College and Research Libraries GILBERT W. FAIRHOLM Essentials of Library Manpower Budgeting Libraries in state-supported institutions find their quest for legislated funds increasingly competitive. New devices are needed .to substantiate their claims for higher budgets. To meet one of these needs, the State University of New York has prepared a library manpower budget formula for its several campuses by developing weighted standard times for accomplishing library operations in various kinds of institu- tions and correlated with various library characteristics. The rationale and methodology of the formula are described. This paper was read to the New York Library Association, College and University Section, in Rochester, on May 3, 1969. BuDGETING FOR HIGHER EDUCATION in state-supported institutions presents unique problems of policy and method- ology. State institutions are dependent on the largess of the legislative body for the bulk of their financial support, and like other state supported public pro- grams, such as welfare and health, must articulate their needs both in terms of academic programs and in a format un- derstood by the governmental budget office. The need for state budget support has been increasing exponentially since 1945. Larger enrollments, expanded functions, inflation, and broadening of the educational objectives of individual institutions are typically part of the man- agement dilemma of harassed universi- ty administrators. Legislators find their involvement with allocation of limited resources, including the exploding high- er education needs, a frustrating one. Their concern has typically been ex- Mr. Fairholm is Principal Budget Exam- iner in the New York State Division of the Budget, Albany. 332/ pressed in terms of requests for more detailed "justification" of academic budg- ets, and to more precise, quantifiable, ob- jective data and methodology. Library administrators have been caught in this squeeze. Often they have been hard-pressed to produce informa- tion in quantitative terms that at the same time reflected accurately their op- erations, activities, and needs. The quan- titative data they have produced has led to the development of formulas of vari- ous kinds to consolidate objective mea- sures of performance. Along with bene- fit-cost analysis and related techniques, these formulas provide both university administrators and state legislatures with a common language and a common knowledge-base upon which to negotiate needed program budgets in an ever-in- creasing competition for limited re- sources. A formula basis for budgeting in uni- versity libraries has many advantages. Formulas are simple and direct, and they emphasize critical measures and provide an objective method of relating one cam- pus to others within a state higher-edu- r cation complex. They reduce the bulk of paper needed, focus on key issues, fa- cilitate comparisons from campus to cam- pus and from year to year, and reduce the element of risk in decision-making ( risk of making mistakes and risk of pro- ducing conflict among competing ele- ments). These advantages, however, are rele- vant only insofar as the formula adopted reflects accurately the real-world situa- tion at campuses. The "multiversity" is becoming the norm in state-supported higher education, and differences among campuses in terms of basic descriptive characteristics must be built into the formula "model" if it is to be of use in a complex multiinstitutional university. En- rollment differences in the academic program-mix, graduate and honors pro- grams, special research goals, and other descriptive characteristics of individual campuses must be considered and prop- erly accounted for in the development and use of budget formulas for higher education libraries. The advantages of a formula that does account for these vari- ables are obvious. This discussion will focus on a formula basis for library manpower budgeting. Library manpower can probably be cor- related to quantitative measures more easily than any other factor of library operation. Especially in large-scale, mul- ticampus academic institutions, library staffing presents problems of rationaliza- tion and equality that require some ref- erences to "standards" or formulas, to some common ground upon which valid decisions and comparisons can be made. When standard-setting authorities in the university fail to set standards in quan- titative terms, budgeting authorities are compelled to do so. The resulting stan- dards usually have the virtue of simplic- ity, but they often fail to meet the real needs of both the institution and the cen- tral management. A unique joint effort at standard-set- Library Man power Budgeting I 333 ting for library staff has been conducted by the New York State Division of the Budget and the State University Sys- tem (SUNY). Formulas to assure ade- quate staffing for the burgeoning librar- ies were seen as the logical solution to this problem. However, formulas had to be developed which recognized widely different organizational and procedural characteristics of over thirty campus li- braries, and which accommodated the decentralized management philosophy of SUNY. As a first step, available rules of thumb and formulas developed for use in other libraries in the United States were analyzed. 1 They were found to be unacceptable for New York State pur- poses because of the unique place of li- braries in the rapid growth phase of SUNY libraries. Accordingly, an attempt was made to develop formulas which accommodated the key variables of li- braries in the system. The result, while admittedly not the last word in formula- based library manpower forecasting, pro- vides a quantifiable basis that could re- solve many key staffing decisions in this area. EssENTIAL AsPECTS Essentially the formula which was de- veloped relies on a correlation of work effort in libraries to five forecastable characteristics (or "descriptors") of an academic library. These five descriptors include holdings, acquisitions, full-time equivalent users, headcount students, and full-time equivalent faculty and se- lected professional staff of the univer- sity. Work effort was defined by refer- ence to an abstract prescriptive model that utilized essential library functions and quantified output in a series of "standard times" related to each unit of the five characteristics. Totals of these five characteristics and their standard output units can be forecasted for plan- ning and budgetary purposes and to- 334 I College & Research Libraries • September 1970 gether are indicative of library activity. The library descriptors reflect two characteristics-holdings and clientele. An academic library's manpower re- quirements are directly related to the size, character, educational goals and mission of the institution it serves. These institutional characteristics are reflected in several quantitative factors that de- scribe tlie library: 1. Holdings-In any library, the size of the collection directly affects staff. Stated simply, large libraries require more staff to perform tasks directly related to books on shelves than small libraries. 2. Acquisitions-Acquisition of new ma- terials has a direct impact on the volume of technical services. The larger the number of volumes added, the larger the staff requirements. The larger the potential clientele of the library, the more staff is required to service it. In academic libraries, users are defined as students, faculty, staff, and community. Users affect the library on three levels. 3. FT E Users-The total student pop- ulations of the institution served by the library (including extension pro- grams) equated to full-time equiva- lent students ( FTE) plus the full- time equivalent faculty users consti- tute a user category to which signifi- cant library functions are directly re- lated. 4. H·eadcount-The total number of ac- tual students enrolled in an institu- tion affects the extent of library ser- vices needed. Whether a student takes one course or a full academic course load, his potential use of certain li- brary services is the same. 5. Faculty and Staff-Often the aca- demic faculty and the university pro- fessional-level staff are ignored as po- tential users of the libraries resources . These individuals actually place iden- tifiably special loads on the resources of the library and should be specially considered in any accurate descrip- tion of library users. Implied in these characteristics are the academic and program faculty, and the scope of curriculum and degree programs, which help describe the institution it- self and, therefore, affect the library. For comparative purposes, these five characteristics most accurately describe the nature of the library. The character of the holdings may also be important; but, except for certain obvious situations (foreign language collections, for ex- ample), size of holdings is a more real- istic basis for comparison of library man- power needs. 2 Also, the bulk of most collections is represented by traditional holdings. More importantly, it was found that each of the common library func- tions was directly relatable to one of these five characteristics. Reference has been made to the re- lationship of the descriptors and the com- mon functions of libraries. The functions of library staff (both professional and nonprofessional) that make up the busi- ness of librarianship were synthesized in this study and a listing was developed that represents the central nature of li- brary activity. Since this listing empha- sized what must be done in academic libraries, rather than how it is done, it was possible to ignore the difference in organizations and procedures of individ- ual libraries and thus deal with the heart of librarianship, not its outward tech- nique. In order to produce this kind of a listing of intrinsic library functions , a model of library operations was first de- veloped. This model, an abstract repre- sentation of library operations, focused on the essential functions and responsi- bilities of libraries, and was displayed in a series of linear flow diagrams. These flow diagrams related activities and tasks necessary to achieve certain events or conditions conducive to processing books or responding to user demands. These diagrams were validated by in- ternal checks of consistency and logic and by a panel of senior library profes- sionals.3 uNITS OF MEASURE Once developed and validated, the list of common library functions was used to prepare a list of 59 units of mea- sure of library activity. These units of measure, reproduced in column 1 of Ta- ble 1, are the countable output of library activity and effectively clump the com- mon functions into countable units. There are fewer units of measure than there are library functions. This is be- cause completion of one unit of measure may require library staff to perform sev- eral functions. Thus, the units of mea- sure actually summarize related library functions. For example, when we say it takes x minutes to charge out a book, we really mean it takes a total of x minutes to charge out the book, place a slip in the circulation files , and, if necessary, renew the book and refile the slip. The units of measure then represent clusters of individual functions. uNIT STANDARDS Using the 59 units of measure as a framework, . a median time (in minutes) to perform each unit of measure was de- veloped statistically through a question- naire and interviews. In order to relate similar facilities and make the statistics more useful, the libraries in the sample were broken down into three classes, corresponding to the different missions of SUNY libraries, i.e. , universities, lib- eral arts colleges, and agricultural and technical colleges. 4 Separate standards were developed for each class. To refine further the statistical data collected, a series of workshops of librar- ian specialists was held in which they were asked: 1. to refine the list of typical library Library Man power Budgeting I 335 functions and the units of measurable output; 2. to develop standard times to perform these key work load units of mea- sure; 3. to modify, as necessary, these stan- dard times, taking into consideration factors that go to make up a good library operation. Throughout the workshops the em- phasis was on producing high quality standards that reflect good librarianship and not merely "average current per- formance," that may or may not reflect good practice. The statistical bases were used as a check against the standards and to re- late developed standard times to cur- rent real-world situations in the librar- ies. As a result of these workshops, Unit Standard Times were produced which reflect (for SUNY at this stage of its de- velopment) the standard of performance of library manpower in terms of staff time (in minutes) per unit of output (see column 3 in Table 1). Separate stan- dards were produced for each of the 59 units of measure for each classifica- tion of library. APPLICATION FACTORS It will be readily seen that Unit Stan- dards alone are not sufficient for fore- casting library staff needs. While they provide an effective measure of indi- vidual performance of a given task, they do not reflect total manpower needs since not all functions are performed for every book or for every potential user. The relationship between functions per- formed and manpower needs was math- ematically derived by relating actual performance of specific functions to one or another of the five key characteristics isolated previously (i.e., holdings, acqui- sitions, FTE users, headcount, and facul- ty and staff). Common library functions were grouped according to these basic characteristics and a mathematical fac- 336 I College & Research Libraries • September 1970 tor derived by dividing the appropriate output data (by class) by the appropri- ate key characteristics. Data used came from the sample of eleven libraries for which data were collected. The result- ant factor represents the frequency with which a given function is performed on an average unit of each of these five characteristics. Column 2 of Table 1 lists these applications factors. WEIGHTED STANDARD TIMES Once the standard time for perform- ing a given function had been deter- mined and the frequency of its applica- tion in the library to a given characteris- tic had been established, simple multi- plication produced a "weighted stan- dard time," or the average time re- quired to perform the function (or clus- ter of functions) for each unit of each (Column 1) Units of Measure Function TABLE I WEIGHTED STANDARD TIMES CLASS I (Ag. and Tech.) LIBRARIES Unit (Column 2) Application Factor (Column 3) Standard Times (Min.) Technical Services Standards-Based on Volumes Added Selection of monographs & sets Titles .759 3.0 Selection of series Titles .07 4 5.6 Search of holdings Searches 1.494 3.0 Typing of cards Number typed 5.000 3.0 Typing of add-ons Add-ons .060 2.4 Typing of cross-references Cross-references · .100 3.0 Revision of cards Titles .100 2.4 Total Weighted Time Technical Services Standards-Based on Total Holdings Inventory Binding Withdrawal of materials from collection Guidance to reader Reference questions Interlibrary borrowed Volumes .906 1.2 Volumes .015 18.3 Volumes .017 15.0 Total Weighted Time Reader Services Standards-Based on FTE Users Referrals 6.116 1.0 Questions 6.790 20.0 Number borrowed .120 42.0 Total Weighted Time Reader Services Standards-Based on Head Count Charge-outs of materials Volume 6.993 2.0 Return of materials Volumes received 9.154 4.2 Preparation of overdue notices Overdues 1.141 7.2 Total Weighted Time Reader Services Standards-Based on Faculty & Staff Abstract prepared Bibliographies prepared Interlibrary loans Abstracts .050 15.0 Bibliographies .200 450.0 Number loaned .250 22.0 Total Weighted Time (Column 4) Weighted Standard Times (Min.) 2.277 .414 4.482 15.000 .144 .300 .240 120.348 min. 1.087 .274 .255 3.627 min. 6.116 135.800 5.040 157.295 min. 13.986 38.446 8.215 60.647 min. .750 90.000 5.500 96.250min. of the five key characteristics. The total of all of these average times (i.e., Weighted Standard Times) indicates the staff time required for each unit for each of the five basic library descriptive char- acteristics. These totals, when multiplied by the forecast quantities of each char- acteristic for a given year, indicate the total staff time required for that library to perform routine library operations (see Column 4, Table 1). Table 1 summarizes this data for Ag- ricultural and Technical Colleges within the SUNY System in 1968. In column 1 the subdivisions list the units of measure isolated from all library functions. Col- umn 2 shows the application factor or the average frequency of performance of a given unit of measure within SUNY li- braries. Column 3 carries the unit stan- dard time or the established time re- quired to perform each unit of measure and column 4 has the weighted stan- dard time or the average staff time re- quired to perform each unit of measure in terms of each of the five key char- acteristics of a library. These totals con- stitute the standard requirement for li- brary manpower for each increment of the five characteristics. Thus, for each unit of holdings, acquisitions, weighted users, head count, students and faculty and staff forecast by a library, the total staff time (in minutes) is shown as re- quired to properly staff the basic func- tions of that library. ADMINISTRATION To this point library management has been left out of the data collected as a basis for the standard times. Administra- tive overhead for the total library has been considered separately. A percent- age factor was developed from the sta- tistical data and modified by the work- shops to produce a management factor ( or a manpower p ercentage ratio) for Library Manpower Budge.ting I 331 each class of library, i.e., I, university; II, liberal arts college; and III, agricul- tural and technical college. These ad- ministrative overhead rates are included in Table 2. TABLE 2 AnMINISTRA TION STANDARDS-BASED ON PERCENT OF TECHNICAL SERVICE AND READER SERVICE TIMES BY CLASSIFICATION Administrative Functions Interlibrary coopera- tion Professional meetings Committee assign- ments Personnel administra- tion: (recruitment, training) Percentage of Total Library Staff Time Devoted to Adminis. Functions Class Class Class I II III Budget administration 20.18% 14.22% 10.00% Reporting and statistics Public relations Mail Control Computer services and applications Planning and man- agement Other administration FORMULA A simple formula has been created to relate this quantitative data in a mean- ingful way, and to facilitate budget fore- casting for library staff. This formula will produce a figure that represents the total staff needs of a library for routine performance of library functions. The figure includes temporary services, staff, student assistants, and other staff mem- bers. The formula involves four steps. 1. Technical Service Man-Year Require- t ( Ts ) _ Aa + Bb men s - 96 ,000 min. 2. Reader Services Man-Year Require- ments ( RS ) = Cc + Dd +. Ee 96,000 mm. 338 I College & Research Libraries • September 1970 Total Weighted Standard Times Where: Class Class Class Descriptors I II III A Forecast Holdings a 3.627 2.768 2.437 B Forecast Volumes Added b 120.348 120.831 142.258 c = Forecast Weighted Users c 157.205 120.308 165.161 D Forecast Head Count Students d 60.647 139.214 207.517 E = Forecast Faculty and Staff e 96.250 148.156 174.174 Where: X = % of TS & RS STAFF Required for Administration 20.81% 14.22% 10.10% And Where: Number of minutes per man-year = 96,000 (one man-year = 1,600 hours x 60 min. = 96,000 min.) 3. Administration Man-Year Require- ments (Adm) = S (TS + RS) 4. Total Library Man-Year Require- ments = TS + RS + Adm LIMITATIONS OF THE FORMULA It must be kept in mind that the for- mula represents staff needs for routine continuing library functions and not necessarily for the total activity of a li- brary. As with any standard or formula methodology, this method of determin- ing library manpower is useful only for those operations common to all libraries in the SUNY system. Unique features of an individual library must be considered separately from the formula. Analysis will show that a few library operations have been omitted from the units of measure. Unique physical char- acteristics of the individual library and special programs of the library or the academic institution also affect library operations and therefore staff needs. These factors, described in more detail below, may require staff in addition to the formula computations. Nevertheless the formula was formed to reflect 85 to 95 percent of all library functions in SUNY and therefore provides a valuable and highly efficient tool for both analy- sis and comparison of libraries. SPECIAL FUNCTIONS A few library functions do not relate directly to any one of the five key char- acteristics identified. They should, of course, be reflected in final requests for library manpower, but they must be jus- tified outside of the formula technique. Some of these functions are common to all or most libraries while others are found in only one or a few institutions. Examples include: machine maintenance and operation, reproduction of general materials, selection of audiovisual ma- terials, and security guard functions. Unique and unusual physical char- acteristics of the environment may have an impact on library manpower needs. Factors which may require manpower beyond mere output levels are: several manned service points on different floors or in different buildings, branch or di- visional libraries, and unusual or exten- siv~ hours of opening. 5 Special programs of the library may require staff beyond the formula levels. These include: responsibility for audio- visual materials, special collections, and special research facilities. The academic programs of the parent institution generate activities in academ- ic libraries. Unusual programs and pro- grams not found in other institutions may occasionally generate an unusual library manpower need. Among those to be con- sidered are: extraordinary foreign lan- guage/ subject matter programs and un- usual research programs. These unique or extraordinary circum- stances may require separate justifica- tion, and when approved will necessi- tate increases in the formula staffing pat- terns. The flexibility thus assured pro- tects individual libraries from a strict uniformity of staff that would stifle ef- fective response to special and evolving situations both in librarianship and in the Library Manpower Budgeting I 339 parent institution. Given this ability to deviate from formula standards, the stan- dards themselves become more mean- ingful as a tool in library manpower de- terminations for higher education in the twentieth century. The library is both an educational in- stitution and a business operation. Man- agement of the library must include evaluation of performance, the fulfill- ment of responsibilities, the accomplish- ment of purposes, the effective use of available resources, and the attainment of long-range objectives. These manage- ment activities must be interwoven in the constant process of serving the reference and research needs of the university. A formula-based realistic and flexible man- power forecasting system such as the one described here facilitates both academic and management goals in the library and within the total university complex. REFERENCES 1. Examples of these 1ules of thumb in- clude the American Library Association «standard" of budgeting for libraries as ccfive percent" of the total university budget. More specific rules of thumb for staffing had been used in certain areas of library operation. For example: a cataloger can process approximately 800- 850 books per year. While they may be pertinent to a specific library or case in point, there is no justification, statistical basis or other rationale, that would sup- port or substantiate their use in the SUNY library system or elsewhere. 2. A few library functions, while common to most libraries, showed no direct cor- relation with one of the five key char- acteristics. (Reproduction of general ma- terials is an example.) These library tasks are excluded from the formula and must be included as a separate treatment of any unique aspects of the total collection. 3 . The Chancellor's Advisory Library Com- mittee reviewed and modified early drafts of these models. Throughout the study, librarian specialists participated by completing questionnaires, discussing their operations with the team, and at- tending workshops. 4. Library studies were analyzed accord- ing to ten descriptive characteristics: holdings, acquisitions, students, total users, faculty, academic programs, ILL materials borrowed, ILL materials loaned, bibliography preparation, and circulation rates. When the characteris- tics for each library were arranged in rank order, they clustered into three groups that reinforced and verified the 340 I College & Research Libraries • September 1970 present groupings of SUNY libraries ac- cording to institutional class-i.e., agri- cultural and technical colleges, liberal arts colleges, and university centers. 5. Under the formula a certain level of manpower can be established based on output or productivity of library staff. Conditions like those shown may re- quire an additional segment of library staff time merely to man an open sta- tion. The incremental difference here must be computed in addition to the formula. For example, a reference sta- tion may require, on the basis of output forecasts, five man-years. The policy on hours of opening mandates that the ref- erence desk be manned for a total of six man-years of time. The one-man-year differential is required on the basis of the policy of hours of opening, not on work load, and should be justified according- ly.