NORMS AND TRENDS IN EXPENSES 93
During the successive pairs of years, 1919-1923, the ratios in-
creased 39 and decreased 33 times. What is the relation of these
direction-changes to the positions—high (+) or low (-) relative
to the district levels—which the ratios held in the first of each
pair of years? The answer to this question, found in Table 65, is
that the prevailing tendency, for ratios which were high in a
given year, is to decrease; and for those which were low is to
increase in the following year. This tendency, however, is not
so marked for ratios of total expense to gross earnings as it is
for those of total expense to earning assets.? To the inverse rela-
TABLE 64
NUMBER OF DISTRICTS WITH RATIOS OF
Total EXPENSE TO Gross EArN-
INGS INCREASING OR DECREASING
FROM YEAR TO YEAR. 10T0-TOAC
ATq =.
—— -
YEARS
TABLE 6%
COMPARATIVE POSITIONS AND YEAR-TO.
YEAR DIRECTIONS OF CHANGE IN Dis-
TRICT RATIOS OF ToTAL EXPENSE
To Gross EARNINGS
Positions Relative
. tothe
Distri-* ¢ «rere.
ne
19010-1020...
1920-1021...
1021-1022...
1022-1923...
1023-1024...
1024-1025...
tion there are 30 exceptions in the former case; in the latter, there
are 12.
While the relation between the positions of the ratios, relative
to their own district levels and the directions of change from year
to year, is inverse, that between the percentage amounts of devia.
tion from type and net percentage amounts of change is direct.
This fact is illustrated in Table 66. While the net percentage
change of the ratios which are above the average is upward, it is
downward in the majority of the cases, and varies directly with
the percentage amounts of dispersion. For the ratios which are
below the average, the direction of change varies inversely and
the net percentage amounts of change vary directly with the de-
gree of dispersion.
The ratios in the different districts may be further compared.
If the yearly percentages for the country are taken as a base and
the ratios in each district each year are computed as plus or minus
? See Table co.
