This item is only available as the following downloads:
The Future of the Library A Report to the Provost University of Florida December 2006 Committee: John Cech Amy Mashburn David Denslow Carol Murphy Daniel Driscoll Robin Poynor Mary Ann Eaverly Betty Smocovitis Stephanie Haas Michele Tennant Willard Harrison Eric Triplett Tess Kulstad Robert Zieger
Summary The somewhat daunting task posed by the Pr ovost was to consider the future of the academic library, both in t he national sense and, more spec ifically, at the University of Florida. Her interest in these issues was sparked by the rapid changes occurring in libraries over recent years, changes that present challenges and opportunities. Given the importance of libraries in the academic community, are the UF libraries responding accordingly or poised to respond? She a sked the committee to consider the next decade: what critical issues should Cent ral Administration focus upon for maximum benefit to the library? And of more imm ediate moment, the impending retirements of the library directors in both the Smather s and Health Center libraries mean new leadership, with potentially new directions. Our charge was to prepare a repor t with recommendations. The committee began by educating itself on t he current status of academic libraries by reading the latest literature, gaining input from k nowledgeable faculty, staff, and students, and consulting with outside experts. From that study, five issues surfaced for detailed consideration. The body of this report offers details, data, and rationales that flesh out the following notes: Comparative Analysis detailed comparisons of UF libraries with relevant peers indicate UFs spending lags theirs by r oughly a third, illustrating a need for increased operating funds, as well as ce rtain types of new space. Librarians and Library Services the increasingly more complex job duties of librarians calls for hiring those trained as experts in new technology, the recognition of this professionalism, and the introduction of pr ograms to gain full use of such talents. Collection Balances the inevitable shift toward di gital acquisitions should not mask the need for continued attention to printed materials, particularly in the humanities. Role in the Capital Campaign using as models the library targets in other academic capital campaigns, UF should assume an appropriate funding go al (suggested $25 million) that recognizes the cent ral importance of the library. Integration of the UF Libraries after a comparison of many important issues, advantages and disadvantages, the current separate system is favored. The committee comes away with a r enewed appreciation for the absolute importance of the library in a major research university, wher e it should occupy a central role. Indeed, the library has been called the DNA of the academic institution, determining our quality and directions. UF is fo rtunate to have a library that is vital and valued, but like many libraries, it stands in a precarious position. We believe that attention to the recommendations of this Report could strengthen the UF library system in these changing times that pres ent both challenges and opportunities. 2
Table of Contents Summary.......... ........... 2 Introduction 4 I. Institutional Comparisons.. 5 II. Role of the Library and Library Services.............17 III. Balancing the Collections. 22 IV. Role of the Library in UFs Capital Campaign.. 27 V. Integration of the Libraries... 32 Appendix 1. 42 Appendix 2. 46 3
Introduction The Library of the 21 st Century: An Evolving Institution The library of the 21 st century builds on a number of age-old and enduring traditions. As an institutional site, it continues to se rve traditional roles such as the safekeeping, organization, classification, and dissemination of knowledge. Knowledge is to be understood as any potentially useful informati on that may find expression in traditional sources, such as print media (publishe d and unpublished, privat e or government in origin), non-traditional sources, such as film, or audiovisual materials (records and sound tapes), historic or ar chival documents, and publications in print or digitized format. The library of the 21 st century also serves a number of new and ever-expanding roles enabled by a diverse array of new technologies. It is therefore simultaneously physical and virtual. As a physical place, it should serv e as the intellectual heart of the modern research university, marked by the appropriate architectural signs of its importance. The centrality of the library is material (b ricks and mortar) and symbolic. It signals the commitment of the instituti on to academic excellence, c onfirming and rendering visible the Universitys missions in research, educat ion, and public service. It is a set of buildings (consisting, at UF, of the main libra ry, the Health Sciences Center Library, the Lawton Chiles Legal Information Center, and the branch libraries), but also a configuration of internal spaces within t hese facilities designed to respond to the changing nature of academic enquiry. Such sites might include, for example: spaces that foster interdisciplinar y and collaborative research activities; spaces that offer university-wide wireless access; spaces that accommodate lectur es, exhibits, and small colloquia; spaces that facilitate the activities of an innovative Library Support group, such as the Howe Society. Consultant Ken Frazier, Directo r of the Libraries at the University of Wisconsin, pointed out the importance of comm unication between libraries and other spaces/places of learning on the campus. At UF, there is potential for collaboration between the library and, for example, the Florida Museum of Natural History, the Samuel P. Harn Museum of Art, and the Genetics Inst itute, among other sites. As a virtual space, the library connects researchers and their communities through an ever-expanding global network. A modern resear ch library facilitates new opportunities for research by integration of traditional sources with the oppor tunities made possible by technological advances. An excellent example of the enhancement of library research by technology can be seen at the University of North Carolina at Chapel Hills Health Sciences Library which enjoys a strong part nership with their Renaissance Computing Institute ( http://www.hsl.unc.edu/Co llaboration/index.cfm ). At UF, possibilities for virtual enhancement of the library exis t with, for example, the REVE ( Research, Education and 4
Visualization Environment ) and the NAVE (New Automated Virtual Environment) Lab of UFs Digital Worlds Institute in the College of Fine Arts. The library of the 21 st century is thus a time-honored site enabled by a diverse set of technologies that expresses and facilitate s humanitys need to order and understand its world by recovering, analyzing, archiving, pr oducing, and transmitting knowledge in both local and global contexts. I. Institutional Comparisons In considering what the future holds for UF libraries, it was first important to determine where we now stand, establishing a baseli ne for recommendations in this report. Traditionally, academic libraries have been ranked by the Association of Research Libraries (ARL). Currently, the ARL bases it s rankings on five measures: volumes held, volumes added, current serial titles, total operating expe nditures, and number of professional and support staf f. As examples, the 20032004 rankings show Michigan State ranked as 39, Florida 38, Arizona 30, Oh io State 21, Wisconsin 11, Berkeley 4, and Harvard 1. Libraries with top ARL rankin gs are usually at large institutions and have histories of stabl e and adequate funding. In this report we compare the University of Floridas library funding, both total and by major category, to that at other libraries. We choose two comparison groups. The first is ARL libraries at five large land-grant inst itutions(a peer group) with medical and law schools, and the second is all American universit ies that report data to ARL. Universities in the first group are close to us in size and mission, so that unadjusted expenditure comparisons are meaningful. That is not the case with the second group, which contains over eighty universities with more widely varying size and missions. To make comparisons meaningful, statistical methods are used to adjust expenditures for the number of students, t he number of faculty, the pres ence of law and medical schools, and U.S. News & World Report rankings. For both comparis ons, the latest statistics available represent t he 2003-2004 academic year. The institutions selected for comparison to Florida in the peer group are: Michigan State, Ohio State, the University of Arizona, the University of Minnesota, and the University of Wisconsin. For some comparisons, Arizona and Michigan State are omitted because their colleges in the health sci ence fields are very different from ours. Data Sources for Comparisons Unless otherwise noted, the library data pr esented is from the 2003-2004 statistical data collected by the Association of Research Libr aries. It is available at an interactive search site maintained by the University of Virginia Library at http://fisher.lib.virginia.edu/cgi-lo cal/arlbin/arl.cgi?task=setupreport The health science center library data is from the 2003-2004 ARL Academic Health Science Library 5
Statistics available at http://www.arl.org/stat s/lawmed/medindex.html Data for Michigan State does not include health science library statistics. Library-Institutional Factors In comparing libraries at the five peer institutions, we first present their major library user populations: undergraduates, graduate students, and facult y. Second, we compare their libraries resources measured by resear ch and instructional staffing, facilities, collection building, and expenditures. Finally, we normalize expenditures by degrees granted and by programs offered. User Populations In 2003-2004, the Association of Res earch Libraries statistical report [http://fisher.lib.virginia.edu/arl/ index.html] indicated that the user populations for each university were: Table 1. User populations at each institution, 2003-2004 Arizona Florida Michigan State Minnesota Ohio State Wisconsin Full time students 30,149 42,042 38,355 35,071 43,796 36,672 Full time graduate students 5,932 10,825 7,657 8,898 10,212 9,493 Full time faculty 1,428 2,865 1,972 1,626 2,967 2,076 Total 37,509 55,732 47,984 45,595 56,975 48,241 Ohio State and UF have the highest major user populations, over 55,000 each, divided in similar proportions among undergraduates, graduat e students, and faculty. Michigan State, Minnesota, and Wisconsinthough also very large have user populations of fewer than that 50,000. Library Staff involved in Research and Instruction Table 2 shows the library professional and s upport staff at each unive rsity. In academic libraries, professional and support staff are the most heavily involved in providing instructional, reference and research serv ices. The counts include staff at various branch libraries. Florida includes the st affing of all ten branches in its count. 6
Table 2. Library Staff supporting instructional and research Arizona (L, M) Florida (B,L,M) Michigan State (L) Minnesota (L,M) Ohio St. (L,M) Wisconsin (L,M) Professional Staff* 100 112 81 98 121 217 Support Staff 159 208 123 200 159 185 Total 259 320 204 298 280 402 Includes: B (branches), L (law), M (medical) Of note here is the fact that Wisconsin has al most twice Floridas professional staff, although UF serves approximately 7,500 more users. Aside from Wisconsin, UFs professional staffing is roughly in line with that of the other institutions, with a relatively large support staff. Facilities Only two data factors related to facilities are touched upon in this report: branch libraries and square footage of the main library. Branch libraries ar e important because of their staffing and collection implications. Table 3. Number of branch libraries at each institution Arizona Florida Michigan State Minnesota Ohio State Wisconsin Branch libraries 3 10 14 5 20 40 With the acceleration in the use of elec tronic resources, squar e footage has become less useful as a measure of collection richne ss. Nonetheless, it does provide a heuristic measure of collection breadth and depth. It is also related to Fraziers concept of an academic library as a social-academic collabor atory, or workspace. The size of each main library is given in Table 4. Table 4. Square footage of main libraries Institution Square Footage Arizona 263,696 Florida 177,000 Michigan State 313,936 Minnesota 382,213 Ohio State 299,695 Wisconsin 363,085 7
Collection Building by Book and Byte Another section of this report ( III Balancing the Collections ) analyzes the collections at UF. Here, we compare collection sizes and expenditures, though the availability of electronic resources has made total volume counts less meaningful for some purposes as we migrate to electronic access. This is especially true for some of the sciences. For many fields in the humanities and social sciences, however, access to large print collections still determines the richness of a library collection. Table 5 below shows total collection sizes reported in 2004. Table 5. Total volumes held in universities and total medical volumes Collection building in 2003-2004 Table 6 shows collection building during 200 3-2004. Florida purchased fewer current serial titles (25,330) than any of its five peers. Ohio State, at 35,561, was next lowest. Wisconsin bought over 55,000 current serial titl es. In terms of mo nographs purchased, Florida bought approximately 47,500 (third lo west); both Ohio State and Wisconsin each purchased some 20,000 more. Though we cannot say for sure without a more thorough look at monograph purchases, it is likely t hat the low monograph and serial purchases have the greatest impact on t he humanities and social scienc es. Electronic journal subscriptions may partially balance the serials situation portrayed here, but generally electronic journals are more closely aligned with science and technology. Collection Building Expenditures Institution Total Volumes Total Medical Volumes Arizona 5,201,065 (L,M) 233,682 Florida 4,075,290 (L,M,B) 348,046 Michigan State 4,747,959 (L) Not included in ARL reports Minnesota 6,374,293(L,M) 477,861 Ohio State 5,809,505 (L,M) 214,766 Wisconsin 7,807,097 (L,M,B) 315,657 Includes L=law, M=medicine, B=branch libraries Table 6. Total current serials and monographs purchased, 2003-2004 Arizona (L,M) Florida (B,L,M) Michigan State (L) Minnesota (L,M) Ohio St. (L,M) Wisconsin (L,M) Current Serials 36,060 25,330 37,880 35,801 35,561 55,164 Monographs purchased 56,666 47,528 30,693 37,161 67,671 68,483 Includes B (branches), L (law), M (medicine) 8
The expenditures listed below (Table 7) show the main classes of materials that were purchased in 2004. The total materi als expenditures include formats beyond monographs and serials, but these are comparably insign ificant. The total materials figure includes electronic jour nals, computer files (literat ure collections, periodical back files, etc.), and har dware and software. Table 7. Monographs, serials, and total materials expenditures for 2004 Arizona Florida Michigan State Minnesota Ohio State Wisconsin Monographs $2,683,130 $1,695,403 $1,802,797 $2,280,527 $3,507,436 $2,638,090 Serials $8,874,139 $7,217,600 $6,570,003 $9,291,150 $9,113,712 $6,727,764 Total materials $12,638,919 $10,167,169 $8,778,721 $11,847,047 $12,621,148 $10,596,306 Electronic vs. Print Expenditures Because electronic resources are often bundled fo r sale, it is extremely difficult to pull out individual number count s. Often several hundred elec tronic books and journals are available under one licensi ng agreement. However, some expenditure comparisons (Table 8) offer further insight into electronic vs. print formats. Of interest in the serial situation is the fact that Florida and Wisconsin each spend less than two million dollars on electronic journals, while Arizona, Minnesota, and Ohio State spend about four million each. Initially, it appears that bo th Wisconsin and Florida may fail to acquire a large number of electronic journals; however, it is possible that consortial agreements have result ed in significant cost savings. It is unlikely, however, that Florida and Wisconsin have been in str onger bargaining positions than their peers. Table 8. Comparison of total serial to electronic serial expenditures for 2003-2004 Arizona Florida Michigan State Minnesota Ohio St. Wisconsin Table 9 shows that Florida spends more on computer file s than Ohio State and Minnesota, and less than Arizona, Michigan St ate, and Wisconsin. With respect to spending on computer hardware and software, Florida lags behind all of the peer Total Expenditure for Serials $8,874,139 $7,217,600 $6,570,003 $9,291,150 $9,113,712 $6,727,764 Expenditures for Electronic Journals $3,905,137 $1,801,347 $2,581,546 $3,946,325 $4,247,736 $1,605,662 9
institutions and very far behind Wisconsin, s pending only 24% as much as the average of the five peers. Even excl uding Wisconsin, Florida spends only 36% as much as the average for the other four. Table 9. Expenditures on Computer files and hardware/software for 2003-2004 Arizona Florida Michigan State Minnesota Ohio St. Wisconsin Computer Files $488,191 $236,314 $387,800 $131,403 $31,837 $433,597 Hardware/software $544,980 $227,038 $778,925 $676,862 $553,375 $2,168,246 Overall Expenditure Table 10 compares overall expenditures. The striking difference in staffing by Wisconsin noted earlier is also reflected in the sala ry category. Wisconsin spends more than any other university on salaries, nearly $10 million more than UF. Table 10. Expenditures by main categories, 2003-2004 Arizona Florida Michigan State Minnesota Ohio State Wisconsin Total Salaries $11,150,833 $11,711,598 $10,379,727 $15,770,857 $12,880,990 $21,584,660 Total Library Materials $12,638,919 $10,167,169 $8,778,721 $11,847,047 $12,621,148 $10,596,306 Other Operating Expenditures & Miscellaneous Materials $3,694,102 $3,086,012 $3,444,330 $3,798,669 $2,731,127 $7,095,014** Total Expenditures $27,064,875 $25,112,380 $22,557,590 $31,640,604 $28,509,784 $39,251,812 Notes: Includes L=law, M=medical, B=branch *No university spent more than $550,000 on miscellaneous materials. **Wisconsin indicates that a substantial amount of Other Operating Expenditures was prepaid in 2002-2003. Discussion As noted earlier, the purpose of our comparison with a lim ited number of peer institutions is to avoid complications that arise from the differing needs of institutions with large differences in size and mission. To focus even more strongly on institutions that closely resemble Florida and have co mparable medical schools, we now further limit the comparison group to Minnesota, Oh io State, and Wiscons in, with the caveat that Minnesota and Wisconsin have noticeab ly fewer undergraduates than either Ohio State or Florida. They do have more faculty, however. We use the more restricted peer group to co mpare spending relative to the number of degrees granted and to the number of degree programs, which serve as additional 10
measures of user needs. In 2003-2004, UF granted more total degrees than Minnesota, Ohio State or Wisconsin. It lagged Mi nnesota and Ohio State in one category, professional degrees. Table 11. Degrees granted 2003-2004 Florida Minnesota Ohio State Wisconsin Bachelors 8,574 6,049 8,288 6,156 Masters 3,022 2,677 2,606 1,968 Doctorate 694 592 560 627 Professional* 539 715 782 631 Total 12,290 10,033 12,236 9,382 *includes dentistry, nursing, public health, law, medicine, veterinary medicine, and pharmacy Also shown is a comparison of the majors offered for advanced research degrees. Counts were taken from the 2003-2004 graduate ca talogs for each of the universities, when available online. The registrars office at Ohio State was contacted directly. The Table 11 counts do not include professional degrees in law and medicine, but do include the doctorates associ ated with health sciences. Because Florida offers almost 100 more res earch majors at the masters level than any of the other three universities, we expect ed that the collection breadth and depth and the library services would be larger and be reflected in higher levels of funding. This is not the case, however. Floridas spending is actually lower. Consequently, the spending per program at Florida is far below that at the ot her institutions. Table 12. Majors offered for advanced research degrees, 2003-2004 Florida Minnesota Ohio State Wisconsin M.A./M.S. 233 112 127* 155 Ph.D. 124 101 95 108 as of Fall 2005 Table 13 below shows funding available to support the doctoral degree research programs, assuming that under graduate needs are sublimat ed, and expenses for the law and medical libraries are removed. Fo r Florida, before computation we have subtracted five doctoral research programs se rved by the medical library. That is, the funds available for non-law and non-medical pr ograms are divided by 119 instead of 124 to obtain the $153,655 average per program for Florida. Fo r Minnesota, Ohio State, and Wisconsin, we have not subtracted progr ams that may be served by their medical libraries. Consequently their numbers may be biased downward. Even so, they average 68% more funding per program than Florida. 11
Table 13. Potential monies available to support Ph.D. research programs, 2003-2004 Florida Minnesota Ohio State Wisconsin Total library expenditures 25,112,380 31,640,604 28,509,784 39,251,812 Subtract Medical library expenditures -4,340,931 -5,044,351 -3,413,349 -3,805,672 Subtract Law library expenditures -2,486,555 -3,414,887 -2,443,223 -983,935 Monies available for non-law/nonmedical 18,284,894 23,181,366 22,708,171 33,002,917 Potential expenditures per program $153,655 $229,518 $239,033 $305,583 The Longitudinal View To add perspective to our analyses for 2003-2004, we present tota l inflation-adjusted expenditures for the year s 1999-2000 to 2003-2004 in Table 14 Table 14. Total library expenditures 2000-2004 (thousands of constant 2003-04 dollars) Fiscal Year Arizona Florida Michigan State Minnesota Ohio State Wisconsin 1999-00 $24,870 $25,597 $20,268 $32,963 $28,876 $34,390 2000-01 $25,296 $30,330 $20,100 $32,159 $30,149 $35,574 2001-02 $26,528 $26,190 $20,298 $33,839 $29,019 $35,046 2002-03 $26,272 $26,746 $21,112 $32,167 $27,694 $40,224 2003-04 $27,064 $25,112 $22,557 $31,641 $28,510 $39,252 Expenditures become more meaningful wh en compared to the potential user populations, as shown in Tabl e 15 for the same years. 12
Table 15. Changes in user populations, 1999-2000 to 2003-2004 Arizona Florida Michigan State Minnesota Ohio St. Wisconsin Total 1999-2000 33,781 43,574 42,693 37,636 51,040 45,416 Total 2000-2001 33,937 45,408 44,216 39,618 46,659 46,400 Total 2001-2002 35,796 52,910 45,058 41,428 52,606 47,258 Total 2002-2003 36,846 54,025 46,766 43,926 54,445 45,250 Total 2003-2004 37,509 55,732 47,984 45,595 56,975 48,241 Increase in populations 2000-2004 3,728 12,158 5,291 7,957 5,965 2,825 Total fulltime undergraduates, graduate students, and faculty From Tables 14 and 15 we calculate that wh ile Floridas user population rose by 28%, more than at any of the other institutions, its inflation-adj usted total library spending fell slightly, by 2%, so that real spending per user fell by 30%. Minnesota and Ohio State also experienced spending decreases, of 3% and 1%, while their user populations rose, by 21% and by 12%. Thus they had real de clines of approximately 24% and 13% per user (vs. Floridas 30%). M eanwhile, Arizona, Michigan State, and Wisconsin had real spending increases of 9%, 11%, and 14%. Their user populations rose by 11%, 12%, and 6%. Thus, Florida had a greater drop in library spending per user than any of the comparison institutions. Comparison to a Larger Set of Universities Though comparison to a select group of inst itutions has the st rong advantage of requiring less worry about just how library ne eds vary according to such university differences as size and mission, there remain s the concern that t he results may depend strongly on the particular institutions sele cted. For this reason we include, in an appendix (See Appendix 1), a statistical analysis of library spending at more than eighty research universities. For the anal ysis, we augmented the ARL data from two other sources: the 2006 U.S. News & World Report rankings and spreadsheets on the web site of The Center, a research group at the University of Florida that compares research spending and other measures across a large number of uni versities. From results of four sets of r egressions reported in the appendix, as well as many others run as checks of robustness and not reported, a conclusion that emerges is: Floridas total library spending is lower than expected for a university of its size and quality, and spending on total electronic materials is remarkably low. Another finding is striking: An 18-point higher U.S. News rating is associated with about 36% more library spending, controlling for size and the presence of medical and law schools. Eighteen points represents t he difference between, say, 59 (Florida, Texas, UC Santa Barbara, UC Davis) and 77 (Berkeley, Carnegie Mellon). The result is very similar using peer evaluations instead of the US News total 13
Since 1986, the library has never receiv ed the funding that would be expected, considering the size and mission of the Univer sity of Florida, and has more often than not received less than 80% of the amount that would put it in line with other institutions. Inadequate funding has been an underly ing factor throughout the history of the library system at UF. In 1982, Dr. Robert M autz chaired a fact-finding Task Force on Evaluation of the Libr ary Management Policies and Practices. He reported: Most of the problems the library faces arise from the fact that operati ng resources have not been commensurate with the growth of the faculty, student body, collections, complexity of information, automation, nor to the addition of constituencies not previously served. Those findings are equally valid twenty-four years later. The funding issues are exacerbated by the soaring costs of some electronic resources, by the development of a plethora of di screte information services on campus and beyond, and by changes in research, teaching and learning modalities in the university environment. These factors are impacting all facets of t he university, not just the library. Thus, summary recommendations are: 1) Increase library funding to appropriate levels to support programs, incl uding new IT and digital functions 2) Support the development of a UF institutional repositor y by the library system as a university-wide service, and as an open access alternative for distributing research results. 3) Revisit space issues, as current space does not appear adequate for collection growth or to create the social-collaborative workspaces suggested by Frazier. 15
The Health Science Center Library The coverage and services of health science libra ries are extremely di fficult to compare. For that reason, they have been omitted from the analyses de scribed above. That is an important omission, and as partial compens ation we add Tables 16 and 17 showing enrollments and expenditures pe r student for Florida, Mi nnesota, Ohio State, and Wisconsin. Table 16. 2003-2004 enrollment for health related colleges Florida Minnesota Ohio State Wisconsin Dentistry 373 556 617 Medicine 906 1969 2,288 1322 Nursing 936 821 1,028 689 Pharmacy 1855 581 788 641 Public Health 1560 610 Veterinary Medicine 511 425 668 371 Total 6141 4962 5389 3023 Ohio State and Wisconsin have Colleges of Medicine and Public Health; Wisconsin has no college of dentistry Table 17. Expenditure per student in health science center libraries, 2003-2004 Florida Minnesota Ohio State Wisconsin Total Expenditures 4,340,931 5,044,351 3,413,349 3,805,672 Total Student Enrollment 6141 4962 5416 3023 Expenditure per student 707 1017 630 1259 In 2005, UFs total enrollment in health re lated colleges was 6914, up 773 students from three years before. With a current budget of $3,774,948, the average student expenditure has now fallen to $546, which represents a 23% decline from 2003-04. Indeed, even without adjustment for enrollment or inflation, the decline in spending is 13%. Though we have not explored the causes of the decline and the categories in which it has occurred, we speculate that it would be difficult for the health science center libraries to improve or even maintain the quality of their services for long in this situation. The University of Florida Health Science Libraries recognize a different set of institutions as peers, based size, diversity of colleges, and other factor s. Information on those peers may be found in the Integration of this report. 16
II. The Role of the Libr ary and Library Services Librarians are essential to the research, te aching, and service missions of the University of Florida. Through innovative services, librarians provide access to information: they evaluate, acquire, organize, and annotate resources; they retrieve, evaluate, and proce ss information for their clients; they teach faculty, students and staff to retrieve, evaluate, and use information to create knowledge; and they build new tools and resources to solve problems of information retrieval, storage, and management. Liaison Librarian Programs: New directions in academic libraries include t hat of the liaison librari an; that is, librarians who are expert in a particular subject area and provide focused, customized and integrated services as follows: they develop and teach course-integrated and stand-alone courses in their subject areas, perform collection develop ment, expert searching, and subjectspecific information consultations; they partner with the academic faculty to create cohesive information-related instruction and resources; and they integrate into their subject-rela ted academic departments while remaining housed in and salaried by the library. Links with departments can include the following: joint appointments; attendanc e at faculty and departmental events; service on curriculum or dissertation commi ttees; or, in the case of clinical or bioresearch information speciali sts, integration into clin ical and research teams. Recommendations to expand liaison library services: that the UF libraries institute a liaison librarian program for the major subject disciplines covered at UF; that the HSCL explore the clinical and bioresearch information specialists in context models, and potential sources to fund such services; and that the UF libraries explore other areas in which discipline experts (librarian or non-librarian) could provide useful non-liter ature-based services to clients, and potential funding sources. Digitization services: Digitization of resources is an essential acti vity for the library. Digitization requires librarians who are expert in preservation; numerous non-professional staff who perform the scanning; experts in meta-data and inde xing who make the digitized material 17
searchable and findable; and the necessa ry equipment and space to perform the digitizing. Recommendations to impro ve digitization services : that the UF libraries continue to apply fo r funding to expand its current digitization efforts; that digitization efforts c enter on the unique UF specif ic collections currently housed in UF collection; that UF libraries be more proactive in terms of gathering and digitizing UFs history. Subscription to electronic journals: Electronic journal prices are rising each year at a rate higher than inflation; UF (or any institution, actually) is unlikely to ever have a budget that can provide online access to every journal of interest to its clients, but efforts should address this problem. Recommendations to increase access to electronic journals: that the HSCL and Marston Science Li brary teach clients how to submit manuscripts to PubMed Central NIHs fr eely accessible online repository of full text article; that the HSCL and Marston Science Librar y provide the service (actually submit the manuscript to PubMed Central for the client) as a convenience for its client; that the UF Libraries educate its us ers on open access and the relationship of tenure and promotion expectations and journals. Development of new information tools : At many academic institutions (particularl y academic medical centers) librarians are involved in the development of new informa tion tools often partnering with informatics or computer science faculty. Librarians are also engag ed in designing better searching systems, the electronic patient record, interactive online tutorials, etc. Recommendation to facilitate new search tools: the exploration of partnerships with relev ant academic faculty, such as computer science or informatics faculty, or increase the hiring of librarians who are trained in such areas. 18
A Commentary on the UF Library Website In this electronic age, a librarys website ca n play an important role in user perception and evaluation. We offer the following observa tions in an informal comparison between UFs Library homepage and those of some of our peer libraries (University of Michigan Ann Arbor; University of Wisconsin-Madison; Ohio State University; University of Minnesota; University of Illinois-Urbana-Ch ampaign; Pennsylvania State University; University of Georgia): UFs Library webpage is too utilitari an (opens up on the catalog with links to mission, history, collections, etc.); It sets a negative tonethe first item on the home page is the announcement of journal cancellation; Elsewhere, the tone is often apologetic, as in the description of our collections: our library is not comprehensive; We do not sufficiently emphasize our str engths, for example our collections, such as the Latin American Studies collection, or the Baldwin Collection of Childrens Literature; We do not include visuals on our webpage, and the colors could be more attractive. Recommendations : that we state the number of volumes in our library (as do all our peers); that the mission statement be re-written to convey more forcefully the importance of the library to the university and its academic mission. See, for example, the mission statement of the Wisconsin librari es, as well as the mission statement and vision of excellence found on the Ohio State pages; that a Welcome from the Dean be added, accompanied by a picture, as in the Penn State library pages; that we more emphatically feature our st rengths. For example, in a rubric that could be entitled Overview of the UF Li braries, more detailed prose about our holdings. The University of Illinois is a good model here; that we make effective use of webpage visu als, as in the University of Georgia library webpages. For example, we coul d include pictures of the new Library 19
West, the Special Collections Room in Li brary East, etc., along with pictures that convey the natural and architectural beauty of our campus. Librarians as Faculty Tenure and promotion criteria: At the University of Florida, librarians ar e faculty, subject to the same tenure and promotion process as academic faculty. Librarian ranks Assistant University Librarian, Associate University Librarian, and University Librarian are equivalent to Assistant Professor, Associate Professor, and Professor. Like academic faculty, librarians have three criteria for tenure and promotion: 1. J ob performance; 2. Research; 3. Service; but the details and we ight differ. Distinction must be achieved in criterion 1; 2 and 3 are considered to have equal weight, with distinction required in one area and good performance required in the other. Recommendation to increase research productivity: that the UF Libraries strengthen the research criterion, such that in order to be considered good, research results mu st be disseminated externally through publication or pr esentation. Hiring policies and procedures: An expert workforce is required for the University of Florida Libraries to successfully move forward into the future. Recommendations: that the UF Libraries strive to hire es tablished experts into their departments in non-administrative roles. Currently, the majority of non-administrative hiring is done at the Assistant University Librarian level, regardless of library/discipline need; that UF Libraries strive to retain its es tablished experts. UF often loses its best people; the libraries appear to lack either the will or resources (or both) to make matching or better offers to these indi viduals, as might be done in an academic department. 20
Administrative positions: Although librarians have facult y rank at the University of Florida, the libraries run on a much more administrative and hierarchica l model than academic departments. This arrangement is typical at most academic libraries; however, such arrangements can lead to stagnant administrations, a lack of vision, and an underutilization of individuals who are not management. Recommendation: that UF Libraries consider a new leadersh ip model, in which department chairs are not permanent, but instead, rotate as they do in many academic departments. The current system fosters an env ironment of us versus them. If department chairs considered themselves more faculty than administration, and knew that they would return to the faculty after their stint as chair, it could bring vibrancy and cohesiveness to the departments and library as a whole. Status of library faculty: The majority of library employees are not faculty (librarians), but instead are staff, including library technical assistants and st udent assistants. In most libraries, the person checking a book out to the user, or th e person shelving the materials, is not a librarian, but is a hardworking and knowledgeable paraprofessional, trained to perform particular duties in the library but lacking t he graduate level degree in library science. Although the libraries could not function without these paraprof essionals, it is important that library faculty be recognized as facult y first and foremost, equal to and partners with academic faculty. However, libraries often do wnplay the faculty status of their librarians (e.g., Smathers library faculty are required to wear badges that identify them as staff a clearly non-faculty connotat ion that can be detrimental to building collaborative relationships with academic faculty.) Recommendation: that librarians (at least public services librarians) explore ways to effectively reach out to academic facult y as collaborators and partners, rather than solely as service providers; that they integrate mo re fully into the academic programs at UF and the university community at large. 21
III. Balancing the Collections The University of Florida is a comprehensive university offering 357 graduate degree programs 1 distributed among the sciences, agricultur e, the professions, the humanities, and the visual and performing arts. We have operated on the pr emise that the University will continue to develop its broad-based research pr ograms in both the sciences and the humanities to enable it to join the ranks of the elite public universities. To achieve this, the library system must support a range of research requirements providing strong holdings in each area of specialization, and it must do so in a way that satisfies the changing needs of diverse users. Thus, a careful balance must be struck in the development of collect ions and the way users access the librarys resources. Digital Resources As a major research institution Florida has an obligation to make its resources available as widely and as efficiently as possible. The internet provides an excellent and costeffective way to fulfill this obligation. The digitization of collections and online access to resources is imperative. While the University of Florida libraries have been productive in digitizing unique collections on campus efficiently, they lag behind the libraries of our peers in the acquisition of electronic re sources. We thus endorse the previous recommendations in this report (p. 18) that urge the library to apply for funding to expand its current digitization efforts; to center digitization efforts on UFspecific and unique collections; and to be proactive in gat hering and digitizing sources relating to UFs history. As we move forward in digiti zation, we must do so thoughtfully While it is crucial that we catch up with peer institutions in some aspec ts of the digitization process, we must also attend to the need for preservation of the physical resources entrusted to the university. Indeed, the physicality of info rmation sources is important in many disciplines, especially in the arts, social sci ences, and humanities. While we thus speak for the need to forge ahead in purchasing an d producing electronic m edia, at the same time we stress the importanc e of print resources. Content balance Although some suggest the par ing back of programs in the arts and humanities (see Reflective Note, p. 25 ), it is clear that the outstandi ng programs in the humanities and arts are crucial to the attainment of t op ten status. In 2005 the University Senate established an Arts and Humanities Work ing Group as an ad hoc committee to report on the state of the arts and humanities at the University of Florida and to make 1 Florida offers 233 M.A. or M.S. degrees and 124 PhD degrees. See Institutional comparison section. 22
recommendations for the arts and humanities for incorporation into the Universitys strategic plan to move into the ranks of the nations elite public universities. 2 The working group was organized both in response to the AAU report Reinvigorating the Humanities: Enhancing Research and Education on Campus and Beyond which describes national concerns about the declini ng role of the arts and humanities at AAU institutions, 3 and in response to perceptions among many UF faculty in the arts, humanities, and other di sciplines. The Senate ad hoc co mmittee report states that, while the arts and humanities are crucial to the mission and vitality of the University as a whole, this segment of t he Universitys intellectual infrastructure has received inadequate investment even relative to loca l standards, and that this must be redressed for the University to attain its stated goals. 4 In another study, William G. Bowen, Martin A. Kurzweil, and Eugene M. Tobi n state that The leaders of Americas major research universities, such as the members of the Association of American Universities, recognize clearly the import ance of the humanities and cont inue to encourage efforts to improve teaching and scholarship in these critically important fields. 5 They further state that Academic libraries and their attendant resources in the United States have expanded greatly since the middle of the 20 th century. Although these libraries devote significant resources to the sciences and engi neering, their role in supporting the arts and humanities is paramount. 6 Keeping in mind that the liber al arts are central to postsecondary education and that the best students gain from teachers who are also re searchers, whether t hat research is in the sciences or the humanities or production in the arts, it is essential that balance be maintained in the developing of the libra ry. The recognized collections at UF are strongest in the humanities, including, for ex ample, the P. K. Y onge Library of Florida History, the Judaica Collection, the Latin Amer ica Collection, the Baldwin Collection, the Proctor Oral History Collection, the Belk nap Collection, and the Africana Collection. Continued support of such collections and special libraries into the 21 st century is critical to the role of the university as a comprehensive university. Challenges Funds for purchasing and staff and space for maintenance and processing are all necessary to meet the demands being placed on the library for ongoing acquisition and preservation of both printed and electronic medi a. In order to assess the way in which these concerns will make an impact on the Un iversity of Florida Library system, we have interviewed both library users and lib rary professionals, reviewed appropriate literature, and drawn upon our own experiences in order to make the following 2 Arts and Humanities Working Group, Preliminary Report, 4/18/2006. 3 AAU Task Force on the Role and Status of the Humanities, Reinvigorating the Humanities: Enhancing Research and Education on Campus and Beyond (New York: AAU, 2004). Available on-line at http://www.aau.edu/issues/HumRpt.pdf. 4 Arts and Humanities Working Gr oup, Preliminary Report, p.1. 5 William G. Bowen, Martin A. Kurzweil, and Eugene M. Tobin, Equity and Excellence in American Higher Education University of Virginia Press, Charlottesville and London, 2005, p. 63. [Bowen is the president of the Mellon Foundation and former president of Princeton.] 6 Ibid. p. 62. 23
recommendations. Recommendations: 1) The Library must increase expenditure on electronic journals and data bases. Todays students and scholars expect and should receive the most current information via electronic media. In a 2004 Graduate Student Task Force Information Gathering Meeting 7 students who responded indicated that they characteristically connect to the library from their homes and rely heavily on electronic access to journals and other serials. 2 ) The library must continue the digitization of UF-specific collections. The University of Florida Digital Library is al ready well on its way in this complex process. Currently, it is said to be the most cost-effective producer of digitized material in the United States. 8 As of the date of this repor t, the University of Florida Dig ital Collections (UFDC) holds 65,000 bibliographic units, the size of a sm all branch library, of which 20,604 items are available to the public. 9 The library should continue sympathetic exploration of collaborative digitalization arrangements with Google, the Open Content Alliance, and/or other commercial entities. In negotia tions with these commercial entities the UF library should retain a copy of the digitized material and the library should have free access to it. Despite the UFDC efforts to produce digital material, several problems impinge on their continuing effectiveness, most notably, s hortages of full-time staff and necessary dependence on grant funds to carry out digitization projects. 3. The Library staff should be encouraged to continue and expand effort s to enter into shared purchasing arrangements with respec t to shared-access materials with other libraries and institutions. 4. Acquisitions of big ticket collections s hould be integrated into a carefully designed profile as defined by use patterns, mission, and relationship to existing holdings. 5. Close liaison between professional library acquisitions staff and faculty needs encouragement and perhaps to be put on a firmer and more explicit institutional footing than currently is the case. 10 In particular, it is important that faculty have ongoing knowledge of library holdings, services, and capabilities with respect to both the research and teaching functions of the faculty. 7 Graduate Student Task Force, Information Gat hering Meeting, Summary, May 26, 2004, p. 2. 8 Erich Kesse, personal communication, October 17, 2006) 9 Ibid. 10 See page 17, Liaison Librarian Program. 24
6 The library must maintain the unique collections that are already in place at the University of Florida. Specia l collections such as P. K. Yonge Library of Florida History, the Judaica Collection, the Lat in America Collection, the Ba ldwin Collection, the Proctor Oral History Collection, the Belknap Collecti on, the Latin American Collection, and the Africana Collection play a unique role at a university that has the announced ambitions for elite status repeatedly proclaimed by the Univ ersity of Florida. They are at the heart of scholarship in history, the social sciences literature, religious studies, area studies, and the humanities generally. They have a distinctive claim on the Universitys resources. In addition, they have unique issues involving future collection, digitization, on-line access, and related matters. Digit ization and on-line access should be encouraged (is proceeding with segments of the Baldwin and PKY collections; is virtually complete with the Proctor Oral His tory Collection) but these collections, along with University Archives, will have ongoing needs for labor-intensive curation and processing. 7. The library must be careful about going to an all-electronic policy with respect to journals. Journals with significant graphic content, for example, need to be acquired in hard copy. A policy of negotiating with v endors for hard-copy, storageable back-up should be pursued. This is especially so because of the problem of losing back runs if vendors go out of business or if other probl ems arise with vendors that may lead us to cancel contracts or switch vendors. It is cr itical that acquisition policies be sensitive to the particular needs of the various academic units. Thus, for example, the importance of acquiring hard copy issues of journals in the visual arts must be appreciated We are cognizant of the problem that continuing need for hard copy of books and journals poses. The physical volume of our library materials is increasing perhaps 3% a year. In 25 years will we have twice the current space for storage and access? 8. The library must continue to purchase books. With the current technology, electronic-only copies of st andard monographs and other scholarly books will not do the job. In the Graduate Student Task Fo rce Information-Gathering Meeting 11 students emphasized the current limitations of relianc e on on-line copies of books, a view that book-reliant faculty overwhelmingly endorse. Electronic copies of books and monographs are useful as back up but not a substitute. Reflective note on the character and direction of the univer sity and top ten status We assume that the University will continue to regard the Humanities as integral to its efforts to join the ranks of the elite public universities. However, the Future of the Library (FOL) Committee should take note of an alternative view that appears to exist among some influential donors a nd political leaders. It would be unwise to ignore the apparently widespread feeling among thoughtful people who are strong supporters of higher education, as indicated by their gifts or by their 11 Graduate Student Task Force, Information Gather ing Meeting, Summary, May 26, 2004, p. 2. 25
votes in the legislature, that in view of the fact that UF has a more limited budget than most flagships, severe prioriti zation is called for. This vi ew usually translates into the idea that business and the Humanities should devote themselves more to teaching than to research, since research in those areas generates little outside funding. The research effort, from the hiring of stars to the funding of research faci lities and library acquisitions, must be concentrated on the hard sciences. A continuation of that line of argument w ould be that with a limited budget we need to bias our scarce library funds toward the di gital resources needed by the sciences, not to the physical resources needed by researchers in the humanities. A rela ted point is that since the physical volume of our library mate rials is currently increasing perhaps 3% a year and since the bulk of that physical gr owth is in holdings associated with the Humanities and related fields, reductions in Humanities-related library materials will serve the dual purpose of recognizing the ne w, reduced role of the Humanities and easing what promises to be ongoing space pressure. The perception that this in fact is the Universitys trajectory is widespread among students and faculty in the Humanities. The recent response to budgetary overruns and the controversial Five Year plan in the Colle ge of Liberal Arts and Sciences (CLAS) are seen as pointing in the direction of dimi nishment of the Humanities. Recent communications from the science chairs quite understandably call for enhanced funding, calling attention to the sciences role in grant ( and hence overhead) acquisition. Given CLASs budgetary problems and the st ress on outside funding, many faculty see the distinct likelihood of the University reverting to a form of the old University College model in which Humanities units are regarded as service units. Centers as the Center for Latin American Studies and the Center for African Studies are recognized as premier world programs for area studies, and it is t he humanities that ess entially drive these centers. In a sense, this issue raises a fundamental question for the Future of the Library Committee. The character and direction of th e library of the future depends on the future character and direction of the University as a whole. Talk of building a Top Ten University, joining institutions such as Michigan, UC-Berkeley, UNC, UVA, and the like is misleading if indeed the Un iversity administration and trus tees plan to move in the direction outlined above. In this case, the Universitys outstanding and unique collections become something of an embarrassment rather than an asset. In order to continue to be nationally and internationally regarded as a research center in the Humanities, special collections such as the P.K. Yonge Library, the Latin Am erican Collection, and others will require substantial infusion of funds for ongoing and backlog processing and digitalization. Yet it makes little sense to devote these resour ces to these collections if in fact the University is otherwise pulling back from its earlier commitment to build outstanding Humanities and Social Science programs, sinc e it is the faculty and graduate students in these areas who use these collections. Digitization would of course make these collections available to scholars elsewhere, but it is unlikely that a BOT and University 26
27 administration engaged in limiting the role of the Humanities would be willing to support the special needs of these collections in t he spirit of general scholarly benevolence. IV. Role of the Library in UFs Capital Campaign Like most major academic uni ts, libraries at public univers ities are funded primarily by state allocations and secondarily by gifts and contributions. Adequat e state funding is by far the more important of these two s ources, and private fund raising can never be thought to compensate for shortcom ings in public support. However, the acquisition of private funding often serves as a margin of excellence for outstandi ng libraries. State funding provides the basic day-to-day operati ng resources; successful private support permits those special initiatives otherwise not possible. Adequate becomes excellent. For the UF libraries to reach their full potent ial, it is imperative that both funding channels be enhanced. At present, the privat e initiative offers special promise. A propitious confluence of opportunities exists for the UF library system. First, a Capital Campaign is already under way in its silent phase targeting an ultimate $1.2 billion goal. Second, a search has begun for new l eadership, a Dean of t he Libraries, whose duties will include a major effort in fund raising. These are complementary opportunities to be seized and exploited for the future of t he library. The library must have a position of first rank at the Capital Campaign table. Its role must be clearly stated and well publicized. Droppings from higher priority targets will not suffice. The library must be a priority, with well documented goals and aspirations. Who speaks for the library? It must start with the president, because without his support, goals are not met. Of course, lib raries are always supported by the administration. Who can not be on the side of the angels? But successful campaigns show that this support must be more t han lip service; insightful presidents put themselves on the line when speaking with al umni, friends, and corporate donors. And when they do, people respond. It will be up to the president to make the library an important priority and to speak out in support throughout the campaign. It is axiomatic in the development profession that libraries are a hard sell. Numbers sometimes are used to back this up, but it can also be a self-fulfilling prophecy. While libraries do not win SEC championships or cu re important diseases, the general public can appreciate the importance of books in t he academy. Successful library campaigns are often followed by comments like They said it couldnt be done, but we did it. We should focus less on the difficulties and mo re on those campaigns that have been successful, using these as models. How were these campaigns successful? They started with strong support at the top and involved innovative strategies to focus attention on this critical co mponent of t he university.
28 Library Needs How can a Capital Campaign make a diffe rence, even produce a major impact on library quality? Todays library needs includ e those of longstanding duration, as well as new high technology concerns unimagined onl y a few years ago. Some donors still see libraries as they were and can help meet still existing needs. Other donors understand, or can be brought to understand, the truly remarkable transformations taking place in libraries today. Libraries have a gift need fo r every taste. How can donors help? bricks and mortar libraries never have enough space, and new facilities designed for todays needs are always in demand; renovations converting outdated buildings to handle the technologydriven library; endowments libraries have critical needs for endowment funds (gifts that keep on giving by an annual return of interest), which can meet targeted needs such as special collections; parti cularly valuable are endowments that offer flexibility for greatest need; digital technology costs of computers, scanning hardware, and trained personnel are rising rapidly; special needs libraries always have new n eeds and opportunities to discuss with potential donors (see below, Thinki ng Beyond the Ordinary), Library Strategies The new dean will not need this committee to describe the elements of a successful fundraising campaign. She/he will have had experience elsewhere to serve as a base for developing a unique strategy at UF. But for whatever small value these may be, we include a few desiderata that have arisen during committee discussion. an activist Dean of Libraries with s ubstantial fundraising experience this is not the plac e or time to learn on the job; a dedicated and well chosen Library Capital Campaign Committee to include at least a few people with Florida name recognition and
29 substantia l personal resources; one 7-figure lead gift, with several 6-figure complements; presidential mandate to the F oundation re library importance; highly visible library cam paign throughout the state and beyond library should be in the news, for all the right reasons, not just to announce cuts in books and journals; fundraising dinners at Library West small and carefully selected groups, hosted by the president or the provost to showcase what is and pitch for what is yet to be; exhibitions tied in with lectures open to the campus and the community spotlighting our collections. Sunday a fternoon collections talks to draw potential donors with specialized interests. Campaign Goal It is not the charge or the prerogative of this committee to suggest a library goal in the UF Capital Campaign. Still, it is difficult not to speculate. In Fall, 2007 the University of Florida will begin its public phase of the Ca pital Campaign, with an anticipated goal of $1.2 billion. While this seems substantial, billion dollar campai gns among universities have become rather common. Reviewing info rmation about campaigns that are either beginning or have recently been completed, the following cases may be useful to consider. As one bright example, the University of Virginia, much smaller than UF, is in the midst of a $3 billion campaign with a libra ry target of $100 milli on. Other campaigns showing the campaign goal and library target include: University of Michigan $2.5 billion/$52 million Penn State University $1.3 billion/$36 million University of North Caroli na $2 billion/$35 million University of Maryland $1 billion/$20 million University of Kentucky $1 billion/$64 million University of Pittsburgh $0.93 billion/$33 million Brown University $1.4 billion/$30-35 million Duke University $2 billion/$60 million University of Miami $1.25 billlion/$55 million
30 So what might be a worthy target for the library system at UF within the $1.2 billion goal? What is the value of a library? For a prospective Top-10 public university, what percentage of a funding campaign would se em appropriate? Some would contend that a library might constitute 10% of university value, which would project a campaign goal of $120M. Too much? Perhaps, but we believe that a $25M target (2%) is highly appropriate, indeed almost modest, compared to our peer examples Given our highly public aspirations, we should not lower our si ghts here. A bold campaign for the library, enhanced further under the state matching program, will greatly aid in meeting suggested goals in this report. The type of gifts received can affect the net success of fundraising. Donors will, of course, give what and how they wish. Most valued (and most rare) are those gifts that allow use by a Dean of Libraries for whatever is the greatest need. More problematic can be gifts in kind, such as collections of books, films, papers, etc. These may greatly strengthen library holdings, but they normally do not include money for processing the materials. Indeed, it may in volve added cost to already scarce resources. However, the assessed value of such gifts, which may be substantial, counts against the campaign goal and can make a unit appear to be better funded than is the case. Raising the library endowment, now at an inadequate $9.2M, is qui te important, for which hard funding (e.g., cash, real es tate, stock, etc.) is required. Fundraising for the Health Center Librar ies has not been included in the Smathers operation. Please see the Integr ation section of this report. Thinking Beyond the Ordinary The main library of a major research univers ity should contain both external and internal spaces that focus the attent ion of the public on the centra l importance of the library. Thus, it is in keeping with the significance of the library, on both literal and symbolic levels, for it to have a dramatic and in spiring presence, one that announces the prominence of this enterprise to the beholde rs, whether they are a current part of our university community, alumni of the university, or visitors to the core campus. If we are to be perceived by both the outside and insi de worlds as a place where important intellectual work is being done, it is necessa ry to have a presence that is, in its own way, as impressive as those t hat house our athletic facilities. In short, the main library should become both a focal and a destinatio n point on the university campus, the signature place that embodies its academic values. Currently, the appearance of the library comple x is uninspiring, to say the least. The main entrance of the newly-reno vated Library West, for some inexplicable reason, is almost completely obscured from view; and the facade of the newly-renovated Library West is crossed by a rectangular grid of al uminum I-beams that seems visually devoid of any architectural or aest hetic purpose. It appears as though an attractive, finishing
31 facade of some kind may have been meant to be hung from this arma ture, but the work has been abandoned. A crumbling, mildew-s tained colonnade stretches in front of Library West, obscuring the entrance to the building. Trash containers, often overflowing with debris and foraging squirrels, are haphazardly placed within the colonnade which, in its turn, is lined with garishly li ghted soft-drink machines, battered newspaper boxes, and bulletin boards with countless strata of stapled, fluttering announcements. Move back into the Plaza of the Americas, the setting for the libra ry, and one finds oneself on broken, mis-matched sidewalks, di rt paths, or untended grass that has been extinguished by countless frisbee games and t he mid-day traffic from the daily crowds that come to the Plaza for a free lunch provi ded by one of the town's religious groups. A brick seating area off the colonnade, next to Library East, has gone fallow as well. It is unshaded, unkempt, and unused save for the occasional staff member who braves the sun and one of the splintered benches to find a place of refuge in the Plaza. With this in mind, we propose that the main branch of the library feature a dramatic, architecturally significant facade and entrance that celebrates the academic mission and intellectual aspirations of the University of Florida. This project should involve an entire redesign of the Plaza of the Americas, to include an equally stunning landscape design that would be respectful of the historical dy namics of the Plaza, wh ile at the same time creating a unified space that would be singula rly attractive, inviting and useable and that would provide a meeting place for t he entire university community. Such an amplification of the librarys footprint on the terra in of the university should be one of the projects included in the Capital Cam paign. It would become a very attractive development possibility if a major, public arch itect like, say, Frank Gehry who is most notable for his design of t he Guggenheim Museum in Bilbao, Spain were to be approached for such a commission. (It s hould be noted that Gehry is currently designing the new performing arts complex for the University of Connecticut.) The University of Florida should think in larger national, and even intern ational term, than the university has previously done regarding the significance of its academic profile. This proposed new design for the library facade, en try, and the Plaza would provide a highly visible naming opportunity for potential donors, who could participate in the funding of various portions of the entire project. A related project that is desperately needed on the University campus is a research center that would be an extension of the Universitys S pecial Collections and that would provide the University community and visiti ng scholars with a concentration of spaces dedicated to the academic enterprise. Such a state-of-the-art cent er would include a central reading room, meeti ng spaces of various sizes for holding conferences, seminars, and colloquia; exhibition spaces for the on-going display of significant materials in the Universitys library collections; office space for university centers and institutes; and a dining area t hat would serve as a facult y center. Once again, the University of Connecticut provides an example, in their Dodd Resear ch Center, for this kind of facility. A project like this should also be included in the Capital Campaign, and
32 it, too, would represent a superb naming opp ortunity, especially if a renowned architect were to be commissioned to create its design. Marketing the Library The new Capital Campaign and the imminent hi ring of a new Dean of Libraries provide an unprecedented opportunity for the library to achieve its financial goals. To best take advantage of these circumstances, however, the Library needs to market itself better. It should be a given that the Libr ary is central to the Universitys mission of teaching research and service, but this has not receiv ed the attention it deserves. Faculty and students need to be made aware of the centra lity of a good research library to the academic enterprise. The library serves a number of constituencies who have a variety of needs, but these should not be viewed as mu tually exclusive. While undergraduates may prize collaborative space within the library, graduate st udents and faculty may want more access to individual volumes and collections and more private study space. All of these groups need to be made aware of the fact that the current fundi ng structure is not adequately meeting the needs of the library and that many of the things they prize could be lost or restricted if they do not emphasize their need for and support of the library. Because the library belongs to the University as a whole, the President must take the lead in emphasizing its importance in hi s Capital Campaign directives. Strong leadership from the to p will ensure that the Foundation is aware that the library is a priority for a University seeking to enter the tier of top-10 Universi ties. Alumni and the general public also must be made aware of the resources and needs of the library, and greater emphasis placed on the Librarys role as the intellectual heart of the University. V. Integration of the UF Libraries The impending retirements of the Directors of the Health Science Center Libraries (HSCL) and the Smathers Librar ies (Smathers) at the Univer sity of Florida create an opportunity to take a fresh look at the re lationship between the two library systems in conjunction with the process of recruiting and hiring their successors. An obvious component of rethinking the relationship between the two systems is to consider whether they should be more formally integr ated. We explored this question from a number of different perspectives, including: the history and culture of both institutions; the specialized needs of professional school libraries; the opinions and experiences of professionals in the area; fi nancial considerations; and pers onnel management issues. To do this, we read a number of articles on professional school and medical libraries; collected documents concerning HSCL (such as consultants reports, LibQual data, mission and vision statement, and proposals fo r funding); collected data for comparison
33 from a number of authoritative resources (including AAHSLD, AAMC); reviewed the 2004-05 comparative data received directly from peer institut ion libraries; and, finally, interviewed a number of expert consultant s, library and Health Science Center leadership and those librarians who would be affected by a merger. Three potential approaches became apparent: 1. The library systems retain their current level of autonomy and integration, but if this model is retained, it is recomm ended that the Director of the HSC Libraries report directly to the Seni or Vice President for Heal th Affairs (SVPHA) and serve on the SVPHAs Deans Council. 2. The libraries integrate their budgets and reporting structures, but HSCL retains autonomy to serve the needs of its clients It is recommended that this model would include a dotted line reporting relationship between the HSCL Director and the SVPHA, and that the HSCL dire ctor serve on the SVPHAs Deans Council similar to the situation at the University of Minnesota. 3. HSCL is fully integrated into the Smathers system and becomes, in essence, a branch library. We considered the arguments for and against each of these models, in light of what we had learned. We also assessed the probabili ty that the benefits of each model would be achieved against the particular risks that each approach occasions. For reasons set forth below, we recommend that UF retain the first model, but strongly urge the University Administration not to accept the status quo, but to address the existing problems that will diminish t he likelihood that any of the approaches will succeed. In order for HSCL to achieve it s potential and continue to provide necessary services at acceptable levels, adequate re curring funding, well-supported, visionary leadership, and a coherent reporting structure will be required. We believe that if progress is made on these fronts, the seco nd model may also be viable. Even with adequate funding and new leadership, however, we do not believe that the third approach is desirable or workable because of t he specialized mission of HSCL, its size, and the vulnerability of its autonomy in a fu lly integrated model with as yet unknown leadership. Background The HSC libraries are as large and varied as their mission. They comprise the main HSCL on the Gainesville campus, as well as the Borland Library 12 at UFs urban 12 The Borland Library in Jacksonville serves primarily the clinical missions at the urban campus, including residency programs and the hospital.
34 medical campus in Jacksonville. The Gainesv ille library meets the information needs of research and clinical faculty, students (including professional, PhD, MS and undergraduates), post-doctoral associates, st aff and administrators from the six HSC colleges (Dentistry, Medici ne, Nursing, Pharmacy, Public Health and Health Professions, and Veterinary Medicine); re sidents from three colleges (Dentistry, Medicine, and Veterinary Medicine); and three teaching hospitals (Shands at the University of Florida, Shands at AGH and Shands at Jacksonvi lle) with 5 additional affiliated hospitals 13 A total of 7,225 students 14 and 2,377 faculty 15 work at the HSC. HSC brings in 52% of the research funds at the University of Florida and 54% of the indirect costs from grants. About 50% of medical libraries at universitie s throughout the country are fully integrated with the universitys main library system; the others are, to varying degrees, autonomous (e.g., report to the senior leadership of th e academic health center or the College of Medicine). 16 In the ways we describe bel ow, HSCL is mostly autonomous and currently operates separately from Smathers, but is also functionally integrated in several significant respects. The two library systems have completely separ ate budgets. The report ing structures of the libraries are differentthe Director of the Smathers Libraries (D ale Canelas) reports to the Provost, while the Director of HSCL (F aith Meakin) reports to the Assistant Vice President, Administrative Support, who is within the office of the SVPHA, but two-steps below the SVPHA on the organizational chart. Sabbaticals are awarded to the librarians under two separate procedures The Smathers system us es a sabbatical committee composed of senior level librarians, wher eas HSCLs sabbatical s are awarded through the SVPHAs office. Each system administe rs its services separately and each has different policies and procedures that are appropriately based on the unique needs of their respective clients. Examples of t hese services include interlibrary loans, the HSCLs Liaison Librarian Pr ogram, loan periods, hours of operation, etc. Smathers employees are part of UFs bargaining unit, whereas HSCL employees are not. Notwithstanding the foregoing, the two system s are functionally integrated in some significant respects, and their librarians coll aborate with one another to a large extent. The libraries share costs on a number of el ectronic resources, and access is available to clients of both systems. The material s owned by HSCL are included in UFs online library catalog and UF library users may c heck out materials from either systems. Duplicate journal subscriptions have been c ancelled to conserve funds. Strong collaboration exists among librarians from the two systems. HSCL librarians are part of 13 Shands at Vista, Shands Rehab, Shands at Lake Shore, Shands at Live Oak and Shands at Starke 14 1,585 undergraduates, 1,567 MS/PhD, 3, 007 professional and 1,066 residents 15 1,956 traditional faculty, 235 postdoctoral fellows and 186 adjunct faculty 16 Because of the difficulties in finding true comparators for the HSC at UF and the s ubtleties of integration and variations in reporting structures, this statisti c may not be particularly meaningful or helpful to our inquiry.
35 the Smathers mentorin g program. The faculty from both library systems go through the tenure and promotion process together, through the virtual College of Libraries. Under the current system, however, HSCL is guaranteed a separate seat on this Colleges Tenure and Promoti on Committee each year. In the 2005-2006 academic year, for the first time, an HSCL librarian wa s voted into one of t he four faculty senate seats for the College of Li braries. A nascent Library Faculty Assembly is being created, which will include librar y faculty from Smathers, the Legal Information Center and HSCL. A separate HSCL faculty assembly is being created, however, to deal with issues of interest only to HSCL. Issues and Analysis An understanding of the following issues is a necessary prerequisite to the integration decision and critical to an accurate asse ssment of the impact of merging the two systems. 1. Indirect Costs and Funding The issue of indirect costs is a complicated one. Currently the libraries receive a total of 1.53% of the universitys in direct cost returns, with 16% of that total going to HSCL and 84 % going to Smathers libraries. The current distribution scheme raises some fairness concer ns. The majority of the indirect cost returns received on campus (54%) come from the Health Science Center. The current distribution of DSR funding is based, not on student FTEs, but rather on the number of PhD research programs. The HSC obviously has a vibrant research program, but its reason for being is to teach students 7,225 of them of which 1,567 are MS/PhD degree program students. By its very nature, the HSC generates a great deal of revenue through its research programs, but because the majority of its educational programs are professional, rather than research oriented, its library does not reap funding rewards at an equitable level. An unfortunate irony is that even though the majority of students at HSCL are professional students (who ar e left out of this indirect fund equation), HSCL is nonetheless responsible for meeting the research needs of the thousands of research faculty, post-do cs, and graduate students that it serves. 17 A further irony is that the journal collecti ons of Smathers and HSCL are so tied together that a reallocation of indirect costs from Smathers to HSCL would end up hurting HSCL in the end. If Smathers cancels journals due to lack of funding, HSCLs users lose them as well. If HSCL then chooses to pick up cert ain individual journal titles that have been canceled, not only will HSCL have new expens es for these titles, but these titles will actually cost more than if Smathers had been able to keep the package; once titles are purchased on a title by title basis rather than in bulk, they cost substantially more. A 17 HSCL splits the molecular biology, biochemistry, ge netics, and other basic science material purchases with Smathers, but receives little DSR money to fund these materials.
36 greater percentage of indirect funds for HSCL may be justified, but it is unclear whether this would happen under any approach. If the Committee believes the current distribution may not be fair, one determination to be made is which integration model is most likely to produce greater equity. Because the two systems collections ar e so integrated and duplication has been reduced wherever publishers allow, it does not appear that any type of monetary savings would be realized thr ough integration. Dale C anelas, Faith Meakin, Beth Layton (Deputy Director, HSCL ), and Linda Watson (Director, HSCL, University of Minnesota) believe that a merger would not reduce the need for a Director of HSCL at the same level of experience and salary as the current HSCL Di rector, and given the state of the library, it is reasonable to be lieve that any candidate worth hiring would require a larger salary (not to mention increased budget and staffing). Furthermore, Dale Canelas made it clear t hat, in her opinion, the size and diversity of the main UF campus and the HSC would make it nearly impossible for the Dean of Libraries to be much concerned with the issues or management of HSCL. Moreover, duplication in the administration of certain services (Inter library Loan and Technical Services) are tied directly to clinical needs, where speed and accuracy can actually mean the difference between a good or poor clinical outcome. 2. Shands Health Care Shands has resources availabl e for distribution to HSCL and the perception (among those at both library systems) is that Shands should play more of a role in resolving some HSCLs funding pr oblems. The hospital and its personnel rely upon HSCL. Shands uses HSCL for JCAHO and resident accredita tion purposes. HSCL reports that t hose associated with Shands use HS CLs library collections and services extensively, but, when the questi on of support arises, Shands discounts the extent to which its employees use the library Smathers is unlikely to provide increased monetary support for HSCL under a merged system, in part, because of the perception that the medical community on campus, could, but is not, t aking care of its own. Similarly, in a merged system it may be less likely that Shands Health Care could be convinced to fund the HSCL appropriately. HSCL could provide expanded services to Shands and be an integral partner in te rms of ensuring evidence-based medicine, quality assurance, etc, if additional library funding were provided. Resolving this problem, however, will likely require some type of intervention from the highest levels of leadership at UF, irrespective of whet her the libraries are integrated. 3. Cap i tal Campaign A new or remodeled HSCL pr ovides a high-profile naming opportunity, and should not be overlooked in the universitys campaign effort Additional HSCL opportuni ties for the campaign include endowed librarianships, purchasing and licensing of expensive bioi nformatics and genomics resources, and numerous equipment needs. Fund-raising efforts for the libraries are not integrated. HSCL will not be included in Smathers capital campaign. The current plans of the
37 SVPHA do not include assisting HSCL with fund-raising. Although the SVPHA, has indicated that HSCL may hire its own develop ment officer if it wishes, HSCLs budget does not make such a hire possible at this time. In a mer ged system, HSCL might provide potential donors and benefit from the public relations and fund-raising efforts of a combined system. 18 4. Services. Academic health sciences libraries serve the same clients as those of other campus libraries (students, faculty, researchers, ad ministrators) with the addition of the clinical enterprise. Services such as Interlibrary Loan and loan periods of materials must reflect the ti me-sensitive nature of patient care. Cataloging is carried out under the National Library of Medicines system, which is built for health-related collections, unlike that used by most non-medical academic libraries (Library of Congress system). Public services, such as reference and instruction, must be tied closely to the needs of the clients, and require an integration into the academic departments and programs that extend far beyond those of traditional subject bibliographers in most non-medical academic libraries. Libraries that support professional programs, such as health or law, take a bout ique approach to services integrated, personal, customized, responsive. HSCL currently has a Liaison Librarian Program that is a model for academic health center libraries throughout the country. HSCL was one of the first medical libraries in the country to provide library-based bioinformatics services and this program is al so a model. While these approaches are specific to the HSCL and meet the needs of its clients, there is also substantial cooperation among the librarians who work in the two systems, particularly between those at Marston Science Library and HSCL Cooperation in collection development, instruction, and reference services has fostered innovative services, allowed librarians from both systems to learn from each ot her, and provided opportunities for shared assistance to multi-disciplinary researchers. 5. Tenure and Promotion Librarians at HSCL (as at Smathers) undergo the midcareer review, tenure and promot ion processes as part of t he College of Libraries. Although the three basic criter ia for tenure and promotion Job Performance, Research, and Service are the same for librarians w hether they are part of Smathers or HSCL, the way these criteria manifest are very different. As de scribed above, librarians that serve academic health centers must integrate very closely with their clients research and educational programs, especially in terms of their public services librarians. To be successful, they ideally spend less time on library-related committe es, and more time on focused, specialized, personal services with t heir clients. Those who work at academic medical centers have a service responsib ility to local and regional professional societies, as well as those at the national level, and librari ans are credentialed through the Academy of Health Information Profe ssionals (AHIP; currently required by the 18 If the libraries are not integrated, it is reco mmended that the SVPHAs office fund a part-time development officer for the HSCL.
38 HSCL). The focus of the HSC Librarians is significantly different from those of their Smathers counterparts and it is, therefore, desirable to ensure that HSCL is guaranteed a seat on the Tenure and Promotion committee each year. Obviously, the integration decision may be made from a number of problem-solving approaches: management theory/ideology; historical/organic/cultural; a quasicost/benefit analysis with the option of quantif ying the costs and benefits; and finally, a pragmatic approach which borrows and blends from all of the other approaches. It may also be made through inaction, by not making any decision or attempti ng to analyze the problem in a coherent way. We chose to be pragmatic, and thus the question became: given what we have learned, what are the potent ial benefits and detriments of greater integration? Some of the potential benefits of integration are: Integration may provide greater opportunity for innovation and revitalization; May allow an Information Czar structuring of UFs libraries; Interaction and collaboration ma y be enhanced among librarians across disciplines; interdisciplinary research may be facilitated. Public relations and fund-raising effort s could be combined, including capital campaigns; The ability to use the possibility of integrating and rest ructuring the two systems in the process of recruiting the new Dean; HSCL might have a better advocate in the new Dean of Libraries (i.e., HSCL would be more of a presence on the collective radar); The tenure and promotion process might be more streamlined and have more oversight; The starting salaries and other salary inequities for HSCL librarians may attain parity with those of Smathers. 19 Some factors arguing against a greater degr ee of integration bet ween the two systems include: Potential negative impacts upon HSCLs funding and staffingmerger, in and of itself, is unlikely to improv e the financial posture of HSCL; Having the HSCL Director report to the new Library Dean may diminish the close working relationship between the Director and the Health Center deans; HSCL may have less autonomy the library may become less able to meet the unique needs of their professional users and may move from boutique services to a Wal-Mart approach; 19 This could require increased funding to Smathers; if the HSCL received increased recurring funding it could raise the salaries regardless of integration.
39 HSCL might lose the ability to co ntrol how its resources are managed and how its services are prioritized; HSCL librarians and other staff may be forced to become more librarycentered rather than user-centered and expected to work on more library committees and spend less ti me with their clients; A major decrease in travel and CE fundi ng for HSCL librarians is likely (HSCL librarians get $1500/year for trav el to conference and $500/yr for CE; whereas Smathers librarians compet e for $400/year for conference attendance); such travel and continuing ed ucation are important for librarians to meet AHIP credentialing standards; The Liaison Librarian Pr ogram, a model for medical libraries throughout the country, might be imperiled; HSCL may lose its seat on t he tenure and promotion committee; and Moving faculty from non-bargaining uni t to bargaining unit may introduce additional complexities. Conclusions Wendy Lougee (Minnesota) identified three conditi ons that she believes are critical to a successfully integrated university library system: the support and cooperation of the Provost; adequate funding for the libraries; and fi nally, forward-looking, positive, and accountable leadership. We trust that the first of these conditions will be met, but have concerns about the latter two. Long-term, chronic under-funding of the UF libraries in general, and HSCL, in particular, has taken it s toll and will likely continue to do so. While it is difficult to compare budgets on a do llar by dollar or program by program basis due to the complexity of funding systems, number of colleges and campuses served, and a number of other factors, 2004-2005 data suggest that UFs HSCL is severely underfunded compared to its peers (University of Minnesota, the University of Michigan, the University of Maryland-Baltimore, the University of North Carolina Chapel Hill and the University of Washington.) These institutions have been chosen because they are public institutions that have representations of colleges closest to those of UFs HSCL. In terms of number of Colleges, only the Un iversity of MarylandBaltimore has less than six; the University of Mi nnesota is the only peer to have a College of Veterinary Medicine; and although UF does not have a separate College of Public Health, its combined College of Public Health and Health Professions ensures that it is the only one of these institutions to serve these seven disciplines Allied Health (Health Professions), Dentistry, Medicine, Nursi ng, Pharmacy, Public Health, and Veterinary Medicine. The UF HSC serves approximately 2500 more students than its closest peer (University of Maryland-Baltimore) and has a budget fully $1,000,000 less than the University of North CarolinaChapel Hill and almost $500,000 le ss than the University of Washington.
40 Based on 2004-2005 statistics reported by our peers, the University of Florida HSCL has the smallest budget per student: University of Florida $545.99 University of Maryland-Baltimore $901.95 University of Michigan $927.23 University of Minnesota $874.43 University of North Carolina Chapel Hill $1,258.26 University of Washington $975.16 UFHSCL again falls toward the botto m when number of students/librarian are calculated: University of Florida 494 University of Maryland-Baltimore 211 University of Michigan 281 University of Minnesota 300 University of North Carolina Chapel Hill 156 University of Washington 296 The University of Florida library systems have evolved separately and each has its own adaptive mechanisms to cope with scarce resour ces. In such an environment, merging the two systems is likely to exacerbate anxie ty about the fair distribution of resources and power. In the absence of trustworthy leadership, competitive attitudes may dominant and undermine any positive effects of merger. While it is clearly possible to recruit such leadership, this dispositive factor is currently an unknown variable in the integration equation. One value that must be fixed and heavily weighted in this formula, however, is the operating autonomy of HSCL. Colleges with t he responsibility for training and serving professionals require a high (or, at leas t, particular) degree of autonomy. The ABA Standards for accrediting law schools are in structive. Standard 601 provides, in relevant part: A law librarys effective support of the schools teaching, scholarship, research, and services programs requires a direct, continuing, and informed relationship with the faculty, students, and administ ration of the law school.
41 Standard 602: (a) A law school shall have sufficient administrative auto nomy to direct the growth and development of the law library and to control the use of its resources. Interpretation 602-1: This Standard recognizes that subs tantial operating autonomy rests with the dean, the director of the law library and the faculty of a law school with regard to the oper ation of the law school library. The Standards require that decisions that materia lly affect the law library be enlightened by the needs of the law school educat ional program. This envisions law library participation in university libr ary decision that may affect the law library. While the preferred struct ure for administration of a law school library is one of law school adminis tration, a law school library may be administered as part of a general unive rsity library system if the dean, the director of the law library, and faculty are responsible for the determination of basic law library policies. We conclude that given UFs history and cu rrent culture, and irrespective of formal structures, HSCL should retain a high degr ee of operating autonomy. Ken Frazier (Wisconsin) cautioned that we should not sacrifice substance to fo rm and that there are no necessarily structured solutions to the challenges posed by integration opportunities. He also suggested that co mbining one stressed system (Smathers) with one that is even more stressed (the HSCL ) is no key to success. The penultimate inquiry for the Committee is whether we can achieve benefits from integration (efficiency, lack of redundancy, better serv ice for clients, improved support for personnel, etc.) without compromising the expert form of timely hands on service HSCL librarians currently provide. At this time, the benefits of increased integr ation seem idealized and generalized. The potential detrimental effects, on the other hand, are quite specific and in sharp focus. 20 Therefore, the ultimate query may be whet her those benefits might also be achieved through greater functional integration wit hout enduring the risks and potential upheavals of a more formal, structural merger. 20 This may be an understandable consequence of, in essence, comparing a known present to an unknown (and given the degree of sc arcity and neglect, unknowable) future.
42 Appendices Appendix 1: Comparative Library Regressions In the text of the report, we compare the University of Floridas libraries on several dimensions to those at six peer institutions, all large land grant universities with law schools and medical schools. As additional s upport for the findings reported there, in this appendix we broaden t he comparison group to include all of the American institutions of higher education whose librari es gather expenditure and staff figures for the Association of Research Libraries. We use the most recent nu mbers, those for the 2003-2004 academic year. We augm ent the ARL data from two other sources: the 2006 U.S. News & World Report rankings and The Center spreadsheets at the University of Florida. Ideally, we would compare the library services provided by the University of Florida to services provided by other institutions: how effectively compared to others do the libraries at UF aid our teaching, research and service missions? Since we know of no practical way to do that, we look at inputs instead. We estimate what might be roughly thought of as demands for various library i nputs as functions of various university characteristics. The characteristics are the number of students, th e number of faculty, the presence of a medical school, the presence of a law school, and, as a crude measure of quality, the U.S. News & World Report scores. We expect the demand for the various library inputs to vary pos itively with each of these measures. The input measures we use are total library spending, total library salaries, the number of professional library staff, spending on el ectronic materials including software, all library spending on materials, spending on journals (serials), and spending on monographs. Our method is ordinar y least squares regression. A ll of the variables are in logarithms, except for the US News rating, and dichotomous indicators of the presence of a law school or a medical school. Prim arily for ease of exposition, we add a dichotomous variable UF, which takes the value one if the observation is for the University of Florida and zero otherwise. We report results from four se ts of specifications of the regression equation. In the first set the independent variables are the log of the number of student s, the log of the number of faculty, the U.S. News rating (on a scale of one to 100), and the dichotomous variables for medical school, law school, and UF With this specification, we have data for 82 institutions (only 79 for total spending on electronic materials). As a minor variation, we impute values for the U.S. News ratings for an additional 20 institutions, using a standard imputation method that avoids inducing bias.
43 Next, returning to the original 82 observation s, we assume that in providing library services, universities experience constant returns to scale in the sense that an institution with 40,000 students and 3,000 faculty requires t wice the library services of one with 20,000 students and 1,500 faculty. This assumpti on is counterintuitive, especially with respect to acquisitions and undergraduates. Few academics would encourage a high school graduate to go to LS U instead of Swarthmore because of LSUs larger holdings. Even for faculty, the increase in their number will result in more people per discipline, not just more disciplines. Our first and second sets of regressions strongly reject constant returns, and suggest that a doubling of university size is associated with an increase in total library sp ending of about sixty percent. Since UF is a very large institution, not allowing for economies of scale makes our spending look especially low relative to our peers. We incl ude this specification in case economies of scale are being rapidly eroded by changes in information and communications technology and by the determination of the p ublishers of commercial journals to seek perfect price discrimination, to charge what the market wil l bear, which may rise in proportion to size. Finally, we go to the other extreme, and add a host of new variables to the first specification for the 102 inst itutions. These variables, from The Center, are (in logarithmic form) the number of postdoctoral fellows, annual giving, the number of members of the National Academy of Sciences, and total research spending. We also add the peer component of the U.S. News rating. These extra variables generally have t-statistics less than one, indicating they do not belong. The inclusion of this extra noise biases the coefficient of UF toward zero, making it appear t hat UF is closer to average spending holding other things constant. Even with this method, the University of Floridas total library spending is 18% below the amount predicted. The conclusions that emerge mo st strongly are that (1) Flor idas total library spending is lower than expected for a university of its size and quality, and (2) spending on total electronic materials is remarkably low. Spending on monographs is also very low, but the number of monographs purchased is not mu ch below the number expected. Spending on serials, in contrast, is close to the expected value (exc ept for the probably inappropriate constant-returns-to-scale specif ication), perhaps reflecting the ability of the publishers of commercial journals to pr ice discriminate against (extract a lot of money from) large institutions. The results of our preferred set of specifications are summarized in Table 1. The column for total spending serves to illustrate how to interpret the re sults. It suggests that: 1. A 10% higher number of students is associated with 4.8% more library spending, other things the same. The coefficient is ov er five times its esti mated standard error (0.09), suggesting it is estima ted with reasonabl e precision.
44 2. A 10% higher number of faculty is associated with 1% higher library spending, other things the same. The effect is not only sma ll but also statistically insignificant. The estimated coefficient is only slightly larger than its standard error. 3. An 18-point higher U.S. News ra ting is associated with about 36% more library spending, controlling for size and the pres ence of medical and law schools. Eighteen points represents the differenc e between, say, 59 (Florida, Texas, UC Santa Barbara, UC Davis) and 77 (Berkeley, Carnegie Mellon). The result is very similar using peer evaluations instead of the US News total. 4. Having a medical school is associated with 8% higher spending. 5. Having a law school is associated with 18% higher spending. 6. UFs total spending is 30% lower than would be expected for a university of its size and rating with law and medical schools. The coe fficient is not very precisely estimated, however. The calculation is 1/e^.36 = .70. Table 1: Regressions Explaining Total Librar y Spending and Categories of Library Spending by the Number of Students, the Number of Faculty, the U.S. News & World Report Rating and the Presence of Medical and Law Schools. Variable Spending Salaries Staf f Electronic Materials Serials Monos Students 0.48 0.53 0.33 0.41 0.36 0.22 0.67 (0.09) (0.10) (0.12) (0. 17) (0.09) (0.08) (0.20) Faculty 0.10 0.16 0.23 0.04 0.09 0.12 -0.05 (0.09) (0.10) (0.11) (0. 17) (0.09) (0.08) (0.19) USNews 0.020 0.022 0.018 0.017 0.016 0.010 0.028 (0.002) (0.002) (0.002) (0. 004) (0.002) (0.002) (0.004) Medical 0.08 0.05 0.08 0.16 0.08 0.10 -0.02 (0.06) (0.06) (0.08) (0. 11) (0.06) (0.05) (0.13) Law 0.18 0.20 0.20 0.01 0.18 0.13 0.16 (0.06) (0.06) (0.08) (0. 11) (0.06) (0.05) (0.13) UF -0.36 -0.40 -0.26 -0.55 -0.29 -0.11 -0.67 (0.22) (0.25) (0.30) (0. 43) (0.22) (0.19) (0.50) Constant 10.04 8.26 -1 .78 9.28 10.67 11.74 6.49 0.59 (0.67) (0.81) (1.16) (0.60) (0.51) (1.34)
45 N 82 82 82 79 82 82 82 R 2 0.78 0.77 0.64 0.35 0.33 0.69 0.49 Note: Data are from the report of the Association of Research Libraries, described in the text, and from the U.S. News & World Report web site (subscription required). The dependent variables are (across the columns): the logarithm of total library spending, the logarithm of the number of total library staff, the logarithm of all spending on electronic materials including software, the l ogarithm of all total library spending on materials, the logarithm of spending on se rials, and the logarithm of spending on monographs. Three institutions did not report their spending on electronic materials separately. The independent vari ables (down the rows) are th e logarithm of the number of students, the logarithm of the number of faculty, the U.S. News rating, the presence of a medical school, the presence of a law school, an indicator for UF, and a constant term. Standard errors are in parentheses. To differ significantly from zero in the standard statistical sense, t he magnitude of an estimated coefficient should be about twice its standard error. The standard errors fo r UF are relatively large partly because Florida is much larger than most unive rsities, which is a reason we emphasize comparisons with other very larg e institutions in the text. Table 2: The University of Floridas Share of Its Predicted Measures of Library Inputs, Using Four Sets of Specifications Input Short 82 Short 102 Long 102 Constrained 82 Total Spending 70% 76% 82% 52% Salaries 67% 74% 80% 54% Staff 58% 83% 89% 57% Electronic 58% 58% 54% 39% All Materials 75% 82% 86% 51% Serials 89% 96% 102% 57% Monographs 51% 56% 58% 39% Notes: The columns represent the various se ts of regression specifications. Short 82 has short sets of regressors represent ing the number of students, the number of faculty, the presence of medical and law schools, and the U.S. News & World Report
46 ratings. Short 102 has the same regressors ex cept that the number of observations is expanded to 102 by including imputed ratings for 20 institutions. Long 102 is the same as Short 102 except that regressors r epresenting the number of postdoctoral employees, the number of me mbers of the National Aca demy of Sciences, total research funds, total giving, and the U.S. News peer rating are added, imputed when necessary. Constrained 82 is the same as Short 82 except that the coefficients of Log Total Students and Log Faculty are constrai ned to sum to one, thus imposing the assumption of consta nt returns to scale. The 70% in the cell represent ing total spending and Short 82 sa ys that the University of Floridas total library spending is only 70% of what the specification Short 82 predicts it would be based on the other 81 institutions. Appendix 2: The Health Science Center Library at UF has made some comparisons with medical libraries identified as serving similar client ele and colleges: University of MarylandBaltimore, University of Michigan, University of Minnesota, University of North CarolinaChapel Hill, and University of Washington. Data on health science college enrollments is taken from enrollment figures from each of the universities: Office of the Regi strar, University of WisconsinMadison http://registrar.wisc.edu/students/acadrecor ds/enrollment_reports/Stats_all_20032004Fall.pdf ; University of Minnesota, Inst itutional Research and Reporting http://www.irr.umn.edu/sti x/fall03/f03stixtab1.pdf ; Ohio State figures are available at http://www.ureg.ohio-state.edu/our web/srs/srscontent/intadobe.html ; and University of Florida Enrollment http://www.ir.ufl.edu/factbook/i-02_hist.pdf Facility size and branch data were taken from the web sites of the universities, building data collected in 1993, and email responses from appropriate staff at the universities. Degrees granted information came for the University of Minnesota came from 20052006 University Plan, Performan ce, and Accountability Report [ Retrieved 10/22/2006 at http://www.academic.umn.edu/a ccountability/reports/20042005.html ]; Ohio State: Degrees Awarded by College Offering Major, Summer 2003-Spring 2004 [ http://www.ureg.ohiostate.edu/ourweb/srs/srscontent/d egree/majcollege/Adobe/cnts0304.pdf ]; Wisconsins 2003-2004 figures [Retrieved 10/22/2006 at http://www.wisc.edu/about/facts/#degrees ; and UFs figures from Degrees, Grades, & Graduation/Retention Rates [Retrieved 10/22/06 http://www.ir.ufl.edu/factbook/degree.htm ] Other library data for this section was obtained from:
47 Association of Academic Health Science Libraries Data for FY 04/05; personal communications with Jane Blumenthal, Director, University of Michigan Health Science Library; Linda Watkins, Director, University of Minnesota Health Sciences Library, Carol Jenkins, Director and Steve Squi res, Library Administration, University of North Carolina Chapel Hill Health Sciences Library, Terry Ann Jankowsk i, Head of Information and Education Services, University of Washingt on Health Sciences Library, and from personnel at the University of Maryland Baltimore Health Sc iences Library.