Introduction

No librarian charged with stewarding collections has ever had too large a budget. However, many in the last decade have survived with far too little. The University of Kansas Libraries (KU) has had a flat collections budget since 2008. Making sure Library patrons have immediate access to the correct materials has necessitated extensive and complex evaluation of what to retain and what to add. Through assessment and hard work the Content Development Department stretched the budget for a number of years without making glaringly obvious cuts. However, eventually KU reached the Rubicon – the items canceled would be obvious and their absences would affect teaching and research. Due to a lack of other options, important databases (that were once considered essential) and eventually a big deal journal package went to the chopping block. The outcome of our negotiations to retain some of these titles remains to be seen.

An all or nothing world

Canceling materials to balance the budget was not a new activity for KU. Recessions and other financial impediments in the past necessitated print cancelations. However, evaluating print materials differed significantly from today’s publishing patterns. Print journal subscriptions were individual commitments – nothing came as a package beyond parent/child titles, making it possible to cancel one title without affecting the rest of the collection. In the print world, journals were usually stewarded by one librarian, and retention and renewal decisions were theirs alone. Reference materials, such as large indexes, might have required the agreement of more than one librarian, but even then it did not necessitate buy-in from all librarians with collection responsibilities or feedback from faculty across campus. The advent of electronic information changed this independence as many publishers only sell materials in packages. Because of this, librarians no longer have the flexibility of choice. It is an all or nothing world.

Stop-gap measures

In late 2007 KU canceled individual print subscriptions that were duplicated in aggregator databases. The purpose of this cancelation project was to decrease redundancy, so it was not an activity precipitated entirely by financial worries. However, this undertaking decreased expenditure sufficiently that, in 2009 when the collections budget did not increase from 2008, there was enough money left to cushion the lack of additional funds. That year’s routine budget analysis concluded that, even with a flat budget, everything would be okay for two fiscal years. However, in the third year and final projection, KU’s collections allocations faced a significant deficit. Unfortunately, these budget projections only included the costs to maintain current acquisitions – there were no funds to purchase new materials or fees beyond anticipated inflation. Because of the desire to add new products and ensure the collection did not become stagnant, we had to free up funds for the future.

Therefore, in 2009 we began a never-ending collection review, which initially focused on quick and easy fixes that would not be blatantly obvious to Library users. With help from our vendor, KU’s robust approval plan went under substantial review. By decreasing the publisher lists, removing more types of materials from automatic shipment, and other refinements, KU enjoyed significant savings. A print demand-driven acquisitions (DDA) program for select publishers in the sciences and social sciences was launched as part of this overhaul. Eventually, this program expanded to include both print and electronic titles across all disciplines. Fine-tuning the approval plan released funds quickly and bought time for more thorough evaluation in other areas with greater savings potential.

Amending the approval plan was far easier than the decisions that lay ahead. So began the arduous task of cutting as many resources as possible that would not be missed. To do this, we focused on identifying duplication and low use. For a portion of the collection, we took a subject approach (following the method used by KU for invoicing and tracking Library materials) and for the rest, the assessment was made by format. Individual journal subscriptions and standing orders at KU are organized by subject and called continuous payment histories (CPH). Starting in 2010, staff scrutinized these lists title by title. Data used in decision making included:

  • use
  • cost
  • full text availability elsewhere
  • lengths of embargoes
  • shift to open access
  • status of title (especially for monographic serials and titles with fascicles)
  • relevance to KU’s curriculum and research.

Although time consuming, this proved productive as it became obvious from these lists that many aspects had not been evaluated for some time. There were many dead and/or completed titles whose records simply needed to be closed. Many other titles were ones KU could simply do without. Although each individual subject CPH review did not accumulate very large savings, when pooled together the funds helped delay the coming doomsday a little longer. Analyzing these lists became an ongoing task for those assigned collection development responsibilities.

While the review of CPHs presented us with the opportunity to cancel individual journal titles outside of packages, we chose to look at some of these titles again, simply because the publisher permitted cancelations on a title-by-title basis. We scrutinized these titles twice, in 2013 and 2014. To evaluate these subscriptions, we gathered significant amounts of data. Our spreadsheet contained the following for each title:

  • cost
  • yearly use for three years
  • average use for three years
  • how often cost-per-use at $50 each exceeded the subscription fee
  • cost-per-use in the previous calendar year
  • other access points and length of embargo in aggregator databases
  • number of uses that fall in embargo
  • cost-per-use during embargoed period.

In particular, we tried to determine how many times an article was used while it was embargoed in aggregator databases. With usage by publication year, we calculated the costs using a copyright fee for each use of $50. Granted, this calculation did not account for the staff time it took to process a request, but that was a fixed cost and now many go through unmediated services like RapidILL (a resource-sharing system). Many titles used during the embargo would cost far less in copyright fees than a subscription. This activity caused a shift in how we thought about subscriptions. No longer could KU pay high rates for one year or 18 months of access. We needed to strive to pay for a title only once, and it became obvious that at times paying the copyright fee for usage while waiting for an embargo to lift was more affordable.

Evaluating electronic resources generated the majority of the work as well as the greatest potential for savings. At KU the classification of an electronic resource includes products that are not journal packages, hence this category includes A&I databases, aggregator databases, reference materials, and other electronic materials. Assessment of these materials began in 2011 but, as the reality of the budget worsened, it became an even more important activity. By 2013, reviewing electronic resources practically became one librarian’s full-time job.

Understanding the potential savings, we decided we needed to be systematic in our approach. Working with our renewals unit, we generated a title list of these materials arranged by renewal date. This document became our bible for cancelations. We worked ahead of cancelations and planned accordingly to have reviews done prior to deadlines for decisions. In this set of evaluations, we aimed to cut material that we classified as ‘low-hanging fruit’ that would not be missed by many. This group of cancelations resulted in the cutting of numerous A&I databases whose use has plummeted as patrons increasingly utilize alternative discovery tools such as Google Scholar. We scrutinized the most used content in full-text aggregators and realized that the same high use materials in one aggregator were usually also available in other products that contained unique content. We canceled the databases without one-of-a-kind materials but kept those that added material to the collection. We had been paying more than once for the same titles. While it was difficult to cut standard titles, such as Biological Abstracts and Factiva, the data proved that use and content did not warrant continuing these subscriptions. Cancelations of these types of products resulted in significant savings. And surprisingly, lack of access to the materials did not generate the concern from faculty and students that we anticipated.

The time had come…

After many years of picking the low-hanging fruit for cancelations, in 2016 we realized the need to find at least $200,000 in cuts. It became apparent that it would be nearly impossible to patch together enough cuts without reviewing a large journal package. Luckily, one of our largest packages was coming up for renewal at the end of 2016, giving us a full year to review titles and usage and collect enough data to make a case for cancelation. This package cost approximately $350,000 in 2016, so it would give us the savings we needed, while leaving some funds for retaining the most important journals in the package.

This particular journal package was purchased through a consortium, whereby all of the schools that subscribed combined their title lists to make a large package of 1,700 titles available to all of the schools in the consortium. In many ways, it still seemed to be a great deal for KU, but upon the initial review, we soon realized that only 765 titles had received 20 uses or more in 2015, and previous years had similar usage. Of the top 500 used titles, 261 were available in aggregator databases with a one-year embargo. Not only were we paying for the journal package, but we were paying twice or more for access through one or more aggregator database. We also calculated the inter-library loan (ILL) costs for the use of the most recent year of publication in 2015 and found that it would cost almost $100,000 less to rely on ILL. In addition, given that cancelations do not transfer to ILL, our ILL costs would be even less. The calculations took into account that five copies of articles from a single title could be obtained for free before a $40 fee would be charged for copyright.

After this initial review, we decided in the late spring of 2016 to look more closely at the 500 top used titles in the package. We filled a spreadsheet with the following data:

  • number of aggregators providing access to a single journal title
  • dates of aggregator provided access
  • impact factors
  • cost-per-use based on list price
  • cost-per-use of embargoed year based on list price (for those titles available in aggregator databases).

We also requested a list of faculty who had self-reported their service on editorial boards from the KU Faculty Professional Record Online office. We then used the list to match titles in the journal package. We checked each journal’s editorial board on their website to make sure the faculty members were still members before we finalized the list. Although this list did not play a key factor in our cancelation decision, it did help us with retention decisions for the titles we planned to retain at list price. Content Development Librarians decided to allocate $100,000 for retaining journals from the package on a title-by-title basis. With the deadline for renewal decisions looming, KU Librarians contacted the consortium to communicate their decision not to renew the package. With all of this data collected, we were prepared to make a case to faculty for canceling this journal package.

Communicating and engaging faculty for transparency

After most of the data was collected, Content Development and Acquisitions staff met with the new KU Dean of Libraries, Kevin Smith, and Scholarly Communications staff to make their case for canceling the package. The Dean was on board and made the decision to approach the new KU Provost, Neeli Bendapudi, who was also supportive of our decision. She asked the Dean to communicate with the Deans of the Schools and Colleges across the KU campus before we made our final decision. Dean Smith sent an e-mail message to the administrators during the summer of 2016, warning them that our important decision was imminent. The feedback from the Deans was supportive and, when the faculty had all returned to campus for the fall semester, Dean Smith wrote another e-mail message to the KU faculty listserv and provided a link to a web page we had developed with information regarding our decision-making process as well as FAQs. On the web page we provided a total list of the titles in the package and a list of the top 500 used titles. Faculty were asked to use a web form, also provided on the web page, to provide feedback and a list of titles they wanted to retain.

Within a matter of minutes, on the Friday afternoon the message was sent out, more than 50 faculty members had responded with lists of journal titles to retain. Over the upcoming days and weeks, faculty continued to send in their comments and retention requests. Most of the feedback was supportive, with some faculty asking how they could become involved in communicating with this publisher to express their grievances concerning this unsustainable publishing model. Several faculty members wanted to learn more about the open access (OA) movement and how they could support a more sustainable scholarly communications model. The opportunity to convey a message about scholarly communication and the OA model was addressed in collaboration with the libraries’ Scholarly Communications staff.

Content Development Librarians compiled a list of titles from the web form to use in developing the final retention list. The goal was to choose approximately $70,000 worth of individual titles for retention with the hopes that the remaining $30,000 allocation could be used to pick up stray titles as more feedback was provided over the course of the remaining part of the fiscal year.

Developing a retention list

Next came the onerous task of developing a list of journals from the package for retention. Parameters for selecting which titles to keep had to be defined. As is typical in such decision-making practices, usage data was an important tool for deciding which journals to retain. First, cost-per-use for each of the top 500 journals was calculated. If a title that was selected by faculty did not appear on the top 500 list, it was not accepted for retention. Next, the cost-per-use for titles that were embargoed for a year in an aggregator database was calculated for usage of the most recent year. If the cost-per-use for a title during the embargoed year was more than the $40 ILL cost, it was not selected for retention. Another metric that was taken into consideration was how many times faculty had flagged a title for retention. Finally, titles that were not available in aggregator databases were considered for retention when their cost-per-use was $20 or less.

Once all of the data had been collected for the retention list, Content Development Librarians began responding to faculty who had provided feedback. Retention decisions were communicated by providing the institutional list price for each title with corresponding cost-per-use data. Faculty who responded were very understanding of the tough choices we were having to make and voiced their support for our decisions. A list of 75 titles was finalized in mid-November and representatives of the publisher were contacted to begin the process of negotiating the cost of a smaller package. During the first meeting with the publisher, they told us we would have to pay the institutional list price for the titles we selected, but they would agree to a price cap if we signed a multi-year license. They also agreed that we could have unlimited swapping as long as we kept the initial spend. We sent them our title list in mid-November and asked them for a three-year deal with a 3% increase price cap. As of early December, they had not responded.

Finally, in late December, we received their response. Due to the nature of the titles in our list, they [the publisher] reneged on offering us a multi-year deal with a price cap. They claimed that the average inflationary increase for the titles we wanted to retain was 5.3% and they were unwilling to give us a lower price cap. In the past, with our consortium deal with this publisher, we had been allowed to provide access to all titles in the package to the KU Medical School, but now, the publisher insists that we must pay 150% to provide access to all of the titles for both campuses. This last demand from the publisher was untenable for KU. The only way for us to continue negotiations was to review the list of titles once more and cut even more.

Negotiations will likely go on past our renewal deadline. There is also concern that another journal package tied to this publisher will be at stake. This other package is also shared on the KU and KU Medical School campuses and gets extremely high use. The future for both title lists is uncertain and very worrisome for librarians on both campuses.

No end in sight

The rest of our large journal packages come up for renewal in 2018 after multi-year deals, and Content Development Librarians will need to scrutinize these packages as well. Ongoing review of all electronic resources along with review of CPHs will continue. There appears to be no end to the review process in sight as we anticipate a flat budget in the foreseeable future, possibly in conjunction with budget cuts. These are challenging times at KU, but we will continue to work to provide our patrons with the information they need.