The General Theory of Employment, Interest, and Money

Chapter 21 
THE THEORY OF PRICES 
I 
So long as economists are concerned with what is called the theory of value, they have been 
accustomed to teach that prices are governed by the conditions of supply and demand; and, in 
particular, changes in marginal cost and the elasticity of short-period supply have played a 
prominent part. But when they pass in volume II, or more often in a separate treatise, to the theory 
of money and prices, we hear no more of these homely but intelligible concepts and move into a 
world where prices are governed by the quantity of money, by its income-velocity, by the velocity 
of circulation relatively to the volume of transactions, by hoarding, by forced saving, by inflation 
and deflation 
et hoc genus omne
; and little or no attempt is made to relate these vaguer phrases to 
our former notions of the elasticities of supply and demand. If we reflect on what we are being 
taught and try to rationalise it, in the simpler discussions it seems that the elasticity of supply must 
have become zero and demand proportional to the quantity of money; whilst in the more 
sophisticated we are lost in a haze where nothing is clear and everything is possible. We have all of 
us become used to finding ourselves sometimes on the one side of the moon and sometimes on the 
other, without knowing what route or journey connects them, related, apparently, after the fashion 
of our waking and our dreaming lives. 
One of the objects of the foregoing chapters has been to escape from this double life and to bring 
the theory of prices as a whole back to close contact with the theory of value. The division of 
economics between the theory of value and distribution on the one hand and the theory of money on 
the other hand is, I think, a false division. The right dichotomy is, I suggest, between the theory of 
the individual industry or firm and of the rewards and the distribution between different uses of a 
given
quantity of resources on the one hand, and the theory of output and employment 
as a whole
on the other hand. So long as we limit ourselves to the study of the individual industry or firm on 
the assumption that the aggregate quantity of employed resources is constant, and, provisionally, 
that the conditions of other industries or firms are unchanged, it is true that we are not concerned 
with the significant characteristics of money. But as soon as we pass to the problem of what 
determines output and employment as a whole, we require the complete theory of a monetary 
economy. 
Or, perhaps, we might make our line of division between the theory of stationary equilibrium and 
the theory of shifting equilibrium—meaning by the latter the theory of a system in which changing 
views about the future are capable of influencing the present situation. 
For the importance of money 
essentially flows from its being a link between the present and the future
. We can consider what 
distribution of resources between different uses will be consistent with equilibrium under the 
influence of normal economic motives in a world in which our views concerning the future are 
fixed and reliable in all respects;—with a further division, perhaps, between an economy which is 
unchanging and one subject to change, but where all things are foreseen from the beginning. Or we 
can pass from this simplified propaedeutic to the problems of the real world in which our previous 

---

147
expectations are liable to disappointment and expectations concerning the future affect what we do 
to-day. It is when we have made this transition that the peculiar properties of money as a link 
between the present and the future must enter into our calculations. But, although the theory of 
shifting equilibrium must necessarily be pursued in terms of a monetary economy, it remains a 
theory of value and distribution and not a separate 'theory of money'. Money in its significant 
attributes is, above all, a subtle device for linking the present to the future; and we cannot even 
begin to discuss the effect of changing expectations on current activities except in monetary terms. 
We cannot get rid of money even by abolishing gold and silver and legal tender instruments. So 
long as there exists any durable asset, it is capable of possessing monetary attributes and, therefore, 
of giving rise to the characteristic problems of a monetary economy.
II 
In a single industry its particular price-level depends partly on the rate of remuneration of the 
factors of production which enter into its marginal cost, and partly on the scale of output. There is 
no reason to modify this conclusion when we pass to industry as a whole. The general price-level 
depends partly on the rate of remuneration of the factors of production which enter into marginal 
cost and partly on the scale of output as a whole, i.e. (taking equipment and technique as given) on 
the volume of employment. It is true that, when we pass to output as a whole, the costs of 
production in any industry partly depend on the output of other industries. But the more significant 
change, of which we have to take account, is the effect of changes in 
demand
both on costs and on 
volume. It is on the side of demand that we have to introduce quite new ideas when we are dealing 
with demand as a whole and no longer with the demand for a single product taken in isolation, with 
demand as a whole assumed to be unchanged.
III 
If we allow ourselves the simplification of assuming that the rates of remuneration of the different 
factors of production which enter into marginal cost all change in the same proportion, i.e. in the 
same proportion as the wage-unit, it follows that the general price-level (taking equipment and 
technique as given) depends partly on the wage-unit and partly on the volume of employment. 
Hence the effect of changes in the quantity of money on the price-level can be considered as being 
compounded of the effect on the wage-unit and the effect on employment. 
To elucidate the ideas involved, let us simplify our assumptions still further, and assume (1) that all 
unemployed resources are homogeneous and interchangeable in their efficiency to produce what is 
wanted, and (2) that the factors of production entering into marginal cost are content with the same 
money-wage so long as there is a surplus of them unemployed. In this case we have constant returns 
and a rigid wage-unit, so long as there is any unemployment. It follows that an increase in the 
quantity of money will have no effect whatever on prices, so long as there is any unemployment, 
and that employment will increase in exact proportion to any increase in effective demand brought 
about by the increase in the quantity of money; whilst as soon as full employment is reached, it will 
thenceforward be the wage-unit and prices which will increase in exact proportion to the increase in 
effective demand. Thus if there is perfectly elastic supply so long as there is unemployment, and 
perfectly inelastic supply so soon as full employment is reached, and if effective demand changes in 
the same proportion as the quantity of money, the quantity theory of money can be enunciated as 
follows: 'So long as there is unemployment,