College and Research Libraries


By G. A. HARRER 

Library Expenditures: An Examination of 

Their Distribution 

I T IS PROBABLY an understatement to say that library budgets have received a 
lot of attention. Investigations have been 
on both broad and narrow scales, by real 
and pseudo-statisticians, based on valid 
and invalid figures and premises. But 
interest continues and not without cause. 
In his desire for guides, for standards of 
operation in a field combining so pecu-
liarly scholarship and business, romance, 
and cold, hard cash, the librarian con-
tinues to examine all facets of his activ-
ity for aid in developing a sound and 
justifiable modus operandi. With all the 
logic, clear and concentrated thought, 
theorizing, and so forth that have been 
applied to the problem in the past, few 
areas within the compass of the library's 
operations have been defined in terms of 
measurable standards, and when it comes 
to explaining with sound calculations 
the requirements of the library budget, 
we are truly fortunate that many con-
trolling authorities simply accept our 
requests. And if the users be satisfied 
with the servic~, we have the strongest 
justification for our actions. 

Among the factors which may be ques-
tioned in a budget is the internal break-
down, the percentage of the total budget 
allotted to the various major and minor 
subdivisions. "Am I allowing too much 
for salaries, or too little?" or "Why so 
much for salaries? What about books?" 
And the question immediately comes to 
mind: "Is there any typical distribu-

Dr. Harrer is chief acquisition librar-
ian~ Stanford University Libraries; he 
was formerly associate order librarian~ 
University of Tennessee. 

tion?" It is often assumed by those deal-
ing with the subject that the larger the 
library the greater the cost of adminis-
tration. In the larger libraries, then, we 
would expect a larger percentage of the 
budget to be allotted to salaries and less 
to books, binding, and periodicals. If 
there is such a fluctuation, it might seem 
hardly worth while to look for a typical 
breakdown. But an examination of the 
figures, I 'believe, produces some inter-
esting results. , 

I have taken as a basis for calculation 
the expenditures of the "Group I" li-
braries as published in the January.. · 
1955, C&RL~ Vol. 17, No. 1, pp. 58-61. 
There are 107 libraries listed. Unfor-
tunately, some have not supplied com-
plete statistics and cannot, therefore, be 
used here. For purposes of determining 
the distribution of funds into the three 
principal pigeon holes, "Salaries," 
"Books, etc.," and "Other," all libraries 
have been included except Yale, which 
supplied only the first two items. For 
investigation of the subbreakdowns, eight 
other libraries must be discarded. It was 
also felt justifiable to rule out one other 
library because of an extremely abnor-
mal distribution of funds, showing al-
most as much spent on "Other" as on 
"Salaries." 

First, take the ninety-eight complete 
sets of figures, add the columns, and de-
termine the typical percentage distribu-
tion of funds. We arrive at the figures 
shown in Table I. 

Then, an analysis of the three large 
subdivisions of the expenditures of li-
braries grouped by size gives the figures 
in Table II. 

210 COLLEGE AND RESEARCH LIBRARIES 



TABLE I 

TYPICAL PERCENTAGE DISTRIBUTION OF FUNDS FOR g8 GROUP I LIBRARIES 

Staff Student Grand 
Salaries Service Total Books Binding Total Other Total 

•• •• 63.4 •• •• 31.9 4.7 - 100.0 
85.9 14.1 100.0 84.2 15.8 100.0 . ,., •• 

TABLE II 

No. of 
Bookstocks Libs. 

Up to 350M 42 
350M- 600M 24 
600M- 900M 19 
900M-1400M 10 

1400M-3000M 8 

Several observations may be made. 
First, there is a fluctuation, but not a 
steady increase. Administration, as rep-
resented here by salaries, apparently 
costs more in the smaller libraries of this 
group as well as in the larger, although 
not as much more. Secondly, more may 
be spent on "Books and Binding" in 
libraries of 600,000 to 900,000 bookstock, 
where there is the lowest cost of adminis-
tration. There is also an increase in 
"Other Operating Expenses" for this 
group, the reason for which is open to 
speculation. 

An analysis of the "Salaries" figure 
shows that the two components, "Staff 
Salaries" and "Student Service," split the 
total as follows: 

Staff Student 
Bookstock Salaries Service 

350M 84.62 15.37 
350M- 600M 86.55 13.44 
600M- 900M 86.63 13.36 
900M-1400M 84.01 15.98 

1400M -3000M 84.41 15.58 

The figures show a tendency toward a 
smaller proportion's being spent on stu-
dent service in the middle range. An at-

MAY .. 1957 

Books& 
Salaries Binding Other 

64.21 31.58 4.19 
63.33 31.82 4.84 
61.87 32.96 5.15 
64.43 30.01 5.55 
67.59 27.81 4.59 

tempt to correlate this with a similar 
trend in the proportion of student hours 
to staff hours worked in this group was 
inconclusive because of scanty figures. 
The cause of this trend, then, is yet to be 
determined. 

The "Books and Binding" group 
breaks down as follows: 

Books and 
Bookstock Periodicals Binding 

350M 84.78 15.21 
350M- 600M 86.59 13.40 
600M- 900M 86.68 13 .31 
900M-1400M 83.67 16-32 

1400M-3000M 82.42 17.57 

Here again we see a general trend from 
the smaller libraries to the middle group 
with its low proportion spent on binding 
and then back in the case of the largest 
libraries. (An interesting coincidence is 
the similarity of these figures to those 
in the breakdown of "Salaries.") Wheth-
er this figure would correlate with a sim-
ilar trend in the number of periodicals 
acquired as compared to books in these 
libraries is, as above, uncertain because 
of incomplete figures on which to base 
such a calculation. 

211 



Now, all the above figures are of inter-
est for several reasons. In the first place, 
definite progressions in the distribution 
of funds are apparent in the various size 
groups. And, in all but one instance, the 
medium-size libraries show one extreme, 
while the smaller andj or larger libraries 
show the other. The group with between 
600,000 and 900,000 bookstock is able 
to spend the most on materials, and, 
within that figure, more proportionately 
on books. In the "Other Operating Ex-
penditures," however, the high is not 
reached in this group, but in the next 
larger. 

A further question immediately pre-
sents itself-a question of real com-
plications and ramifications. What fea-
ture or features can be determined to be 
the primary causes of this fluctuation in 
distribution? This is, of course, a matter 
of correlating the fluctuation of some 
other measurable feature of the library 
situation with the changes in the distri-
bution of funds within the library 
budget. 

Is it, for instance, the size of the book 
budget or the size of the book collection 
which causes libraries to have similar 
expenditure distribution? Then a fur-
ther thought occurs: "Is it possible that 
the relation of these two features to each 
other would have an effect on the inter-
nal distribution of funds?" For instance, 
if a library had a large collection and a 
small book budget, would this cause the 
librarian to allocate more of his funds 
to staff salaries and less to building the 
collection? We could easily use the 
amount (in dollars and cents) spent per 
book in the collection as a measure of 
this relationship, but a further question 
also presents itself: "How large is the 
public which is served?" 

The larger the student body the great-
er amount, very probably, that would 
have to be spent on staff, all other fac-
tors being equal. It is obvious at this 
point that we are getting into an almost 

insoluble problem. To mention briefly 
other factors which would no doubt 
have an effect on the internal distribu-
tion of funds, we could bring up the 
composition of the student body, i.e., 
percentage of graduate students and 
number of faculty; the number of de-
partmental libraries (as suggested above); 
the types of courses offered-whether 
primarily in sciences, agriculture, engi-
neering, the humanities; the availability 
of the books-whether to a large extent 
on open stacks or closed, whether pri-
marily on reserve for reading courses or 
not; and finally, the quality of service 
offered by the library, a product of all 
the above features. For this last, I can 
see no way to produce a statistical fig-
ure. But it may, in the final analysis, be 
the most important. 

These figures, however, demonstrate 
one other very interesting fact: that 
there is actually little variation in the 
various size groups. The largest variation 
is in the figures in the "Salaries" column 
between the 600,000 to 900,000 and the 
1,400,000 to 3,000,000 groups, a differ-
ence of 5.72 per cent. And the progres-
sion from one figure to the other may 
be expected to be regular. 

This sort of breakdown can be used 
for general checking purposes against in-
dividual budgets. If any figure in a given 
budget is found to be more than, say, 
eight to ten per cent from the general av-
erage, there may be a reason for it; but 
an investigation would at least bring 
such deviations to light and allow for 
questions. If it is discovered, as in the 
case of one library on the list, that only 
20 per cent of the total budget was spent 
on "Books and Binding" (as compared 
to a norm of 31.88 per cent for the size 
group) and, of that, 90 per cent is spent 
on books, leaving binding with only ten 
per cent (as compared to the normal 
split of 85-15 per cent), it is possible that 
investigation will show a justification for 
redistribution of funds. 

212 COLLEGE AND RESEARCH LIBRARIES 

I 

I 

1