College and Research Libraries JOSEPH J. KOHUT Allocating the Book Budget: A Model Inflation is currently affecting library book budgets, particularly with respect to the acquisition of serials. A model is proposed which would balance the purchase of serials against the purchase of monographs by individual funding units within the academic library. Special con- sideration is given to inflation as a cost factor affected by both the form of publication and the subject matter. Applying the model to a specific example demonstrates its use in providing control over col- lection development and allowing for equitable distribution of book funds among funding units. INTRODUCTION CoNCERN ABOUT BOOK BUDGETS is acute particularly in light of the current in- flation and proliferation of new serial titles. If the current trends. continue, within a few years the total book bud- get of many academic libraries faced with constant or decreasing book bud- gets will be exhausted by serial subscrip- tion costs. However, as the serial budget is only one aspect of the total acquisi- tions problem, no recommendations re- garding serials can be made without con- sidering the total acquisition policy I book budget of the library. Because of the apparent lack of workable models in the literature, the author formulated basic goals and as- sumptions for setting budgetary pol- icies. It is imperative to note, however, that the proposed model incorporates decisions that are in fact now being made either explicitly or implicitly by Mr. Kohut is science librarian at Port- land State University, Portland, Oregon. The author gratefully acknowledges the as- sistance and criticism of his colleagues in the preparation of this article. 192 I all academic libraries with limited book budgets. Its major contribution is to clarify the context in which these deci- sions are being made and to define the underlying principles which should guide these decisions. Because the pro- posed model is fundamentally expressed as proportions of library-resource units it is applicable regardless of annual fluctuations of the total book budget. BALANCING SERIAL WITII MoNOGRAPH AcQUISITIONs A Model Implicit in the concern about serial costs is the assumption that too much of the book budget is devoted to serials. Most libraries could balance the acquisi- tion of monographs with serials so as to maximize their potential to serve the university's information needs. Al- though such balancing is customarily ex- pressed in terms of dollars, the value of the library to the community of users is defined by numbers of library-re- source units. 1 To the user, therefore, it is the proportion of actual monographs to serials that is critical. To discuss col- lection development balancing in terms of monies allocated is misleading be- cause of differential rates of inflation for each form of publication (e.g., seri- als vs. monographs) and for each · dis- cipline (e.g., biology vs. geography). Assuming there is an optimum bal- ance between the number of mono- graphs and number of serials that should be received within a given year, this ratio could be best expressed in li- brary-resource units rather than money allocations. In addition, because every discipline has its own optimal balance between serials and monographs, the op- timum ratio for the entire library be- comes a composite of the ratios for all the disciplines the library supports. These disciplines may constitute fund- ing units; in .turn, the general frame- work of funding units should for the most part reflect the academic structure of the university.2 Each funding unit should be assigned a proportion of the total number of li- brary-resource units to be acquired. These proportions can be adjusted to re- flect significant changes in academic pro- grams, which are affected by the number of disciplines studied, the level of pro- grams (graduate, undergraduate), and other factors. At any given time, then, there is a multitude of internal con- straints which control the collection de- velopment for each academic program. Each library must, therefore, apportion library-resource units according to its own objectives. Although criteria are not set here for identifying an optimum proportional distribution of library resources among funding units, certain factors external to particular institutional situations can be considered. For example, the annual output of all subject -literatures varies from subject to subject and statistics can be applied to adjust support for each funding unit according to the size of its respective literature. However, in applying a correction for literature size it must be assumed that each library re- source in : ~me funding unit equals that Allocating the Book Budget I 193 in another in terms of its information value (e.g., a mathematics serial or monograph is as potentially useful to a mathematician as a psychology serial or monograph is to a psychologist), an assumption which may not be necessari- ly valid. Publication quality control in one field may differ from that in anoth- er subject. Given these variations, the following model involves two fundamental con- siderations: ( 1) the proportion of li- brary-resource units to be allocated to each funding unit, and ( 2) the mono- graph-serial balance within each fund- ing unit. 1. Let a book budget ( GT = 100 per- cent) be apportioned among a giv- en number of funding units ( N) so that A + B + C ... + N = GT. 2. Let each funding unit (A, B, C, . . . , N) be subdivided into a cer- tain percentage of monographs ( m) and a complementary percent- age of serials ( s). 3. The matrix shown in Example 1 ~~ ~ GT .. Tm Bs .. Ts s Example 1 results from ( 1) and ( 2), where Am+ As= A, Bm + Bs = B, ... , Nm + N8 = N; Tm equals the total per- centage of the budget spent for monographs; and Ts equals the to- tal percentage of the budget spent for serials. The proportional allocation of li- brary-resource units among funding units is made prior to, and independent- ly of, any consideration of the desired monograph-serial ratio for each fund. Each fund must subsequently be divid- ed between monographs and serials, both of which are expressed as percent- 194 I College & Research Libraries • May 1974 ages of the total number ( 100 percent) of library-resource units to be acquired. The monograph-serial balance, however, is taken into consideration when cor- recting for inflation, but only because the rate of inflation of monographs and serials within a funding unit or among funding units is rarely the same. Inflation If there were no variation in infla- tion among forms of publication-seri- al vs. monograph-or among publica- tions of various funding units-history, psychology-then proportions of li- brary-resource units could translate di- rectly into dollars. However, compara- tive studies show that a constant book budget is unevenly eroded by inflation which varies both with Jorm of publi- cation and from one discipline to an- other. A collection that was in balance in 1967, for example, is now greatly out of alignment. Tables 1-3 demonstrate the degree to which inflation aHected costs for period- icals during the period 1967 I 69 through 1972. Chemistry I Physics (Table 1) and Art (Table 2) were selected as examples because they represent the extremes of inflation. Table 3 shows average period- ical costs I inflation. Only periodicals published in the United States are repre- sented by these figures. Comparison of Tables 1-3 indicates that: 1. Chemistry I Physics subscription costs have risen at a more rapid rate than the average, whereas the Art subscription costs have risen at a slower rate. The differences in rates (price indexes) are substantial. 2. Given a constant periodicals bud- get, by 1972 Chemistry /Physics could continue only 54 percent of its 1967 I 69 periodical subscriptions; that is, out of every 100 periodical subscriptions in 1967 I 69, 46 would have had to have been cancelled by 1972. Art would have been forced into 20 periodical cancellations per 100; that is, a reduction in period- ical titles of 20 percent. By the end of this relatively brief interval TABLES 1-3 INFLATIONARY EFFECTS ON ACQUISITION OF SELECTED PERIODICALS Year TABLE 1 CHEMISTRY /PHYSICS (1) (2) (3) Year 1967-1969 100 1967-1969 TABLE 2 ART (1) 100 112 (2) (3) .90 -10 1970 137 .73 -27 1970 ~19~7~1----------715=7~----~.6~4~-----~3~6 -19~7~1----------~~----~------~ 122 .82 -18 1972 186 .54 -46 1972 126 .80 -20 -------------------------------- -------------------------------- TABLE 3 AVERAGE Year (1) (2) (3) 1967-1969 100 1970 120 .83 -17 1971 135 .74 -26 1972 153 .65 -35 (1 ) Price index (base year = 1967 ) . ( 2) Percentage of periodicals that can be purchased in terms of 1967 dollars. ( 3) Number of periodicals per 100 lost to inHation ( 1970-72). Data from: "Price Indexes for 1972; U.S. Periodicals and Serial Services," Library Journal 97:2356 (July 1972). Chemistry I Physics would have been compelled to cancel more than twice as many periodical subscrip- tions as would Art. 3. The rate of inflation within each discipline varies from year to year. The price index for Chemistry I Physics jumped by twenty between 1970 and 1971 and by twenty-nine between 1971 and 1972. Corre- sponding figures for Art are ten and four. The average rate of in- flation was relatively constant (ca. 15 percent annual increase). Be- tween 1970 and 1972, therefore, the rate of increase in periodical costs accelerated for Chemistry I Physics but decelerated for Art. If these trends continue, the dispar- ity between Chemistry I Physics and Art periodical subscriptions will in- crease at a more rapid rate than in the past. Although all funding units were Allocating the Book Budget I 195 faced with cancellations due to infla- tion, the original ( 1967 I 69) balance of the periodicals collection has been gross- ly distorted. Not only is budgeting ac- cording to a proportional distribution of periodicals money misleading, but an across-the-board percentage change in fund money would have only perpetuat- ed this growing imbalance. Tables 4--7 show the effect of inflation on monographs. Table 5 (Science), 6 (Art), and 7 (Average) were chosen for comparison with the section on period- icals above. Sociology/Economics (Table 4) is added because it displayed the greatest amount of inflation during the study interval. All figures are solely for United States publications; base year is 1967. In contrast to periodicals, mono- graph costs are not cumulative. How- ever, losses due to inflation can be cal- culated for each year and summed over the selected interval to obtain a total loss figure for each fund. A constant TABLES 4-7 INFLATIONARY EFFECTS oN AcQUISITION OF SELECTED MoNOGRAPHS TABLE 4 TABLE 5 SociOLOGY /EcoNoMics SciENCE Year (1) (2) (3) Year (1) (2) (3) 1967 100 1967 100 1970 153 .65 -35 1970 115 .87 -13 1971 216 .46 -54 1971 123 .81 -19 1972 209 .48 -52 1972 124 .81 -19 Total -141 Total -51 TABLE 6 TABLE 7 .ART AvERAGE Year (1) (2) (3) Year (1) (2) (3) 1967 100 1967 100 1970 130 .77 -23 1970 138 .72 -28 1971 132 .76 -24 1971 157 .64 -36 1972 120 .83 -17 1972 154 .65 -35 Total -64 Total -99 ( 1 ) Price index (base year = 1967 ) . ( 2) Percentage of periodicals that can be purchased in terms of 1967 dollars. ( 3) Number of monographs per 100 lost to inflation ( 197(}-72). . Data from: "1972 U.S. Book Industry Statistics: Titles, Prices, Sales Trends," Publuhers' Weekly 203:49 (5 Feb. 1973). Table A-Index of Prices (Per Volume) of Hardcover Books, By Category, 1967 and 197(}-72. 196 I College & Research Libraries • May 1974 monograph budget is assumed in the in- terpretation made below. 1. Inflation .as measured by the price index is variable. The general trend is upward, but actually showed a slight decline in most dis- ciplines between 1971 and 1972. Only Science showed an increase in cost from 1971 to 1972, but it was not significant. The average monograph cost somewhat less in 1972 than in 1971. 2. Although all disciplines suffered, the effect of inflation differed dra- matically between fields since 1967. Sociology I Economics could pur- chase slightly less than half the number of monographs as in the base year, whereas Science and Art could purchase about 80 percent. 3. In terms of monographs lost ( un- purchased) due to inflation, Sci- ence . fared best (ca. 50 percent fewer losses than average); Art, next ( ca. 35 percent fewer losses than average); and Sociology I Eco- nomics, much worse (·ca. 40 per- cent more losses than average). Inflation has taken a toll which varies both with form of publication and from .one discipline to another. Given constant monograph or periodical bud- gets for each funding unit, a collection that was in balance in 1967 is now great- ly out of alignment. Proportional dis- tribution of library-resource units must be translated into a proportional distri- bution of the book budget by annually taking into account current inflation fig- ures. In ·practice monograph inflation ad- justments would lag a year behind those for serials. Because serials acquire cum- ulative costs, serials budgets must be ad- justed for inflation and projected for the coming year. The degree of accu- racy in the serials budget estimate de- pends, of course, on the ability to an- ticipate the rate of serial inflation for each funding unit. This rate must be estimated. For monographs no inflationary pro- jections are ordinarily made. Though generally upward, the rate of inflation of monographs is variable from year to year, even decreasing for some disci- plines in some years. A projected esti- mate would, therefore, be tenuous at best. In fact, it is unnecessary. To de- rive an inflation correction for mono- graphs there is no reason why data for the current year could not be compared with data for the past year. The following example introduces a method for taking inflation into ac- count when balancing serial with mono- graph acquisition by funding units. The example is purposely ·kept simple for reason of illustration. 1. Let · the total number of library- resource units to be acquired be di- vided equally between two fund- ing units (A, B ) . Fund A consists of 25 percent monographs/ 75 per- cent serials; and fund B, 75 per- cent monographsl25 percent seri- als, thus: A B M 25 75 s 75 25 Because each fund is 50 percent of the total, the total balance is shown as: A M 12.5 s 37.5 50.0 B 37.5 12.5 50.0 50 50 100 2. Let the following rates of infla- tion apply: fund A-serials= 20 percent, monographs = 20 percent; fund B-serials = 10 percent, mono- graphs = 10 percent, thus: A B M 20 10 s 20 10 Multiplying these inflation percent- ages by the figures for total hal- ance ( 1 ) yields: A B M 2.5 3.75 6.25 s 7.5 1.25 8.75 10.0 5.00 15.00 Note that for every 100 library-re- source units fund A loses 10 and fund B loses 5 for a total of 15 library-resource units lost to infla- tion. Fund A will lose 100 percent more library-resource units than fund B, thereby upsetting the 50- 50 desired balance unless inflation corrections are made. 3. Multiplying the total inflation loss ( 15 percent) by the percentages given in ( 1) yields the following: A B M 1.9 5.6 7.5 s 5.6 1.9 7.5 7.5 7.5 15 Instead of a 10 and 5 loss as in ( 2) , above, each funding unit should lose 7.5 library-resource units. Fund A, therefore, wiil lose 2.5 li- brary-resource units less; and fund B, 2.5 library-resource units more, if no inflation correction were ap- plied. 4. Subtracting corresponding figures in ( 3 ) from those in ( 2) yields: M s A B .6 -1.9 1.9 - .6 2.5 -2.5 -1.3 +1.3 0 5. By addition, applying the correc- tions in ( 4) to the first total bal- ance ( 1 ) produces the following corrected model: A M 13.1 s 39.4 52.5 B 35.6 11.9 47.5 48.7 51.3 100.0 By allocating the book budget ac- cording to these figures the desired balance of library-resource units Allocating the Book Budget I 197 ( 1 ) will be maintained. Both funds will receive 7.5 library-re- source units less, thereby retaining the initial 50-50 distribution. Note that 52.5 percent of the total bud- get would be spent by fund A as op- posed to a 47.5 percent allocation for fund B. Slightly more would be spent for periodicals than monographs in this particular case. Application All decisions that determine the pro- portions in the proposed model are, in fact, currently being made in academic libraries. For the most part they are im- plicit and uncontrolled. However, all the data are generally available and may be compiled in such a format as to be used in the model. To demonstrate how the balancing model could be applied to a university library, ten funding units, each servic- ing a particular discipline-oriented col- lection of the Portland State Universi- ty Library, were investigated. This ex- amination showed that the model to control collection development can be realistically applied to an academic li- brary and that disparities . exist between actual collection development in terms of library-resource units and apparent collection development based on dollar allocations. Data were compiled for monographs on the basis of orders placed mainly during the first half of fiscal year 1972- 73. As sufficient monographs had already been ordered, a significantly large sam- ple was available to calculate an average price per item (Table 8). Also listed in Table 8 are Bowker's 1972 prices cited in Publishers' Weekly. A comparison of the Bowker figures with those generated internally reveals substantial discrepan- cies. Only for two fields-Physics and Psychology-are the figures comparable. Only in one field-Art-have books on the average cost substantially more than the Bowker average. In all other disci- 198 I College & Research Libraries • May 1974 plines average costs of books purchased by Portland State have been substantial- ly less expensive than indicated by Bow- ker. Sociology and Economics book costs have been roughly half of the average cost cited by Bowker. For whatever rea- sons these discrepancies arise, it appears that the Bowker average is inapplicable to this particular book budget. It is con- cluded, therefore, that inflation correc- tions can be best determined from inter- nal data; that is, by comparison of aver- age cost increases or decreases for each funding unit from one year to another. Utilizing internal costs to set inflation corrections introduces a weighted vari- able into the budget allocation process. This factor may work to the detriment of desired collection development, but if properly controlled is potentially beneficial. If, for example, a number of very expensive items were purchased out of a particular fund, this would boost the fund inflation correction. In the following year-assuming a con- stant budget-this fund would benefit at the expense of other funding units in terms of actual dollars available to spend. Ostensibly, the fund would be rewarded for making expensive pur- chases; yet in reality this fund's propor- tion of the total book budget in the model would be unchanged, and the fact that each serial subscription would become a permanent commitment by its fund should encourage discrimination in ordering. Without this weighted vari- able an inordinately expensive serial would seriously limit the total number of library-resource units that fund could acquire, thereby stunting develop- ment of the collection it supports. This example balances the impor- tance of a very expensive commitment to a particular community of patrons against total library needs. With a limit- ed budget such considerations must be made whether the model is adopted or not. The model merely provides a mech- anism for distributing the costs of such burdens. From the data in Table 8 an estimat- ed number of monographs ordered/ to be ordered in fiscal year 1972-73 was cal- culated for each funding unit. These values along with the number of an- nual/ irregular serials and periodical subscriptions for each funding unit were used to construct Table 9. All fig- ures are percentages of total library-re- source units acquired by the ten fund- ing units analyzed. It is a working stan- dard based on fiscal year 1972-73. The growth rate of the collection in 1972/73 was highly variable from one form of publication to another and especially TABLE 8 CoMPARISON OF PoRTLAND STATE UNIVERSITY LmRARY AVERAGE CosT/MoNOGRAPH ( 1972173 FIScAL YEAR) WITH BoWKER's FIGUREs ( 1972 CALENDAR YEAR) Allocation Average Cost Average Cost Fund Sample Size (Monographs) (PSU) (Bowker) 0 D ifference Applied Science 234 $2,763.56 $11.85 $16.11 -$4.26 Art 241 4,353.79 18.07 14.94 + 3.13 Business Administration 397 3,025.24 7.62 12.45 - 4.83 Economics 339 3,221.12 9.50 16.93 - 7.43 Education 578 3,306.97 5.72 10.26 - 4.54 English 762 6,950.17 9.12 12.03 - 2.91 History 723 8,481.22 11.73 14.92 - 3.19 Physics 219 3,554.81 16.23 16.05 + 0.18 Psychology 301 3,076.78 10.22 10.44 - 0.22 Sociology 422 3,616.11 8.57 16.93 - 8.36 o Data from: "1972 U.S. Book Industry Statistics: Title, Prices, Sales Trends," Publishers• Weekly 203:49 (5 Feb. 1973). Table A-Index of Prices (Per Volume) of Hardcover Books, By Category, 1967 and 197Q-72. >( 'y Allocating the Book Budget I 199 TABLE 9 MoDEL BAsED ON TEN FuNDING UNITS FOR THE PoRTLAND STATE UNIVERSITY LmRARY Serials Fund Monographs ( Annnal/Irregular) Periodicals Total Applied Science 3.9 0.3 2.3 6.5 Art 3.6 0.3 1.1 5.0 Business Administration 6.6 2.7 5.6 14.9 Economics 5.1 1.1 2.1 8.3 Education 10.3 1.2 3.5 15.0 English 10.4 1.1 3.6 15.1 History 10.6 1.4 1.8 13.8 Physics 2.9 0.9 1.6 5.4 Psychology 6.3 0.5 1.8 8.6 Sociology 6.3 0.3 1.2 7.8 TOTAL 100.4 0 Total percentages ( 100 percent) slightly inaccurate because fund figures rounded to one decimal place. among funding units. It is obvious that Business Administration, Educatimi, En- glish, and History actually obtained a much greater percentage of library-re- source units than Applied Science (En- gineering), Art, and Physics. This pat- tern of collection development was com- pletely unexpected when compared to the relative distribution of money among these funds. CONCLUSION The proposed model does not, of course, solve budgetary problems. Through appropriate inflation correc- tions, however, it provides a mechanism for equitable distribution of book budget funds. An ancillary benefit is that it clearly maps the general direc- tion in which the collection is develop- ing and allows better control over col- lection development. Given this model, adjustments necessary to steer the collec- tion toward selected goals could be iden- tified and made despite fluctuation in the total book budget. The model is also flexible in that it can incorporate any number of funding units (e.g., Physics, English) and subdivisions thereof (e.g., periodicals, monographs). REFERENCES 1. One volume as defined by and reported to the U.S. Office of Education in the annual Higher Education General Information Sur- vey, or one reel of microfilm or eight Micro- cards or microfiche as reported on the same survey. For reporting purposes, a volume is a physical unit of any printed, typewritten, handwritten, mimeographed, or processed work contained in one binding or portfolio, hardbound or paperbound, which has been classified, cataloged, and/or otherwise pre- pared for use. Include bound periodical vol- umes . Include government documents that have been classified and cataloged, counting as a volume such material as is contained in one binding or portfolio. 2. It is recognized that although funding units will, for the most part, be defined by dis- ciplines, exceptions are necessary. For exam- ple, those acquisitions that are necessary but so interdisciplinary as to require a special fund (e.g., General Fund) and those more reasonably grouped by form of publication (e.g., Newspaper Fund).