08

MIGRATION AND BUSINESS CYCLES
In Chart 15 we have the relation of factory employment to male
immigration in the three years 1893 to 1895 presented in four different
 ways. These four charts, though based on the same fundamental
 data, do not look closely similar nor do they convey the
same Impression.
In Fig. A, the series are expressed as percentage deviations from
their computed trends. The impression received from this section
of the chart is that fluctuations in employment are relatively minor
as compared with those of immigration. This is literally correct,
but the resulting impression is misleading, for the chart conveys no
suggestion of the fact that a one per cent fluctuation in employment
involves a much larger number of men than a one per cent change in
immigration. Nor is it easy to determine from Fig. A whether the
fluctuations of the two series are substantially similar in timing
and direction. Ease of comparison in timing and direction of movement
 is obtained by presentation in the form shown in Fg. B;
that is, with the data expressed as deviations from their trends
measured in terms of the typical deviation of each respective series.
This latter method, which has been used in several of the charts in
this book, has the distinct advantage of throwing the curves close
together and thus facilitating comparison of their changes in direction,
 but, to avoid false impressions, it should be noted that the
numerical significance of a given change is almost entirely concealed.
On such a curve the change in immigration may appear exactly
equal to that in employment, but we cannot tell from the curve
whether the number of men represented by the change in employment
 is equal to the number affected by the change in immigration
or, possibly, one hundred times as great.
For reasons which will be more obvious as we note the many
possible bases of comparison, it is practically impossible to select
scales which will give a precise and unquestionably true impression
of the relative numerical importance of the changes in employment
and immigration. However, if we are turning our attention to the
relation of changes in the volume of employment and immigration,
in terms of the number of persons affected, rather than to the
timing and direction of such changes, then a more accurate impression
 is probably obtained by the use of charts similar to those found
in Fig. C and Fig. D of Chart 15.
In Fig. C the fluctuations in employment are emphasized by the
use of a larger scale than that used for immigration, so that a deviation
 of one per cent in employment appears as great as a ten per