The New Supply Chain and Its Implications for Books in Libraries

Over the past decade or more, academic libraries have become increasingly sophisticated in managing their workflow, ingesting huge amounts of content with relatively low administrative costs as they struggled with the growth in the quantity of scholarly publications and the tightening vise of budget contractions. A relatively new innovation in this collection building is patron-driven acquisitions, familiarly known as PDA, which has now moved from the experimental stage into a more central part of the practices of many libraries.

Also called DDA for demand-driven acquisitions and even PIA for patron-initiated acquisitions, PDA provides access to electronic and (sometimes) print books through a library catalog but delays purchase of a book until a patron actually uses it. At least in the case of PDA e-books, the patron has no way of knowing whether the title in the catalog has already been purchased by the library. PDA can be seen as the logical extension of a shift in research library collection development, ongoing since at least the mid-1990s, from "just-in-case" to "just-in-time" acquisition models. Although the scope and intensity of all collection building is determined by the collections budget, academic research libraries traditionally aimed to collect as "comprehensively" as funds would allow, with the goal of anticipating the short- and long-term needs of faculty and students. But pressures on collections budgets have increasingly pushed just-in-case approaches out of reach for even the largest research libraries.

For the most part, a library purchases books from a vendor, via approval plans and firm orders, immediately after the books are first published. These books are then entered into the catalog and put on the shelves—or, in the case of e-books, are catalogued and made available through the vendor's software platform. It is entirely possible that these titles may be part of the library's collection for many years before a patron actually checks out one of the books.

By some estimates, as many as 40 percent of the books in academic libraries are unlikely to circulate at all.¹ In fact, the number of non-circulating books varies widely with the type of library. Public libraries often see their books circulate so often that the books fall apart. But academic libraries, especially the largest of them, collect highly specialized material, books that may sell under 1,000 copies total, and they may collect books published in hundreds of languages. It is not surprising that such titles may not circulate often. A rigorous 2010 study of print monograph circulation at the Cornell University Library found that only around 45 percent of all its print books published since 1990 had circulated at least once during the twenty-year period. Circulation was significantly higher for English-language books, at 61 percent in Cornell's main library for the humanities and social sciences. Bracketing the most recent acquisitions to focus on English-language books published between 1990 and 1999, circulation by 2010 rises to nearly 73 percent for the same library unit.² Although the 40 percent figure is a useful shorthand to indicate those parts of a collection that are not actively accessed, at least over a relatively short period of time, it does not reflect the diversity among libraries and among individual books.

A library that begins to work with a PDA vendor typically has three goals: to reduce costs; to align the library collection more closely with current users' immediate needs; and to present a much broader array of titles for patrons. This last point is particularly intriguing. In a traditional model, a library may be presented with, say, 100 titles and may purchase 25. In a PDA model, the library purchases none of the titles initially but places records for the entire set, all 100, into the catalog, where patrons can browse for them and (at least in the case of e-books) gain immediate access. The financial question for publishers and librarians alike, who sit on opposite sides of this relationship, is whether 25 titles is more than or less than what libraries will purchase if patrons have all 100 titles to choose from.

PDA poses questions about the availability of books over the long term. The traditional focus of libraries on just-in-case purchasing was based on the realities of the book marketplace when books were exclusively available in print. A publisher would print a certain number of copies and sell them until the warehouse was empty. Unless the sales came about fairly rapidly, after that one printing the books were likely to go out of print. For a librarian, a book not acquired soon after publication represented a gap that was difficult to fill retroactively, when patron demand might assert itself. Thus libraries typically purchased books at the time of publication in order to be sure that they would have the book when a patron needed it. Sometimes librarians made the right choice, and sometimes they didn't.

Theoretically, e-books will always remain available for purchase. But how effective will PDA prove to be over the long term? After two decades of licensing access to remote electronic resources, libraries and publishers have clarified many questions surrounding ownership and access for licensed resources, including e-books. Good progress is also being made in the preservation and archiving of e-books, with trusted third parties that emerged in the e-journal context (e.g., Portico and LOCKSS) now archiving e-books from numerous publishers. It seems likely that libraries will be able to count on the persistence of the digital files made available via PDA and on the licenses ensuring perpetual access to materials purchased via PDA. Less clear is whether "latent" PDA records in a library catalog will initiate an automatic purchase years into the future. The Cornell print monograph usage study suggests that academic patrons continue to discover books on library shelves after more than a decade. If PDA is to function over the long term as a viable substitute for prospective collection development, then libraries, PDA vendors, and publishers need to agree on mechanisms and best practices to ensure that PDA records lead future patrons to content and not to broken links.

PDA is a disruptive practice. If it were implemented widely, especially if it were to become the primary way that academic libraries purchase books, it would lead to the restructuring of the academic book-publishing industry. At least over the short term, libraries would save money and publishers would lose money. How much would be lost by publishers depends on many things, but the principal question is how important libraries are to a publisher's sales mix. For example, a major trade publisher such as Random House would barely notice if sales to academic libraries dropped to zero (public libraries are another matter). But some academic publishers such as university presses, with a sizable proportion of their sales going to libraries, could be severely impacted by a widespread shift to PDA. In some instances, PDA would lead to lower sales forecasts overall, which in turn would mean that certain books would not get published in the first place.

Whatever pressures PDA may put on publishers, it's difficult to argue with the logic of a librarian who does not want to collect books that will not circulate. Subsidizing the publication of a book with weak sales prospects may be good for the academic community overall, but it is not necessarily good for the library, which would in effect be providing that subsidy. The challenge is how to embrace the new—the innovations of PDA and all its efficiencies—without giving up necessary and critical aspects of the old. In July 2012 the University of Missouri announced that its press would be phased out of current operations. The university felt that it could no longer provide the $400,000 annual subsidy to keep the press afloat. PDA did not lead to the downfall of that press, but the question remains:  Who should underwrite the creation and dissemination of scholarship that can never hope to reach a mass (hence, profitable) audience? The University of Missouri Press published 2,000 books since its founding; without some form of subsidy from somewhere—from its university parent, from donors, and/or from those libraries that historically funded the non-circulating 40 percent of books—those cultural artifacts would never have come into existence.

There are bigger issues here than the fortunes of any single press or the policies of any particular library. The critical question is: Have we decided that no underwriting is needed because in the future we will not want books such as the 2,000 published by the University of Missouri Press? Is the future of scholarly communications something that will just happen—something outside the scope of our attention and the range of our abilities to foresee? Ultimately, PDA raises questions about how the academy will support and sustain the long-form scholarship that serves as the bedrock of many academic fields.

The research for a study of patron-driven acquisitions and their impact on university presses was made possible through the generous support of the Andrew W. Mellon Foundation.

1. The figure comes from a 1979 University of Pittsburgh study, which found that approximately 60 percent of books acquired in 1969 had circulated by the end of 1975. See Allen Kent et al., Use of Library Materials: The University of Pittsburgh Study (New York: M. Dekker, 1979).
2. Kizer Walker, Richard Entlich, et al., "Report of the Collection Development Executive Committee Task Force on Print Collection Usage, Cornell University Library," November 22, 2010.