The General Theory of Employment, Interest, and Money

e
d
, 
e
w
, 
e
e
and 
e
o
upon which the effect on prices of changes in the 
quantity of money depends, 
e
d
stands for the liquidity factors which determine the demand for 
money in each situation, 
e
w
for the labour factors (or, more strictly, the factors entering into prime-
cost) which determine the extent to which money-wages are raised as employment increases, and 
e
e
and 
e
o
for the physical factors which determine the rate of decreasing returns as more employment 
is applied to the existing equipment. 
If the public hold a constant proportion of their income in money, 
e
d
= 1; if money-wages are 
fixed, 
e
w
= 0; if there are constant returns throughout so that marginal return equals average return, 
e
e
e
o
= 1; and if there is full employment either of labour or of equipment, 
e
e
e
o
= 0. 
Now 
e
= 1, if 
e
d
= 1, and 
e
w
= 1; or if 
e
d
= 1, 
e
w
= 0 and 
e
e
× 
e
o
= 0; or if 
e
d
= 1 and 
e
o
= 0. 
And obviously there is a variety of other special eases in which 
e
= 1. But in general 
e
is not unity; 
and it is, perhaps, safe to make the generalisation that on plausible assumptions relating to the real 
world, and excluding the case of a 'flight from the currency' in which 
e
d
and 
e
w
become large, 
e
is, 
as a rule, less than unity.
VII 
So far, we have been primarily concerned with the way in which changes in the quantity of money 
affect prices in the short period. But in the long run is there not some simpler relationship? 
This is a question for historical generalisation rather than for pure theory. If there is some tendency 
to a measure of long-run uniformity in the state of liquidity-preference, there may well be some sort 
of rough relationship between the national income and the quantity of money required to satisfy 
liquidity-preference, taken as a mean over periods of pessimism and optimism together. There may 
be, for example, some fairly stable proportion of the national income more than which people will 
not readily keep in the shape of idle balances for long periods together, provided the rate of interest 
exceeds a certain psychological minimum; so that if the quantity of money beyond what is required 
in the active circulation is in excess of this proportion of the national income, there will be a 
tendency sooner or later for the rate of interest to fall to the neighbourhood of this minimum. The 
falling rate of interest will then, 
cet. par
., increase effective demand, and the increasing effective 
demand will reach one or more of the semi-critical points at which the wage-unit will tend to show 
a discontinuous rise, with a corresponding effect on prices. The opposite tendencies will set in if the 
quantity of surplus money is an abnormally low proportion of the national income. Thus the net 
effect of fluctuations over a period of time will be to establish a mean figure in conformity with the 

---

154
stable proportion between the national income and the quantity of money to which the psychology 
of the public tends sooner or later to revert. 
These tendencies will probably work with less friction in the upward than in the downward 
direction. But if the quantity of money remains very deficient for a long time, the escape will be 
normally found in changing the monetary standard or the monetary system so as to raise the 
quantity of money, rather than in forcing down the wage-unit and thereby increasing the burden of 
debt. Thus the very long-run course of prices has almost always been upward. For when money is 
relatively abundant, the wage-unit rises; and when money is relatively scarce, some means is found 
to increase the effective quantity of money. 
During the nineteenth century, the growth of population and of invention, the opening-up of new 
lands, the state of confidence and the frequency of war over the average of (say) each decade seem 
to have been sufficient, taken in conjunction with the propensity to consume, to establish a schedule 
of the marginal efficiency of capital which allowed a reasonably satisfactory average level of 
employment to be compatible with a rate of interest high enough to be psychologically acceptable 
to wealth-owners. There is evidence that for a period of almost one hundred and fifty years the 
long-run typical rate of interest in the leading financial centres was about 5 per cent, and the gilt-
edged rate between 3 and 3½ per cent; and that these rates of interest were modest enough to 
encourage a rate of investment consistent with an average of employment which was not intolerably 
low. Sometimes the wage-unit, but more often the monetary standard or the monetary system (in 
particular through the development of bank-money), would be adjusted so as to ensure that the 
quantity of money in terms of wage-units was sufficient to satisfy normal liquidity-preference at 
rates of interest which were seldom much below the standard rates indicated above. The tendency of 
the wage-unit was, as usual, steadily upwards on the whole, but the efficiency of labour was also 
increasing. Thus the balance of forces was such as to allow a fair measure of stability of prices;—
the highest quinquennial average for Sauerbeck's index number between 1820 and 1914 was only 
50 per cent above the lowest. This was not accidental. It is rightly described as due to a balance of 
forces in an age when individual groups of employers were strong enough to prevent the wage-unit 
from rising much faster than the efficiency of production, and when monetary systems were at the 
same time sufficiently fluid and sufficiently conservative to provide an average supply of money in 
terms of wage-units which allowed to prevail the lowest average rate of interest readily acceptable 
by wealth-owners under the influence of their liquidity-preferences. The average level of 
employment was, of course, substantially below full employment, but not so intolerably below it as 
to provoke revolutionary changes. 
To-day and presumably for the future the schedule of the marginal efficiency of capital is, for a 
variety of reasons, much lower than it was in the nineteenth century. The acuteness and the 
peculiarity of our contemporary problem arises, therefore, out of the possibility that the average rate 
of interest which will allow a reasonable average level of employment is one so unacceptable to 
wealth-owners that it cannot be readily established merely by manipulating the quantity of money. 
So long as a tolerable level of employment could be attained on the average of one or two or three 
decades merely by assuring an adequate supply of money in terms of wage-units, even the 
nineteenth century could find a way. If this was our only problem now—if a sufficient degree of 
devaluation is all we need—we, to-day, would certainly find a way. 

---

155
But the most stable, and the least easily shifted, element in our contemporary economy has been 
hitherto, and may prove to be in future, the minimum rate of interest acceptable to the generality of 
wealth-owners. If a tolerable level of employment requires a rate of interest much below the 
average rates which ruled in the nineteenth century, it is most doubtful whether it can be achieved 
merely by manipulating the quantity of money. From the percentage gain, which the schedule of 
marginal efficiency of capital allows the borrower to expect to earn, there has to be deducted (1) the 
cost of bringing borrowers and lenders together, (2) income and sur-taxes and (3) the allowance 
which the lender requires to cover his risk and uncertainty, before we arrive at the net yield 
available to tempt the wealth-owner to sacrifice his liquidity. If, in conditions of tolerable average 
employment, this net yield turns out to be infinitesimal, time-honoured methods may prove 
unavailing. 
To return to our immediate subject, the long-run relationship between the national income and the 
quantity of money will depend on liquidity-preferences. And the long-run stability or instability of 
prices will depend on the strength of the upward trend ofthe wage-unit (or, more precisely, of the 
cost-unit) compared with the rate of increase in the efficiency of the productive system.