College and Research Libraries - - - --- - ---------------------------------., JOEL A. NACHLAS and ANTON R. PIERCE Determination of Unit Costs for Library Services As with other public service activities, inflationary trends and public opin- ion provide a clear mandate for attempts to control the increasing costs of providing library services. Cost control necessarily requires knowledge of the quantities and sources of costs. A methodology ; known as microcosting, for identifying the unit costs of providing specific services is presented here. The method is designed to enable library managers to identify at a detailed level the resources consumed in providing a particular service. This infor- mation provides'' a quantitative basis for and a monitor of library manage- ment decisions. To illustrate the use of the methodology , it is applied to the determination of the unit costs of tracking overdue materials at a major university library. THE RAPID INCREASE in the costs of provid- ing library services over the past several years is well known. 1- 3 The causes of this increase are difficult to isolate but hold the key to reducing the rate of growth in library costs. While it is a vast undertaking to iden- tify all of the economic, social , and technological factors that have contributed to cost increases, one cause that can be recognized is the lack of impetus for the de- velopment ~nd use of effective cost con- tainment methods. The provision of library services has been viewed by universities as essential for in- structional and research programs and by municipalities as required for constituent cultural opportunity. Library services have been viewed in the same light as medical services. A person who needs medical care asks if it can be provided rather than what it will cost. Universities have thus tended to ask if "adequate" libraries and library ser- vices can be provided rather than how they can be provided in a truly cost-effective manner. This approach may have been ap- joel A . Nachlas is assistant professor of opera- tions research and industrial engineering at Virginia Polytechnic Institute and State Univer- sity , Blacksburg . Anton R. Pierce is planning and research librarian at VPI & SU. 240 I propriate in the past as the financial re- sources for providing library services have generally been available. Recently universities and municipalities have found that their financial resources are more constrained, that there is greater competition for those resources, and that increasing costs for all programs have re- sulted in a decline in what can be obtained with their resources. As a result , libraries have found that funds are harder to get , must be accounted for more carefully, and still buy less. Thus libraries have recognized the need to manage their resources more carefully and to attempt to allocate those resources more efficiently. The key to managing library resources is knowing what the costs of providing library services are . A method that can be used to isolate the costs of providing specific ser- vices is discussed here. An example of how this method has been applied to the iden- tification of the cost of pursuing overdue materials in a university library is de- scribed. Finally, the potential for applica- tion of the methodology to other library service activities is discussed. COST DETERMINATION Cost determination can be divided into two categories, macrocosting and microcost- ing. Both types of cost analysis have been used extensively in production environ- ments in industry. 4 During the past few years, they have also been applied to many service organizations, most notably in hospi- tals . 5 The purpose of macrocosting is to exam- ine a system such as a library in order to determine its present costs . and from these costs to estimate trends and make infer- ences concerning future system costs. For libraries , this could prove to be a very fruit- ful endeavor and appears to have been at- tempted on a limited scale within the con- text of budget preparation. Nevertheless , there is substantial opportunity to pursue macrocosting studies in libraries especially in the areas of long-range planning and of evaluation of proposed program or service changes. The purpose of microcosting is to obtain measures by which specific management decisions may be judged. Microcosts can · be used to evaluate specific changes within a library system. This is because microcosts focus upon fundamental "unit costs" and can be used to d etermin e the cost of providing a specific service or even a single unit of that service. The microcosting technique is explained here by example. For illustration , the appli- cation of th e method to the handling of overdue materials is described step by step . The results of this application are described , and the extension of the method to other li- brary services .is discussed. MICROCOSTING The underlying framework for microcost- ing is conceptually simple. The first step is the construction of a flowchart describing the specific tasks and events that constitute the provision of a particular service. Figure 1 shows an example· of a flowchart for track- ing overdue materials using the Virginia Polytechnic Institute and State University (VPI & S U) automated Circulation and Finding System. 6 The most important attribute of the flow- chart is that it represents all of the events that may occur in the tracking of an overdue item. Thus the flowchart schematically de- scribes the order in which the various tasks associated with tracking an overdue item Determination of Unit Costs I 241 may be performed. The possibility that cer- tain steps may be repeated is represented by a "feedback" or return loop. Once the flowchart has been constructed, it is used to design a work sampling study to identify the library resources that are used to track overdue materials. This work sampling study is described in detail below. First, the pertinent costs and the resulting motivation for the work sampling study are discussed. The total cost of providing a service is composed of four components: direct labor, indirect labor , materials and equipment , and overhead. Personnel time is a limited resource , the consumption of which carries an associated cost. The personnel time expended in the specific tasks that constitute the provision of a service is called the direct labor commit- m e nt to that service. The direct labor com- ponent of the cost of providing a service is the cost of the pesonnel time allocated to the provision of that service. The first step in determining the direct labor cost for a service is to use the flow- chart to help enumerate the personnel that participate in the provision of the service. In the case of the tracking of overdue mate- rials , the participating personnel are one li- brarian , one library assistant, two level B clerks , one level C cl e rk , on e lev el C clerk typist , and three leve l B cl e rk typists. All of these persons devote part of their time to tracking overdue materials. Identification of the cost of tracking over- due materials requires the determination of what portions of the personnel salaries are appropriately allocated to this service. This determination is made using work sampling. Indirect labor is classified in two distinct activity categories. The first of these is work that is not related specifically to any proce- dure or responsibility. Activities that fall into this category are answering telephones and cleanup. The second category is non- productive time . Nonproductive time in- cludes idle time , standby time , lunch breaks , and personal time for coffee and rest room breaks. The costs of indirect labor time must be apportioned among the vari- ous procedures or responsibilities of the personnel. The actual method of apportion- ment can be determined by existing man- 242 I College & Research Libraries • May 1979 COMPUTER INCLUDES ITEM OH OVERDUE LIST YES NO PRINT AND MAIL OVERDUE NOTICE YES UPDATE COMPUTER FILE EXCEPTION REPORT AND BILLING PROCEDURE RETURN '11/0 FINE RETURN 'II/ FINE BILLING PROCEDURE PATRON SEARCHES STACKS PROCESS SAYS RETURNED FORM YES UPDATE COMPUTER FILE Fig . 1 Flowchart for Overdue Materials Tracking Tasks agement policy or can be based upon direct labor allocations. In the overdue materials study, indirect labor times were distributed to the task of tracking overdue materials in proportion to the direct labor expended on that activity. Current materials costs can be obtained from the purchasing agent. In cases where materials costs fluctuate, a moving average can be used to estimate costs of particular items. Equipment depreciation for items used in the performance of a task can be ob- tained from the accounting department. Care must be taken to allocate the deprecia- tion cost in proportion to its use for the given procedure. In the case of the tracking of overdue materials, the only significant equipment in use is the computer and the computer terminal. Computer-use charges include depreciation. The depreciation cost for the computer terminal is allocated in the proportion equal to the percentage of direct labor committed to tracking overdue mate- rials. Overhead costs are real costs and must be included in the calculation of the cost of providing a service. Overhead is divided into two parts, departmental and general. Departmental overhead includes the cost of providing supervision and support per- sonnel. These costs are specific to the de- partment but not to any given departmental responsibility. General overhead is the cost allocated to the department by the library and includes housekeeping , building depreciation , and utilities. Both types of overhead are specified by accounting , and both can be distributed over department responsibilities in propor- tion to the direct labor allocated to de- partmental procedures. At the VPI & SU li- brary , the circulation department is respon- sible for the tracking of overdue materials . However, by university policy , departmen- tal and general overhead rates are aggre- gated and valued at 63 percent of labor costs (direct and indirect) . All four categories of the costs of provid- ing a service have been· linked to the direct labor time expended in providing that ser- vice . It therefore remains to determine the direct labor commitment to the service. An effective method for doing this is work sampling. 7--9 WORK SAMPLING Work sampling is a technique of ran- domly observing personnel perform their defined tasks and inferring from those ob- servations a profile of the average labor allo- cations to the various employee respon- sibilities. By random observations is meant that the times at which observations are made are selected by use of a random de- vice (e .g., a table of random numbers) as opposed to their being scheduled at regular intervals . This randomization of observation times guarantees that representative and unbiased data are obtained. Determination of Unit Costs I 243 The key to the accuracy of the inferences is to make the observations without disturb- ing the work environment. Proper introduc- tion of the study and its objectives to the staff prior to execution of the study contrib- utes significantly to minimizing the distur- bance of the work environment . An im- portant aspect of this introduction is the emphasis of the focus of the study upon the cost of the service rather than upon individual worker productivity. To construct the work sampling study, a set of task cate gories is define d using th e procedure flowchart. In addition , general work categories are defined to represent tasks associated with services other than the one under study. Categories corresponding to indirect labor activities are also defined. Figure 2 shows the task categories used in the overdue materials study. Note that each task category is defined and has been assigned a code. To execute the work sam- pling study, observation times are selected randomly and recorded on the data sheet shown in figure 3. Then , at each of the selected observation times, a "snap shot" of the department is taken. By "snap shot" is meant that an observer locates each of the departmental personnel and enters on the data sheet the task code corresponding to the activity in which the person is engaged. Figure 3 shows an example data sheet in which task code e ntri es have be e n mad e . In the overdue materials study, 1,013 such individual data entries were accumulated. These data are then analyzed to determine the quantity of direct labor consumed in the provision of the given service. To analyze the accumulated data , the fractions of available employee time com- mitted to the various tasks are tabulated. Then , the following variables are defined: Tj the total available work time for employee j , t · J the direct labor time expended by employee j in the provision of the service under study, the total indirect labor time for employee j , r = the aggregate overhead rate (63 % in the present study), d the computer terminal depreciation rate , 244 I College & Research Libraries • May 1979 Task Codes Overdue Materials Tasks SS - stack search M - mailing overdue notices T - computer terminal RF - returns and fines E - exception reporting SR - says returned reporting C - complaints RP - running overdue programs 0 -other Non-Overdue -Related Tasks CHK- checkouts and check-ins HR - holds and recalls Res - reserve Sup - supervision UP - user problems 0-11 - other-11 Additional Designations P- personal L- lunch I- idle Task Definition for Overdue Materials Stack Search. Canvassing the shelves to be sure material listed as overdue has not been re- turned directly to the shelves by the borrower. Mailing Overdue Notices. Any activity requir- ing the handling and/or sorting of overdue notices for the purpose of mailing them. This is for the 8-, 15-, and 30-day notices that are automatically generated by the computer. Computer Terminal. Entering overdue update information obtained from stack search into the computer via the terminal or running the com- puter programs to generate the overdue materials list. Returns and Fines. Processing the returns and collecting fines submitted for overdue items. Exception Reporting. Typing or processing ex- ception reports for overdue materials. This task is specific to items more than 30 days overdue . Says Returned Reporting. Preparing entries for the "says returned" file. Complaints. Dealing with patrons who indicate that an overdue item has been returned . Running Overdue Programs. Running the computer programs to obtain listing for stack search. Other. Any other task that does not fit into one of the above categories but does relate to overdue materials. Any such entry should be explained on the back of the data form . Task Definitions for Responsibilities Other Than Overdue Materials Checkouts and Check-ins. Working at checkout desk. Holds and Recalls. Processing or working on holds and recalls. Reserve. Working on reserve book activities . Supervision. Directing or supervising student employees. User Problems . Responding to patron requests for assistance that are not related to overdue ma- terials . Other-ll. Any other task that does not fit any of the above definitions . Should be accompanied by a category of task on reverse side of the data form . Additional Designations Personal. Individual is involved in a personal activity such as a coffee break or a visit to the rest room . Lunch. Individual is on a lunch break. Idle. Individual is at a work station but is idle. Fig. 2 Task Categories e the equipment and materials cost of providing the service under study (compu ter costs for the present study), and cj = the salary rate for employee j . Then, with the use of these definitions, the total cost of providing the library service is expressed by equation 1, where N is the number of employees that participate in the provlSJon of the service. In this expression that total cost of providing a service is de- fined in terms of the direct and indirect labor expenditures that are, in turn, ob- tained from the work sampling study. RESULTS Table 1 shows the percentage of available employee time expended in the various di- rect labor activities associated with tracking (1) Determination of Unit Costs I 245 ANALYST JOE ~ s 7:42 8 :12 8 :35 8 :52 9 : 16 9 :38 9 :50 10 :11 I0:2E 10:48 JOHN CHK HR CHK HR HR HR HR p HR CHK (Clerk Bl MARY HR HR HR I RES RES RES p RES RES (Supervisor) BILL I T ss ss ss ss p RES CHK I (Clerk B) SUE RF RF SUP I UP M E E p p (Clerk-Typist C ELLEN I T ss ss T CHK CHK p CHK CHK (Clerk Bl TOM RES HR RES RES c SUP HR RES CHK SUP (Library Asst) JILL CHK CHK HR HR E E I HR M M (Clerk-Typist B RICK HR HR ss ss ss ss p CHK M M (Clerk C) Fig. 3 Example Data Sheet overdue materials and the resulting total di- rect labor fraction commitment to this ser- vice at the VPI & S U library. These figures represent the values . of the quantities t/fj and are presented as more informative than the actual tj values. These values are com- puted using the work sampling data ac- cumulated as are the corresponding frac- tions of indirect labor times shown in table 2 . . The entries in table 2 represent the total indirect labor fraction and the fraction of the available employee time consumed by indi- rect labor that is allocated to the tracking of overdue materials. As an example, note that for the super- visor 5.4 percent = [17.7/(100.0 - 23.5)] [23.5]. By adding the allocated personnel times listed in tables 1 and 2, the total frac- tion of available employee time allocated to the tracking of overdue materials is ob- tained. These allocations are shown in table 3. After obtaining computer costs from the computer log , terminal depreciation from accounting, and personnel salary rates from the personnel office, the total cost of track- ing overdue materials is computed using the expression defined above, and these costs are distributed to the overdue materials tracked. Using the tabulated data , it is found that the average cost of tracking TABLE 1 Activity Code ss M T RF E SR c RP 0 Direct Labor Fraction (tjffj) PERCENTAGES OF A VAILABLE EMPLOYEE TIME EXPENDED ON DIRECT LABOR ACTI VITIES FOR OVERD UE MATERIALS TRACKING Library Clerk Job Class Clerk Clerk Supervisor Assistant Typist B Typist C B 3.9 1.0 11.1 10.2 18.9 0.0 3.4 0.3 0.0 0.0 0.0 0.5 1.0 0.0 0 .0 0.0 2.9 2.8 0.0 7 .8 0.0 0.0 0.0 0.0 0.0 2. 1 0.4 0.0 0.0 1.9 0.0 1.0 0.3 0.0 0 .9 3.9 9.8 0.3 0 .0 0 .0 7 .8 2.4 1.1 0.0 4 .1 -- 17.7 21.4 16.9 10.2 33.6 C lerk c 1.0 0.0 0.0 0.0 0.0 0.9 1.0 0.0 0.0 2.9 246 I College & Research Libraries • May 1979 TABLE 2 PERCENTAGES OF A VAILABLE EMPLOYEE TIM E EXPE OED ON INDIRECT LABOR AND ALLOCATE D AS I DIRECT LABOR TO TRAC KING OvERD UE MATERIALS j ob C lass Supervisor Library Assistant Clerk Typist B Clerk Typist C Clerk B Clerk C Indirect Labor Fraction (Perc•en t) 23.5 25.7 30.6 25.9 20.6 23.8 Indirect Labor Allocation (Percent) 5.4 7.4 7.4 3 .6 8.7 .9 overdue materials at \PI & SU is $.22 per overdue it em. At VPI & SU this cost amounts to approximately $18,330 per year. Upon completion of the study, the stack search activity was examined. It was found that fewer than 3 percent of the items listed as overdue had been returned to th e shelves without being discharge d . As a re- sult of this finding, the stack search activity has been eliminated fi·om the set of overdue materials tasks routinely undertaken , Exam- ining the study results indicates that th e s.tack search activity is time-consuming. In fact , this task requires an average of approx- imately .61 pe rson-days per day . Eliminating the staek search frees this time fur allocation to other tasks. Including consideration of ~ob category, this time rep- resen ts $8,000 of the $18,330 annual cost of tracking O\ erdu e materials. Now it should h e noted that total library operating costs are not necessarily reduced by this amount. lns.tead, th e work avoided by eliminating th e stack search creates additional available employee tim e that is committed to other activities. A follow-up study is plann ed to dete rmine how much of this additional tim e is used to provide user sen·ices. TABLE 3 TOTAL FRA CTIO NS OF A\'AILABLE EMPLOYEE TIME ALLOCATED TO TRACKING OVERD UE MATERIALS j ob C lass Supervisor Library Assistant Clerk Typist B Clerk Typist C Clerk D Clerk C Total Lal>Or Fraction (Percent ) 23.1 28 .8 24.3 13.8 42.3 3.8 CONCLUSIONS The use of microcosting has permitted the iden~ification of the unit cost of providing a specific type of library service, the tracking of overdue items. Obvious uses for this in- formation are the formulation of overdue re- turn policies and overdue fine schedules. The unit costs can be combined with incen- tives for efficient materials use to develop operating policies that recover library costs while promoting more responsible utiliza- tion of library collections. It should be noted that the methods by which the indirect and overhead costs are obtained are specific to the particular library being studied. However, it is quite easy to computerize the data manipulation proce- dure , This permits easy alteration of cost computation formulas and can also be used as a format for refining the method de- scribed. The key point is that unit costs of providing user services can be identified using this method. The method described for microcosting does not require sophisticated equipment or highly trained personnel. In the overdue materials study, library employees with no previous exposure to the technique col- lected the data. The actual design of the study and the analysis of the data should be entrusted to an industrial engineer. Nearly all university libraries can find an industrial engineer on campus. Urban-based munici- pal libraries can locate an industrial engi- neer at a nearby industrial plant or through the state society of professional engineers. Thus the resources required to undertake a microcosting study are likely to be readily available. A variety of other library services, both routine and special, are well suited to mi- crocos.ting analysis. Reference services interlibrary loans, photocopying services: acquisitions and cataloging, and check-in and return services are all reasonable sub- j~cts for mi.crocosting studies. In fact, any hbrary service or activity that requires the use of personnel time can be investigated u~ing t.he microcosting technique. For any gtven hbrary , the method will have greatest utility in examining services over which some decision-making control is possible. In summary, a method has been de- scribed that will enable library managers to identify the costs of providing specific ser- vices. The method was illustrated by appli- cation to overdue materials but can be applied to nearly any library activity. In addition, the method can be standardized and loaded on a computer so that it can be applied throughout the library. The value of the method is that it provides a sound ap- proach to determining at a microscopic level how library resources are being consumed. The information derived can be used to make decisions or formulate library policy. Determination of Unit Costs I 247 One approach to taking advantage of the power of the method is to use it both before and after a decision or policy is im- plemented to determine the eflect of the change. For example, the impact of a very stringent overdue fine schedule could be determined by executing the overdue mate- rials study again after the institution of such a schedule. Thus microcosting is a tool that library managers, as they face increasing costs and shrinking budgets, can use to learn about "where the money goes" and how to stretch it further. REFERENCES l. William J. Baumol and Matityahu Marcus , Economics of Academic Libraries (Washington: American Council on Education, 1973). 2. Richard De Gennaro, "Austerity, Technology, and Resource Sharing : Research Libraries Face the Future," Library journal 100:917-23 (1975). 3. Office of the Executive Director of Univer- sitywide Library Planning, The Unicersity of California Libraries: A Plan for Decelopment, 1978-1988 (Berkeley: Systemwide Administra- tion of the University of California, 1977). 4. Clarence B. Nickerson, Managerial Cost Ac- counting and Analysis (New York: McGraw- Hill, 1962). 5. American Hospital Association, Hospital Engi- neering Handbook (Chicago, The Association, 1974). 6. Systems Development Department, Virginia Polytechnic Institute and State University, De- tailed Documentation of the Library Circula- tion and Finding System (Blacksburg, Va. : Virginia Polytechnic, 1978). 7. Association of Research Libraries, Determining Indirect Cost Rates in Research Libraries, SPEC Kit No.34 (Washington, D.C .: The As- sociation, 1977) . 8 . John S. Goodell, Libraries and Work Sam- pling (Littleton, Colo.: Libraries Unlimited, 1975). 9 . Lesley Gilder and J. G. Schofield, Work Measurement and Library Management : Methods of Data Collection, LMRU Report No.2 (Cambridge, England: University of Cambridge, Library Management Research Unit, 1977).