Page 77 - clra62_0019-(GIPE)

ON INTEl\NATIONAL STATISTICAL COMPARISONS 71
menl, no proper comparison can be made. This applies
specially to a comparison between wages in out-of-door
trades in a country like the United States, with a severe
climate, and wages in the same trades in England.
\Vages in the former country may well be higher per
nominal day or week of actual work, and yet the differ-
ence not be so great when the earnings and hours of
labour of the whole year in England are reckoned.
What I would most desire to direct attention to, how-
ever,-is the statistical importance of a somewhat differ-
ent point. This is the distribution of the population
according to remuneration. I t is quite conceivable that
in one of two countries the earnings, and still more the
nominal wages, may be higher than in the other in every
single employment which can be enumerated and com-
pared, and yet the average earnings of the average
wages-earning man may be higher in the latter country
than the former, the reason being the different distribu-
tion of the people according to earnings. This can be
shown very clearly in a theoretical comparison. Take
first a community of 1,000 wages-earners, with the
people distributed according to earnings, in the folIow-
ing classes-A, B, C, D, and E-as follows:

Firsl Community.
Class. Per Annum. Nos. Total.
A. Earnings £5 0 5 00 £25,000
B. 60 200 12,000
C. " 7 0 100 7,000
D. " 80 100 8,000
E. " 9 0 100 9,000
"
Total • 1,000 £61,000
----
Average per head, £61.
And compare this with another community of equal
numbers, in which there are also five grades, each re-
munerated at a lower rate than the corresponding grade
in the first community, but in which the average ofthe