NORMS AND TRENDS IN EXPENSES 101 
But the second question asked in the paragraph above requires 
an answer inasmuch as the marked uniformities given in Table 70 
do not necessarily indicate agreement by districts as to the years 
which are high or low. Specifically, the question is: Do the 
ratios for the respective types of banks, when paired by districts, 
agree in giving a year a particular position in the scale of “low,” 
“moderate,” “high”? The answer is found in Table 71, from 
which it is apparent that except for 1921, and possibly 1923, the 
ratios by districts for both groups of banks were high or low, 
relative to their respective levels, at the same time. 
Not only, for given years, are the district ratios for the twe 
types of member banks, in general, similarly placed with respect 
to their own levels, but the greater the percentage deviation for 
the national, the greater the percentage deviation for the state 
banks. That is, if in a given year ratios of total expense to gross 
earnings for national banks deviate widely, plus or minus, from 
the district levels, ratios of expense for state banks, similarly 
measured, deviate widely, the percentage amounts in both cases 
varying directly with each other. Abnormal conditions making 
for high or low total expense in terms of gross earnings affect 
similarly both types of banking institutions. This fact, of 
course, is to be accounted for partly by the conditions, common 
to all banks, which affect interest rates—the dominant part of 
gross earnings—and by compe- 
tition and regulation, applying 
as they do to interest on de- 
posits, interest and discounts on 
borrowed money, reserve re- 
quirements, and so on. 
So much for the relation of 
the ratios, for the two banking 
groups—national and state 
members—to their own seven- 
year average levels. 
The ratios, being above or 
below their own district levels 
at different times, change from 
year to year. There are year- 
to-year oscillations and long- 
time trends. Are these alike in