THE PRE-WAR QUARTER CENTURY 111 
emigrants over immigrants is concurrent with an increase in the 
number unemployed—then migration is clearly an alleviation rather 
than a primary direct cause of fluctuations in unemployment in the 
host country. On the other hand, if the net migration of workers 
exceeds in number the employment change, or is contrary in direc- 
tion, it is clearly a disturbing factor. 
Net migration is obviously the most significant basis for volume 
comparisons; but, particularly where emigration is not known, 
the volume of arrivals is important as indicating the magnitude of 
the absorption task as compared with the current tendencies in 
employment. Even if the number of arriving immigrants is balanced 
by an equal number of departing emigrants, it is scarcely to be 
assumed that the necessary employment adjustments are made 
without considerable loss of time to the worker and disturbance to 
industry. 
But over what period shall the change in employment and mi- 
gration be compared? A week, a month, a quarter year, a year, a 
decade, the period of decline in employment, or the duration of a 
depression? The answer will depend upon the particular purpose to 
be served. If the purpose is to show the adjustment of migration 
to seasonal variations in employment, or to the combined effects of 
cyclical, seasonal, and other forces, a month-to-month comparison 
may be pertinent. If it is desired to eliminate in part the erratic 
month-to-month fluctuations, and yet to restrict the analysis to the 
effect of relatively current immigration, a three-month comparison 
is appropriate. If from January to March, inclusive, employment 
falls off 50,000, and 100,000 working immigrants arrive, it is obvious 
that their arrival is not well timed and apt to aggravate the unem- 
ployment situation. To eliminate seasonal factors, comparisons 
over twelve-month periods are suitable. In considering a given 
depression period, it would appear worth while to ascertain the 
cumulative volume of migration either during the period of decline, 
or, to change the point of view somewhat, during the depression 
period—defined herein, ordinarily, as the period during which 
employment is below its computed trend. It would be somewhat 
unduly dogmatic to insist that any one of the bases of comparison 
just mentioned is, in all cases, the most logical. It is worth noting 
what results are obtained from the use of each of them, and con- 
sequently, in the various volume comparisons in this chapter, no 
invariable basis of cumulation has been adhered to. Accordingly,