College and Research Libraries Collection Growth and Expenditures in Academic ·Libraries: A Preliminary Inquiry Richard Hume Werking These data, from a group of liberal arts college libraries as well as from the much more scruti- nized ARL libraries, raise important questions about certain articles of faith in academic librari- anship. These questions relate to: the "doubling time" of library collections; the "60-30-10" division of library expenditures; the growing robustness of materials expenditures as a percent- age of total expenditures, especially among the college libraries; and the phenomenon of in- creases in total expenditures considerably exceeding increases in major price indexes. ''That most librarians dislike statistical rec- ords is patent. But without figures capable of intelligent interpretation, we are seriously handicapped indeed. William Thomas Kelvin expressed the need adequately and suc- cinctly, ' ... when you can measure what- ever you are talking about, and express it in numbers, you know something about it.' " -Lawrence S. Thompson, 1945. "It is essential that more be known about the present use and management of library budgets ."-WarrenJ. Haas, 1986.1 Generalizations about academic li- braries in the United States are fre- quently based on data from or experi- ences in those libraries which are members of the Association of Research Libraries. In order to examine data re- lated to collection growth, expenditures, and automation, I wanted to give my at- tention to another group of libraries which have collected and shared data for more than twenty years. These are the schools on the so-called ''Bowdoin List'' of liberal arts college libraries, a group . taking its name from the institution whose library director has compiled the statistics since 1967. Examining data and trends among these college libraries should be useful not only in itself, but also in carefully generalizing about other groups of academic libraries, and in comparing trends with the ARL li- braries. In time, perhaps, other re- searchers will study other groups of aca- demic libraries. These studies will lessen our dependence on the ARL Statistics for generalizing about aspects of academic librarianship. This article is divided into several parts. Sections I through V present the statistical data from the Bowdoin List li- braries and compare them with ARL data, both to illustrate and to serve as the basis for discussing significant trends in two important sectors of academic li- brarianship. Section I covers collection growth, while Sections II through Vex- amine data related to various categories · of expenditures, both for the ARL and the college libraries. A subsequent arti- cle will report on the state of automation among this group of college libraries . THE BOWDOIN LIST LIBRARIES From 1943 until1960, the Association of College and Research Libraries pub- lished library statistics for colleges and universities. The statistics for 1958/59, published in 1960, comprised the last such compilation because ACRL turned the task over to the federal government Richard Hume Werking is Director of Libraries at Trinity University in San Antonio, Texas 78212 . 5 6 College & Research Libraries and the HE GIS reports. 2 Soon after- wards, in 1962, the Association of Re- search Libraries began the annual publi- cation of its members' statistics, and five years later a group of college libraries be- gan to share their statistics with one an- other.3 In 1967, Richard Harwell, Librarian at Bowdoin College in Brunswick, Maine, prepared a list of thirty-seven college li- braries from which he solicited annual statistics to compile and share with the contributors. The first Bowdoin List of li- brary statistics covered the 1966/67 aca- demic year. 4 It has since been continued annually, with Arthur Monke assuming responsibility for its compilation after he succeeded Harwell as director at Bow- doin. Over the years the list grew to in- clude forty-two institutions (see figure 1). The colleges on the Bowdoin List are widely recognized as among the most prestigious liberal arts colleges in the country. They are all private institu- . tions, are primarily undergraduate, ex- ercise a high degree of selectivity in ad- missions, and are nonsectarian. They are also relatively small; in the first year of the Bowdoin List, enrollments ranged from 1,865 at the largest school to 842 at the smallest, with a median of 1,267. Twenty years later the range was be- tween 3,453 (for Bucknell, which had not been on the list at the outset) and 479, with a median of 1,532. As one di- rector commented to me, "It is not an objectively determined list, but it is a very useful list, convincing to adminis- trators and faculty.'' Thus, the Bowdoin List college li- braries constitute a fairly homogeneous, self-identified group. No attempt is made here to claim that they are '' typi- cal" academic or college libraries. Stud- ies of groups of libraries in addition to those which are members of the Associa- tion of Research Libraries, the Bowdoin List, and the relatively new "ACRL Uni- versity Libraries" list would likely give us a fuller understanding of the various sectors in academic librarianship. METHODOLOGIES After securing a complete set of the January 1991 Amherst* Antioch Bates Beloit* Bowdoin* Bryn Mawr Bucknell* Carleton* Colby* Colgate* Connecticut* Davidson* Dickinson Earlham* Franklin & Marshall* Grinnell* Hamilton* Haverford* Hollins Knox Lafayette* Lawrence* Macalester* Middlebury* Mills* Mount Holyoke* Oberlin* Occidental* Randolph-Macon* Reed* Smith* Swarthmore* Trinity (Connecticut)* Union* University of the South* Vassar* Wabash* Washington & Lee Wellesley* Wesleyan* Wheaton* Williams* *Indicates a response to the survey. FIGURE 1 The Bowdoin List Institutions Bowdoin List data since 1966/67, a data sheet was prepared for each library, fill- ing in for each the collection size, ex- penditures for salaries and wages, for materials, total expenditures, "other" expenditures (the total less the sum of salaries/wages and materials), and for size of staff. 5 A questionnaire was also prepared to elicit any additions or cor- rections to the data (a substantial amount of each was received), as well as information concerning: how various expenditure categories were reported; the status of automation or plans for au- tomation; how automation was being fi- nanced; and how the directors felt about the shifts in categories .of expenditures. 6 After the survey was "piloted" with several library directors and other indi- viduals, it was sent to the directors of the forty-two Bowdoin List libraries. Thirty- five were returned, for a response rate of 83%. In addition to the survey, I received a considerable amount of information during personal interviews of library di- rectors at twenty-two of the colleges. It is evidence of their willingness to be help- ful, and perhaps to some extent of their interest in the project, that not a single director declined to be interviewed or was unavailable because of scheduling conflicts. To describe statistically the "typical" library for any given variable (rate of col- lection growth, materials expenditures as a proportion of the total, etc.), the me- dian, that point on an arrayed scale where half the observations fall above it and half below, was chosen as the mea- sure of central tendency. This has been the method used by the Association of Research Libraries for many years. The median was also supplemented with the "interquartile ranges," those points which lie halfway in each direction be- tween the median and the farthest ob- servation. Hence, readers can quickly determine the values which incorporate three-fourths of the observations, from an (unknown) end point value through the value expressed by the quartile on the opposite side of the median. Because it was desirable to include the 1960s within the coverage of this study and because neither the Bowdoin List nor the ARL Statistics existed at the be- ginning of that decade, other sources of information had to be consulted in order Collection Growth and Expenditures 7 to capture the data for 1960/61. For the colleges, I relied on the American Library Directory, 1962 and obtained data for thirty-three of the forty-two Bowdoin List college libraries in 1960/61.7 In that same volume, five other colleges on the list reported data for 1959/60 and four for 1961/62; these were not used. For infor- mation about collection size among the research libraries, a list of the forty-two largest university libraries in the country was used, compiled by staff at Princeton University and entitled "Statistics for George Piternick's sensible observation is worth repeating: 'Statistical inference always involves risk; it is essential, there- fore, that any inference be made with much care and some humility.' College and University Libraries for the Fiscal Year 1960/61."8 Because total li- brary expenditures were not provided in the Princeton statistics, this article con- tains no 1960/61 financial data for the forty-two research libraries. A CAUTIONARY NOTE ABOUT LIBRARY STATISTICS Library statistics can be misleading and need to be approached cautiously. Those used in this article are certainly no exception. George Piternick' s sensible observation is worth repeating: ''Statis- tical inference always involves risk; it is essential, therefore, that any inferences be made with much care and some hu- mility.''9 One problem with statistics is the like- lihood of errors, ranging from minor and occasional to major and frequent. These can occur at the time of the initial count- ing, or when first recording the count, or when the number is transcribed at any of several stages, including the final com- pilation within the library or the compi- lation by the organization or individual issuing the statistics for a group of li- braries. For example, in one edition of 8 College & Research Libraries the ARL Statistics a library's expendi- tures are recorded as follows: $738,188 for materials and binding; $1,088,292 for salaries and wages; $34,819 for other op- erating expenditures; and total expendi- tures of $1,123,101. 10 It is clear that an er- ror was made somewhere. When errors are noticed subsequent to publication, errata sheets are sometimes issued. In addition to errors is the more subtle issue of definitions and categories, over space and over time. Within a group of libraries there will be, at least initially, different opinions about what items should be included in a given category. For instance, in reporting the number of volumes held, should the figure be the bibliographic or the physical count? Should the total reflect just the number of books and bound periodicals, or should it also cover government docu- ments, microform pieces or volume equivalents, or other formats? Should the figure for total expenditures include fringe benefits (which appear on the li- brary's budget sheets at some institu- tions but not at others)? If so, should the fringe benefits be included as a portion of the reported expenditures for salaries and wages? Not only will these practices o:a.• tecording and reporting data vary somewhat between libraries, but over a period of time they may well vary even at the same library, either with changes in administrators or with the same ad- ministrator deciding (or complying with the request of the extramural compiler) to report the figures differently. The college library statistics, like the.ir well-studied ARL counterparts, do re- flect some differences of definition. The data from several of the libraries over time have shown considerable fluctua- tions in the numbers of volumes re- ported. These fluctuations reflect, at least in part, not only weeding (a prac- tice rarely found to a significant degree in research libraries) but also redefini- tion of what to include in the volume count. Moreover, of the thirty-four li- brary directors responding to a question about reporting fringe benefits, seven- teen do not presently include fringes in total expenditures . Of those seventeen January 1991 who do, seven report them as part of the salaries and wages expenditures (thereby obtaining a larger percentage for that category of expenditure and a smaller percentage for "other"). There are also significant differences between institutions in terms of what benefits they offer. The important point to make, however, is that few of the libraries ap- pear to have changed the way they han- dled fringe benefits or student wages be- tween 1967 and 1987. Hence, it is doubtful that such changes have had much impact on the trends described in this article. Beginning with the 1987/88 compilation, however, the Bowdoin List library directors were asked by the com- piler of the statistics to include their stu- dent wages as a portion of their regular salaries and wages, with the result that salaries/wages as a proportion of total expenditures rose from a median of 42.5% in 1986/87 to 44% in 1987/88, while the "other" category declined from 18% to 17.5%. Materials remained unchanged at 38%. One change I made involved the num- ber of staff reported for the ARL libraries for some of the years. Before 1974/75, the ARL statistics for staff excluded student workers; in that year they were included and have continued to be. The Bowdoin List data have always excluded student workers from the staff count, capturing their contribution in an ''hours of stu- dent assistance" category. Hence, for the earlier years of the ARL statistics, FTE student workers were added to the staff figures, resulting in an adjusted fig- ure that makes those years comparable with later ones. 11 A common problem in analyzing data from a group of institutions over a pe- riod of time is that in one year some insti- tutions are included and in another year they are not. The result is that, in effect, one is comparing different groups of in- stitutions. Thus for each of the tables in this report, data for an institution are in- cluded only if that institution's data are also included for each of the years being compared in that table. Consequently, I am not including any library that joined ARL after 1967, which can have an im- pact on the results one obtains and per- haps on the conclusions one reaches. For example, the median total expendi- tures figure for sixty-eight ARL libraries grew by 463% between 1967 and 1987. When the 1967 median expenditure is compared to the median expenditure of all106 ARL libraries in 1987, the increase is only 377% . There were seventy ARL li- braries in 1967, sixty-nine of which have retained that status. Alabama Arizona Boston U. British Columbia Brown* California/Berkeley* California/Los Angeles* Chicago* Cincinnati* Colorado* Columbia* Connecticut Cornell* Duke* Florida State Florida University* Georgetown Georgia Harvard* illinois* Indiana* Iowa State* Iowa University* Johns Hopkins* Joint University* Kansas* Kentucky* Louisiana State* Maryland M.I.T.* McGill Michigan State* Michigan University* Minnesota* Missouri* *Indicates inclusion on the 1960/61list. Collection Growth and Expenditures 9 I. COLLECTION GROWTH It has been forty-six years since the ap- pearance of Fremont Rider's The Scholar and the Future of the Research Library, in which the author observed that research libraries seem to double every sixteen years or so. Rider's thesis has enjoyed a durable and tenacious credibility; as re- cently as 1985 Warren Seibert referred to Rider's "near-venerable findings. " 12 Nebraska* New York/Buffalo New York University* North Carolina* Northwestern* Notre Dame Ohio State* Oklahoma University Oregon Pennsylvania State Pennsylvania University* Pittsburgh Princeton* Purdue* Rochester* Rutgers* St. Louis University Southern California Southern illinois Stanford* Syracuse Temple Tennessee TexasA&M Texas University* Toronto Tulane Utah Virginia* Washington State Washington University, Missouri* University of Washington* Wayne State Wisconsin* Yale* FIGURE2 The ARL Institutions in 1966/67 10 College & Research Libraries Although virtually all of the subse- quent literature on collection growth has focused on the larger university li- braries, Rider himself was not so limit- ing, notwithstanding his book's title. In the book, the first table records collec- tion growth in ten American men's col- lege libraries (including Wesleyan, Amherst, and Bowdoin), while the sec- ond provides similar information for five "Unless a college or university is willing to be stagnant, unless it is willing not to maintain its place in the steady flow of educational development, it has to dou- ble its library size every sixteen years." libraries at American women's colleges (Smith, Vassar, Wellesley, Bryn Mawr, and Mt. Holyoke); thirteen of these fif- teen are today Bowdoin List libraries. And just several pages later the author stated categorically: ''In fact, this may be asserted as almost axiomatic: unless a college or university is willing to be stag- nant, unless it is willing not to maintain its place in the steady flow of educa- tional development, it has to double its library in size every sixteen years, or thereabouts. " 13 By this exacting stan- January 1991 dard, a number of institutions have fallen short. Data on collection growth between 1967 and 1987 were obtained for thirty- eight of the Bowdoin List libraries, by taking those data from the annual com- pilations and also by receiving additions and corrections from many of the thirty- five directors who responded to the sur- vey. These libraries ranged in size in 1967 from 636,437 volumes for the larg- est to 92,892 for the smallest; by 1987, the figures were 996,222 and 151,989 re- spectively. Table 1 provides a summary of the size of collections. 14 In the twenty years between 1967 and 1987, ten of the thirty-eight college li- braries doubled or more than doubled the size of their collections (including the library whose collection grew by 99%). As shown below, the median of the increase in collection size over the twenty-year period was 74.5%. For the first of the two decades, the growth was slightly greater than in the second, with median percentage increases of 33.5% and 30% respectively. Table 2 summa- rizes the data. Calculating from the beginning of the 1960s adds considerably to the number of college libraries which at least dou- bled the size of their collections by 1987. If two libraries that increased by 98% TABLE 1 NUMBER OF VOLUMES, 1967 TO 1987, THIRTY -EIGHT COLLEGE LIBRARIES 1966- 67 317,342 222,051 173,172 TABLE2 1976- 77 417,920 309,299 231,017 PERCENTAGE INCREASES IN NUMBER OF VOLUMES 1967 TO 1987, THIRTY-EIGHT COLLEGE LIBRARIES 1966-67 49 .5 33.5 26 1976-77 35 30 18 1986- 87 530,327 395,021 309,115 1986-87 97 74.5 54.5 Note : For this and subsequent tables showing percentage increases, the procedures followed were the same : calculating the per- centage increase for each library for the indicated period; arraying the percentages in descending order for each period; identifying the median of the array, and the third and first quartiles. When a mid-point falls between two data points, the value is reported as the average of those two points . By comparison, the median collection size, as opposed to the median percentage of growth, rose by 78% over the twenty years, 30% in the first decade and 28% in the second. and 99% are counted, there are twenty- one of them, or about two-thirds. (Data for six of the thirty-eight libraries de- scribed in tables 1 and 2 were not avail- able for 1960/61.)" (See table 3.) Naturally, research libraries add many more volumes each year than do college libraries. As shown by these data, their collections also have tended to grow at a more rapid rate. This result is, or course, more difficult with a larger number of volumes on hand at the beginning of the measurement period. (As one college li- brary director stated, "Of course we dou- bled in size over that period of time; we didn't have very much to start with.'') Of sixty-nine ARL libraries, thirty-six grew by 100% or more between 1967 and 1987, while thirty-three did not. Tables 4 and 5 provide summaries. Naturally, research libraries add many more volumes each year than do college libraries. Collection Growth and Expenditures 11 The increase between 1967 and 1977 was considerably greater than in the subsequent decade. Going back to 1960/61, and to a smaller group of the forty-two largest research libraries, all but five of them doubled the size of their collections by 1986/87; of those five, Harvard grew by 65%, Yale by 87% and the other three by between 91% and 95% (see table 6). It is worth noting that the collections of the ten college libraries which at least doubled between 1967 and 1987 (about one-fourth) grew at a faster rate than thirty-three of the research libraries (about half) during the same period. For 1961 to 1987, the ten fastest-growing col- lege library collections (about one-third) increased faster than twenty of the re- search library collections (about hal£). 15 II. 110THER" EXPENDITURES Library expenditures have for many years been divided into three general categories: materials (traditionally- books, periodicals and other serials, usually binding, and often ''other mate- TABLE 3 Q3 Median Q, Q3 Median Q, NUMBER OF VOLUMES, 1961, AND PERCENTAGE INCREASES IN NUMBER OF VOLUMES, 1961 TO 1987, THIRTY-TWO COLLEGE LIBRARIES Vols %!ncr . %!ncr. % !ncr. %!ncr. 1961 1961-67 1967- 77 1977- 87 1967-87 258,556 41 46 34 93 184,500 22.5 32.5 28.5 73.5 134,160 15 25 19 54 TABLE4 NUMBER OF VOLUMES, 1967 TO 1987, SIXTY-NINE ARL LIBRARIES 1966-67 1976-77 1,863,233 2,910,461 1,213,855 1,852,841 982,860 1,446,011 TABLE 5 PERCENTAGE INCREASES IN NUMBER OF VOLUMES 1967 TO 1987, SIXTY -NINE ARL LIBRARIES 1967-77 68 52 33 1977-87 42 32 25 %!ncr. 1961-87 165 124 82 1986-87 3,881,945 2,484,152 1,950,400 1967-87 125 102 69 12 College & Research Libraries January 1991 TABLE6 NUMBER OF VOLUMES, 1961, AND PERCENTAGE INCREASES IN NUMBER OF VOLUMES, 1961 TO 1987, FORTY-TWO RESEARCH LIBRARIES Vols %Incr . 1961 1961-67 Q3 1,652,521 40 Median 1,113,122 32.5 Ql 911,248 25 rials"), salaries and wages, and "other" (everything else). Conventional wisdom has been that the normal division among the three categories was ''60-30-10'': 60% for salaries and wages; 30% forma- terials; and 10% for ''other. ''16 This third aggregation has long been a catch-all for supplies of various kinds, noncapital equipment and equipment mainte- nance, telephone charges, travel ex- penses, interlibrary loan charges, and the like. More recently it has (usually) in- cluded monies for payment to biblio- graphic utilities. Because many libraries report their fringe benefits and student wages expenditures but do not include them under the "salaries and wages" category, these therefore become, de facto, part of the ''other'' category of ex- penses. Still, the smallest of the three catego- ries, "other" expenditures in the Bow- doin List colleges in 1986/87, ranged from a high of $623,670 (and 29% of total expenditures) to a low of $38,079 (and 7%). Not surprisingly, perhaps, this is the category which over the last two dec- ades has experienced the largest relative growth, as shown in table 7. In 1966/67 the median college library spent 8% of its budget on costs other than salaries and wages or materials; twenty years later, it was spending 18%. A subset of this group of the college li- braries for which there are 1960/61 data % Incr . % Incr. % Incr . % Incr . 1967-77 1977-87 1967-87 1961-87 56 38 110 191.5 48 27.5 88.5 161.5 32.5 24 66 120 demonstrates the same overall trend (see table 8). As shown in table 9, the research li- braries display this same general trend, rising from a median expenditure of 6% for "other" in 1966/67 to 13% in 1986/87. Because of differences between the two groups of libraries in terms of what is in- cluded in which expenditure categories, readers should be very cautious about comparing this 13% figure with the 18% figure for the median college library. What is significant, and common to both groups, is the growth of "other" as a proportion of the total. (Because the 1960/61 data for the re- search libraries did not include data on "total expenditures," this article does not provide a second table covering these forty-two libraries in the several sections dealing with expenditures). If significantly larger portions of li- brary expenditures are going to ''other,'' they must be coming from one or both of the remaining two budget cat- egories. The chief contributor, and the only one in the case of the college li- braries, has been the salaries and wages category. III. SALARIES AND WAGES Although still the largest of the three categories, salaries and wages have de- clined sharply as a percentage of total ex- TABLE 7 Q3 Median Ql PERCENTAGE OF TOTAL EXPENDITURES DEVOTED TO "OTHER," 1967 TO 1987, THIRTY-EIGHT COLLEGE LIBRARIES 1966- 67 11 8 4.5 1976- 77 17.5 14 9 1986-87 21.5 18 11.5 Collection Growth and Expenditures 13 TABLES PERCENTAGE OF TOTAL EXPENDITURES DEVOTED TO "OTHER," 1961 TO 1987, TWENTY-EIGHT COLLEGE LIBRARIES 1960- 61 13 9 6 1966- 67 13 8 6 1976-77 18 15 9 1986-87 22 18 12 TABLE9 PERCENTAGE OF TOTAL EXPENDITURES DEVOTED TO "OTHER," 1967TO 1987, SIXTY-SEVEN ARL LIBRARIES 1966-67 8 6 5 penditures. Between 1967 and 1987, among the Bowdoin List libraries the median expenditure for salaries and wages fell from 55% to 42.5%, as shown in table 10. Data from the Bowdoin List subset, which includes 1960/61, indicate that for the colleges this trend began earlier. In fact, the median library in this group matched exactly the 60% funding level for salaries and wages found in the 60- 30-10 guideline, as shown in table 11. The picture for the ARL libraries like- wise shows a decline in the salaries and wages percentage since the 1960s, but not nearly so great a decline, and one which occurred only after an increase between the mid-1960s and the mid-70s. Table 12 summarizes the data. Although by 1987 both the research li- 1976-77 10 8 6 1986-87 17 13 11 braries and the college libraries were spending a smaller proportion (and for the colleges a significantly smaller pro- portion) of their budgets on salaries and wages, they were not spending those dollars on fewer people. Both sets of li- braries experienced growth in the num- ber of emp1oyees over the course of these twenty years, the median college library by 25% and the median ARL li- brary by some 37%. Consequently, al- though the numbers of staff in ARL li- braries are much larger than in the college libraries, the rate of increase in the ARL libraries has still been 50% greater than that in the colleges. At the same time, the percentage increase in the number of librarians has been greater among the college libraries (see table 13). TABLE 10 SALARIES AND WAGES AS A PERCENTAGE OF TOTAL EXPENDITURES, 1967 TO 1987, THIRTY-EIGHT COLLEGE LIBRARIES 1966-67 60.5 55 .5 51 TABLE 11 1976-77 51.5 47 44 SALARIES AND WAGES AS A PERCENTAGE OF TOTAL EXPENDITURES, 1961 TO 1987, TWENTY-EIGHT COLLEGE LIBRARIES 1960-61 64 60 51 1966-67 60 55.5 50 1976-77 51 46 44 1986-87 49 .5 42.5 38 1986-87 48 43 38 14 College & Research Libraries January 1991 IV. MATERIALS EXPENDITURES ited the same general trends (albeit to varying degrees )--an increase in the first and a decline in the second. It is in the case of the third category, materials ex- penditures, that they part company. For Thus far, for the "other" and the "sa- laries/wages" categories, both the col- lege and the ARL libraries have exhib- TABLE 12 SALARIES AND WAGES AS A PERCENTAGE OF TOTAL EXPENDITURES, 1967 TO 1987, SIXTY-EIGHT ARL LIBRARIES 1966-67 1976-77 Q3 60 63 Median 55 58 Ql 52 53 TABLE 13 NUMBER OF STAFF, 1967 TO 1987, THIRTY-FIVE COLLEGE LIBRARIES 1966-67 1976-77 1986-87 Libns . Total Libns. Total Libns. Q3 9.8 23.4 10 25.6 11.8 Median 7 17.5 8 22 10 Ql 5 11.5 5.9 12.9 6.4 1986- 87 54 51 47 Total 32.5 23.7 17.3 Note : Numbers are for full-time equivalent staff . Data for the colleges do not include student workers . Because there are data for only sixteen of the college libraries for 1960-61 and each of the other years reported in these tables, no attempt is made to compare college library staffing in 1960-61 with subsequent years . TABLE 14 PERCENTAGE INCREASES IN STAFF 1967TO 1987, THIRTY-FIVE COLLEGE LIBRARIES 1967- 77 1977- 87 1967- 87 Libns . Total Libns . Total Libns . Total Q3 41 .5 38 27 .5 23 .5 71 70.5 Median 13 20 15 9 40 25 Ql -2 5.5 0 1 5.5 7.5 TABLE 15 NUMBER OF STAFF, 1967 TO 1987, SIXTY-FIVE ARL LIBRARIES 1966- 67 1976- 77 1986-87 Libns. Total Libns . Total Libns . Total Q3 85 312 104 406 113 428 Median 64 213 73 262 87 321 Ql 44 167 54 205 61 255 Note : Numbers are for full-time equivalent staff. Data for the research libraries include student workers, calculated at 1,800 hours per year equalling one full -time staff member. See ARL Statistics for 1966-67. TABLE 16 PERCENTAGE INCREASES IN STAFF 1967 TO 1987, SIXTY-FIVE ARL LIBRARIES 1967-77 1977- 87 Libns . Total Libns . Total 42 13 0 45 19 9 28 10 -2 28 14 4 Libns. 56 30 7 1967- 87 Total 72 37 22 the colleges, the increase in the "other" category as a proportion of expenditures has come entirely from the reduction in the salaries/wages portion. Indeed, the materials expenditures portion has even increased over the years, as seen in table 17. (The median amount expended for materials was $31,000 in 1960/61; $69,000 in 1966/67; $189,000 in 1976/77; and $520,000 in 1986/87-all rounded to the nearest thousand.)(See table 18.) The subset of college libraries with 1960/61 data shows the median library with materials expenditures accounting for 30.5% of the total in that year. When taken together with the information from tables 8 and 11, the median library in each of the three groups shows 60% going toward salaries/wages, 30.5% for materials, and 9% for other, conforming almost exactly to the time-honored 60- 30-10 breakdown. The research libraries, on the other hand, show a decline over the years, Collection Growth and Expenditures 15 with only a partial recovery between 1977 and 1987, as table 19 demonstrates. Another way of looking at the growth of materials expenditures for the three sets of libraries is to compare it with in- creases in the prices of books and peri- odicals. Tables 20, 21 and 22 provide such a comparison. 18 They show that de- spite the concern expressed in recent years about the soaring prices of library books and periodicals, the prices of books and periodicals published in the United States grew much more rapidly between 1967 and 1977 than during the ensuing decade. Moreover, for the most part, the materials expenditures of these college libraries kept pace with those price increases although they certainly fell behind the proliferation of book and journal publishing. Typically, these col- leges spend a considerably larger pro- portion of their materials budgets on books than on journals. 19 The typical re- search library spends over half its mate- TABLE 17 MATERIALS EXPENDITURES AS A PERCENTAGE OF TOTAL EXPENDITURES, 1967 TO 1987, THIRTY-EIGHT COLLEGE LIBRARIES 1966-67 40 35.5 31 1976-77 42 38 35 1986- 87 43 38 35.5 Note : All the data pertaining to " materials expenditures " reflect the inclusion of binding expenditures, which is the traditional approach. It is the one still used among the Bowdoin List libraries and was used for the research libraries until the 1985- 86 ARL Statistics. They do not include the category of " Miscellaneous Materials Expenditures," dollars for which in fact go not for materials, but instead for " expenditures for bibliographic utilities, literature searching, security devices, memberships for the purposes of publications, etc ." (See ARL Statistics)1 7. TABLE 18 MATERIALS EXPENDITURES AS A PERCENTAGE OF TOTAL EXPENDITURES, 1961 TO 1987, TWENTY-EIGHT COLLEGE LIBRARIES 1960- 61 36 30.5 28 1966- 67 41 34.5 31 TABLE 19 1976-77 42 37.5 33 MATERIALS EXPENDITURES AS A PERCENTAGE OF TOTAL EXPENDITURES, 1967 TO 1987, SIXTY-EIGHT ARL LIBRARIES 1966- 67 41 38 .5 34 1976-77 37 32 29 1986- 87 43 38 36 1986- 87 38 34 30 16 College & Research Libraries January 1991 TABLE 20 PERCENTAGE INCREASES IN MATERIALS EXPENDITURES 1967 TO 1987, THIRTY-EIGHT COLLEGE LIBRARIES 1%7- 77 1977-87 1967-87 Q3 213 187 719 Median 152.5 148 518.5 Ql 96 112 390 U.S. Book Prices 130 86 325 U.S. Periodical Prices 207 190 790 TABLE 21 PERCENTAGE INCREASES IN MATERIALS EXPENDITURES 1961 TO 1987, THIRTY-THREE COLLEGE LIBRARIES 1961- 67 1967-77 1977- 87 1961-87 Q3 158 215 182 1,828 Median 119 155 147 1,399 Ql 74 91 118 1,019 U.S. Book Prices 44 130 86 513 U .S . Periodical Prices 42 207 190 1,168 TABLE 22 PERCENTAGE INCREASES IN MATERIALS EXPENDITURES 1967 TO 1987, SIXTY-EIGHT ARL LIBRARIES 1%7- 77 Q3 149 Median 104.5 Ql 64 U .S . Book Prices 130 U.S . Periodical Prices 207 rials budget on journals. For the twenty-year period and the 1967-77 decade, materials expenditures for the median college library rose con- siderably more than for its ARL counter- part. For the 1977-87 decade, the me- dian ARL library was slightly ahead. V. TOTAL EXPENDITURES In addition to the growth and decline of different budget components is the issue of total library expenditures. It is likely that many, if not most, academic librari- ans share the oft-cited view that library budgets in higher education have long been anemic. For example, in a recent ar- ticle in College & Research Libraries, Bar- bara Moran refers to the ''stringent budgets of the '70s and '80s. " 20 "Strin- gency,'' of course, is in the eye of the be- holder, although there is no question that during the 1970s and 1980s, particu- larly when measured in terms of con- stant dollars, library budgets did not 1977-87 1967-87 185 519 160.5 406 115 321 86 325 190 790 sustain the growth they had experi- enced in the 1960s. Table 23 summarizes total library ex- penditures for the Bowdoin List libraries over a twenty-year period. The data from both sets of libraries, Bowdoin List and ARL alike, record a sig- nificant increase in total expenditures for the years under consideration. For com- parative purposes, increases in the Con- sumer Price Index and the Higher Educa- tion Price Index are also provided. The latter index, which is concerned with the prices of those goods and services pur- chased by colleges and universities, has grown at a significantly faster rate than the Consumer Price Index. Nevertheless, the data in tables 24, 25, 26 and 27 show that percentage increases in total expend- itures for both the college and the re- search libraries, even for some libraries in the lowest quartile of each group, have considerably outstripped price increases as measured by the HEPI. 21 Collection Growth and Expenditures 17 TABLE23 TOTAL EXPENDITURES, 1967 TO 1987, THIRTY-EIGHT COLLEGE LIBRARIES 1966-67 240,860 199,786 143,202 To underscore the relative prosperity of the 1960s for academic libraries, table 25 shows that for the median Bowdoin List library total expenditures rose al- most nine times faster between 1961 and 1967 than the Consumer Price Index and more than three times faster than the Higher Education Price Index. For the next two decades, the differences are not nearly so great (and not nearly so great between the CPI and the HEPI, either). For both sets of college libraries, the median library experienced a greater in- crease in total expenditures than the me- dian ARL library, particularly from 1977 to 1987. For the most part, the college directors were not concerned about the shift in ex- penditures to "other." To the question of how they viewed the significantly more rapid growth of the ''other'' ex- penditures category, first in terms of their own library, and then in terms of academic librarianship as a whole, thirty-five directors provided thirty- seven and thirty-eight responses respec- tively. The breakdown of their re- sponses was as follows: Own In Library(#) General(#) ''Very concerned'' 2 2 "Somewhat concerned'' 6 5 ''Neutral'' 2 3 "Fairly satisfied" 2 1 1976-77 574,616 448,911 308,552 ''Very satisfied'' ''As irrelevant, since what is important is having enough money for materials, staff, and 'other' regardless of their relative proportions'' ''As irrelevant for other reasons" "Not the trend here" 1986-87 1,590,942 1,213,180 853,778 Own In Library (#) General(#) 2 1 22 1 37 24 2 38 In terms of their own libraries, eight of the directors (between one-fourth and one-fifth) expressed concern, while four were satisfied. Regarding this trend in the profession, seven were concerned and two satisfied. In both theaters, of course, the great majority of respon- dents considered this relative growth in the ''other'' category of expenditures to be irrelevant. The college library directors were di- vided in their responses to several ques- tions related to collection growth which were raised in the course of the inter- views, and they were unanimous in their responses to one other. Eleven of the directors thought that the number of volumes their library was adding each year would remain constant, six that they would increase, and four that they TABLE 24 Q3 Median Ql CPI HEPI PERCENTAGE INCREASES IN TOTAL EXPENDITURES 1967 TO 1987, THIRTY-EIGHT COLLEGE LIBRARIES 1%7-77 1977-87 184 174 142.5 151.5 107.5 120.5 78 90 89 102 1%7-87 612 505.5 384 238 278 18 College & Research Libraries January 1991 TABLE 25 TOTAL EXPENDITURES, 1960-61, AND PERCENTAGE INCREASES IN TOTAL EXPENDITURES 1961 TO 1987, TWENTY-EIGHT COLLEGE LIBRARIES Total Expend . %!ncr . %!ncr. %!ncr . % lncr . 1960-61 1961-67 1967-77 1977- 87 1961-87 Q3 133,466 126 189 178 1,505 Median 100,797 97 147.5 159.5 1,122 Ql 75,123 83 110 123 934 CPI 11 78 90 274 HEPI 29 89 102 386 TABLE26 TOTAL EXPENDITURES, 1967 TO 1987, SIXTY-EIGHT ARL LIBRARIES 1966- 67 1976- 77 1986- 87 Q3 2,799,073 6,406,850 13,967,683 Median 1,777,012 4,174,622 10,564,074 Ql 1,314,158 3,309,771 7,772,439 TABLE27 PERCENTAGE INCREASES IN TOTAL EXPENDITURES, 1967 TO 1987, SIXTY-EIGHT ARL LIBRARIES 1967-77 Q3 170 Median 135 Ql 98 CPI 78 HEPI 89 would decrease. At the same time, twelve of the directors believed that the number of added volumes could decline to some extent because of telefacsimile, other delivery mechanisms, special ar- rangements with other libraries, etc. Seven thought that the number could not decline, one responded "possibly" and another did not know. Also, eleven of the directors believed that ownership was significantly less important than it used to be in terms of providing access, while seven thought it was not. Taken together, these responses indicate a combination of two factors: that the di- rectors are more willing to contemplate such a decline than are other influentials on campus, and that they believe that such a course is more practicable once ef- fective resource-sharing mechanisms become more common. Finally, not one of the directors re- sponded affirmatively to the following 1977-87 1967-87 161 549 141.5 455 118 361 90 238 102 278 question:" Are we approaching a time of 'no-growth' collections, and, hence, can we stop worrying about in- creasing the amount of space devoted to library materials? Or at least a time of very slight collection growth?'' Sev- enteen directors responded "no," three "not now, but in the foreseeable future," and one director thought that the number of volumes would con- tinue to grow, but in formats that would not require much additional space. One director responded: "No. Show me one no-growth library.'' An- other commented: "The number of volumes and titles will grow, but not in a way that will require much more space. Information will be coming in compact forms. In twenty years most back issues of periodicals will be on disk; presently we devote a lot of space to periodical backfiles. Supplement- ing this development are weeding and the use of compact shelving. We've put our pre-1970 bound periodicals into compact shelving." CONCLUSIONS This study is both heuristic and empir- ical. It may raise as many questions as it answers. Among the most important conclusions are the following: Between 1967 and 1987, about one- fourth of the college libraries in this study doubled the size of their collec- tions; during the same period, about half the libraries belonging to the Associ- ation of Research Libraries grew by at least that same rate. Conversely, three- fourths of these college libraries and half the ARL libraries failed to double the size of their collections in this twenty- year period. It would seem, therefore, that there are by now enough exceptions to the 1 'doubling-every-sixteen-years'' rule for academic libraries to render it highly suspect as a general expectation in the last years of the twentieth century. Although the rate of collection growth is probably slowing, none of the college library directors interviewed believes that he or she is presently facing a 'I no- growth" library situation. The inter- views with directors revealed that many are still very collections conscious. Only four directors think that the number of volumes they are adding each year is likely to decrease in the near future. Notwithstanding the concern ex- pressed in recent years about the soaring prices of library books and periodicals, the prices of books and periodicals pub- lished in the United States grew much more rapidly between 1967 and 1977 than during the ensuing decade. A cor- ollary finding is that, for the most part, the materials expenditures of the college libraries included in this study kept pace with those price increases. Indeed, ex- penditures for materials as a percentage of total expenditures have risen in the college libraries over the last twenty years. However, they have declined in the research libraries over the same pe- riod. The increases in total expenditures for these college libraries and for the ARL li- Collection Growth and Expenditures 19 braries from the 1960s to the 1980s have significantly exceeded the increases in both the Consumer Price Index and the Higher Education Price Index, between 1977 and 1987 as well as between 1967 and 1977. The college libraries have fared better than the ARL libraries. This phenomenon of expenditures rising considerably more than inflation is likely related to the competition among col- leges and universities for better students and faculty and for enhanced reputa- tions.22. There are by now enough exceptions to the 'doubling every sixteen years' rule for academic libraries to render it highly suspect as a general expectation in the last years of the twentieth century. The 60-30-10 rule, which reflected real- ity in the "typical" Bowdoin List library in 1960, certainly no longer applies ei- ther in the group of colleges studied here, or in the ARL libraries. As of 1986/87, the "typical" library showed a division closer to 40-40-20 in the former group, while in the ARL libraries the cor- responding division is closer to 50-35-15. Kendon Stubbs explicitly, and Jerry Campbell rather more implicitly, have already called our attention to this shift away from 60-30-10 for the ARL li- - braries. 23 The trends recorded here contradict Richard Talbot's contentions in 1984 that II the pattern of library budgetary alloca- tion remains unaffected, '' that salaries and wages as a percentage of library ex- penditures have remained at 60% II since at least 1960," and that "the percentage of the library internal budget for acquisi- tions is fixed.' ' 24 They also demonstrate that Herbert White was in error when he recently asserted (without documenta- tion) that there has been a "transfer of funds from all other sources to the aca- demic library materials budget over the last fifteen years." 25 Conversely, these 20 College & Research Libraries findings also raise questions about as- sertions that libraries generally have funded automation by taking funds from acquisitions. It would be surpising if the college li- braries were able to sustain this high a percentage for materials during the next decade, as they spend more on maintain- ing automated reference products and other automated library systems. In the college libraries studied here, the proportion of expenditures going to the ''other'' category has grown enor- mously, from 9% in 1960/61 and 8% in 1966/67, to 18% in 1986/87. Contrary to authorities such as Barbara Moran and Charles Churchwell, and contrary to the initial supposition of this study, this growth has generally not come at the ex- pense of the materials budget. 26 Instead, expenditures for materials have grown as a proportion of total expenditures, from 31% in 1960/61 to 38% by 1976/77 and holding at that percentage a decade later. (Indeed, data just received for the Bowdoin List libraries in 1988/89 show a 39% figure for the median library.) Rather, the relative decline of sala- ries/wages expenditures has accompa- nied the increase in the other two cate- gories although the numbers of both professional and support staff have grown. The explanation for this set of circumstances is likely that costs for ma- terials, and for items in the ''other'' cate- gory, have risen more rapidly than have the costs of people. Most consumers, in- cluding college and university adminis- trators, will buy goods and services with an eye on economizing, and the services of library workers have been obtainable at a lower rate of dollar increase than have books, journals, supplies, mainte- nance, etc. This phenomenon is likely true of most categories of workers in the United States during recent decades, and it would seem to merit further study. January 1991 Among the college libraries, the growth in materials expenditures as a percentage of total expenditures is likely understated when the investigator takes into account the situation on many col- lege or university campuses regarding audiovisual centers for housing films or videotapes, records, and slides, and for distributing audiovisual equipment around the campus. During the past twenty years or so, a number of audiovi- sual centers were either created within the administrative/budgetary structure of the library or were moved there. Such entities are generally more staff- and equipment-intensive than they are materials-intensive. To the extent that ARL libraries have come to contain me- dia units, their materials expenditures as a proportion of the total are likewise probably understated. 27 Conversely, another factor serves to inflate the reported materials expendi- tures of the ARL libraries. It is widely known that these expenditures include significant amounts for bibliographic utilities and other nonmaterials costs, thus exaggerating the amount actually spent on library materials. One librar- ian, from a medium-sized, non-ARL li- brary, explained his library's practice of charging computerized cataloging costs to its materials budget as follows: "Our 'other' budget categories have not re- ceived the support for growth that our materials budgets have, so we find it log- ical to charge this major expense to ma- terials. " 28 In recent years, the ARL Statis- tics have included ''Miscellaneous Materials Expenditures" (in addition to the more traditional''Other Library Ma- terials") as a separate category to cap- ture these expenditures, but it is likely that the new category does not presently include all nonmaterials costs reported as materials expenditures. As for the col- leges, in only three instances did the Bowdoin List directors indicate that sig- nificant portions of materials funds were spent for electronic services, such as OCLC charges. Several more indicated that they were including as a part of their reported materials expenditures funds for online computer searching (ranging between $2,000 and $9,000 annually). It would be surprising if the college li- braries were able to sustain this high a percentage for materials during the next decade because they spend more on maintaining automated reference prod- ucts and other automated library sys- tems. Some of them may be tempted to follow the lead of other libraries by ''burying'' some of their automation costs in what has traditionally been the materials budget. A subsequent article in this journal will discuss the state of automation within these college libraries and will make certain connections with the findings and opinions reported here. The rate of increase in materials ex- penditures was substantially greater in the college libraries than in the ARL li- braries between 1967 and 1977 and was also well ahead of that in ARL libraries for the 1967-87 period. For 1977-87, those increases were slightly greater in the ARL libraries. For both sets of li- braries between 1967 and 1987, rates of growth in materials expenditures con- siderably outpaced the increases in U.S. book prices, but they fell considerably short of rising prices for U.S. periodicals (with a commensurately heavier burden on the research libraries, which have been devoting a larger proportion of their materials expenditures to periodi- cals than have the college libraries). 29 At the same time, to underscore once again the relative prosperity of the 1960s for academic libraries, between 1961 and 1987 the median increase in materials ex- penditures among thirty-three college li- braries was 1,399% (see table 21), far out- stripping even the 1, 168% increase in U.S. periodicals prices for the same pe- riod. U.S. book prices increased by a comparatively modest 513%. If, in fact, the prices of books and jour- nals rose at a much faster rate between 1967 and 1977 than they have since, and if the rate of increase in materials ex- penditures during these decades has Collection Growth and Expenditures 21 significantly exceeded the increases in book prices, why all the concern and frustration during the 1980s about in- adequate acquisitions budgets? Two rea- sons appear to be especially germane, and they are quite familiar to collection development librarians although proba- bly still not to many college and univer- sity administrators. One has been the rapidly rising prices of scholarly jour- nals, with the bulk of the impact (though by no means all) felt by the research li- braries. These libraries generally serve institutions which are relatively more research-oriented than the colleges and, hence, are more journal dependent than the college libraries. Consequently, they spend not only many more dollars on journals but also a higher proportion of their materials budgets on journals than the college libraries. The other reason, somewhat more subtle, is that scholarly publishing con- tinues to grow, so that even if academic libraries' acquisitions budgets kept pace with price increases, those acquisitions would continue to constitute, each year, a diminishing fraction of the world's output of recorded information. At the same time, it is not at all clear that this is a new problem; the topic warrants an in- depth study . The number of book titles published in the United States appears to have grown by some 77% between 1966 and 1986 (30,000 titles to 53,000), yet grew by 100% during the much shorter period between 1960 and 1966 (15,000 to 30,000). 30 In 1974, the Faxon Company's database held 38,000 serial titles as "active" and available for pur- chase; by 1988 that number had grown to 105,000 such titles. 31 Perhaps it is time that more academic librarians occasion- ally adopt the skepticism articulated by the director of one major research library early in 1990 at a public forum: "Perish the thought that any academic thought will go unpublished and that we will fail to store it. " 32 22 College & Research Libraries January 1991 REFERENCES AND NOTES 1. Lawrence Thompson, "Suggestions for Statistical Records, I," College & Research Libraries 6:210 (June 1945); Warren J. Haas, "Foreword," in Martin M. Cummings, The Economics of Re- search Libraries (Washington, D.C.: Council on Library Resources, 1986), p.8. 2. College & Research Libraries 4:153 (March 1943), and 21:316 (July 1960). 3. Jim Skipper, executive director of ARL, to directors of ARL libraries, July 26, 1963; copy in au- thor's possession. 4. Richard Harwell, library director at Bowdoin College, to thirty-six college library directors, Oc- tober 23, 1967, copy in author's possession. 5 . The author is much indebted to Arthur Monke, library director at Bowdoin College, for provid- ing a complete set of the data. 6. Readers interested in seeing a copy of the questionnaire should consult Richard Hume Werk- ing, "Collection Growth, Expenditures, and Automation in Academic Libraries: A Prelimi- nary Inquiry," Educational Resources Information Center, Document #318482, 1990. A report on the automation portion of the project is forthcoming in the next issue of this journal. 7. The American Library Directory, 1962 (New York: Bowker, 1962). 8. The author is indebted to Robert Molyneux of the Graduate Library School, Louisiana State University, and to Kendon Stubbs, Associate Library Director at the University of Virginia, for supplying a copy of this list. 9. George Piternick, "ARL Statistics-Handle With Care," College & Research Libraries, 38:419-23 (Sept.1977). See also Kendon Stubbs, "Apples, Oranges, and ARL Statistics," Journal of Aca- demic Librarianship 14:231-35 (Sept. 1988); Robert M. Hayes, Ann M. Pollack, and Shirley Nordhaus, "An Application of the Cobb-Douglas Model to the Association of Research Li- braries," Library and Information Science Research . 5:297-306 (1983); Eli M. Oboler, "The Accu- racy of Federal Academic Library Statistics," College & Research Libraries 25:494-96 (Nov . 1964). 10. ARL Statistics for 1976/77. 11. The figures for student work were always supplied with the ARL Statistics. The change in 197 4/75 amounted to including an FTE equivalent of student workers as part of a ''Total Library Staff" figure. 12. Warren F. Seibert, "How Libraries Grow: A Brief Look Backward (and Forward)," Journal of Academic Librarianship 11:22 (Mar. 1985). 13. Fremont Rider, The Scholar and the Future of the Research Library (New York: Hadham, 1944), p.3-5, 9 (emphasis in the original). 14. The author's ERIC document, cited above, contains more complete data for each of this arti- cle's tables . 15. Ibid., Appendix D. 16. Richard J. Talbot, "College and University Libraries: Lean Years and Fat Years-Lessons to Be Learned," The Bowker Annual of Library and Book Trade Information (New York: Bowker 1984), p .77-81; Kendon Stubbs, "Introduction," ARL Statistics, 1987-88 (Washington, D.C.: ARL,1989), p.8; Barbara Moran, "The Unintended Revolution in Academic Libraries : 1939 to 1989 and Beyond," College & Research Libraries 50:30 (Jan. 1989); Jerry D . Campbell, "Academic Library Budgets: Changing 'The Sixty-Forty Split,' "Library Administration & Management 3:78 (Spring 1989). · 17. ARL Statistics, 1986-87, p.56. 18. The price data for books and periodicals are from The Bowker Annual of Library and Book Trade Information for the following years: 1963, p. 95-96; 1968, p.103, 105; 1978, p.318, 320; 1979, p.337; 1988, p.426-27. This information about price increases for books and periodicals is of- fered as a comparison to the materials expenditures increase. Because the data are for U.S. publications only, they do not capture price trends during these years for foreign publications, of particular significance for the research libraries. To provide a price for the cost of books in an academic year, an average price was derived for the two relevant calendar years. Periodicals, on the other hand, are paid for in advance for an ensuing calendar year; hence, price informa- tion for the appropriate calendar year was used. 19. Data concerning serials expenditures as a percentage of materials expenditures for a group of sixty-two liberal arts college libraries, in the author's possession. · 20 . Moran, "Unintended Revolution," p.29. Collection Growth and Expenditures 23 21. Data from which calculations were made for both indexes may be found in U.S. National Cen- ter for Education Statistics, Digest of Education Statistics, 1988 (Washington, D.C .: U.S. National Center for Education Statistics, 1988), p.35. The HEPI data for 1986-87 were obtained by tele- phone from the National Center. See also D. Kent Halstead, Inflation Measures for Schools and Colleges (Washington, D.C.: National Institute for Education, 1983), p.50-51. 22. President Hanna Gray of the University of Chicago recently noted that in the years ahead, universities and colleges will face ''sharpened competition for students, faculty and dollars at all levels." The New York Times, December 17, 1989, I, p.1. See also William 0. Beeman," As- sessing Intensive Computing on the College Campus,'' in Integrated Planning for Campus Infor- mation Systems, Daphne N . Layton, ed. (Dublin, Ohio: OCLC, 1989), p.12. 23. Stubbs, "Introduction," p.8; Campbell, "Academic Library Budgets," p.78. 24. Talbot, "College and University Libraries," p.77, 80. 25. Herbert S. White, "Pseudo-Libraries and Semi-Teachers," American Libraries 21:105 (Feb. 1990). 26. Moran, "Unintended Revolution," p.30; Charles Churchwell, remarks at a conference of the Florida Chapter of the Association of College and Research Libraries, November 4, 1988, sum- marized in ''The Academic Library Is More than an Information Center: Report on the Confer- ence," by Betty D. Johnson, in CLS Newsletter (Spring 1989), p .6. 27. MichaelS. Freeman, library director at Haverford College, provided author with this insight. 28. Sherman Hayes,"Budgeting for and Controlling the Cost of Other in Library Expenditures: The Distant Relative in the Budgetary Process," Journal of Library Administration 3:129 (Fall/Winter 1982). 29. See note 18, above. 30. The Bowker Annual of Library and Book Trade Information, 1962, p.59; 1967, p.45; 1988, p.403. To be sure, the proportion of these titles that would be appropriate acquisitions for academic libraries is not readily apparent. 31. Rebecca T. Lenzini, "Serial Prices: What's Happening and Why," Collection Management 12:23 (1990). 32. Discussion sponsored by the Financial Management Committee, Library Organization and Management Section, Library Administration and Management Association, Chicago, Janu- ary 8, 1990. Statement of Changes Beginning with this issue, College & Research Libraries appears in a slightly different format. The size of the journal is now smaller to take advantage of standard paper size economies.