OPPORTUNITIES FOR IMMIGRANTS 7
fluctuations in the numbers of railway trackmen, other than section
foremen, on June 30th of each year from 1889 to 1914.
Coal Mining.
The United States Geological Survey has published statistics of
the movement of men employed and of days worked in anthracite
and bituminous mining, respectively, for the years 1890 to 1921,
and statistics of the production of coal are available from 1870 to
date. Based upon a careful study of the returns filed with them,
the Survey reaches the conclusion that the figures of the average
number employed represent “not the average number of men
actually working at any one time, nor the aggregate number of men
who have worked at any time during the year, nor the absolute
average number on the payrolls, but rather the number of men
commonly dependent on the mine for employment.” Hence, by
multiplying the average number of men employed in each year by
the average number of days worked, we obtain a figure which
affords a better index of employment conditions in the mines than
the average number of employed. To illustrate, the reported average
number of men employed in bituminous mines is even greater in
the depression year of 1908 than in 1907, but the number of days
worked was but 193 as compared with 223 in 1907.
The resulting estimate of employment was compared with the
statistics of bituminous coal production, which are available for a
longer period, and the two series were found to agree so closely that
the longer, or production series, has been used for an indicator of
probable conditions of employment in the bituminous coal industry.
In like manner, the production of anthracite coal is used as an ap-
proximate index of employment in that phase of mining.
Construction.
Especially valuable for our purposes would be a comprehensive
index of the number of men, particularly of common laborers, em-
ployed upon new construction—buildings, sewers, railways, and
streets and highways—but unfortunately no such index is available.
Fragmentary evidence is furnished by statistics of building permits,
miles of railroad constructed, and building contracts awarded, but
none of these series is both comprehensive enough and available
over a sufficiently long period to afford an adequate index of em-
ployment conditions over the period in which we wish to study the
relations of migration and employment. Statistics of the miles of
5%