. MIGRATION AND BUSINESS CYCLES 
tions. Owing to the fragmentary nature of the data available, 
there is necessarily a considerable margin of error in these estimates, 
and hence the comparisons made with their aid must be interpreted 
as giving roughly approximate rather than closely accurate results. 
A second estimate of the actual numbers represented by fluctua- 
tions in employment is found in the study made by Dr. W. I. King 
for the 1921 depression period and described more fully in Volume 
V of the publications of the National Bureau of Economic Research, 
Employment Hours and Earnings wn Prosperity and Depression, 
1920-1922. Based upon returns from a large number of employers 
in various lines of industry, estimates were made of the changes in 
numbers employed from the first quarter of 1920 to the first quarter 
of 1922, inclusive. From these estimates, which are given by in- 
dustries, we have selected, in Chapter VI, those industries which 
are most significant from the point of view of employment oppor- 
tunities for immigrants and made comparisons with the number of 
immigrants and emigrants during the period covered by the es- 
timates. 
CHAPTER SUMMARY 
The direct and indirect indices of employment conditions to be 
utilized in the following chapters include (1) for the entire period 
over which immigration statistics are available, the annual statistics 
of imports of merchandise; (2) for the decades between the Civil 
War and 1890, annual statistics of pig iron production and quarterly 
statistics of imports of merchandise; (3) for the period beginning 
in 1890, estimates of factory employment and of pigiron production, 
by months, and (4) particularly in the post-war years, various 
short-period indices of employment conditions, the description of 
some of which is deferred to the chapters in which they are used. 
In this chapter we have noted the nature of the major series of 
statistics of economic conditions to be used, made some comparisons 
between these indices and other evidences of economic activity, 
and indicated the methods used in putting these employment data 
into convenient form for statistical comparisons. The subsequent 
chapters are devoted chiefly to the analysis of fluctuations in im- 
migration with the aid of the employment indexes to which attention 
has been directed in this chapter. 
76