nces in the cost of living when wage standardization
has been extended. No constant tendency, for example, can be found in
the agreements made by different local branches of the same national
trade union to build up a wage scale in accordance with differences in
the cost of living at different points.[94]

The most complete body of material on the subject is contained in the
report of the Investigating Commission of the Board of Trade (Great
Britain) on Working Class Rents, etc., in the United States (1911). This
commission studied the wage schedules of skilled men in the building,
engineering and printing trades in twenty-eight of the large cities of
the United States and compared these wage schedules with the calculated
cost of food and rent in these towns--weighing food three times as
heavily as rent. The results are presented by single cities, by
geographical groups, and by population groups--i.e., cities grouped in
accordance with size of population. _Real_ wages tended to be more equal
as between population groups than between geographical groups. The range
of the index number between geographical groups is from 85 to 104 (New
York is taken as 100); between population groups from 89 to 100 (New
York, 100). They reveal a tendency for money wages and living costs to
be high in the largest cities, and for both money wages and living costs
to decline in the cities making up the smaller population groups. No
correlation can be found between living costs and money wages as between
individual cities, however.

The argument for variation or limitation because of differences in the
cost of living is a two-fold one. Firstly, it may be argued that such a
policy is calculated to maintain industrial activity in the smaller
centers, where the cost of living is usually lower, in the face of the
competition of the larger centers, in which the cost of living is
usually higher. Secondly, it may be argued, that variations in the cost
of living at different places are indications of the fact that at some
places the economic essentials can be procured with a smaller
expenditure of human labor and capital than at other places (since labor
and capital can move between them) and, therefore, it is to the general
interest to encourage industrial development at the points where the
cost of living is relatively low.

As to the first argument, it seems to me that there is considerable
wisdom in the wish to encourage a diffusion of industria