F.— ECONOMIC SCIENCE AND STATISTICS 155 



ing assumptions ; cf. the frictionless surface, etc., in which rates of increase 

 will themselves figure as unknown terms. 



One reason for holding development along these lines to be needed 

 is the unsatisfactory condition of the theory of interest in static economics. 

 I refer now not to the results reached by Mr. Keynes in his important 

 study of the dual nature of capital supply (waiting and liquidity-sacrifice), 

 but to a still more fundamental difficulty, i"- Using the assumptions 

 required for static price determination, namely persistence of tastes, 

 technology and supply of factors unchanged, the demand for new saving 

 at any given rate of interest is zero, since so long as the fundamental 

 conditions and the equilibrium are maintained, the volume and method 

 of production must be unchanged. To put the same thing in other words, 

 the static equations determine the price of capital and the quantity of it 

 which will be used. It is the quantity of capital in use which, along with 

 the quantity of land and labour in use, remains unchanged throughout 

 the maintenance of a given equilibrium. But if the quantity of capital in 

 use is the same the rate of saving is zero. I have the impression that 

 writers, other than the most careful, tend to get one dimension wrong at 

 this point, and suppose that the ' laws of supply and demand ' (static 

 theory) may determine not the quantity of capital but the amount of 

 saving, i.e. rate of increase in the quantity of capital at a given level.'^^ 



That it is possible to reach interesting conclusions on the basis of the 

 static assumption of no saving may be seen from Mrs. Robinson's article 

 on the ' Long Period Theory of Employment.' The paradoxical air of 

 that essay may well be due precisely to her strict adherence to the static 

 assumption. The fact that she quite properly compels us to consider the 

 true effect of any change in the light of its consequences in the state of 

 equilibrium only reached when all saving has fallen to zero, suggests 

 that it would be expedient to tackle the problem more directly. In 

 place of a succession of static equilibria we need the concept of motion 

 under the influence of steadily operating forces. 



" I regret that it is not possible within the scope of this paper to consider, 

 from a methodological point of view, the great contributions to thought recently- 

 made by Mr. Keynes. My division into sections was necessarily guided by 

 reference to economics as a whole, and his contribution, although internally 

 highly coherent and constituting a unified structure, belongs in part to all my 

 divisions, so that a full discussion would not be wholly relevant to and would 

 unduly swell any one. See Econometrica, January 1937 ; R. F. Harrod, 

 Mr. Keynes and Traditional Theory. 



1* We might imagine a static state as follows. People would save out of earned 

 income in their early years and invest in life annuities such sums as would make 

 their income rise at a rate which would make its marginal utility fall at a rate 

 equal to the rate of interest. Meanwhile the rate of interest would be fixed at a 

 critical level, sufficient to make them hand on their inherited capital intact, 

 despite their inferior regard for their heirs. These conditions would, on the 

 assumption of a stationary age distribution, make saving equal to zero. If their 

 regard for their heirs happened to be as great as their regard for themselves, then, 

 with a positive rate of interest and supposing the state of bliss described by 

 Ramsey in his well-known article not to be reached, there would be some positive 

 saving, and the assumptions of the static theory would be mutually inconsistent. 

 Similarly a socialist state in conditions otherwise static should arrange for 

 positive saving. 



