The General Theory of Employment, Interest, and Money

mutatis mutandis
to the case of a boom. 
It follows that with the actual practices and institutions of the contemporary world it is more 
expedient to aim at a rigid money-wage policy than at a flexible policy responding by easy stages to 
changes in the amount of unemployment;—so far, that is to say, as the marginal efficiency of 
capital is concerned. But is this conclusion upset when we turn to the rate of interest? 

---

132
It is, therefore, on the effect of a falling wage- and price-level on the demand for money that those 
who believe in the self-adjusting quality of the economic system must rest the weight of their 
argument; though I am not aware that they have done so. If the quantity of money is itself a function 
of the wage- and price-level, there is indeed, nothing to hope in this direction. But if the quantity of 
money is virtually fixed, it is evident that its quantity in terms of wage-units can be indefinitely 
increased by a sufficient reduction in money-wages; and that its quantity in proportion to incomes 
generally can be largely increased, the limit to this increase depending on the proportion of wage-
cost to marginal prime cost and on the response of other elements of marginal prime cost to the 
falling wage-unit. 
We can, therefore, theoretically at least, produce precisely the same effects on the rate of interest by 
reducing wages, whilst leaving the quantity of money unchanged, that we can produce by increasing 
the quantity of money whilst leaving the level of wages unchanged. It follows that wage reductions, 
as a method of securing full employment, are also subject to the same limitations as the method of 
increasing the quantity of money. The same reasons as those mentioned above, which limit the 
efficacy of increases in the quantity of money as a means of increasing investment to the optimum 
figure, apply 
mutatis mutandis
to wage reductions. Just as a moderate increase in the quantity of 
money may exert an inadequate influence over the long-term rate of interest, whilst an immoderate 
increase may offset its other advantages by its disturbing effect on confidence; so a moderate 
reduction in money-wages may prove inadequate, whilst an immoderate reduction might shatter 
confidence even if it were practicable. 
There is, therefore, no ground for the belief that a flexible wage policy is capable of maintaining a 
state of continuous full employment;—any more than for the belief that an open-market monetary 
policy is capable, unaided, of achieving this result. The economic system cannot be made self-
adjusting along these lines. 
If, indeed, labour were always in a position to take action (and were to do so), whenever there was 
less than full employment, to reduce its money demands by concerted action to whatever point was 
required to make money so abundant relatively to the wage-unit that the rate of interest would fall 
to a level compatible with full employment, we should, in effect, have monetary management by the 
trade unions, aimed at full employment, instead of by the banking system. 
Nevertheless while a flexible wage policy and a flexible money policy come, analytically, to the 
same thing, inasmuch as they are alternative means of changing the quantity of money in terms of 
wage-units, in other respects there is, of course, a world of difference between them. Let me briefly 
recall to the reader's mind the four outstanding considerations. 
(i) Except in a socialised community where wage-policy is settled by decree, there is no means of 
securing uniform wage reductions for every class of labour. The result can only be brought about by 
a series of gradual, irregular changes, justifiable on no criterion of social justice or economic 
expedience, and probably completed only after wasteful and disastrous struggles, where those in the 
weakest bargaining position will suffer relatively to the rest. A change in the quantity of money, on 
the other hand, is already within the power of most governments by open-market policy or 
analogous measures. Having regard to human nature and our institutions, it can only be a foolish 
person who would prefer a flexible wage policy to a flexible money policy, unless he can point to 
advantages from the former which are not obtainable from the latter. Moreover, other things being 

---

133
equal, a method which it is comparatively easy to apply should be deemed preferable to a method 
which is probably so difficult as to be impracticable. 
(ii) If money-wages are inflexible, such changes in prices as occur (i.e. apart from 'administered' or 
monopoly prices which are determined by other considerations besides marginal cost) will mainly 
correspond to the diminishing marginal productivity of the existing equipment as the output from it 
is increased. Thus the greatest practicable fairness will be maintained between labour and the 
factors whose remuneration is contractually fixed in terms of money, in particular the rentier class 
and persons with fixed salaries on the permanent establishment of a firm, an institution or the State. 
If important classes are to have their remuneration fixed in terms of money in any case, social 
justice and social expediency are best served if the remunerations of 
all
factors are somewhat 
inflexible in terms of money. Having regard to the large groups of incomes which are comparatively 
inflexible in terms of money, it can only be an unjust person who would prefer a flexible wage 
policy to a flexible money policy, unless he can point to advantages from the former which are not 
obtainable from the latter. 
(iii) The method of increasing the quantity of money in terms of wage-units by decreasing the 
wage-unit increases proportionately the burden of debt; whereas the method of producing the same 
result by increasing the quantity of money whilst leaving the wage-unit unchanged has the opposite 
effect. Having regard to the excessive burden of many types of debt, it can only be an inexperienced 
person who would prefer the former. 
(iv) If a sagging rate of interest has to be brought about by a sagging wage-level, there is, for the 
reasons given above, a double drag on the marginal efficiency of capital and a double reason for 
putting off investment and thus postponing recovery. 
III 
It follows, therefore, that if labour were to respond to conditions of gradually diminishing 
employment by offering its services at a gradually diminishing money-wage, this would not, as a 
rule, have the effect of reducing real wages and might even have the effect of increasing them, 
through its adverse influence on the volume of output. The chief result of this policy would be to 
cause a great instability of prices, so violent perhaps as to make business calculations futile in an 
economic society functioning after the manner of that in which we live. To suppose that a flexible 
wage policy is a right and proper adjunct of a system which on the whole is one of