Building a Sustainable Research Collection
Challenges and Strategies
A “research collection,” is not just what lives, physically, on the library’s shelves, but the whole array of resources, print and online, to which we provide access: resources that faculty and students depend on in teaching, learning, and producing scholarship. What do we mean by “a sustainable collection”? Books and journals, to the researcher, might seem pretty inert -- not things that require a lot of care (unlike lab equipment, computers, etc.). Library resources represent, generally speaking, the final, published outcome of the research process, which lends them an image of stability. But it’s as inputs to the research process that we are here concerned with their sustainability.
Research evolves, and it’s important that a research collection keep pace, constantly growing in order to represent the latest research while also making available that portion of the usable past that remains relevant. Because of this requirement, the majority of library resources -- considered as a portion of the Libraries’ total spending on collections -- are what we call continuations. Continuations are simply resources for which we incur an ongoing cost for access, journals being the most obvious example. Likewise, most of our databases -- whether they index the published literature or provide other kinds of information -- require a subscription. And even where we own the content, if this content is online, it typically resides on the publisher’s servers, for which we pay an ongoing access or maintenance fee. In this context, “sustainable” means “a collection that keeps pace with researchers’ needs within the constraints of a limited budget.”
What (some) e-resources really cost
Focusing in on one resource provides some perspective on the cost of access to the most current research. Web of Science -- which is a cross-disciplinary index of scholarly publications that is most heavily used in the sciences -- has an annual subscription cost equal to the 2012 median price of a single-family home in the United States. Web of Science isn’t an aggregator of content, like JSTOR. It’s primarily an index: it aggregates citations, but not the full text of the publications themselves. It does some other important things, too, like providing cited reference searching and article-level metrics. Certainly, none of us want to have to cancel Web of Science. But it’s a big elephant in a room of elephants, when it comes to the hidden cost of scholarly research.
While Web of Science is a dramatic example, it’s just an example. The figure below gets closer to the heart of the problem. In a nutshell: library budgets remain stationary or grow slowly, while inflation on scholarly resources increases at an alarming rate. Whether this cost increase is due to increasing profits from some publishers, increasing volumes of research output needing to be published, or increasing costs of digital publishing, the result for the researcher is the same. A flat budget for the purchase of scholarly output results in the purchase of a declining percentage of scholarly output.
Scholarly articles are not subject to the same rules of competition as other consumer products so publishers do have, in effect, a monopoly on the articles they supply. As with the "competition" among the telecom giants, it's hard to understand the problem by analogy with the market for other consumer goods. One might suppose that a library should be like a well-stocked larder - whatever you need, it's there - or like a supermarket. But there's an important difference. The dizzying array of high-priced chocolates that Whole Foods makes available is designed to capture a wide spectrum of consumer preferences. The diversity of resources in an academic library, however, is different. The latter is a productive diversity. It represents the diversity of perspectives from the conjuncture of which new ideas emerge. It represents the possibility of error-correction and reproducibility that remains the bedrock of the scientific method. It represents the accommodation of emergent objects and fields of study, which don't supplant but coexist with and complicate the old.
All of this makes the choice between Journal A and Journal B unlike a choice between Coke and Pepsi. In choosing between two journals in a given field, we have criteria we can apply, like usage over time, impact factor, and most important, faculty input. But in most cases, we should subscribe to both journals and publishers know that. They set prices appropriate to an inelastic demand. It is important to note that even in lean times, we remain committed to the principle of intellectual diversity. Although we do have to cancel journals when their cost-per-use exceeds a certain threshold, faculty and students can still access articles from those journals via Interlibrary Loan at no charge. And that is because GW Libraries absorb the Interlibrary Loan costs, which amount to more than $30 per article.
Journal Package “Deals”
Another important feature of the market for academic resources is the consolidation of journals into packages that publishers offer libraries at a per-title discount. Journal packages provide access to a high volume of journals across the disciplines. Our cost-per-use analysis tells us that in most cases, these packages are cost effective (compared with the cost of Interlibrary Loan).
But packages are not a free lunch. They represent a constraint on how much money we can save by cancelling individual journal titles, because the publishers’ contracts typically prohibit dropping titles from the package (and they provide only limited scope for “swapping out” subscribed titles for others that we might prefer). As a result, packages constrain our ability to acquire new resources. More troubling still is the following fact: our dependence on the package-model for access to a substantial amount of the published content in many disciplines leaves us even more vulnerable to predatory inflation by the publishers. This year, for instance, Nature Publishing raised our subscription price for its Physical Science collection – an important and popular journal package – by 85 percent! In such cases, we have only two choices: 1) Break up the package and subscribe to (some of) the journals separately. 2) Keep the package and accept the increase. In this case, our analysis showed that Option 1 would entail a 35 percent increase over last year's cost, even if we subscribed to only the most heavily used journals in the package. Reluctantly, we chose Option 2. What this story illustrates is how libraries -- and by extension, faculty and students -- are caught in a zero-sum game.
Books and Readers
Subscriptions represent the bulk of our commitments for collections every year, but books are important, too. The scholarly monograph remains the de facto standard of scholarship for disciplines in the humanities and edited volumes are an important venue for scholars across the humanities and social sciences.
From a collection development perspective, books pose a special set of problems. Though GW is a research university, we don’t have the kind of library budget that allows us to purchase comprehensively across the disciplines. It’s important that we focus our selection on those books that will likely prove important to our faculty and students. While our usage is better than some average estimates, the figure below shows that even among books that we purchased six years ago, a significant number remain unused. The data suggest that the percentage of books never used tends to level out over time, so that there will always be a remainder that, in a manner of speaking, goes to waste.
In relation to journals, we have seen how the publishing market creates conflict between the Libraries’ commitment to the principle of intellectual diversity and the constraints of our budget. There’s also a certain cultural ideal that construes the library as an institution dedicated to preserving culture. From that perspective, perhaps unused books shouldn’t trouble us, but it’s important to consider the previous figure in juxtaposition with this one. For the same period and set of titles, the figure to the left presents a measure of the uniqueness of our holdings. The 18% are titles with zero overlap -- meaning no other copies are available in the Washington Research Library Consortium (WRLC). The majority of books that we acquire are owned by at least one other library in the Consortium, and many are owned by three or more other libraries.
Building a Sustainable Collection for the 21st Century
Our membership in the Consortium is one of our most important resources. It represents one way that we can realize the principle of intellectual diversity in spite of the constraint of individual institutional budgets. Our resource-sharing staff work hard to make sharing books and articles among Consortium members (and even beyond) as seamless and efficient as possible. It’s routine to get a journal article or book chapter -- in PDF format -- the same day that one submits the request. At the level of Consortial governance, the member institutions are working more closely together than ever before to coordinate and collaborate on things like subscription licensing, preservation, and collection development. Working as a Consortium gives us leverage with publishers, for instance, that we lack as separate institutions.
In terms of the book collection, there’s another development that’s working in tandem with the Consortial arrangement to change the way we build collections. That’s called Demand Driven Acquisition, or DDA. The premise is simple: we should focus our buying on what our faculty and students want. With print books, this always posed a challenge. But with the growing prevalence of ebooks, it’s a different story.
This electronic edition of a recently published book is waiting in the catalog for a reader, but unlike the books on our shelves, we won’t actually purchase this book until someone uses it. A certain threshold of use is required to trigger a purchase-- opening the book and skimming the table of contents, for instance, is free, but if a GW faculty member, student, or staff member spends time reading this book online – or downloads it to a tablet or personal computer -- the book will be invoiced and paid for out of a library fund set aside for that purpose. (At that point, we “own” the book, and all subsequent uses are free.) In the short term, at least, DDA has proven to be quite cost-effective (given the percentage of books in the collection that never get used). It also has the advantage of making a lot of content available for potential use, while we pay only for what actually gets used. How well DDA performs over time is an open question; we are closely monitoring cost and use, and these models are still in flux, as publishers, ebook vendors, and libraries play tug of war over what’s in the best interest of our constituents. All that complication is, thankfully, hidden behind the scenes. To our readers, it’s just a library book.
We are also sensitive to the fact that an ebook doesn’t work for everyone or for every purpose so we’ve included an option which lets you, the user, request that the library buy a print copy of this book, if that’s what you really prefer to use (and if there’s not a print copy already available in the Consortium for request). While we have our own DDA program, the Consortium has also entered a DDA agreement with JStor, which now offers recent titles from a variety of high-quality academic publishers. This program has been very successful -- with a high volume of usage since we launched the program last year -- and it has been, from a cost-per-use perspective, very effective. We will probably see many more of these Consortial agreements and programs in the years to come.
The GW Libraries mission remains unchanged: to provide you untrammeled access to the materials that you need, whenever possible. In order to fulfill that mission, we have to juggle a lot of challenges. We in the Libraries like to see ourselves as partners with you, GW’s faculty and students, in this process. We don’t only value your input on the Libraries’ collections, we need it. You can help us keep track of the emerging priorities in teaching and scholarship. You can help us see what parts of the process need improvement from a user’s perspective. And you can help us learn what’s most important to advocate for, in our ongoing negotiations with publishers and vendors over everything from pricing models to content to digital rights management. Staff throughout the Libraries are working hard to make our collecting practices more collaborative, evidence-driven, transparent, and sensitive to our users’ needs. We value the participation of faculty and students in this process. Please contact your collection development librarian with questions, ideas and comments. If you are unsure of your librarian, please email firstname.lastname@example.org and we will be happy to connect you.