SUMMARY 241 
is not seen for almost a year afterward, but in other instances the 
fluctuations in employment and migration appear to be substantially 
concurrent. The more common lag in the migration fluctuations is 
from one to five months. 
Relative Violence of Cyclical Fluctuations. 
On the whole, the changes in migration are more erratic and 
more violent than those in industry. The seasonal variation in 
migration is more marked and the amplitude of the cyclical move- 
ments is, as a rule, greater than that of the corresponding fluctua- 
tions in employment. This comparison, however, refers to devia- 
tions in terms of percentages, and not to the number of persons 
affected by fluctuations in employment and migration respectively. 
An industrial depression usually brings a sharp decrease in immi- 
gration; but, owing to the larger total number of persons involved, 
a decline in employment which in percentage terms appears relative- 
ly less than the concurrent decline in immigration, may affect a much 
larger number of persons. However, as explained more fully in 
Chapters V and VI, numerical comparisons between migration and 
employment are most appropriate when cumulative migration is 
compared with changes in the number employed. 
Effect of Migration upon the Cycle in Employment. 
The demonstrated sensitiveness of immigration and emigration to 
employment conditions may lead to an exaggerated estimation of 
the efficacy of migration as a safety-valve for an overcrowded labor 
market in depression periods. We have seen that depression retards 
immigration and accelerates emigration, but the weight of evidence 
is, not only that these compensating movements are often not 
numerically adequate to decrease the number of workers in this 
country in a period of depression, but that, on the contrary, even 
in periods of low employment net immigration is sometimes steadily 
adding to the supply of workers. While immigration falls off 
materially when employment is slack, it never ceases entirely, and 
a considerable number of new workers arrive even during a de- 
pression period. True, there is at the same time an exodus, a 
movement which occasionally has exceeded immigration in volume, 
but, judging from the few depression periods for which complete 
statistics are available, there is, when the entire duration of the 
period of dull employment is considered, always a net immigration. 
(See Directors’ footnote “a”, p. 120). 
~ + 
A