BANKING STANDARDS 
account must be taken of the relative influence exerted by varying 
operating conditions, years, and districts, in so far as they bear 
upon both gross earnings and total expense, the relations between 
which determine the net earnings. All of this has been done in so 
far as the nature of the data permits. It is of interest, apropos of 
what has just been said, to review briefly the manner in which this 
has been accomplished. 
The analysis began with the presentation, for a variety of 
series of data, of the actual ratios for member banks, by districts 
and years. Diversity was immediately apparent—differences re- 
lating to the various years and to the several districts. In order 
to secure a basis for the study of the ratios, an average or a series 
of averages, therefore, was selected from which the ratios were 
measured ‘as percentage deviations. Moreover, the percentage 
changes from year to year for the ratios in each district for each 
series of data were also calculated. These, together with the 
percentage differences from type, were the data out of which, for 
each series, norms and trends were determined. But the dis- 
cussion of such differences and trends related specifically to the 
individual series; each was treated independently, the relations 
between them being left for subsequent study. 
What in general resulted from the first type of analysis? In 
spite of diversity of district location, of number of banks, and of 
yearly and district variations in the ratios, each series of data 
revealed norms and trends. 
[t was patent, however, that the different series were not in- 
dependent of each other. Even the barest knowledge of the func- 
tions performed by banks and of banking operations suggested 
this fact. But in what way and to what degree they were related 
was not, on an @ priori basis, apparent; analysis of a type designed 
to reveal these facts was necessary. Accordingly, later chapters 
of the discussion were concerned with this problem. It is un- 
necessary to repeat in detail the findings in these respects, save 
certain of those which have to do with ratios of gross earnings and 
of total expense, the relations between which determine the ratios 
of net earnings, and to which attention is called inasmuch as they 
have a direct bearing on the matter which follows. 
In reviewing our findings, so far as correlation between series 
is concerned, it is appropriate to begin with the subject of gross 
earnings. Other things being equal, it is to be remembered, the 
254