H.W. Arndt - The Importance of Money: Essays in Domestic Macroeconomics, 1949-1999

This essay formed part of a review article on The New Economics: Keyness Influence on Theory and Public Policy, edited by Seymour E. Harris (Knopf 1947). The essay drew on six chapters of the book to sum up the prolonged discussion of the theory of wages implicit in Keyness General Theory.

Originally published as Recent Discussion of Keyness Theory of Wages: A Review, in the Economic Record, December 1949.

In a symposium on Keynes Influence on Theory and Public Policy1 which was published in the United States in 1947 both to mark the occasion of the tenth anniversary of the General Theory and as a Memorial to Lord Keynes, one aspect of Keynesian theory receives, apparently by accident rather than design, an unexpectedly large amount of attention. Of the twenty-five or so new essays of substance contained in the volume, at least six deal wholly or in part with Keyness theory of wages.2 These six essays do not in themselves form a well-arranged symposium. But if one takes the trouble to piece the arguments together, one finds in them a useful starting point for a summing-up of the prolonged discussion which this part of the General Theory provoked.

The task which Keynes set himself in his treatment of wage theory in the General Theory was to refute the implication of classical theory that general wage deflation was an appropriate remedy for general unemployment. His special problem was to perform this task while retaining one of the basic assumptions of classical theory, the assumption that employment is inversely related to the level of real wages (from which it follows that unemployment is involuntary only if labour is willing to accept work at a lower real wage). His solution was expressed in two propositions. First, wage deflation is difficult because labour resists cuts in money wages even though it would be prepared to accept an equivalent reduction in real wages through rising prices (and would, therefore, qualify as involuntarily unemployed in the classical sense). Secondly, even if labour were willing to accept reductions in money wages, this would not increase employment since prices would fall proportionately, leaving real wages (and therefore output and employment) unchanged.

Both propositions and particularly the more far-reaching second one, were qualified by Keynes himself in the General Theory and have been subject to criticism ever since. Much of the discussion in these essays summarises and elaborates these qualifications and criticisms.

Keyness own qualifications of the second proposition related to four main points. (a) The argument applies only in a closed economy. In an open economy, a reduction in money wages will improve the balance of trade and thus increase employment, although these effects may be nullified by similar wage reductions (or tariffs) abroad or by changes in exchange rates. (b) A general fall in money wages and prices, by releasing cash balances from the transaction sphere, may reduce the rate of interest and thus stimulate investment. But this effect, on which Keynesone cannot help feeling, with his tongue in his cheekplaced chief emphasis, is likely to be nullified either by the liquidity trap (at low interest rates) or by the tendency for investment to be interest-inelastic. At best, monetary management by the Trade Unions would be a singularly clumsy solution. (c) A reduction in money wage rates may have a favourable effect on business confidence. But if the general thesis is correct, any additional investment stimulated by the favourable impression made by cost reductions is bound to lead to losses, even if the wage cut is expected to be once-for-all and permanent. If, as is more likely because of resistance of labour, it is merely one stage in the process of a sagging wage level the effect will certainly be unfavourable. (d) Lastly, Keynes considered possible effects on the propensity to consume, through redistribution of income from wage-earners to other prime factors or from entrepreneurs to rentiers. The second of these Keynes thought doubtful and unimportant, while the effects of the first would be unfavourable.

At least four of the six contributors to the symposium on Keyness wage theory seem prepared to follow Keynes up to this point. The chief exception is Professor Haberler who is his generally critical essay makes a root-and-branch attack on Keyness position.3 Professor Haberler is primarily concerned to demonstrate that Keyness theory of under-employment equilibrium depends on his postulate of (downward) rigidity of money wages. He has little difficulty in showing that a completely unstable Wicksellian system could hardly be said to be in underemployment equilibrium; though, as Professor Hansen points out, that is hardly relevant to Keyness main problem. Professor Haberler also advances two arguments against the thesis that employment is independent of the level of money wages. The first, to the effect that a continuous rise in the value of money would increase the real value of cash balances held in the form of gold and may thus raise the consumption function, is not taken very seriously by himself. The second would seem to be a mere reassertion of the classical position against Keynes. No Keynesian denies that a reduction in the cost of certain consumption or investment goods may well stimulate demand for them; but when Professor Haberler adds and for consumption and investment as a whole he surely begs the central question. The argument which he advances in support of this proposition is little more convincing. Assume that the elasticity of demand for some of these things, and therefore incidentally for labour, is unity. Then the wage bill remains unchanged and there are no adverse effects through a fall of consumption demand of the workers. Then employment will clearly rise (p. 172). This assumes that the additional output (investment) will generate equivalent additional income before cost reduction has any effect on consumption; it is merely one hopeful version of the argument that cost reduction will increase business confidence, the marginal efficiency of capital and investment. Professor Haberlers further point that the additional output must be interpreted as due to a rise in the marginal efficiency of capital or the consumption function, according as to whether the newly produced goods are consumption or investment goods, is even more difficult to accept. Professor Haberlers oblique claim, later in the essay (p. 175), to have shown that Keynes was unable to reconcile a competitive system (flexible money wages) with the existence of unemployment can hardly be substantiated.