122 GEOEGE HAGUE ON STATISTICS, Etc. 



than to exaggerate it. This is a long settled tradition, well fixed in the national character. 

 The tendency to exaggerate wealth is foreign to the English habit of mind. The people, 

 as a rule, are secretive in that respect. There is private emulation and rivalry about 

 wealth, as is well known ; but whenever the matter of making statements, filling up 

 returns, and giving material for large and general estimates come into play, there is a 

 tendency to made things less rather than more. If a number of landowners were asked 

 the value of their property a majority of them would without doubt state it at less, rather 

 than more, than what they could sell it for, or were williug to be taxed for it. 



18. Taking, then, these two opposite tendencies in England and the United States 

 respectively, there would be two elements of error in the great comparison attempted to 

 be instituted ; a universal tendency to exaggeration on the one side, and a considerable 

 tendency to diminution on the other. It is easy to see how fallacious any comparison 

 would be under such circumstances. 



19. To sum up the whole matter : — 



In all statistics the moral and metaphysical element must be taken into consideration. 

 This is especially the case where, from an immense number of small data, calculations 

 with respect to large areas and large numbers have to be determined. Small initial errors 

 in such cases may lead to a gigantic error in the final result, just as an error in the 

 size of an angle, which may be almost imperceptible at the beginning, may become mil- 

 lions of miles when the lines are carried out to astronomical distances. 



The principle above enunciated has a wide application. It has to be borne in mind 

 in estimating the value of all statements as to the population and revenue of cities, states, 

 and nations, the amount of their crops, the product of their industries, and estimates of 

 their military strength. It comes into play also in reading of the number of soldiers on 

 each side in a battle or campaign ; the cost of certain wars and, most of all, the amount of 

 national wealth. In all these matters there may be room for various modes of computa- 

 tion, various methods of valuation, according to the state of mind of the individuals con- 

 cerned, as the object aimed at may be of one or another kind, and with widely varying 

 results. 



It follows, therefore, that statistical totals should always be received with caution 

 under the following circumstances : — 



1. When, from the nature of the case, anything approaching to absolute accuracy is 

 impossible of attainment, and large totals are built up by multiplication of uncertain 

 initial figures. 



2. When political objects are to be obtained by a statement in one direction or 

 another. 



3. When national pride or vanity is concerned, or where figures are taken as the 

 foundation, or support, of theories, economical, political, social, or ecclesiastical. 



