General Similarity. The cyclical curves for quarterly imports of men and of merchandise, exhibited in Chart 12, reveal marked similarities in their major swings. Both show a marked boom in the early seventies, TABLE 19.-CYCLES IN MALE IMMIGRATION AND IMPORTS OF MERCHANDISE, BY QUARTERS, 1868-1889 Percentage deviations from trend, corrected for seasonal variation and expressed in terms of the standard deviations (29.77 per cent for male immigration and 12.05 per cent for imports). 1869 1868 -0.24 -0.59 +0.21 +0.42 +0.10 -0.11 1870 1871 -0.57 -1.21 -0.27 +0.50 +0.14 -0.05 -0.69 -0.88 -0.76 -0.44 -0.04 -0.89 -0.35 -0.10 +0.56 +0.66 +0.61 +1.05 +1.24 1872 +1.13 +0.64 +0.57 +1.04 +1.24 +2.32 +2.20 +1.43 1873 +1.42 +1.98 +0.28 +0.30 +1.82 +0.75 +0.80 1874 -0.40 +0.29 -0.50 +0.47 +0.38 +0.11 -1.31 -1.00 -1.20 -0.17 -0.90 -0.80 -0.15 +0.26 ■Computed from quarterly statistics of male immigrants in Table I of appendix, and from statistics of imports published by the U. S. Bureau of Foreign and Domestic Commerce in the Monthly Summary of Commerce and Finance. For methods of adjusting for seasonal variation and trend, see Chapter IV. followed by a long decline; another boom in 1880, followed by a short recession and a new peak in 1882; then a decline to a low in 1885 and a recovery in the late eighties. The import recovery in 1877, however, finds no comparable movement in immigration. The Lag. The movements in quarterly male immigration and merchandise imports are not, however, exactly coincident. From 1869 to 1873 immigration clearly lags from two to four quarters behind the turns in imports. The lag, if any, in the depression of the seventies is not obvious; and thereafter, while evident at times, does not appear to be so great as in the first part of the period, ranging from no lag in the short depression of 1881 to about three quarters in the depression of 1885. CHAPTER SUMMARY In this chapter we have first taken a bird's-eye view of the relation of immigration to industrial activity by comparing the annual statistics of immigration with those for imports of merchandise over the entire century beginning in 1820, and with the annual production of pig iron in the period beginning with 1860. From these comparisons it is clear that, particularly after the Civil War, the cyclical fluctuations in immigration are to a large extent a reflex of industrial conditions in the United States, the effect upon immigration evidently becoming apparent in something less than a year. Then we have subjected the period from 1868 to 1889 to a somewhat more detailed scrutiny by comparing quarterly cycles of male immigration and imports of merchandise, both corrected for their typical seasonal variations. This comparison strengthens our preliminary conclusions based upon annual data, and indicates a lag in the effect of industrial conditions upon immigration of from two to four quarters in the early part of the period and a somewhat shorter lag in the latter part of the period. In the following chapter we turn our attention to an examination of the more detailed data available for the years following 1889. CHAPTER V THE PRE-WAR QUARTER CENTURY: 1890-1914 Characteristics of the Period. The twenty-five years from 1890 to the outbreak of the World War include a most fascinating period of American industrial history, and one which in many ways affords the richest field for study of the relations of migration to industrial activity. The decade of the eighties witnessed the virtual passing of the frontier with an abundant area of free and fertile land available for the homesteader. While railroad construction activities did not in the following decades reach the magnitude which they had attained in the eighties, and though the first years of the nineties were characterized by prolonged depression and business uncertainty, toward the end of the decade there began a sharp recovery in industrial activity, accompanied by an equally remarkable increase in the volume of immigration, which reached its peak in the calendar year 1913 with a recorded total of 1,387,318 immigrant aliens and 229,585 nonimmigrant aliens. It will be remembered, also, that it is in the nineties that the "old" immigration from northern and western Europe ceased to be the predominating element in the immigrant stream, yielding in numbers to the rising tide of immigrants from southern and eastern Europe. Data Available for Quantitative Analysis. This quarter century also affords much more adequate data for the purposes of our study than are available for the earlier decades. While the years during and following the Great War have witnessed a remarkable development in the variety and adequacy of statistics of production and employment, even in the preceding quarter century to which we now wish to turn our attention we find at hand statistics which facilitate the close study of cyclical movements in industry. Monthly estimates of pig iron production, usually considered a good index of industrial activity, are available in some form throughout the entire period; and, as noted in a previous chapter, it has been possible to weld together a monthly index of factory em ployment in one or more states which are notable immigrant centers. Monthly statistics of immigration, classified by country of origin beginning in July, 1888, and by sex from July, 1892, are available; and toward the end of the pre-war period the monthly immigration statistics become more and more detailed, so that for the last five or six years of the period they are available by race, country of origin, occupation, and other bases of classification. Also, beginning in July, 1907, monthly statistics of emigration were published, so that it becomes possible thereafter to give a relatively complete picture of the net movement of migration. With its relative abundance of immigration statistics, accompanied by reasonably adequate measures of industrial activity and with little in the way of war or legal restriction to interfere with the free interplay of industrial forces and immigration, the quarter century from 1890 to 1914 affords an exceptional opportunity for the analysis of these phenomena. Method of Analysis. The approach in this chapter is, first, by comparisons over the entire quarter century between the cyclical fluctuations in the monthly statistics of male immigration on the one hand and pig iron production and factory employment on the other. Then, to facilitate the study of certain significant details which are apt to be unduly subordinated in comparisons covering as long a period as a quarter century, and particularly to make possible the satisfactory analysis of emigration series which are not available prior to July, 1907, the entire period from 1890 to 1914 has been broken up into shorter segments, each of which includes at least one major or minor industrial depression and one or two years of the preceding period of prosperity and of the succeeding period of recovery. These selected depression periods are: the severe depression of 1894, the depression of 1904, the major depression of 1908, the minor depression of 1911, and the decline beginning in 1913. This concentration upon short periods facilitates the focusing of attention upon certain details in the reaction of migration to employment which are apt to be overlooked in the more inclusive picture. In the last three of these short periods we introduce comparisons with emigration and with the net results of immigration and emigration. The analysis, however, of the movement of various separate elements in the immigrant current, such as studies by race or occupation, is largely postponed to a subsequent chapter. As will be noted more in detail in connection with the immediate discussion of each period, the method of analysis differs somewhat from period to period in order to make the most profitable use of data available. QUARTER-CENTURY COMPARISONS In Chart 13 we have depicted the fluctuations of male immigration and factory employment. Both series represent deviations from computed trends with the normal seasonal movement eliminated, and hence represent the cyclical fluctuations to the extent that these can be statistically isolated. In plotting these curves the scale unit for each curve is the typical measure of its fluctuations, or the standard deviation, so that the curves are brought into convenient form for comparison of the timing of the cyclical fluctuations. It should be remembered, however, in interpreting these curves, that the method used conceals the fact that the fluctuations in the migration curve are relatively more violent. An approximate measure of the relative violence of fluctuation of the two series is found in their average deviations from trend, which are, respectively, 24.74 per cent for male immigration and 3.40 for employment. Immigration Fluctuations Lag Behind Employment Changes. These series and others subsequently discussed have been examined for the degree of consistency in their timing by two methods. In the first place a graphic comparison was made by plotting the curves on separate sheets and superimposing them over an illuminated chartbox with varying degrees of lag assigned to the migration curve. Then, in significant cases, coefficients of correlation have been calculated as one means of testing the conclusions reached from the graphic comparisons. It is quite obvious upon a brief examination of Chart 13 that in general contour the fluctuations of male immigration and factory employment bear a marked resemblance. Both show a decided depression in 1894, a checkered recovery through the late nineties, a mild depression in 1904, a boom in 1906 and 1907, followed by the severe depression of 1908, a new high in early 1910, and a sharp decline in 1914. There are also a few striking differences which challenge attention. The sharp drop in immigration in the latter part of 1892, a temporary crest in the latter part of 1904 and the first months of 1905, |