has been nothing in these comparisons to indicate how the volume of immigration or of emigration compares in number of persons with the corresponding change in the number of persons employed To the extent that migrants are members of the working class, the number of arrivals less the number of departures represents a net addition to the number of workers seeking employment. Unless this net addition is accompanied by an increase in the number of persons employed, the necessary result is an increase in the total number of unemployed persons in the United States. If, in a given month, the immigration of workers exceeds the emigration of workers by 50,000, and the increase in the number of employed in the United States is only 30,000, it is obvious that there has been a net increase of 20,000 in the number of the unemployed.

Fully satisfactory data for making comparisons of seasonal net migration and changes in employment are not available, but we have made the best approximation we could, in the following manner. In the first place, for the several industries which have been selected, for reasons previously indicated, as particularly significant when studying employment opportunities for immigrants, we have computed an estimate of the typical number of persons employed in each month of the year in the pre-war period. Statistics for the year 1909 were used in determining the average number of workers to be assigned to each industry. This computation yields an estimate of the typical month-to-month change in the number employed in factories, bituminous and anthracite coal mining, railway track maintenance, and construction work, when the cyclical tendencies have been as far as practicable eliminated, leaving the joint effect of the trend and seasonal factors. Inasmuch as the typical net migration, by months, represents a corresponding increase or decrease in population, it is appropriate to compare therewith the typical change in employment which results from the combined influence of the growth and seasonal elements.

The results of the employment estimates appear in Table 59 and Chart 55.

For the net migration to be used in comparison with the typical month-to-month change in employment, we have selected the excess of arriving over the number of departing male aliens. This group includes those male aliens who are officially classified as temporary migrants-that is, as nonimmigrants or nonemigrants. Many of these come for employment purposes, and hence it ap

Thousands of Persons

CHART 55

THE PRE-WAR NET MIGRATION OF ALIEN MALES AND THE TYPICAL MONTH-TO-MONTH CHANGE IN THE NUMBER EMPLOYED"

JAN FEB. MAR. APR. I MAY JUNE JULY AUG. SEPT. OCT. NOV.
Numerical data in Tables 55 and 59.

DEC

peared advisable to include them, as well as those officially listed as immigrants, in calculating the volume of net migration.

In Chart 55, Fig. A, this net migration series is compared with the month-to-month change in all the selected industries. A bar

above the zero line indicates the estimated increase in the number employed as compared with the preceding month; a bar below the zero line, an estimated decrease from the preceding month. The curve represents the typical pre-war net migration of males in the given month.

TABLE 59.-ESTIMATE OF THE SEASONAL DISTRIBUTION OF EMPLOYMENT IN SELECTED INDUSTRIES IN THE PRE-WAR PERIOD

(Adjusted for cyclical variations but not for trend)

These figures represent an estimate of the joint effects of seasonal and trend factors. The indices of seasonal variation given in the preceding tables were applied to the numbers employed in the respective industries in 1909, as recorded in the Census or estimated from other sources, and then the results were adjusted by adding the estimated effect of trend movements as indicated by data for the years 1907 to 1914. bIncludes factory employment, bituminous and anthracite coal mining, railway maintenance, and consConstruction and railway maintenance.

truction.

dComputed from Columns A and B, with allowance for trend in computing the December to January

change.

Assuming that our estimates present a true picture of the typical changes in employment and in net male migration prior to the quota restrictions, we observe from Chart 55, Fig. A, and the tables upon which it is based, that in January there is a small net excess of alien male arrivals over departures, to the extent of about 5,500 persons, while employment in the selected industries decreases about 137,000. In the following eight months an excess of arriving over departing alien males is, in each case, accompanied by an increase in the number employed. In all these months but April, July, and Sep

tember, the increase in employment exceeds the net volume of male arrivals. In October and November a decrease in employment is accompanied by a small net immigration, and in December a heavy decrease in employment is accompanied by a small excess of departing over arriving male aliens.

In other words, decreasing employment in January and November is aggravated by a small net excess of arrivals, and in October by net arrivals to the number of about 32,000. Also, in April, July, and September, the increase in employment is not sufficient to absorb the new arrivals.

Only in December, and then only to a small extent, is the slack created by a decrease in employment taken up in part by a net outgo of male aliens.

It is true that in five months-February, March, May, June, and August-the number of workers employed is increasing faster than the net inflow of male aliens, and if there chances to be a shortage of resident workers in these months, immigration may be looked upon as alleviating this shortage. On the other hand, if in these months the increase in employment is in fact not adequate to relieve an existing unemployment situation, then the net inflow of alien workers merely acts to check the decrease in unemployment.

In summarizing the above comparison of the typical net movement of alien males with the month-to-month change in employment ascribed to the growth and seasonal factors, it should be noted that the evidence presented should at best be taken as suggestive rather than conclusive. The data upon which the estimates are based are too fragmentary, and the margin of error involved in the computations too large, to justify treating the computed relations as more than rough approximations. Here, as in the greater part of this chapter, we are dealing with pre-war, and hence pre-restriction, conditions.

With the above qualifications in mind, we may summarize the evidence presented in Chart 55 and the accompanying tables as indicating that the seasonal distribution of male immigration and f emigration is such as to aggravate unemployment in six months of the year and to alleviate it slightly in one. In the other five months, being those in which net male immigration is less than the increase in employment, its effect is to alleviate the effects of a shortage of resident workers, if such a shortage exists.

In Fig. B of Chart 55, a comparison similar to that just made for

"all selected" industries is presented for "selected outdoor" industries. Not all of net male migration, of course, goes into these industries, but large numbers of the recent immigrants are employed therein, particularly in pick and shovel work.

Using the same method of interpretation applied to "all selected" industries, and assuming, for purposes of comparison, that the entire volume of net migration is absorbed in these outdoor industries, it would appear that in December a small excess in departures probably lessens slightly a tendency toward increasing unemployment; in six months-March to August, inclusive the increase in employment is greater than the net number of male arrivals and hence male migration in these months is either alleviating a shortage of labor, or if such shortage does not exist, is merely slowing up the decrease in unemployment which would otherwise arise from increasing activity in these outdoor industries. In February the number of net arrivals is greater than the increase in employment; and in four months-January, September, October, and Novemberemployment is decreasing while arrivals exceed departures, though in January and November the excess of arrivals is not enough to be of appreciable significance.

On the whole, the evidence favors the conclusion that in the months from March to August, inclusive, the seasonal distribution of net male arrivals is well adjusted to the changes in employment due to activities in construction and railway maintenance; that in January, November, and December the net movement is too small to be of great significance; and that in February, September, and October the new arrivals must look largely to other industries for employment.

SEASONAL FLUCTUATIONS IN IMMIGRATION UNDER THE QUOTA ACTS

The preceding discussion has referred, in the main, to the relation between the seasonal movements in migration and employment prior to the Great War. The quota acts of 1921 and 1924 caused material modifications in the seasonal distribution of immigration. The act of 1921 permitted up to twenty per cent of the annual quota to enter in any one month. As the immigration year begins July 1st, the effect of this provision was to concentrate the heaviest immigration in the months from July to November, inclusive. In the first year of the operation of the act, beginning July 1, 1921,