College and Research Libraries BLAIR STEW ART Periodicals and the Liberal Arts College Library College libraries are encountering increasing difficulties in meeting their readers' needs for access to periodical literature. Analysis of the periodical holdings of the ten liberal arts college libraries that created the ACM Periodical Bank shows that the "basic list" of periodicals that every such library holds is very short and that these titles are the ones from which the member libraries most frequently requested photocopies. The study also suggests that little or no use is being made of most of the periodicals on college library holdings lists. ONE OF THE MOST DIFF~CULT TASKS faced by the college librarian is that of providing faculty mem hers and students with access to the periodical literature that they seek. Rapidly rising subscrip- tion costs and the increasing numbers of periodicals being published are mak- ing the task more difficult. each year. A major difficulty is that most college li- brarians have very meager information on the use being made of the periodi- cals that they are being asked to acquire. As a consequence, most such libraries are spending thousands of dollars each year on periodicals that no one reads, and still the needs of their readers for access to . periodical literature are not being fully met. Studies of the period- ical holdings of the liberal arts colleges that created the Periodical Bank of the Blair Stewart is research director of the Periodical Bank of the Associated Colleges of the Midwest. The investigation reported here was supporte.d in part by the U.S. Of- fice of Education; however, the opinions ex- pressed do not necessarily reflect the posi- . tion or policy of that agency, nor should of- ficial endorsement by the office be inferred. Associated Colleges of the Midwest and of the use they have made of the Peri- odical Bank, provide some useful in- sights into periodical use patterns and their relation to the periodical holdings of liberal arts college libraries. The creation of the Periodical Bank was based on a number of assumptions. The member colleges of ACM are rela- tively small liberal arts colleges with similar curricula . and, presumably, very similar needs for library services. It was assumed that these needs included in each library current subscriptions and bac:k files for a basic core of something like 500 periodicals. In addition, faculty members and students needed ready ac- cess to a much larger list of less fre- quently used periodicals, a list so large that only the most affiuent college li- brary could hope to hold a substantial part of it. Ready access meant a more - rapid service than could be provided by the ·usual interlibrary loan services, many of which were not available in re- sponse to undergraduate student re- quests. It was proposed that such a ser- vice be provided by establishing a cen- tral periodical library which would sub- / 371 372 I College & Research Libraries • September 1975 scribe to, and maintain back files of, the less used periodicals not included in the basic list. The member college libraries would be connected with the central li- brary by teletypewriter, and photocopies of desired articles would be promptly produced and sent to the ordering li- brary by first class mail. Inherent in the procedures adopted was the assumption that the role of a college library's service in the field of periodicals is to meet the expressed needs of its readers. Substantial sums are spent each year by every library in acquiring, processing, binding, and stor- ing the periodicals designed to meet these needs. Some portion of this ex- penditure is for periodicals that are not used. The omniscient college librarian would presumably not acquire such periodicals, but would only acquire those periodicals, the use of which justi- fied the cost and effort involved in ob- taining and retaining them. It was as- sumed that it was not appropriate to levy a charge on the reader when the li- brarian had not been omniscient and had acquired periodicals that no one read, but could not provide direct .ac- cess to some periodicals that were need- ed. Therefore, no charge was to be made for photocopies of articles to be supplied by the central library. The Pe- riodical Bank service was to be fully in- corporated into the libraries' normal provision of access to periodical litera- ture, although screening procedures might be necessary to avoid abuse of the opportunity to obtain free photocopies of periodical articles. The Periodical Bank was created in 1968 by the ten original members of the Associated Colleges of the Midwest: Beloit, Carleton, Coe, Cornell, Grinnell, Knox, Lawrence, Monmouth, Ripon, and St. Olaf. A central periodical col- lction was established with holdings of some 1,600 periodicals. Soon after the bank began operations in 1969, arrange- ments were made to use the periodical collections of major libraries in the Chi- cago area to supplement the central col- lection. The cooperating libraries are the John Crerar Library, the Newberry Library, and the libraries of the Uni- versity of Chicago, of the University of Illinois-Chicago Circle, of Northwest- ern University, and of the Central Seri- als Service of the North Suburban Li- brary System. After approximately three years of service to · the original member libraries, the bank began to offer its ser- vices to other academic and public li- braries as associate members, but this study is confined to the periodical hold- ings of the ten original libraries and the service to them by the bank from Sep- tember 1, 1969, to June 30, 1972. MEMBER LIBRARY PERIODICAL HoLDINGs Union lists of holdings of the ten member libraries were produced by the Periodical Bank in 1970 and 1971. Anal- ysis of the union list as of June 1971 re- sulted in two surprises that indicated that some preconceptions held by the creators of the bank were invalid. It had been assumed that the lists of the periodical holdings of the member li- braries were pretty much alike and that there was a basic list of perhaps as many as 500 titles that all of the librar- ies would have. As shown in Table 1, the "basic list" in fact consisted of only eighty-five titles. But eighty-five titles on ten holdings lists do not mean that all ten libraries were currently subscribing to these eighty-five periodicals. Included in the eighty-five was the Ken.Yon Re- view, which ceased publication in 1970. In addition, subscriptions to nine of these periodicals had been dropped by one library, and two of them had been dropped by two of the libraries. There were, therefore, only seventy-three peri- odicals to which all ten of the libraries were subscribing in 1971. The second unexpected discovery was the large number of periodicals held by only one library-no less than 2,347 ti- TABLE 1 PERIODICAL TITLES HELD BY MEMBER LIBRARIES, jUNE 1971, CLASSIFIED BY THE NuMBER OF LIBRARIES HoLDING THE TITLE Number of Libraries Holding Title 1 2 3 4 5 6 7 8 9 10 TOTAL Number of Titles Held 2,347 624 315 207 145 127 76 89 91 85 4,106 ties. Here again it is necessary to exam- ine the nature of the titles held. It turns out that 350 of them were files of peri- odicals that had ceased publication, and 755 were remnants of currently pub- lished periodicals to which subscriptions had been dropped. This reduces the number of periodicals to which· only one library was subscribing to 1,242. Even this seems a very large number when compared with the 73 titles to which all ten libraries were subscribing. The 755 back files of periodicals to which subscriptions had been dropped and which were unique to a single li- brary were over half of the 1,367 dropped subscriptions shown on the union lists of holdings. The existence of so many remnants of currently pub- lished periodicals raises a number of questions about periodical acquisition and retention policies. There are, of course, a number of different reasons for dropping subscriptions. The sub- scription rate may have become so high that continued subscription could not be justified, even for a periodical that was sometimes used. Shifts in curricu- lum or faculty interests may reduce or eliminate entirely the demand for a once useful periodical, or it may be de- . cided that the original subscription was a mistake. Where the periodical has con- tinuing usefulness, the retention of the Periodicals I 313 truncated back file may be justified. Where it is not being used, there is nor- mally little excuse for its retention in a liberal arts college library. The back file may be salable, in which case a bet- ter use can presumably be found for the money it will bring. Loading the shelves with unused back files is, in any case, an expensive procedure. The member libraries seem to have had a wide range of policies with re- spect to ceased titles and dropped sub- scriptions. At one extreme were two li- braries that had over 90 percent of their holdings in the form of periodicals to which they were currently subscribing. At the other extreme were two libraries that were subscribing to less than 64 pe~­ cent of the titles on their holdings lists. In the former case the dropped or ceased periodicals retained were presum- ably judged to be useful. Where large numbers of remnants from dropped subscriptions were retained, however, it may be assumed that they include sub- stantial numbers of periodicals that had not been useful enough to warrant con- tinued subscription; in which case a par- tial back file would be even less useful. On this assumption it is of interest to look at the distribution of dropped ti- tles in terms of the number of libraries holding the periodical. This is done in Table 2. This table shows the extreme diversity in the holdings of the ten libraries. Well over half of the 3,553 active titles are uniquely held, and there are only 450 titles held by a majority of the li- braries. It also shows that those libraries that have continued to hold back files of periodicals to which they once sub- scribed have most frequently been dis- satisfied with periodicals to which they were the only subscriber among the ten libraries. As stated above, more than half of the subscriptions dropped were to periodicals now held by only one li- brary. As the number of libraries hold- ing a title . increas~s, the proportion of 374 I College & Research Libraries • September 1975 dropped subscriptions falls. Only 4.3 percent of the subscriptions to titles now held by six or more libraries have been dropped, while the proportion is 23.3 percent for titles held by five or less libraries. To some extent this contrast is an artifact of the policy of some li- braries of disposing of many of the titles to which they have dropped sub- scriptions. The data in Table 2, never- theless, tend to suggest that the number of libraries subscribing to a title is an indication of its probable usefulness for a liberal arts college library. The 450 titles to which a majority of the libraries subscribe, however, cannot be looked upon as a "basic list" to which every liberal .arts college library should subscribe. A total of 4,500 subscriptions would be required to provide all the li- braries with these titles. The number of subscriptions actually maintained for these periodicals was 3,399, or approxi- mately three-fourths of the potential. It certainly cannot be assumed that the 1,101 instances of failure to subscribe or of dropped subscriptions for these 450 periodicals were all mistakes. uSE OF THE PERIODICAL BANK FOR AccEss TO PERIODICALS OWNED BY MEMBER LIBRARIES The second part of this study deals with the requests for photocopies that were filled by the Periodical Bank for the member libraries and the relation of this use to their holdings. This rela- tion is shown in Table 3. This table shows only the requests for periodicals on the holdings lists of member librar- ies filled by the Periodical Bank between September 1, 1969, and June 30, 1972, a period of time that included three academic years. It does not include re- quests filled from the Periodical Bank collection, nor from the cooperating li- braries for photocopies from titles not on the holdings lists of any of the mem- ber libraries. An examination of Table 3, especial- ly of the last two columns on the right, disposes of the idea that the Periodical Bank could ignore a basic list of period- icals which all the member libraries would hold and for which they would supply all of their readers' needs. The heaviest demand, 78.4 requests per title requested, came for photocopies from titles held by all the libraries, followed by the demand for photocopies from titles held by all but one of the member libraries. The least demand, 12.4 re- quests per title, was for periodicals owned by only one of the mem her li- braries. Not shown on the table, but lowest of all, was the demand-7.6 re- quests per title-for periodicals owned only by the Periodical Bank or a coop- erating library. _ TABLE 2 NuMBER OF SuBSCRIPTIONS MAINTAINED AND NuMBER DROPPED, CLASSIFIED BY THE NuMBER OF MEMBER LIBRARIES HoLDING THE TITLE, JuNE 1971 Number of Number of Subscriptions Percent Libraries Titles Maintained Dropped Total Dropped 1 1,997 1,242 755 1,997 37.8 2 524 865 183 1,048 17.5 3 267 659 142 801 17.7 4 188 676 76 752 10.1 5 127 570 65 635 10.2 6 119 661 53 7f4 8.2 7 73 492 19 511 3.7 8 84 636 36 672 5.4 9 90 783 27 810 3.3 10 84 827 13 840 1.5 TOTAL 3,553 7,411 1,369 8,780 15.6 Periodicals I 375 TABLE 3 NUMBER OF REQUESTS FILLED BY THE PERIODICAL BANK, SEPTEMBER 1, 1969, TO jUNE 30, 1972, BY NUMBER OF MEMBER LIBRARIES HoLDING THE PERIODICAL REQUESTED Number of Average Number Libraries Number of Titles Percent of Number of of Requests Holding Not Titles Not Requests Per Title Per Title Title Requested Requested Requested Filled Requested Held 1 310 2,037 86.8 2 222 402 64.4 3 142 173 54.9 4 130 77 37.2. 5 99 46 31.7 6 101 26 20.5 7 69 7 9.2 8 78 ll 12.4 9 83 8 8.8 10 77 8 9.4 TOTAL 1,311 2,795 68.1 The discovery that the demand per title increased directly with the number of libraries holding the periodical called for further investigation. This was done by taking samples of 200 re- quests from each library and classifying the articles requested under three head- ings: ( 1) requests for articles from pe- riodicals not included in the library's list of holdings; ( 2) requests for peri- odicals in the library's holdings list but from issues of the periodical shown by the list not to be owned; and ( 3) re- quests for periodicals that, according to its list of holdings, should be available in the library. The largest number of requests, 55.7 percent, was for articles from periodicals that the library did not have; and 32.5 percent was for ar- ticles from issues not included in the li- brary' s holdings. Thus 88.2 percent of the requests were for photocopies of ar- ticles that the holdings lists showed were not available on campus. Both of these types of demand were expected, but the importance of the second category re- flecting incomplete back files, was cer- tainly underestimated. It was also ex- pected that there would be an occasion- al situation in which an issue of a peri- odical would be at the · bindery, could not otherwise be located, or had been mutilated. There was also the possibil- 3,829 12.4 1.6 3,283 14.8 5.3 3,144 22.1 10.0 3,148 24.2 15.2 3,777 38.2 26.0 3,201 31.7 25 .2 1,947 28.2 25.6 2,326 29.8 26.1 4,34.0 52.3 47.7 6,034 78.4 71.0 35,029 26.7 8.5 ity that screening procedures would not eliminate all requests for photocopies of articles actually available in the re- questing library. Whatever the reason, such requests amounted to 11.8 ·percent of the sample. The tabulating procedures used made it possible to determine what propor- tion of all the requests filled were for titles on the requesting library's list of holdings. This gave some indication of the representativeness of the sample. The proportion of all requests which were for titles not on the requesting li- brary's list of holdings was 57.2 percent, as compared with the sample estimate of 55.7 percent. The sampling proce- dures apparently gave somewhat dispro- portionate weight to the libraries that made more than the average proportion of their requests for titles on their own holdings lists. The estimate that over 88 percent of the requests filled were for photocopies from items not in the requesting li- brary's holdings may, at first glance, seem incompatible with the finding that the heaviest demand per title was for photocopies from periodicals found on the holdings lists of all ten libraries. It should be borne in mind, however, that only 77 of the 1,311 titles requested were in the heaviest demand per title 376 I College & Research Libraries • September 1975 category and that . many of the member libraries' back file holdings of these ti- tles were incomplete. THE uNUSED PERIODICALS Another significant aspect of Table 3 is the large number of the periodicals on the holdings lists of these libraries from which no requests were filled in three years. A total of 2,795 periodicals, or 68.1 percent of the titles owned, were never requested. The proportion not re- quested, no less than 86.8 percent of the titles held by only one library, was dis- tinctly greater for periodicals held by a minority of the libraries than for those held by a majority. There are, therefore, three bits of evidence sug- gesting that the number of libraries holding a title is a useful, if indirect, indication of its probable value in a lib- eral arts college library. They are the higher proportion of dropped subscrip- tions among the less commonly held periodicals, the positive relation be- tween the number of requests per title, and the number of libraries holding it, and the predominance of titles never re- quested among the titles held by only a few libraries. The best measure of the importance of a periodical for a given library is, of course, the amount of use made of it by the library's readers. If the number of requests filled by the Periodical Bank is an indication of a periodical's usefulness, the large num- ber of periodicals which were never re- quested raises serious questions about the usefulness of many of the period- icals held by these college libraries. Furthermore, there is no reason to sus- pect that the periodical selection pro- cesses at these libraries are significantly different from those employed by other liberal arts college libraries. The most extreme situation, of course, is that of the 2,037 periodicals owned by only one library and requested by no other li- brary in three academic years. INTERMEMBER PERIODICAL AccESs It may be noted that the ten libraries listed a total of 4,106 different period- icals in their holdings. This may be con- trasted with the 1,675 titles held by the Periodical Bank in June 1971. The po- tential resource for an interlibrary pe- riodical photocopy service represented by these 4,106 titles was recognized when the Periodical Bank was established, and arrangements were made for an inter- library service paralleling the Periodical Bank service: using the teletypewriter for requests and first class mail for the delivery of photocopies. A summary of the use of this service for one year, July 1, 1971, to June 30, 1972, is given in Ta- ble 4. ~ If one imagines the hypothetical sit- uation in which the existence of a peri- odical on a library's list of holdings means that the library is able to satisfy all of its readers' needs for access to the title, and that all the periodicals are equally useful, we would expect no in- terlibrary requests for titles held by all ten libraries, and approximately nine times as many requests for periodicals owned by only one library as for those held by nine. Table 4 shows that there were 92 ( 18 plus 7 4) requests for titles held by all ten libraries, but there were indeed more requests for uniquely held titles than for those held by nine li- braries; 798 ( 665 plus 133) as compared with 156 ( 134 plus 22). But there were 2,347 periodicals held by only one li- brary and only 91 titles held by nine li- braries. Consequently, there were almost twice as many requests per title owned for the periodicals held by nine librar- ies as for those held by only one library. · This was very different from the 9 to 1 ratio in the opposite direction suggested in the hypothetical situation described above. There is an even greater contrast in the ratio of almost 30 to 1 in favor of the titles held by nine libraries shown by the data in Table 3 for re- Periodicals I 377 TABLE 4 NUMBER OF REQUESTS FILLED BY THE INTERLIBRARY PERIODICAL ARTICLE PHOTOCOPY SERVICE }ULY 1, 1971, TO JUNE 30, 1972, BY NUMBER OF LmRARIES HoLDING THE TITLE Number of Average Libraries Number of Titles Percent of Number of Number of Requests Holding Not Titles Not Requests Per Title Per Title Title Requested Requested Requested Filled Requested Held 1 137 1,894 93.3 665 4.9 0.3 1 plus PB 0 41 275 87.0 133 3.2 0.4 2 80 323 80.1 321 4.0 0.8 2 plus PB 0 55 166 75.1 231 4.2 1.0 3 48 ll8 71.1 127 2.6 0.8 3 plus PB 0 37 112 75.2 119 3.2 0.8 4 23 49 68.1 156 6.8 2.2 4 plus PB 0 47 88 65.2 178 3.8 1.3 5 20 21 51.2 106 5.3 2.6 5 plus PB 0 32 72 69.2 llO 3.4 1.1 6 12 12 50.0 33 2.8 1.4 6 plus PB 0 36 67 65.0 114 3.2 1.1 7 6 5 45.5 19 3.2 1.7 7 plus PB 0 26 39 60.0 92 3.5 1.4 8 8 4 33.3 68 8.5 5.7 8 plus PB0 24 53 68.8 79 3.3 1.0 9 5 6 54.5 22 4.4 2.0 9 plus PB 0 35 45 56.2 134 3.8 1.7 10 6 3 33.3 18 3.0 2.0 10 plus PB 0 32 44 57.9 74 2.3 1.0 TOTAL Not held by PB 345 2,435 87.6 1,535 4.4 0.6 Also held by PB 365 961 72.5 1,264 3.5 1.0 All titles 710 3,396 82.7 2,799 3.9 0.7 0 Titles held by the Periodical Bank and the number of member libraries indicated. quests filled by the Periodical Bank. The interlibrary loan service was par- ticularly useful for the less generally held periodicals. Of the 2,799 requests filled for member libraries by other member libraries, 798, or 28.5 percent, were for titles held by only one mem- ber library, and 2,146, or 77.7 percent, were held by five or fewer libraries. Even so, the proportion of the total number of titles held from which other libraries requested photocopies was low- er for the less widely held titles. This is particularly true of the periodicals held by only one member library. Only 178 of the 2,347 titles in this category, or 7.6 percent, were requested. The cor- responding figure shown in Table 3 for requests to the Periodical Bank was 13.2 percent. Periodicals owned by only one library constituted over 57 percent of the 4,106 different titles held. But 347 of the uniquely held periodicals had ceased publication, and subscriptions to 755 had been dropped. Of these 1,102 dropped or ceased titles, 1,018 were not requested of the Periodical Bank in three years, and 1,061 were not request- ed in one year through the interlibrary loan service. It seems clear that many of the titles held by only one library were of limited usefulness, especially those periodicals that had ceased publication and those to which subscriptions had been dropped. Table 4 shows separately the data for titles held by member libraries and those held also by the Periodical Bank. More than half of the periodicals re- quested were also held by the bank, but the number of requests for these peri- odicals was less than for those not owned by the bank. There are probably a number of different situations in 378 I College & Research Libraries • September 1975 which another library rather than the bank is asked for a photocopy. Among them is the greater proximity of a neighboring college, and the fact that during 1971- 1972-but not subsequent- ly-the member libraries were charged twenty cents per exposure for photo- copies made .at cooperating libraries, but only ten cents per exposure for photo- copies made from the holdings of the Periodical Bank or a member library. If the Periodical Bank was unable to fill a request from its holdings, it notified the requesting library, which then might tum to the interlibrary photocopy ser- vice with the request. ·of the 4,106 different periodicals owned by the member libraries, 2, 795, or 68.1 percent, were never requested in one year from another member library. Measured by these tests, it appears that the member libraries are spending thou- sands of dollars each year on periodicals of doubtful utility. The real test, of course, is the actual use made by the li- braries' readers. The savings promised by the elimination of unused period- icals justifies substantial efforts at every liberal arts college library to discover just how much use is being made of ev- ery title on its list of periodical hold- ings.