The General Theory of Employment, Interest, and Money

cet. par.
, by a rise in its own 
wages, there is naturally for all groups a pressure in this direction, which entrepreneurs will be more 
ready to meet when they are doing better business. For this reason a proportion of any increase in 
effective demand is likely to be absorbed in satisfying the upward tendency of the wage-unit. 
Thus, in addition to the final critical point of full employment at which money-wages have to rise, 
in response to an increasing effective demand in terms of money, fully in proportion to the rise in 
the prices of wage-goods, we have a succession of earlier semi-critical points at which an increasing 
effective demand tends to raise money-wages though not fully in proportion to the rise in the price 
of wage-goods; and similarly in the case of a decreasing effective demand. In actual experience the 
wage-unit does not change continuously in terms of money in response to every small change in 
effective demand; but discontinuously. These points of discontinuity are determined by the 
psychology of the workers and by the policies of employers and trade unions. In an open system, 
where they mean a change relatively to wage-costs elsewhere, and in a trade cycle, where even in a 
closed system they may mean a change relatively to expected wage-costs in the future, they can be 
of considerable practical significance. These points, where a further increase in effective demand in 

---

151
terms of money is liable to cause a discontinuous rise in the wage-unit, might be deemed, from a 
certain point of view, to be positions of semi-inflation, having some analogy (though a very 
imperfect one) to the absolute inflation which ensues on an increase in effective demand in 
circumstances of full employment. They have, moreover, a good deal of historical importance. But 
they do not readily lend themselves to theoretical generalisations. 
(5) Our first simplification consisted in assuming that the remunerations of the various factors 
entering into marginal cost all change in the same proportion. But in fact the rates of remuneration 
of different factors in terms of money will show varying degrees of rigidity and they may also have 
different elasticities of supply in response to changes in the money-rewards offered. If it were not 
for this, we could say that the price-level is compounded of two factors, the wage-unit and the 
quantity of employment. 
Perhaps the most important element in marginal cost which is likely to change in a different 
proportion from the wage-unit, and also to fluctuate within much wider limits, is marginal user cost. 
For marginal user cost may increase sharply when employment begins to improve, if (as will 
probably be the case) the increasing effective demand brings a rapid change in the prevailing 
expectation as to the date when the replacement of equipment will be necessary. 
Whilst it is for many purposes a very useful first approximation to assume that the rewards of all the 
factors entering into marginal prime-cost change in the same proportion as the wage-unit, it might 
be better, perhaps, to take a weighted average of the rewards of the factors entering into marginal 
prime-cost, and call this the 
cost-unit
. The cost-unit, or, subject to the above approximation, the 
wage-unit, can thus be regarded as the essential standard of value; and the price-level, given the 
state of technique and equipment, will depend partly on the cost-unit, and partly on the scale of 
output, increasing, where output increases, 
more
than in proportion to any increase in the cost-unit, 
in accordance with the principle of diminishing returns in the short period. We have full 
employment when output has risen to a level at which the marginal return from a representative unit 
of the factors of production has fallen to the minimum figure at which a quantity of the factors 
sufficient to produce this output is available.
V 
When a further increase in the quantity of effective demand produces no further increase in output 
and entirely spends itself on an increase in the cost-unit fully proportionate to the increase in 
effective demand, we have reached a condition which might be appropriately designated as one of 
true inflation. Up to this point the effect of monetary expansion is entirely a question of degree, and 
there is no previous point at which we can draw a definite line and declare that conditions of 
inflation have set in. Every previous increase in the quantity of money is likely, in so far as it 
increases effective demand, to spend itself partly in increasing the cost-unit and partly in increasing 
output. 
It appears, therefore, that we have a sort of asymmetry on the two sides of the critical level above 
which true inflation sets in. For a contraction of effective demand below the critical level will 
reduce its amount measured in cost-units; whereas an expansion of effective demand beyond this 
level will not, in general, have the effect of increasing its amount in terms of cost-units. This result 
follows from the assumption that the factors of production, and in particular the workers, are 

---

152
disposed to resist a reduction in their money-rewards, and that there is no corresponding motive to 
resist an increase. This assumption is, however, obviously well founded in the facts, due to the 
circumstance that a change, which is not an all-round change, is beneficial to the special factors 
affected when it is upward and harmful when it is downward. 
If, on the contrary, money-wages were to fall without limit whenever there was a tendency for less 
than full employment, the asymmetry would, indeed, disappear. But in that case there would be no 
resting-place below full employment until either the rate of interest was incapable of falling further 
or wages were zero. In fact we must have 
some
factor, the value of which in terms of money is, if 
not fixed, at least sticky, to give us any stability of values in a monetary system. 
The view that