Fifteen years ago, when the Big Deal was really coming into its own as a journal purchasing model, I was working in a library where we embraced it enthusiastically. In fact, we had a pretty standard approach whenever a sales rep pitched a journal to us, or a faculty member requested a new subscription: our first response was usually “Well, how much would it cost to subscribe to the publisher’s whole list online?”
At that time there were two things we knew for certain: on the one hand, subscribing to a publisher’s entire journal list (or whatever approximation of “entire” was on offer) was likely to yield a tremendous value in terms of cost per article; on the other, the Big Deal model was probably not sustainable in the long term. But we embraced it anyway for a couple of reasons: first, because the value proposition was so compelling in the short term, and the short term does matter; second, because it was clear even then that the scholarly-communication ecosystem was in the middle of some kind of fundamental revolution, and there was no telling what it would look like ten or fifteen or twenty years into the future. In light of the massive short-term gain that the Big Deal offered us, we made a calculated bet that things in the publishing world would change sufficiently in the coming years to minimize the long-term risk.
Now here we are, fifteen years later — and fundamental revolution has not arrived, the bill has come due, and tough decisions are now required.
Of course, warning voices have been raised repeatedly since the Big Deal emerged, and with good reason. But that has actually been part of the problem: for years now, libraries and commentators have been crying “Wolf!”, saying that the Big Deal was about to implode, and that libraries were about to start bailing out of them. Individual libraries themselves sometimes warned that they were about to cancel their Big Deals, and every once in a while one of them did. But not very many, and soon a conventional wisdom began to emerge: that some libraries were crying “Wolf!” prematurely and then backing down when their faculties got wind of the impending massive cancellations, while other libraries were crying “Wolf!” as a negotiating tactic and subsequently getting attractive offers from publishers in order to preserve their business, and still other libraries were successfully canceling their Big Deals only to return to the fold a year or two later when either a) it became clear that the content really was essential, or b) the publisher offered an irresistible deal (or both).
In recent years, though, it has started looking to me as if the conventional wisdom might be cracking. Instead of hearing about libraries backing away from their Big Deal cancellations, I was starting to hear about libraries really canceling, and I started wondering how pervasive this phenomenon really was. Finally, I posed a question to a couple of listservs that deal with libraries and scholarly communication. I asked the listserv participants to contact me off-list and share the names of any libraries they knew of that had announced plans to cancel their Big Deal packages. By gathering those names and then investigating, I hoped to get a better sense of how many libraries were really canceling, how many were announcing plans to cancel but then backing down, and how many were canceling but then later coming back. I had no idea what I would find, or whether the resulting data set would lend itself to rigorous and quantitative analysis, but it seemed like it was worth asking and seeing what I got in response.
Indeed, the results were very interesting. Before I proceed to the discussion of what I found, I must offer a few clarifying notes about the data:
- Some of this information may not be completely correct. I worked hard to verify it and deliberately excluded examples in which I didn’t have strong confidence, but there may still be errors. I welcome corrections and updates.
- What I present here are only results for 31 North American libraries. I also heard about a small handful of libraries in Europe and Africa, but since there were fewer than five of them in total (not counting the situation in Germany, where a national deal with Elsevier is still in the throes of contentious renegotiation — see here and here) I decided to leave these out of the discussion in order to avoid giving a distorted impression of what may be happening in those large geographic regions.
- This discussion is only about Big Deals, not smaller subject-based packages. I encountered quite a few dead ends when I was told that “Library X just cancelled their Big Deal,” only to investigate and find that what had actually happened was a journal-cancelation project that targeted smaller packages and individual titles.
- Two libraries reported to me that they were currently in the process of negotiating departures from one or more of their Big Deals. I’ve left these out of the discussion, since their outcome is still uncertain and the process sensitive.
- One library (which shall remain nameless) told me that a publisher (which shall also remain nameless) left campuswide access to its Big Deal in place even after cancellation, apparently in order to preserve its brand presence on campus. This was an intriguing piece of information, but I’ve left this instance out of the discussion below for obvious reasons.
I should also emphasize that what I’m presenting here is “discussion” rather than “analysis.” As I had feared, the data I ended up with — I should probably be putting the word “data” in scare quotes — was too various and spotty to lend itself to rigorous and quantitative analysis or concrete conclusions about the state of the library market overall. However, some broad but significant conclusions can, I think, fairly be drawn. Here’s what I found:
This (Table 1, below) was by far the largest group, and I was a little bit surprised by how large it was. Of the 31 libraries (or library systems) I checked out, 24 of them have cancelled Big Deal packages that, as of this writing and as far as I can tell, remain canceled. Between them, they have cancelled packages from 46 publishers. These libraries serve a mix of research, comprehensive, professional, and liberal arts institutions. Several of them cancelled three or more Big Deals — and two of them cancelled six. Attentive readers will notice that one of these, San Francisco State University (SFSU), is listed as having canceled a Big Deal with an unnamed publisher. The information about their cancellation is derived from a presentation given at the 2016 Charleston Conference by David Hellman, the SFSU library’s collection development librarian. He declined to publicly identify the publisher involved for reasons laid out in his presentation, to which it is apparently not possible to link directly but which is easily findable by searching Google for the terms “cutting cord charleston hellman.”
One other thing is worth noting about these cancellations: 79% of them have taken place since 2015.
Table 1. Real Deal-Breakers
|West Virginia University||Wiley||2017|
|U of Oklahoma||Oxford UP||2015|
|U of Calgary||Oxford UP||2015|
|Taylor & Francis||2015|
|SUNY Buffalo||Taylor & Francis||2012|
|U of Montréal||Springer||2016|
|California State U||Wiley||2015|
|Lister Hill Library (U of Alabama)||Lippincott Williams & Wilkins||2009|
|Kansas State U||Springer||2017|
|U of North Texas||Wiley||2015|
|Institute of Physics||2015|
|American Institute of Physics||2015|
|Amer. Soc. of Civil Engineers||2015|
|UNC Chapel Hill||Cambridge UP||2017|
|Middle Tennessee State U||Elsevier||2016|
|East Tennessee State U||Wiley||2014|
|Medical U of South Carolina||Wiley||2010|
|St. John’s U||Wiley||2011|
|Taylor & Francis||2011|
|Illinois Wesleyan U||Wiley||2015|
|San Francisco State U||(Unnamed)||2016|
|George Mason U||Taylor & Francis||2017|
|Southern Illinois U||Elsevier||2010|
|American Chemical Society||2013|
|Memorial U of Newfoundland||Cambridge UP||2016|
|Taylor & Francis||2017|
These (Table 2, below) are the libraries that cancelled their Big Deals but later subscribed to them again. This is the second-largest group — but it’s much, much smaller. You might notice that the University of Montréal appears on both lists. That’s because they returned to their Big Deal with Wiley, but as of this writing have not resubscribed to the Springer deal.
Table 2. Returners
|Library||Publisher||Date Cancelled||Date Renewed|
|U of Montréal||Wiley||2014||2015|
|U California||Taylor & Francis||2013||2016|
|U New Mexico||Wiley||2006||2016|
These (Table 3, below) are the libraries that walked up to the precipice of Big Deal cancellation, but then backed away from it. In some cases it was because when they looked over the edge, it turned out to be a much longer drop onto a rockier surface than they had anticipated — which is another way of saying “the faculty would have stormed the library and taken library staff hostage.” In other cases it was because the library’s host institution gave the library extra money at the last minute in order to keep the Big Deal going.
Table 3. Second-Thought Havers
|U of Ottawa||Springer||2016||Announced cancellation, but subsequently reached renewal agreement with publishers.|
|Taylor & Francis||2016|
|Lister Hill Library (U of Alabama)||Wiley||2009||Announced cancellation, but subsequently reached renewal agreement with publishers|
|U of North Texas||Elsevier||2015||Announced cancellation, but university president provided emergency funds to allow renewal|
The relative largeness of the “Real Deal-Breakers” group suggests a number of things, but after lots of reading and conversations related to this investigation, I think the most important thing it suggests is that relatively few libraries that actually do cancel their Big Deals end up regretting it. Many libraries seem to have found that demand for the content included in their Big Deals — even what they thought was “core” content — was not nearly as robust as they had believed it to be based on usage data that had been provided by the publishers. How did they know? Two explanations for that lack of regret have come up most often in my discussions with these libraries:
First, despite what they were led to expect by strong usage data over time, in some cases there simply wasn’t the outcry from students and faculty that these libraries anticipated after they cancelled and the content went away. Why might the data have been misleading? There’s any number of possible explanations: the data may simply have been wrong, a significant number of downloads may have been triggered by something other than genuine need, the ease of access afforded by Big Deal packages may have led to lots of casual or even accidental downloading, etc.
Second, in some cases libraries replaced Big Deal access with the promise of effectively unlimited on-demand access to articles from the cancelled journals, and the resulting demand for that service was much lower than expected. Again, it’s important not to jump to conclusions as to what is causing this apparently low demand: it may be that demand was never terribly high to begin with, or it may be that the nuisance of requesting articles on an individual basis is great enough to dissuade library patrons from requesting them (leading them to go without access they genuinely want — or, perhaps even more likely, sending them into the arms of Sci-Hub or other pirate operations).
Despite all the ambiguities, one thing does seem clear to me: the Big Deal is no longer sacred. My prediction is that two converging factors will lead to a growing number of Big Deal cancellations in academic libraries: the continued downward trend in collection budgets, and the growing number of libraries that are demonstrating by example that you can cancel your Big Deal and live to tell the tale.
Update, 11 May 2017: The original version of this posting indicated that the University of Calgary had canceled a Big Deal package with Mary Ann Liebert. Inc., Publishers. I have since been informed by the publisher that they do not offer a Big Deal, and that the cancelation that took place at Calgary was a selective cancelation of titles from a list of individual subscriptions. Table 1 has been amended accordingly.