by Anthony Raymond (Business Librarian, Santa Clara University, Santa Clara, CA 95053)
(This article was originally published in the April 2017 issue of Against the Grain.)
Canceling an individual serial subscription when the journal is available in a third-party aggregated full-text database (AFTD) has been an option for academic libraries since these databases came into wide use in the late 1990s, yet little discussion of this option has taken place in the literature. Third-party aggregated full-text databases refer to products sold by companies that do not themselves publish journals but only distribute journal content — for example, various well known products sold by EBSCO and ProQuest and some open access databases such as Project Muse. This article looks at several case studies that discuss this option at some length and describes Santa Clara University Library’s (SCU) experience employing it. Two of the studies conclude that canceling individual journal subscriptions based on their availability in an AFTD is an acceptable, even desirable, option while two others conclude it far too risky. Considering the many variables involved, this article argues that there is insufficient evidence to make a definitive judgement about whether this option is appropriate for all academic libraries or for all subject areas. It also suggests that fiscal responsibility demands that academic librarians evaluate this option within the context of their institutional and disciplinary circumstances rather than rely on studies and experiences from other libraries that may have little relevance to their specific situation.
Among the numerous articles describing budget-driven serials cancellation projects undertaken by academic libraries since 2000, only a few give serious consideration to the option of canceling serial subscriptions based on their availability in AFTDs. Considering the significant potential cost-savings, it is surprising that more attention has not been paid to this issue. Sprague and Chambers (2000) is perhaps the earliest attempt to understand the implications of canceling individual journal subscriptions based on their availability in AFTD. Their study, conducted at the University of Colorado, Colorado Springs, Kraemer Family Library, compares the content of seventy-nine print journals from a wide range of subject areas to the full-text content of these same journals in five AFTDs that the library subscribed to at the time. They compared the content in terms of currency (access to the latest issues of the journal); coverage (reliable access to all major articles); graphics (the inclusion of all figures, tables, formulas and other graphical information); and stability (availability of journal content over the long term); finding deficiencies with AFTD content in terms in all four areas. Their conclusion was that it was much too risky for libraries to use AFTDs as a replacement for individual journal subscriptions.
Kalyan (2002) is an early description of how this option was employed by a U.S. academic library in response to a budget shortfall. This study identified 461 print journals eligible for non-renewal in a wide range of disciplines based on the fact that the content was sufficiently available in one or more AFTDs that Seton University Libraries subscribed to. Journals with embargoes of twelve months or more were excluded. It was decided that the cost-savings of $83,000 resulting from canceling the 461 journals duplicated in at least one AFTD outweighed the risk of publishers imposing embargoes or completely pulling their content. The Kalyan study was the inspiration for adopting this option for the subject areas of business and economics at SCU in 2005. Nixon (2010) describes Purdue University Libraries’ use of this option in response to yet another budget-driven serials cancellation project. Unlike two previous cancellation projects undertaken in 1992 and 1997, in 2009 Purdue had access to several AFTDs providing a significant increase in the number of full-text journals they had access to and they decided to take advantage of this despite the known risks. Anticipating that content would indeed be pulled, Purdue set aside funds to replace any journals that were dropped by an aggregator. In a follow-up study to Sprague and Chambers (2000), Thohira, Chambers and Sprague (2010) confirmed that the deficiencies found in AFTD content ten years earlier, most importantly content instability, made using AFTD journal content as a replacement for individual journals still too risky to be a viable option at the University of Colorado.
The Santa Clara University Experience
While it is not known if the two libraries above that used this option ended up regretting doing so, SCU’s experience suggests this is probably not the case. Since early 2005, SCU Library has been canceling individual serial subscriptions in the subject areas of business and economics when the serial is sufficiently available in an AFTD. For SCU Library, sufficiently available means no publisher-imposed embargo. Exceptions to the no embargo criterion are made for publications considered of only marginal value to the SCU research community. With Innovative Interfaces, it is a simple matter to generate a list of all active serial subscriptions by call number and, using Proquest Serials Solutions, identify those journals available in one or more AFTD.
The cost savings have been significant: since 2005, seventy-five individual serial subscriptions in the subject areas of business and economics have been canceled, serials supporting the disciplines of Accounting, Finance, Management, Marketing and Operations Management Information Systems (see Table 1). The savings, calculated simply as the cost of the subscription at the time of cancellation, amounts to $22,750 annually, or $227,500 over ten years. Calculating in the annual inflationary cost increase, which is notoriously high for journals, would substantially increase the savings achieved over a ten-year period. This practice was never announced to the university research community because it was known that neither faculty nor students are interested in whether the journal article they want is available directly from the publisher through an individual subscription, or through an AFTD. More than ten years after this practice began, SCU faculty and students have not submitted a single complaint about, or even commented on, the coverage (embargoes or otherwise missing articles) or the quality of the articles (missing tables, charts, graphics, etc.) retrieved from AFTDs. Nor have faculty asked that a single one of these cancelled serials, including those where the AFTD content is embargoed, be re-subscribed to.
The cost savings achieved without any noticeable negative consequences for the SCU research community can only be interpreted as an unqualified collection management success and raises a number of questions regarding the findings of Sprague and Chambers (2000) and Thohira, Chambers and Sprague (2010). Why did the deficiencies identified in these studies — and there is no question that these deficiencies are real and still persist today — have no noticeable impact in SCU’s case? Certainly, it is not because SCU faculty are less concerned about missing data tables, charts, etc., in the articles they access than their counterparts at other institutions. A quick perusal of the Leavey School of Business (LSB) website reveals that LSB faculty are highly productive, internationally recognized scholars, in some cases among the most influential scholars in their field. The reason must be that either the deficiencies identified in the University of Colorado studies do not appear in business and economics journals (highly unlikely), or they do appear but so infrequently and with such little consequence that they are, in practice, insignificant to faculty and students. If this is the case for the areas of business and economics, could it also be the case for other social sciences? Might it be the case for the humanities and sciences as well? Regarding the small number of cancelled journals whose content is embargoed in AFTDs, it appears that these were correctly identified as marginal and neither students nor faculty felt this loss of coverage to be worth bringing to the Library’s attention. Of course, it could also mean that faculty and students switched to requesting articles from embargoed titles using the Library’s article delivery service but this is unlikely because LSB faculty place a premium on immediate access to articles and had the embargoed journals caused significant delays to access, it would certainly have been brought to the attention of the Library.
It is probable that the sample sizes in the University of Colorado studies and in the SCU experience, 79 and 75 titles respectively, are too small to be generalizable, and no definitive judgement can be made based on either case. This may mean that it is far less risky to cancel serials based on availability in AFTDs than assumed in the University of Colorado studies. Institutional and disciplinary circumstances may render deficiencies in AFTD content less significant in practice than they appear to be in theory. Of course, there are common sense considerations that should be taken into consideration but experienced subject specialists should be able to make many non-renewal decisions without having to consult faculty, with the caveat that it is always wise to consult faculty when in doubt about a particular journal. Careful consideration of whether to cancel a journal whose content is embargoed in an AFTD is very important because, in most cases, this will result in an unacceptable loss of content. At SCU, if there is any doubt that an embargoed serial is of only marginal value, it is not cancelled.
Depending on the discipline and the local characteristics of the research community, other factors may also be critical to the success of employing this option. For example, in the case where high quality reproductions of works of art are critical to the reader, unless the quality of the reproductions in the AFTD are known to be of sufficiently good quality, then canceling the journal would be unwise even if it were otherwise sufficiently available. Similarly, for some journals in the sciences and engineering, missing data tables, charts, etc., from AFTD content would be unacceptable to researchers and exercising extreme caution in canceling these journals is called for. In such cases, the AFTD content must be examined carefully to see that it faithfully reproduces the contents of the print or electronic journal it is going to replace. Today, most academic libraries are transitioning away from print and many print journals have already been replaced with online only subscriptions. For example, SCU Library no longer receives print journals in the subject areas of business and economics. However, the fact that an individual subscription is for an e-journal, rather than its print version, does not exempt it from being cancelled if it is sufficiently available in an AFTD.
Considering the significant immediate and long-term cost-savings academic libraries can achieve, it seems a matter of fiscal responsibility that this option be given serious consideration, not only as a response to a budget shortfall, but simply because the money saved can be put to better use. In the worst-case scenario of faculty demanding that an individual journal be re-subscribed to, this can be done easily enough. Large numbers of academic libraries canceling large numbers of individual serials subscriptions, print or electronic, based on their availability in an AFTD would, because of the enormous loss of revenue, surely provoke a response from publishers. If canceling just 75 subscriptions in the subject areas of business and economics saved SCU Library tens of thousands of dollars in the short term, and hundreds of thousands of dollars over the long term, how much more might be saved (and journal publishers lose) if this option was adopted for all subject areas? Multiply this by hundreds, or thousands, of academic libraries across the United States and it is clear that the loss of revenue would be far too significant to ignore. How would journal publishers respond to the mass cancellation of individual serial subscriptions because they are sufficiently available in AFTDs? The obvious option would be for journal publishers to impose long embargoes on AFTD content. After all, the purpose of embargoes is precisely to prevent libraries from canceling individual journal subscriptions. In fact, this is why some academic librarians are reluctant to even discuss this option in public. However, perhaps it is time that academic librarians disrupted the current business and distribution models in the best interests of the research communities we represent. We know that our parent institutions cannot indefinitely continue annual library budget increases that keep pace with the annual cost increases imposed by journal publishers — increases that are routinely three, four or five times greater than the average annual rate of inflation. Nor can we continue to pay for access to AFTDs that publishers claim should be considered only as indexing and abstracting tools and should not be used to replace individual journal subscriptions. It is time for publishers to reexamine the current business and distribution model that forces academic libraries to maintain current individual serial subscriptions while at the same time forcing them to subscribe to very expensive AFTDs with overlapping content. Perhaps there is a better solution, one that is financially sustainable and better meets the requirements of the academic research communities in the 21st century.
Kalyan, S. (2002). Non-renewal of print journal subscriptions that duplicate titles in selected electronic databases: a case study. Library Collections, Acquisitions, & Technical Services, 26, 409-421.
Nixon, J. M. (2010). A reprise, or round three: Using a database management program as a decision-support system for the cancellation of serials. The Serials Librarian, 59, 302-312.
Sprague, N., and Chambers, M. B. (2000). Full text databases and the journal cancellation process: A case study. Serials Review 26(3), 19-31.
Thohira, M., Chambers, M. B., and Sprague, N. (2010). Full-text databases: A case study revisited a decade later. Serials Review, 36 (3), 152-160.
Table 1: Serials canceled because they were available in 3rd party AFTD.