Cost Analysis of an Automated and Manual Cataloging and Book Processing System

A comparative cost analysis of an automated network system (WLN) and a local manual system of cataloging and book processing at Washington State University Libraries indicates that the automated system is about 20 percent less costly than the manual system. A per-unit cost approach was used in calculating the monthly cost of each system based on the average number of items processed per month under the automated system. The process and the results of the analysis are presented in a series of charts which detail the tasks, items processed, unit and total monthly costs of both the manual and automated systems. The higher costs of the manual system were essentially staff costs.


Cost Analysis of an Automated and Manual Cataloging and Book Processing System
Joselyn DRUSCHEL: Washington State University , Pullman.
A comparative cost analysis of an automated network system (WLN) and a local manual system of cataloging and book processing at Washington State University Libraries indicates that the automated system is about 20 percent less costly than the manual system.A per-unit cost approach was used in calculating the monthly cost of each system based on the average number of items processed per month under the automated system.The process and the results of the analysis are presented in a series of charts which detail the tasks, items processed, unit and total monthly costs of both the manual and automated systems.The higher costs of the manual system were essentially staff costs.
The Technical Services Division (TSD) of Washington State University Libraries (WSUL) has had considerable experience in the use of automated techniques in selected areas of technical processing.An in-house automated acquisitions system was developed and implemented in 1967; that in-house system was eventually replaced by the acquisitions component of the Washington Library Network (WLN) .Since November 1977, the Technical Services Division of WSUL has used the WLN bibliographic component for data verification (searching) and cataloging of materials.
Although the Library has generally known its total automation expenditures, it has lacked a more precise breakdown of cost data on automated processing.Moreover, the library has practically no cost data on manual processing.This report deals only with the costs of using the WLN bibliographic system, not the WLN acquisitions component.An analysis was made of the total costs of both the automated and manual book processing systems.The objectives in undertaking the cost analysis were threefold: (1) to identify the essentially unknown costs of manual processing; (2) to provide more exact cost data on automated processing; and (3) to develop comparable data on the costs of each system .Manuscript received October 1980; accepted December 1980.

METHODOLOGY
The methodology used in this cost analysis was a per-unit cost approach.First, each process or task in which the staff were engaged in cataloging and book processing was identified.Second, the per-unit cost-e.g., staff, data base, materials-of each process was calculated.Finally, monthly costs were determined by multiplying the average number of items processed per month by the unit cost per task.The cost analysis charts (tables l(a)-l(e)-manual system; tables 2(a)-2(d)-automated system), which detail the tasks, items processed, and unit and total costs form the body of• the analysis.Equipment costs-purchase, lease, maintenance-were calculated separately, and are included in the summary cost data for each system (table 3).

IDENTIFICATION OF PROCESSES
The staff of the TSD Cataloging and Book Processing Unit perform the following functions: bibliographic verification, bibliographic record production, bibliographic record maintenance, the marking of materials, binding preparation and receipt (for most of the library system), and the preparation of book cards.4 for sample profile sheet).The primary purpose of the profiles was to develop a detailed account of work distribution throughout TSD in order to determine the staffing requirements necessary for each unit to maintain an even workflow.In the cost analysis , the Cataloging and Book Processing (CBP) profile was used to identify each unit process, as well as to provide the basic data on the number and level of staff and the time required to perform each process.Additionally, for the automated system, the CBP profile sheets, together with WLN invoices (see figure • 1 for sample invoice) and WLN monthly activity reports (see figure 2 for sample activity report) were used to determine the average number of items processed per month.For example, since about 85 percent of the cataloging done in TSD is via WLN, it was possible to derive exact figures from WLN invoices for the average number of items cataloged per month.The WLN invoices also differentiated between data-base copy cataloging and original data entry.The CBP profile sheets were used to determine average number of non-WLN items cataloged.
Using a combination of WLN invoice and profile data, a chart was constructed of the average number of items searched and cataloged per month under the automated system (see table 5).In order to make costs comparable, an assumption was made that the same average number of items was searched and cataloged under the previous manual system and a similar chart was made for it (see table 6).In reality, the available staff under the manual system could not process the same amount of material per month.In the cost analysis of the automated system, the monthly wages for staff members of the Cataloging and Book Processing Unit were based on current monthly salaries (as of February 1980) plus estimated fringe benefits (21 percent).The total wages were added together for each level of staff and divided by the number of staff at that level to give an 18 , 750 .00 459.00 810.00 1,003.00378.00 5 , 335 .00 5,54 1. 00 @ 4¢ A TITLE @ 15¢ A COPY @$ 1. 60 RECRD @ 15¢ EACH @ 14¢ EACH @ 6.9¢ EACH @ 5.5¢ CARD TOTAL SERVICES CHARGES TOTAL CHARGES , Fig. 1.Washington Library Network Customer Invoice .(Note: Part of the "no charge" inquiries are generated and used by the Acquisitions Unit and are therefore not included in this analysis.)Although the terminal service and line charges might simply have been added as a total amount to the data-base costs, it seemed more meaningful to distribute these costs on a per-use basis .The method used to distribute these charges was to identify each use of the bibliographic data base, and to divide the total monthly costs of terminals and lines by the total monthly units of use (see table 9).This method of distributing terminal service and line charges not only provided per-unit terminal use costs, but also served to categorize kinds and quantity of data-base use.

SUBSCRIPTION AND MATERIAL COSTS
Subscription costs include only those bibliographic tools purchased for use in TSD for the purpose of bibliographic searching.As a result of the increased growth of the bibliographic data base, fewer tools are being used for searching under the automated system than under the manual system.Prior to the implementation of WLN, the library subscribed to bibliographic data (LC and CIP copy) on microfiche supplied by the Information Dynamics Corporation (IDC).The per-unit costs of all subscriptions are presented in the cost analysis charts (tables 1(a) and 2(a)).
Material costs include only those materials unique to cataloging and book processing; general supplies, such as pencils and paper, are not included.The calculation of the per-unit cost of most materials is generally straightforward.It should be noted, however, that under the automated system, products, i.e., materials, are included in the data-base costs, and only those materials used independent of the data base, e. g., book pockets and book cards, are listed as material costs on the charts.Under the manual system, due to the divisional arrangement of the library system and the number of card catalogs being maintained, the formula for producing sets of cards for a single title was complex.For this reason, the costs and number of cards produced for the titles cataloged per month are listed as a separate line item.
EQUIPMENT COSTS Equipment costs include only equipment unique to cataloging and book processing, i.e., required for processing or products.General equipment, such as desks, book trucks, typewriters, are not included.

Equipment-Automated System
During the period covered by the cost analysis, November 1977 to February 1980, the following equipment was purchased for the automated system: system, eleven more staff and 1,365 more timeslip hours were needed per month to process the same amount of materials as is processed under the automated system.In fact, compared to the staff costs of both the manual and automated systems, the costs of equipment, data-base use (including products), terminal service, and telecommunication lines of the automated system are a relatively small percentage (27 percent) of the total cataloging and book processing costs .This analysis serves to underscore a basic reality of the current library organization: personnel is one of its largest expenditures and staff-intensive systems are very costly.This cost analysis has not directly addressed the issue of the quality of processing and products of either the manual or automated systems.The analysis suggests, however, that the automated system is more efficient in terms of staff time .Moreover, the TSD staff has found that not only can more be done with fewer staff, but the automated system also provides more accurate data and has the flexibility to accommodate with relative ease the many corrections and changes that must be made to the library's bibliographic files.
Joselyn Druschel is assistant director for automation and technical support at the Washington State University Libraries .She is currently chairing a staff task force which is developing specifications for the Libraries' on-line catalog. •

Table I (
c). Cost Analysis: Manual Cataloging and Book Processing System accession # on shelflist; record call # in book; type slip for marking) LT II (.094/min @ 15 min/item) 1.41 1.41 50 71

Table 2 (
d). Cost Analysis: Automated Cataloging and Book Processing System

Table 3 .
Cataloging and Book Processing System: Summary Comparison CostsSince 1978 this unit, as well as all units in the Technical Services Division, have periodically analyzed unit activities, and recorded the data collected on Work Assignment/Staffing profile sheets (see table

Table 4 .
Technical Services Division Work Assignment/Staffing Profile: November 1978 Unit: Cataloging and Book Processing.Subunit: LC Copy Editing .

Table 5 .
Type and Average Number of Items Searched/Cataloged Per Month on Automated System (Based on WLN Invoice Data and CBP Work Assignment/Staffing Profile )

Table 6 .
Type and Average Number of Items Searched/Cataloged Per Month on Manual System (Based on CBP Work Assignment/Staffing Profile)

Table 9 .
Bibliographic Data Base Use Per Month (One Unit = One Access to or Process in Data Base) 15650 = $.06/terminaluse for Cataloging and Book Processing System

Table 10 .
Cataloging and Book Processing System: Summary Comparison Costs by Function (Excluding Equipment Costs )