750 REPORT— 1903. 



TUESDAY, SEPTEMBER 15. 

 The following Papers were read : — 



1. Statistical Methods and the Fiscal Controversy.^ 

 By A. L. BowLEY, M.A. 



Much of the present confusion is due to erroneous uses of statistics in argu- 

 ment. The fault is as often in logical method as in the statistics themselves. 



i. Figures are taken for one year without reference to the series which precede 

 them. Every series has its own characteristics of ' trend ' and ' fluctuation ' ; the 

 effect of a particular event {e.g. the alteration of a tax) cannot he determined 

 without reference to these, and very often cannot he determined at all. 



ii. Figures relating to quantities, which are not similar, are added, and the 

 total used as if it were homogeneous ; e.g. the values of imports and exports are 

 added together. 



iii. The distinction between value and quantity in trade statistics is ignored ; 

 values are generally used, while quantity is ultimately the more important. Pro- 

 gress may often be shown only by increased quantity, while value is stationary. 



iv. The measurement of accuracy is generally ignored. Accurate and in- 

 accurate estimates are added or multiplied, and the sum or product treated as 

 representing an ascertained fact. No sum or product is known exactly, but only as 

 correct to 1, 10, or 50 per cent. Partisan writers by adding different estimates 

 which are mere guesses to quantities accurately known produce opposite results. 



V. In the same way a measurement of the change of a part is taken as an 

 adequate measurement of that of a whole. To test this it is necessary to deter- 

 mine a priori what quantity should be measured, and then (by the criteria of 

 accuracy) to determine what percentage error is involved by only measuring part. 

 Thus home produce and exports must not be separated when productive progress 

 is in question ; and productive progress is not a sufScient measure of general 

 prosperity. In default of complete measurement partial indexes are used, which 

 easily lend themselves to personal bias. A metliod of testing the growth of 

 national prosperity is suggested. 



vi. Statistics, even when accurate, are constantly made to support conclusions 

 which have no logical connection with them. 



This analysis suggests certain rules for criticism and subjects for statistical 

 inquiry. 



2. The Faihire of Free Traders to Attain their Ideals. 

 By W. Cunningham, D.D. 



This paper regards the fiscal question as primarily economic ; there is no reason 

 to believe that a change of policy would create new positive ties between the 

 diflerent parts of the Empire. The paper also assumes universal free trade as the 

 ultimate aim, but discusses why there has been so little success in realising it, and 

 how it may be attained. (1) It appears from the views of Jeft'erson and Hamilton 

 that there was every prospect that the United States would have developed as a 

 free trade country, and the policy of Pitt would have favoured this. But Fox 

 and the English shipowners forced the United States to become protectionists in 

 self-defence. (2) Oobden's anticipations that other countries would follow our 

 example and adopt free trade have been falsified, as each politically distinct 

 nation has preferred to develop an independent and alUround economic life, so far 

 as possible, and the colonies are inclined to pursue a similar course. 



There is a real economic danger to England as a manufacturing community, as 

 she is being increasingly cut off from the opportunity of purchasing raw materials 

 (e.^r. cotton) and food (e.(/. corn). An increased cotton-supply from our colonies 



' Published in the Ecoiwmic Journqli September 1903. 



