Column Editors: Allen McKiel (Dean of Library Services, Western Oregon University)
and Jim Dooley (Head of Collection Services, University of California, Merced)
One constant over the last several years has been library experimentation with various eBook acquisition models. While the majority of these experiments have involved individual libraries, some, most notably by the Orbis-Cascade Alliance, have involved consortia. As a result of the experience of the Alliance with a consortial Demand-Driven Acquisitions (DDA) program, the University of California (UC) Libraries decided in 2013 to implement a systemwide DDA pilot with ebrary and YBP. The pilot began in January 2014 and ended on December 31, 2015.
This column in the June 2014 issue of Against the Grain reported on the first four months of the pilot. Michael Zeoli, in his review of Academic E-Books: Publishers, Librarians, and Users in the December 2015-January 2016 issue of Against the Grain quotes some statistics for the UC pilot from August 2014. In the interest of providing a complete picture, this column will report the results of the full two years of the pilot and discuss next steps.
The details of the structure and organization of the pilot can be found in the June 2014 “Changing Library Operations” column. Briefly, the pilot involved 63 university press publishers and was limited to social science and humanities (not including art) titles with publication years between 2010 and 2015. All UC campuses participated except for the University of California, San Francisco (UCSF) which is exclusively a graduate health and life sciences campus. Central administration of the pilot was performed by the California Digital Library (CDL), a unit of the University of California Office of the President. YBP profiled the titles to be included in the discovery pool and managed the deposit account. Titles were purchased after three Short Term Loans (STLs). When a purchase was triggered either three or four copies of that title were acquired to provide access for all nine participating UC campuses. The number of copies acquired was based on historic average systemwide print purchases per title per individual publisher.
In December 2015 the task force formed to conduct the pilot prepared a detailed assessment of results through October 2015. This column will focus on the results of this assessment supplemented by some additional data through December 2015.
The principal finding of the assessment was that the pilot was successful in testing a systemwide eBook DDA model for the University of California. The nine participating campuses and CDL demonstrated that they could work together to plan and implement such a pilot. Access to titles from 63 university presses was provided to nine UC campuses at a systemwide cost of $27.57 per purchased title, per campus. Beyond experimenting with a particular business model, another purpose of the pilot was to test campus interest in eBooks in the Humanities and Social Sciences. The usage data show that interest is high, particularly in Social Sciences (general) and in History. Although the pilot included titles published between 2010 and 2015, usage was concentrated in titles published in 2013 and 2014.
An initial concern was that a majority of the funds would be expended on Short Term Loans to provide access rather than on purchases to build research collections. Given the way the pilot was structured, this concern proved unfounded. Within the pilot the STL costs were 16% of the budget compared to 84% spent on purchases. This result was skewed to some extent due to the multiplier of three or four when a title was purchased. Presumably a different business model could have produced different proportions of spending on STLs vs. purchases.
During the twelve-month period between September 23, 2014 and September 28, 2015 there were 2,538 STLs and 415 titles purchased out of a discovery pool of 4,378 titles. Purchased titles were 9% of the available titles and 1,412 unique titles (32%) had STL activity. Of the 63 participating publishers, 38 had purchases (60%) and 56 (89%) had STL activity. The average list price was $76 for a single copy and $248 with the systemwide multiplier. From the official start of the pilot in January 2014 (although it took several months to fully implement) through December 2015, 12% (578) of available titles were purchased out of a discovery pool of 4,784 titles.
A number of participating publishers, including New York University Press, University of Chicago Press, and Oxford University Press, significantly raised STL rates during the pilot. The task force decided to keep all participating publishers in the discovery pool for the duration of the pilot. However, if the pilot had continued it is possible that publishers that had significantly raised STL rates would have been dropped. Three publishers instituted STL embargoes of 12 months (MIT Press and Cornell University Press) or 18 months (New York University Press) during the pilot which resulted in their front list titles not being available in the pilot. Again, had the pilot continued a decision would need to be made regarding the inclusion or exclusion of publishers with STL embargoes.
Another frequently raised concern with DDA plans is usage of eBooks after purchase. As stated above, 415 titles were purchased during the assessment period. Total aggregated usage of these titles after purchase was 440,524 uses. Each of the 415 titles had post-purchase usage, ranging from fewer than 100 uses (94 titles) to over 10,000 uses (4 titles). Over half of the purchased titles (224 titles) ranged between 100-500 post-purchase uses. Since the pilot provided a fixed number of copies rather than unlimited simultaneous usage, turnaways occurred when the number of concurrent users was exceeded. During the pilot there were 685 turnaways involving 114 titles.
During the assessment period purchases occurred mainly in Social Sciences (111 titles) and History (81 titles) although purchases also occurred in Literary Criticism, Political Science, Business and Law. Within Social Sciences the most popular subject was Sociology/General followed by Anthropology/Cultural.
Before the pilot actually began, the first issue to be addressed was that of publisher willingness to participate. Of the 193 international university presses available on the ebrary platform in 2013, only 63 agreed to participate in the UC pilot when contacted by ebrary. The task force managing the pilot was explicit that it wanted to test the use of STLs, so a reasonable assumption would be that those publishers that declined to participate did so because of an objection to the use of STLs rather than to DDA in general. This assumption may be tested in future projects employing different models. In any event, slightly less than one-third of the university press publishers contacted agreed to participate in the pilot. A small number of publishers were participating in DDA programs with other vendors and were therefore not available to participate in the UC pilot with ebrary.
Another publisher-related issue is that 30% of participating publishers made less than 50% of their total publishing output available through ebrary. Limited title availability could have contributed to lower activity for those publishers during the pilot. On the other hand, 43% of participating publishers offered 75% or more of their output for the pilot. Less than full availability of publisher output, particularly front lists, through aggregators has long been a problem for libraries extending far beyond particular models such as DDA with or without STLs. Specifically for the pilot, selectors at many UC campuses were unable to determine if particular titles from participating publishers would in fact be available through the pilot or if these titles would have to be firm ordered.
A different type of problem was the ongoing difficulty in receiving MARC records containing OCLC numbers from ebrary, now ProQuest, in a timely manner. Records were harvested from ebrary by the UC Shared Cataloging Program (SCP) and distributed to the participating campuses. The June 2014 “Changing Library Operations” column placed the majority of the blame on OCLC; subsequent events showed that the problem in fact mostly lay with ProQuest. The issue became pronounced in the last quarter of 2015 when the situation deteriorated to the point that the SCP was required to download records directly from WorldCat for distribution.
During the pilot participating libraries decided not to try to de-duplicate print acquisitions against the eBooks in the pilot although this would have been an issue had the pilot continued as a permanent program. Campuses with local eBook DDA plans already in place reported the lowest duplication numbers. This may be due to these campuses having already reduced print acquisitions in favor of eBooks. YBP introduced a method for de-duplicating local print approval plans against the systemwide DDA profile in fall 2014. However, according to YBP, 50% of eBooks currently significantly lag the publication of their print counterparts. Thus it is still possible for a large number of print titles acquired through local approval plans to be received before the electronic version is available. Individual libraries varied as to whether they de-duplicated their local e-DDA plans against the systemwide pilot.
At the conclusion of the pilot the task force managing the pilot made two principal recommendations that have been adopted. First, the scope of the task force has been expanded beyond the implementation of a DDA pilot using STLs to include other models of eBook acquisition; in recognition of this expanded scope the task force has been renamed the Emerging E-Book Models Task Force. This signifies a general recognition within the UC Libraries that eBooks have become sufficiently important to require at least a semi-permanent body to investigate, monitor, recommend, and in some cases implement systemwide eBook programs.
In 2013 the UC Libraries published an E-Book Value Statement http://libraries.universityofcalifornia.edu/groups/files/cdc/docs/UC_Libraries_E-Book_Value_Statement.pdf detailing the aspirational goals toward which the UC Libraries would work in developing systemwide eBook programs. The statement lists many desirable aspects of such a program in the areas of content supporting research and instruction, fair use and scholarly communications, positive user experience, product platforms, and sustainable and fair business models. At this time the offerings of commercial aggregators are not well aligned with many of the principles of the Value Statement. A second recommendation was that the task force experiment with a vendor whose products more closely align with the Value Statement. Preliminary investigations have begun; hopefully, decisions will be made and a new pilot launched reasonably quickly.
The UC Libraries remain interested in DDA as an important mechanism for acquiring eBooks. At the same time publishers have made the use of STLs problematic by significantly raising rates and instituting embargoes on front list titles. The time appears ripe to explore other DDA models. The goal is not to conduct pilots, but for many it is to implement a sustainable, permanent systemwide program to acquire eBooks for the UC Libraries that will most likely utilize some form of DDA. Time will tell if this goal is achievable.