MIGRATION AND BUSINESS CYCLES 
in pig iron production may be explainable largely by the fact that 
he bulk of the immigrants credited to any fiscal year actually 
rrive in the second half of the year, that is, between January an 
une, inclusive, so that, for example, the effect of the industrial 
oom indicated by pig iron production in 1909 would, if there is 
ittle or no lag in its effect on immigration, be evident in the im- 
igration for the year ending June 30, 19009. 
Crop values, as portrayed in the lower section of Chart 47, are 
ore erratic in their fluctuations than pig iron production. Never, 
in the thirty years covered, is the direction of change constant for 
ore than two years. Also, the degree of agreement between crop 
alues and pig iron production is relatively low. Nor is there close 
greement between immigration and crop values.» However, if 
mmigration (fiscal years) is compared with crop values of the year 
erminating six months earlier, there is evident in several instances 
tendency for poor crops to be followed by an increase in immigra- 
ion and vice versa. For example the immigration curve rises in 
890, 1893, 1900, 1903, 1905, 1907, 1909, and 1913, while the crop 
alue curve shows a decline for the preceding year. Also, in 1892, 
897, 1904, 1908, and 1912, the immigration curve declines while 
he crop value curve rises in the preceding year. | 
But are these years in which relatively poor crops are followed by 
n increased emigration, or good crops by decreased emigration, 
Iso years in which immigration from Italy is not well explained by 
he changes in industrial activity in the United States? On the 
ontrary, in all of the fourteen years just mentioned but 1905 and 
909, an increase in Italian immigration to the United States is 
receded by an increase in pig iron production, or a decrease in such 
immigration by a decrease in pig iron production; so that, in view 
f the further fact that in about half of the period under considera- 
ion an increasing emigration to the United States follows relatively 
ood crops, there is scant evidence to support the theory that cyc- 
lical fluctuations in emigration from Italy are largely due to crop 
onditions in the ur 
The coefficient of correlation between crop¥values and pig iron production, when 
oncurrent items are compared, is 4.32 with a “probable error” of 4.11; and between 
rop values in calendar years and Italian immigration to the United States in the res- 
pective years ending six months later is + .31 + .11. 
2The relationship between the three series under discussion may also be expressed in 
ustomary mathematical terminology, that is, by stating the coefficients of correlation, 
which are +.50 + .09 for pig iron production and immigration from Italy in the fiscal 
ear ending six months later; and only 4.31 + .11 for the corresponding comparison 
between Italian crop values and immigration from Italy. 
202