156 SECTIONAL ADDRESSES 



The laws will govern the relation between and determine the mutual 

 consistency of the rates of increase of various magnitudes, e.g. working 

 population, technical powers, quantity of capital, of circulating medium, 

 etc. Some empirical foundation is necessary. Bare study of mutual 

 implications will not yield much, since there is an infinite variety of pos- 

 sibilities. But I have the impression that a few basic empirical laws, of a 

 generality not much inferior to that of the law of demand in statics, may 

 yield, in connection with the study of mutual implications, an elaborate 

 structure of deductive theory. 



An example of a basic empirical generalisation may be found in the 

 proposition put forward by Mr. Keynes in his recent work, that at a given 

 rate of interest people will save a larger absolute amount from a larger 

 income. We could get still further if we could establish, but this is 

 perhaps too audacious for the early stages, that people save a larger pro- 

 portion of a larger income. Both these propositions are clearly open to 

 empirical verification. They will be subject to ceteris paribus clauses 

 regarding the distribution of income and institutional arrangements, but 

 these would probably not impair their high scientific utility. The 

 statistical work of verification required is no doubt substantial, but light 

 compared with that required to fill in the blank forms of the static theory 

 equations. The phenomena are much more amenable to the attainment 

 of reliable results in this field than in that of static supply and demand 

 schedules. The de facto growth of society assists the former while it 

 hinders the latter type of statistical enquiry. 



May I be excused for touching on a theory in which I believe, subject 

 of course to the eroding researches of historians of thought, that I have 

 certain proprietary rights ? If it is true that the most important factor 

 governing the demand for new capital is the rate of growth of the system, 

 and the most important factor governing its supply is the absolute size 

 of the system, then, having regard to the truism that demand must be 

 equal to the supply, a host of interesting conclusions should follow. 

 Premises containing these peculiar mathematical relations should surely 

 be a gift, precious beyond compare, to economists of mathematical bent 

 seeking new conclusions. I risk saying that if, when trade cycle theory 

 comes to be established on firm and agreed foundations, these relations 

 are not judged to have central causal significance, I shall be dumbfounded. 



IV. 



Empirical Studies. 



I now come to the most difficult, the most tentative, and withal the most 

 important section, the search for causal laws outside the realm of deduc- 

 tions from the law of demand or the simple laws of growth. 



Having previously tended to belittle the causal significance of the theory 

 of value and distribution, I should like to pay tribute to the high import- 

 ance of the work of classification, not achieved without much toil and the 

 insight of genius, which is the groundwork of that theory as well as of the 

 analytical map. This is likely to prove a valuable and indeed indis-. 



