by Karen Harker (Collection Assessment Librarian, University of North Texas Libraries, Denton, TX)
and Coby Condrey (Collection Development Liaison Librarian, University of North Texas Libraries, Denton, TX)
and Allyson Rodriguez (Strategic Collections Librarian, Univ. of North Texas Libraries, Denton, TX)
and Emily Billings (Electronic Resources Librarian, University of North Texas Libraries, Denton, TX)
The “great squeeze” of increasing costs and decreasing funding put our library into a series of budget crises which resulted in severe restrictions of library purchasing. Resistance, anger and frustration from faculty upon learning of the cuts was fueled partly by a lack of awareness of issues and technicalities from both sides. Faced with rising faculty push-back, the university president guaranteed a significant allocation of discretionary funding for collections purposes, but only for three years. This has presented the Collection Development Department staff with a dilemma — how to spend a relatively robust but temporary budget, while planning for an unknown future. Through demand-driven acquisition (DDA), in-depth collection assessment, targeted enhancements, Open Access resources, and rigorous evaluation and tracking of current and possible resources we seek to meet the needs of our faculty and students.
Since 2004, the UNT Libraries have been funded nearly completely by a student use fee, with undergraduate and graduate hourly caps. Until 2010, the increase in student enrollment exceeded the increase of costs due to inflation. Student enrollment, however, has plateaued since 2011, reducing the purchasing power of the library. From the 2011 to the 2015 academic years, the UNT Libraries has reduced its collections budget by 36%, accomplished largely by cancelling most approval plans, eliminating dual print and electronic subscriptions, breaking up selected journal “Big Deals,” and reducing firm-order or one-time purchase funds.
Before 2012, collection development at the UNT Libraries was largely decentralized, with subject specialists or Liaison Librarians responsible for spending a set amount annually on one-time purchases. The management of this collection development method was cumbersome, given the innumerable accounts (firm orders, subscription, etc. for each subject). Also difficult to manage were the selectors themselves. Many of the funds were spent appropriately, but there were those that were overspent, underspent or unspent.
Methods for the Three Sets of Cuts
Managing Budget Cuts from 2011-2014 — The first year of cuts, $750,000 for budget year 2011-12, was relatively modest and met by eliminating most approval plans, reducing firm order funds by over 70%, and converting from print-plus journals to electronic only. The next year, another $1 million was eliminated from the collections budget. The focus of these cuts were on the recurring expenditures. The criteria used to evaluate these resources were low overall usage, high cost-per-use, overlap, and access restrictions. In addition, we invited the Liaison Librarians to rate resources specific to their fields based on feedback from faculty. Despite liaisons inability to rate related resources purchased from other funds and challenges with librarian ratings of resources, we were able to meet the budget reductions.
For the 2013-14 year, the UNT Libraries was provided funding to cover inflation, sparing additional cuts. We used this year to prepare, by compiling more longitudinal usage data and to revise our evaluation methods. The article by Foudy and McManus (2005) described a method to rank resources for de-selection purposes, incorporating measures of access, cost-effectiveness, breadth/audience, and uniqueness inspiring our current method.
Measures Used in Evaluation — We used five measures in our evaluation: Overlap, Usage Trends, Cost-Effectiveness, Cost Trends, and Librarians’ Input. For overlap, we examined not only full-text overlap of journals (as compared to subscription access), but also of database indexing. Several tools were used for this analysis for comparing coverage between full-text aggregators, as well as database indexing. We also downloaded journal coverage lists from database providers for manual comparison. This method is limited in that it only includes titles, but lacks date ranges.
For usage, we applied the usage measure closest to the end user’s experience with the resource. This measure would form the foundation of cost-effectiveness (CPU) and vary by resource: abstract/record view for databases and reference resources; full-text downloads for ejournals and eBooks. Thus, the resources were compared with others of the same type. The three-year average usage and average cost-per-use were the measures used in the analysis.
To incorporate our concerns about extraordinary cost increases, we included expenditures over several years. Adjusting for changes in pro-rated costs for missing or overlapped months, we were able to calculate the average annual rate of change in costs.
For evaluating our “Big Deals” we not only examined overall usage, but also the distribution of usage across each of the titles in the package. The greater the distribution of titles used, the more effective the deal was as a package. If a package was considered a poor value, then we calculated the cost-per-use based on the subscription cost of each title, comparing that with the cost of obtaining articles from alternative methods like interlibrary loan.
The final component of our scale would be the more subjective measures of quality, which were provided by the Liaison Librarians. We provided the liaisons with an expanded list of resources that included not only those directly paid for by their funds, but also relevant interdisciplinary resources.
Relative Distribution — The usage, cost-per-use and the change in costs were all continuous measures, each on a different scale, making it difficult to compile into a single scale measure. To avoid the scale being unduly influenced by a single factor, we placed all of these measures on the same scale: their distribution relative to other resources of the same type. Using percentiles, we were able to effectively rank the resources on the same scale for each of the criteria.
Finally, we averaged all three scales (CPU, Inflation, and Liaison Ratings) into one composite score, which ranked the resources within their respective categories from lowest- to highest-performing percentile. Those in the lowest 20th percentiles were considered strongly for cancellation, while those in the highest 50th percentile were renewed automatically. Those performing between these two thresholds were examined more closely, with additional feedback sought from the liaison librarians.
Communicating Cuts and Fallout
Preparing for Cuts — We closely examined only resources costing $1,000 per year or more, and recurring subscription, or subscription-like costs. We then set a series of goals, including the percentage across the board and overall dollar amount of cuts to be made, as well as a timeline for liaisons to select items to drop.
To communicate the data to the liaisons, we simplified the measures previously described. The overlap and usage trends measures were omitted and the remaining measures were relabeled more explicatively. These slightly revamped measures allowed us to organize the data about our purchased resources to more effectively communicate with the liaisons, and the various academic departments, options for a plan of action.
The liaisons reviewed the spreadsheet, which allowed them to see how resources ranked, plus key information about the resources. Resources were identified as irrevocably cancelled, planned to be cancelled (but able to be switched out), eligible to be sacrificed, modified in some way, or retained and unable to be cancelled. Liaisons were directed to share the data with their faculties with an adequate timeline to reach a collaborative decision about how to reduce their resources to reach the budget goal.
Differences in Perception of Need and Desire — As we communicated this information, it became apparent that some highly-prized resources were at risk. Some faculty were especially vocal in advocating the retention of certain subscriptions to the exclusion of other, higher-ranking titles. The data revealed a dichotomy between faculty perceptions of necessity in a research discipline and empirical evidence about what was a good value.
We countered the considerable faculty push-back as positively as possible, turning conversations with faculty into teachable moments. We encouraged the use of interlibrary loan. We informed faculty of the library’s flat materials budget, the impossible inflation rates, and varied pricing structures based on “hidden” factors like potential audience size, simultaneous use, etc. We explained how our usage data was gathered, encouraging faculty to link to an article rather than uploading a PDF copy to learning management systems. We documented our decisions and set up a method to track resources when the performance data did not coincide with expectations: an “endangered species” list of sorts.
The Breaking Point: Big Deal Packages — One of the biggest challenges we faced was dismantling the library’s journal packages, not all of which were up for contract renewal. We applied our same analytical methodology to the titles within our “big deals,” with the additional measure of the distribution of usage across titles within the collection. Then we targeted between 80% and 90% of the titles in the three major journal packages targeted for elimination. We allowed liaisons and faculty to make a case to retain especially important titles, or to swap out other titles, but we ultimately had to excise a major portion of our journal content received via bundled deals.
The following year, 2014-15, the library’s budgetary conditions required more cuts. We turned our focus to the packages up for renewal, notably Elsevier’s ScienceDirect. While the package was not performing particularly poorly, the cost of the subscriptions was untenable. The proposed eliminations in ScienceDirect were severe, and our ability to subscribe to journals with high ratings was limited due to the comparatively high cost of many of the titles. We again made preliminary recommendations to the liaisons and faculty, and allowed a period for feedback and negotiation.
When the faculty realized the impact of a second year of cuts, there was another outcry. Tensions rose leading to the faculty senate discussing the funding problems in open meetings, the local press reporting the story, and university administration becoming concerned about image. The university president intervened allocating funding to stave off the second round of cuts. He also committed to continuing the extra funding for two additional years, giving the university and library administrations time to address the shortcomings of the library’s funding model in a long-term way. For the short term, though, we were able to renew our remaining big deal packages mostly intact.
Summer of Love
The infusion of funding created a different set of challenges for the collection development librarians. Partially in response to expected shortfalls in funding, the collection development department was implementing an entirely new philosophy (ABCD or access-based collection development) (Crawford et al., 2014) that emphasized access over ownership. After conducting so much analysis of resources, it seemed cavalier to ignore our analysis. We refused to reinstate everything we cancelled. We reinstated many titles that were in the moderate ranges, but aggressively pursued better pricing from vendors, modified arrangements with vendors reducing costs, promoted “endangered” titles, and financially and fundamentally supported interlibrary loan.
The plan to use the money as effectively as possible incorporated three elements that the department recently had developed and put in place: the ABCD group, adoption of demand-driven acquisitions programs, and an initiative to “enhance” the collection in targeted disciplines.
ABCD Group — The ABCD group consists of key librarians who collectively oversee the use of the materials budget, investigating resources, and selecting purchases. The department’s leadership and the librarians responsible for assessment, liaison communications and monographs acquisitions, strategic collections, serials, electronic resources, and media meet regularly to plan spending, identify needs and desires, report options from vendors, and make decisions. The ABCD group cohesively operates with a sense of collective purpose, and from a comfortable position of accountability for spending decisions.
DDA — DDA has become an important part of the ABCD plan. We have embraced the new model as an effective way to make a wide range of materials available to researchers without buying all of them up front; we monitor its use regularly to avoid overspending. We have implemented a variety of DDA-like programs for streaming media, some major eBook publishing firms, and through our primary monographs vendor who allowed us to consolidate management of multiple eBook platforms. With an institutional profile capturing the scope of the university’s research interests and automatically loading selected discovery records into our catalog, we feel confident we’re meeting a wide range of needs with reduced intervention by staff while ensuring that the eBooks purchased are ones valuable to researchers.
Targeted Subject Areas — Finally, we have devoted a substantial portion of the new monies to support selected programs or research areas. The collection assessment librarian annually evaluates the support of several specific disciplines on a ten-year, rotating basis. Using these disciplinary collection assessments from the previous year as a starting point, a sub-set of the ABCD group investigates ways to enhance the resources supporting research in the targeted disciplines. Given the large amounts of money, title-by-title procurement is not feasible; instead, the guiding principle is to identify purchases that provided access to many resources at a single pass. The ABCD group reviews and discusses the final recommendations for each enhancement purchase.
Preparing for the Inevitable Drought
Endangered Species — The Endangered Species list was created to monitor our ongoing subscriptions using the criterion created during the budget crisis. These resources, as well as those newly added to our collection, are given three years, during which time the liaison librarians promote or review the resource. The ongoing and open nature of the list offers our liaisons ample opportunities for feedback and awareness not previously provided. The analysis gives us valid reasons for cutting a product and using the money in other, more productive ways.
Open Access — Beyond the current methods outlined above, we must prepare for a lack of funding. Including high-quality, librarian-evaluated Open Access materials in our library catalog is one method of preparing for the future. However, simply pointing a patron to an Open Access resource may not be enough. We are exploring ways to harvest and archive a local copy of Open Access resources through collaboration with our Digital Library. Furthermore, our current budget allows us to support Open Access Initiatives, such as Knowledge Unlatched or Independent Voices, providing current and future access to scholarly materials and ensuring the future of our collection.
As we approach our uncertain future with funding for collection development, our decisions become more important in providing access to quality resources that are in demand now. Cost for materials and demand for items increases every year, but with strategic growth and detailed plans, we are able to transition and remain relevant to our users. We were fortunate to gain a temporary infusion of funding. With these funds the ABCD group, DDA, and targeted subject area enhancements will prepare us for the possible drought. The continued detailed monitoring and analysis of our electronic resource subscriptions and the addition of open access resources will take the collection in new directions.
Crawford, L. S., E. F. Avery, K. Harker, T. Enoch, and C. Condrey. “White paper: redefining collection development at University of North Texas Libraries.” May 2014. http://digital.library.unt.edu/ark:/67531/metadc335286/.
Foudy, Gerri and Alesia McManus. “Using a decision grid process to build consensus in electronic resources cancellation decisions,” The Journal of Academic Librarianship, volume 31, #6, pages 533-538. November, 2005.