College and Research Libraries By ROBERT W. CHRIST Acquisition Work in College Libraries I Mr. Christ ts assistant librarian , Duke Univ ersity. I N THE recent survey, College and Univer-sity Libraries and Librarians hip, pre- pared by the College and University Post- war Planning Committee of the American Library Association and the Association of College and Reference Libraries, the fol- lowing statement is made: The actual procedures and routines for book buying, within the individual library, are in a good many cases, it is suspected, susceptible of simplification and streamlin- ing. . . . Every library might well examine or re-examine its order routines in an effort to cut down as much as possible that delay between the initiating of an order and the availability of the book for use, a delay that is with some justice more or less of a standing complaint of the average faculty member .... We may expect to see , and should encourage, much experimentation . . . in quest of the simplest and most economical way to accom- plish this important library function. 2 Last winter the writer made such an "ex- amination" of acquisition work in ten eastern college libraries to discover what procedures for the book purchasing are cur- rently employed, what results are obtained, what routines may be designated as the most satisfactory, and what differentiation, if any, is made in college libraries between profes- sional and nonprofessional functions of the acquisition department. The results of this investigation are reported in detail in an essay presented for the master's degree to 1 Pape r presen ted a t the College Librari es S ect ion , A.C. R.L. , June r6, 1948, A t lantic City, N .J. 2 C h icago, 1946 . p . 39-4 0. JANUARY, 1949 the School of Library Service at Columbia University. The essay goes one or two steps further than suggested in the above quota- tion and sets up, on the basis of current practices and results in the libraries studied, tentative standards by which the perform- ance of the acquisition department of any college library may be evaluated, and sug- gests a purchasing procedure suitable for all libraries of this type. This paper, which is based on the larger essay, summarizes some of the chief data collected. It touches briefly on the colleges and the libraries, the acquisition departments and their functions, and discusses in some detail the results obtained by the acquisi- tion departments and the tentative performance standards. Material collected on procedures, records, forms, and similar factors is omitted. The investigation was conducted by personal visit~ to ten college libraries, where the librarians and acquisition heads were interviewed, records and files examined, and the workings of the departments observed. The ten colleges are Connecticut College for Women, Trinity (Hartford), Bowdoin, Wesleyan University (Middletown), Wil- liams, Amherst, Mount Holyoke, Vassar, Wellesley, and Smith. They are private, four-year liberal arts colleges, selected be- cause they are sufficiently similar in cur- ricula and objectives to have comparable library programs and similar acquisition problems and requirements. Yet there is a sufficiently ~ide variation in the size of the colleges, the libraries, and their budgets to make up a representative group of college 17 libraries for the purposes of the study. The colleges vary in number of .students ( I946-47 enrolment) from 837 to 2266, not quite a 300 per cent variation. Their libraries vary in number of volumes from I I 7,866 to 338,323, again about a 300 per cent vanatwn. The library budgets run from $25,389 to $98,807, nearly a 400 per cent range, and the library staff from 6 to 29.5 persons. The acquisition budgets (i.e., books, periodicals, and binding), which ranged from $8,88o to $27,323, are allo- cated to departments of instruction and a librarian's general fund in six of the li- braries. In all the libraries the chief source of recommendations for purchase is the faculty and the librarian. The acquisition departments in these li- braries vary from one-half to three and one-half persons or the equivalents. Three have a library school graduate giving full time to acquisition work, and two more have a library school graduate giving half time. In all the libraries the facilities for the acquisition department were adequate, that is, adequate space and suitable location, type- writers, adding machines, and bibliographi- cal collections. Several have superior facilities of this sort. Only four, however, have an acquisition department manual in any form, which suggests that the others place rather unreasonable reliance on there being no sudden personnel changes. The investigation showed that normally the acquisition department of the college library does the purchasing of all types of materials: books, continuations, periodicals, documents, maps, etc. Only three libraries have a separate serials department which takes care of periodical subscriptions. These libraries are the three largest in size of collection, the sixth, seventh, and tenth in size of library staff, and the sixth, eighth, and tenth in total expenditure for acquisi- tions. One of them has a documents department as well, which handles all order- ing and receiving of federal and state docu- ments. The solicitation and disposition of gifts is in all cases handled outside the acquisition department, usually by the li- brarian, but searching and processing is customarily done by the department as for purchases. The same would be true of ex- changes, except that all but two of the libraries reported that because of increased service demands without corresponding in- crease in staff, they were not doing enough exchange work to record it as part of the acquisition department's work load. In none of the libraries does the department have a specific responsibility for the selection of titles, although three of them may make the decision on the edition to be purchased, and only one routes publishers' announce- ments and dealers' catalogs. The preparation of bills for payment, bookkeeping, and budgetary control are the responsibility of the acquisition department except at one library wh~re an executive secretary keeps all financial records. In all but one library the department prepares slips for Library of Congress card orders, though frequently the slips are turned over to the catalog department to send. In the single exception, the catalog department prepares and sends the card orders after a daily ex- amination of outgoing book orders. At two libraries, the acquisition department re- ceives the cards and matches them with the books on receipt. In addition to this regularity of function, a variety of other responsibilities is assigned to the acquisition department in almost all the libraries. In some cases this is doubtless because of the nature of the activity, such as the preparation and distribution of .an accessions list, or the ordering of all library supplies and the distribution and mainte- nance of stock. In other cases, primarily in libraries with small staffs, this is probably 18 COLLEGE AND RESEARCH LIBRARIES more a matter of convenience and available personnel, and such duties as the operation of a special textbook library, maintenance of the accessions register, service as the li- brarian's secretary or administrative as- sistant, or the preparation of materials for binding, fall to the lot of the acquisition department. Criteria for Measurement To measure the results obtained and to judge the quality of service (in 1946-47) several factors, all capable of exact statisti- cal recording, were selected. Principal among these are the following: I. Time lags a. From receipt of recommendation to placing of order b. From placing of order to receipt of book c. From receipt of book to completion of acquisition process 2. Discounts received (on current domestic publications) 3· Bibliographical accuracy a. Number of unintentional duplicates received b. Number of incorrect items received c. Number of items rejected by dealers as not identifiable 4· Percentage of successful orders from dealers' catalogs 5. Number of titles and volumes purchased and otherwise processed 6. Expenditure for books The investigation showed that it would be impossible to make a wholly valid evalua- tion of the performance of the acquisition departments, partly because of the variation in functions performed, but primarily be- cause reliable statistics are not available for most of the test factors. College library administrators have shown regrettably little interest in testing the efficiency of their acquisition departments, and as a result no statistical records of work performed by the departments, even for test periods, are main- JANUARY~ 1949 tained. The discounts currently received on domestic publications were available from all libraries, though even here it was neces- sary actually tq examine invoices to secure these figures in some instances. No library had any record of the number of u~inten­ tional duplicates received in the course of a year, the number of incorrect items sup- plied, or the number of items rejected by the dealers. It was only the investigator's persistence which brought forth even esti- mates for these factors. The same is true of the percentage of successful orders from dealers' catalogs, although every title ordered and not secured represents an actual loss to the library in time spent on searching, preparation of orders, and maintaining records. Time lags within the acquisition depart- ment and in the supplying of books ordered were in all cases only estimates, though made with less reluctance and more con- fidence than the estimates on factors men- tioned above. Since only three libraries count books as they are received in the ac- quisition department, this factor had gen- erally to be recorded on the basis of the count of books cataloged. Although the number of titles is more significant in rela- tion to work load in both acquisition and catalog departments, the majority of the libraries could supply only the number of volumes. Two of the libraries do not keep financial records in sufficient detail to have readily available the total expenditure for books only, but lump books and periodicals. A third has no available breakdown beyond the total expenditure for acquisitions: books, periodicals, and binding. With these very considerable limitations in mind, however, it is possible to make some general observa- tions on the performance and quality of service of the acquisition departments in col- lege libraries as exemplified by this sample group. 19 For time lags, all but one of the libraries made a point of sending rush orders within one day of the receipt of requests. For normal requests, three libraries ordered onee a week, and one only once every two weeks (with foreign orders sent monthly). The average time for the others was about three days. For very small purchasing operations, weekly orders are perhaps economical of staff time since there are real savings in performing certain routines on a number of titles at one time, th~ugh it is doubtful whether such a practice can be defended in terms of service to users. The time required to receive books (do- mestic, in print) after placing the orders was estimated by most libraries as from fourteen days up. One library, however, said it seldom received books in less than three weeks, aoother in less than four weeks. At one library, which estimated fourteen days as the average time lag between the ordering and receipt of books, an analysis of all orders placed for current domestic· publications between September I 5 and Oct. I5, I947, and filled prior to Decembe.r 28, was made on that date. Fifty-six per cent of rush requests had been ordered within one day, and I 2 per cent more within two or three days. (The remaining 32 per cent did not show the date the request was received.) Among normal requests, of which I3.4 per cent did not show the date of receipt, 63 per cent were ordered within one day, and another 20.2 per cent within two or three days. Combining the two types of orders, 80.5 per cent of all requests were ordered within three days, which verified this library's esti- mate of this time lag. The same library's estimate of fourteen days from the placing of an order to the receipt of the book was proved, however, to be far from reliable. Of rush orders, only 28 per cent were received within two weeks, and 25.2 per cent of normal orders in the same time. The next week brought 40 per cent more of the rush books but only 20.2 per cent of normal orders. In four weeks, 72 per cent of rush items and 56.3 per cent of normal orders, or only 59 per cent of all orders were received. Of all orders filled for current domestic publica- tions, only 79.8 per cent had been supplied in six weeks, and the remaining 20.2 per cent required more than six weeks and up to I03 days (i.e., the total time from Sep- tember I 5 to December 28, the day analysis was made). But in addition to the I44 books ordered and received . within this period, I4 additional titles (including one rush item) were ordered and had not been received or reported on. Two more (in- cluding one rush) were ordered and in- correct items supplied, the corrections not having been completed. Thus of I6o titles actually ordered in the test period of one month, I I 5 or 7 I .g per cent were received within 42 days; 29 more or I8.I per cent were received within 43-I03 days; and I6 titles or IO per cent required an indeter- minate longer period. If the error in esti- mating was comparable in other libraries, there is validity in the suggestion made in the quotation with which this paper opened. At one library, where the librarian felt confident of his estimate of a two-week delay from the ordering of a book to its receipt, a number of filled-order cards were examined at random and bore out the assertion that it was exceptional for there to be a longer time lag. A detailed analysis such as described above could not, however, be made at this library. Most of the departments completed their records and moved books on for cataloging in not more than two days. In most cases, invoices are received at the same time or before the books. One library never waits for an invoice (in fact, checks in its books 20 COLLEGE AND RESEARCH LIBRARIES without the invoice even though it has been received), two always wait for invoices, al- though at one of them the invoices are not infrequently received as much as two weeks after the books, which is a considerable de- lay. (Since this investigation was made, this library has revised its procedure and no longer holds books for invoices.) The study of discounts received on cur- rent domestic publications by libraries in this group showed a range from low to high of IO to 40 per cent, but the concentration was heavy at 25 to 30 per cent on trade titles and IO per cent on short discount items. The highest discount was related to slow service, and the lowest to fast service, proving once again (perhaps) that one gets what one pays for. Under bibliographical accuracy, the esti- mates for the number of unintentional duplicates received ranged from 6 to I5, with concentration at IO to I2. In pro- portion as this number is reduced, without loss in other factors of service, the depart- ment's efficiency may be considered to be increased. Three libraries had no recol- lection of incorrect titles supplied, but the others estimated from 5 to 36, or approxi- mately two to six per thousand volumes purchased. These may result from errors on the part of the dealer, or of the depart- ment in preparing orders and supplying bibliographical details. Wherever the re- sponsibility lies, such errors are costly to both parties in time required to make cor- rections. Five libraries had no recollection of titles which their dealers had rejected as not identifiable or because they required bibliographical corrections or additions. The other five estimated from two to twelve a year. The most common statement concerning the percentage of successful orders from dealers' catalogs was, "we lose a lot." When pressed for an estimate, the libraries' JANUARYJ 1949 replies varied from 2 5 to 7 5 per cent, half . of them estimating 50 per cent. The li- brary where the time lag estimates were checked as described above estimated that it received 50 per cent of all orders from dealers' catalogs. To verify this, all such orders placed from July I to Nov. 30, I 94 7, were checked. (This was done on Decem- ber 28, when it was assumed that all suc- cessful orders would have been received.) Of I36 items ordered in this period, 85 were received. This represents 62.5 per cent success on this type of order. The following table gives a picture of the acquisition departments' performance in the ten libraries in terms of the amount spent for books, the number of volumes purchased, and the size of the staff. It shows also certain other factors affecting the work loads of the departments. The apparent net cost, figured simply by divid- ing the expenditure for books by the num- ber of volumes purchased, is included as a matter of interest. The variations and vagaries of library accounting and statistical procedures are so well known that it is unnecessary to do more than point out that this cost may at best be considered only a rough estimate. Any attempt to select from such records the most efficient acquisition department must take all of these factors into account, along with other data (not included here) showing the forms used and records kept at each library, the ·actual purchasing pro- cedures followed, the complete list of functi_ons performed. It might appear that library III has the best performance record, since only half the time of one person was required to spend $8564 for 2839 volumes at approximately $3.02 each, and in addition to handle I968 gift volumes, library book- keeping, and periodical purchasing. But this library does not have a departmentalized allocation of its book funds. Therefore its 21 Table I Acquisition Department Performance, 1946-47 I Book Purchasing Principal Other Functions c: en I ~i. ~~ ~ CIS cu . 0 0 ::l CU·- ::l- $...4·.;:: .!:: ~ ""g .... c: c ;.ao cu·- ""g en~~ ..c.~ C: en c< 0 c: >< 0 ~ z~ w~ CIS w~ I 2 $16,434 $12,357 II 2 888o b III . sa 13,273 8564 IV 2a 19,008 6020 v I I6,842 9996 VI 2.5 23, I I7 15,662 VII 2 16' I39 10,404 VIII 3·5 22,229 d IX 3·5 27,323 16,994 X 2 . 5 26,o66 e a Estimated full-time equivalent. b Not available; for books and periodicals, $7446. ° From the librarian's annual report. "'C en cu ei ::l u - ~ 0 ::l :>P.. 2923 I988 2839 I636f 2267 6238 3130 3223 6I07 7852 Affecting Work Load "'C I ~ '"den .... ~t: --;~ en 0 CIS 0 .... cu bO c: UOJ ~P:; c: bO c: bO ·-en ..., E cu"'C en ·a. -c: oo""; ~ en E cu C: en cu «l.-c: ::l ::l cu .... ~ 0 ~ cu u en c: u cu- ·- 0) ·- c I ::l en en ..!( u 0 CU · - 2:t c:_ o.."'C uoo 0 ~ ~ cu·-....c:ti 0 0 cu ~ u cu 0 cu ::l CIO