26 STATISTICS OF EXPENDITURE AND 
be more readily estimated in a single one of the tables of weights 
and measures. The tables from which the per capita consumption 
of sugar, tea. coffee and dried fruit has been calculated, were 
compiled from the Annual Sessional Papers on Trade and Com- 
merce; the per capita consumption of beer, spirits, ete., is the 
calculation of the inland revenue officials, and may be found in 
Statistical Year Book for the current year.* 
The consumption of these articles is recognized as one of the 
best tests of the prosperity of a country. The middle classes 
everywhere are well provided with the comforts and decencies of 
life, in which class these articles are placed, although sugar is 
rapidly becoming a necessary of life; and an extension of the 
consumption of these goods means that the working classes are 
consuming more, the middle class, it being presumed already, 
using as much as they desire. Ina country like Canada, where as 
we have seen there are few extremes of wealth, an increased 
consumption means that the whole body of the people are con- 
suming more. 
An increased consumption of any article may mean one of 
three things,—(1) it may result from a fall in price, which enables 
the people to consume more without spending more ; (2) it may 
mean a rise in the average income, which enables the people to 
spend more on one article than they have been doing, without 
curtailing their consumption of other articles; (3) it may mean 
simply that the form of consumption has changed and that the 
well-being of society is the same, or but slightly increased. In 
all probability, the increased use of eccoa, from the value 
of $44,249 in 1880+ to $158,849 in 1896 has been due to a 
mere change in the form of consumption ; and the addition of 
this amount to the consumption of the community probably does 
not indicate a corresponding increase of spending power. The 
increase in the use of sugar and tea is due, not to increased 
*Itake this opportunity of acknowledging my indebtedness to the Dominion 
Statistician, Mr. George Johnson, whose work I have freely used in the preparation of 
this paper. 
t+Average of three years. 
