NORMS AND TRENDS IN NET EARNINGS 133
registering this type of change. The greatest percentage decline
for all districts combined, as well as for the majority of the dis-
tricts, came between 1922 and 1923; the greatest percentage rise
came between 1919 and 1920. Between 1919 and 1921, net earn-
ings increased at a diminishing rate; between 1921 and 1923 they
fell at an increasing and between 1923 and 1925 at a diminishing
rate. But the percentage amounts of decline and of rise in the
ratios in the various districts are functions not alone of time but
also of the district levels themselves and of the positions held by
the ratios in a given year relative to these levels. These facts are
apparent from the detail in Tables 84 and 85.
If the yearly district ratios of net earnings are classified into
two groups, plus and minus, according to the positions which
they hold relative to the seven-year district levels, as is done in
the stub of Table 84, and for each class the direction of change
from year to year is determined, as is done in the caption classes
of the same table, it is found that (1) ratios which are low tend
to increase, and (2) those which are high tend to decrease. To
this rule there are 22 exceptions in 72 cases.
It will be recalled that, in general, during 1921 to 1923, ratios
of net earnings to earning assets decreased from year to year.
This is evident from the fact that of the 72 year-to-year changes,
43 were decreases and 29 increases. But of the 72 possible
positions held by the ratios in the first of each pair of years,
relative to their respective district levels, 41 were high, and
of the number so placed, 31 decreased in the following year.
Conversely, of the 31 which
were low in a given year, 19
increased the next year. In
25% of the cases the year-to-
year changes were upward
when the ratios were relatively
high, and in 39% they were
downward when they were al-
ready low. That is, there is
a tendency, so far as direc-
tion of change is concerned,
for net earnings to regress
* In Kansas City and in Dallas, the ratios were the same in 1924 and 1025; in
Atlanta they were the same in 1022 and 1022
