Peach, T. Interpreting Ricardo, 1993, CUP Ricardo, D. Principles (Works, vol I), ed P. Sraffa, 1951, CUP Smith, A. Wealth of Nations (Glasgow Edition), 1976, CUP. ============ FOOTER TO HES POSTING ============ For information, send the message "info HES" to [log in to unmask]