OPPORTUNITIES FOR IMMIGRANTS : 
computed. For the years 1909-1914, this index exhibited fluctua- 
tions considerably more violent than exhibited by the same series in 
1904-1908 or by the census data in 1909 and 1914. Consequently, 
in order to get a consecutive series on a reasonably homogeneous 
basis, the fluctuations of the trade union unemployment data were 
scaled down in the ratio which they bore in 1914 to the fluctuations 
shown by the Census. 
Monthly Production of Pig Iron. 
For evidence supplementary to that afforded by our index of 
factory employment by months, we have used monthly statistics 
of pig iron production. The original figures were adjusted for 
seasonal variation by a method designed to make allowance for 
the tendency of the typical seasonal variation to change over a 
long period of years. The method used is developed by Dr."W. I. 
King in an article published in the Journal of the American Statis- 
tical Association.» His data, seasonally corrected, were used for the 
years 1905 to 1914, and together with figures obtained by similar 
methods for the years 1884 to 1904, were corrected for a computed 
trend based upon a seven-year moving average smoothed to elimin- 
ate minor irregularities. Small fluctuations were then ironed out 
by taking a three-month moving average of the indices obtained by 
correction for trend and seasonal variation. The results appear in 
Chart 14 in Chapter V. 
The Numerical Volume of Employment and Unemployment. 
Information concerning the actual number of workers represented 
by fluctuations in employment or unemployment is scant. We 
have made use, however, of two studies of this nature. The first, 
covering unemployment in non-agricultural occupations during the 
years 1902 to 1917, by months, was made by Professor Hornell 
Hart.: The method used, as described by Professor Hart, was to 
ascertain for each year and month the total number of persons 
normally occupied in non-agricultural pursuits, and to subtract 
from these normal supply figures the estimated ‘connected demand’ 
for labor. This “connected demand’ for labor was determined “by 
a synthesis of widely scattered information on employment fluctua- 
tions,” chiefly from various Federal and State statistical publica- 
2“An Improved Method for Measuringfthe Seasonal ¥Factor,” September, 1924. 
“Hornell Hart, Fluctuations in Unemployment in Cities of the United States, 1902 to 
1917, Studies from the Helen S. Trounstine Foundation, Volume 1, Number 2.7% | _ 
75