670 ■ TRANSACTIONS OF SECTION F. 



repeated periodically so as to admit of historical comparisons must from the 

 nature of the case 'be based on what Sir Eobert Giffen used to call 'common 

 statistics.' The term 'common' is not used in any derogatory sense, but in its 

 strict meaning, viz., statistics not designed or compiled specially for one particular 

 purpose, but destined to serve several purposes in common. The obvious reason 

 for this' is that human beings are not willing to spend their lives in filling up 

 forms of inquiry to suit the needs of every statistical investigator. There being, 

 therefore, limits to the amount of statistical data which can be extracted from 

 the public, it follows that returns filled up primarily for one purpose have to 

 serve several other purposes as well. _ 



A single example of the multiple use of common statistics will suffice, viz., 

 the statistics of foreign trade. 



Primarily the classification of the foreign trade statistics of every country is 

 based on the subdivisions of its Customs tariff, the object being to enable the 

 operation of the tariff to be watched. Of course, in the case of the United 

 Kingdom, where Customs duties are now confined to a small number of articles, 

 the existing tariff classification has but a minor effect on the classification of the 

 trade accounts, though even our trade accounts show abundant traces of the 

 influence of the tariff subdivisions of bygone days. But in protectionist 

 countries the statistics of foreign trade are practically governed by tariff con- 

 siderations, and international statisticians are fully aware of the difficulty which 

 tariff variations place in the way of the attainment of statistical uniformity among 

 the different commercial countries of the world. 



The second purpose the trade accounts have to serve is that of the practical 

 trader, who is concerned not at all with the attainment of statistical uniformity, 

 but very much with the safeguarding of his own particular trade interests, and 

 who therefore wishes for a classification which will furnish him with all the data 

 needed for his business, while suppressing all details that will serve the purpose 

 of his trade rivals and foreign competitors. 



When we have reconciled the claims of the Customs authorities and the 

 practical trader we have to meet the insistent demands and criticisms of persons 

 interested in public affairs who wish to learn from the trade accounts what is the 

 true economic state of the nation, in comparison either with its foreign rivals 

 or with some previous period of its own history. But it is very soon discovered 

 that the requirements of these critics, while not always compatible with those of 

 the practical trader or the financial authorities, are not consistent among them- 

 selves. So far as they wish to make international comparisons they recommend 

 modifications which would assimilate our classification to that of foreign 

 countries, but so far as they wish to make historical comparisons they deprecate 

 any changes of classification that will impair continuity. 



It is not necessary to labour the point further, though illustrations perhaps 

 even of a more striking kind might be afforded by the multiple uses to which tho 

 results of the general census are put. 



Thus those who collect and compile common statistics have to serve many 

 masters — sometimes with the usual result. If every statistical enthusiast had his 

 way and the declarations required from traders and citizens were adapted to meet 

 the precise requirements of each investigator in turn, the schedule to be filled up 

 would be something from which the practical business man would recoil in horror. 

 Hence, in practice, we are driven to a rough compromise between divergent and 

 conflicting aims, relying on the resources of statistical science to enable us to 

 apply the needful qualifications to the necessarily imperfect results. 



So far we have only been dealing with the apparatus and methods of research 

 and exposition, and not at all with the objects to which such research should be 

 applied, still less with the ultimate ends of economic study and conduct. 



The next tendency we have to note belongs to quite another region of ideas. 



This is the growing emphasis laid on ends as distinguished from means as 

 the subject of economic study. 



There used to be some disposition to question whether the economist was at 

 all concerned with ends, whether he had not fully discharged his duty in making 

 a correct analysis of the structure of existing economic society and of the forces 

 acting upon it ; and it was rather the fashion to suggest that when this analysis 



