When Bigger Is Not Better

© 2004 Abby Smith

EDUCAUSE Review, vol. 39, no. 1 (January/February 2004): 52–53.

Recent innovations in information technology have met and—in many cases—exceeded all but the most utopian visions for increasing access to research materials. First in the natural and social sciences, but increasingly also in the humanities, we have seen a rapid adoption of digital technology to create and disseminate scholarly information. Although the costs of the technology are high and the risks of losing access to digital information in the future because of hardware and software obsolescence are now widely recognized, the compelling advantages of digital technology make digital the default mode for the creation of text, images, and sound.

A little-remarked but potentially troubling "unintended consequence" of this rather rapid development, however, is that the large, often comprehensive book and journal collections that characterize noted research libraries may be in danger because they no longer provide a strategic advantage to their home institutions. This is true even though the collections represent an almost incalculable, and certainly irreplaceable, investment in information—the fruit of centuries of labor by authors, publishers, and librarians. Some far-sighted campus leaders—librarians foremost among them—are asking themselves how they can best manage the risk to that long-term investment. How can they best conserve these irreplaceable assets, how can they integrate these assets into an increasingly digital service environment, and how can the research community as a whole manage this heritage of intellectual and cultural achievement cost-effectively? One thing is clear: We cannot let this bedrock on which present and future scholarship is built erode through time and neglect. And so, two primary challenges confront us.

The first challenge is to find the right approaches and incentives to continue the investment in assets that are kept, by and large, "just in case" at a time when the emerging paradigm for allocating resources for information is "just in time." Imprints can be managed cost-effectively to optimize fitness for use over time through secondary storage and an array of low-cost preventive preservation measures. But the fact remains that the overwhelming majority of books and journals in libraries are rarely, if ever, used by contemporaries—hence the phrase "just in case." The assessment of their value to the research community as a whole must be pegged to a time scale that differs from so-called Internet time. These resources are best thought of as a public good, with their continued well-being to be supported by all in the research community. This makes any local investment in them look "iffy," of course, or at least not immediately compelling during tight budgetary times.

The best approach to these collections, as it must be to all public goods, is to share the costs broadly across the community and thereby affect economies of scale. This could be rather easily achieved through regional repositories that store low-use materials, the kinds that are already being sent off-site to make room for new acquisitions. In addition, an increasing number of heavily used journals are available online, and their hard-copy counterparts are in less demand. And whole runs of journals that are, for example, available through JSTOR are being sent to remote storage. Aggregating these little-used but high-value imprints into a few central repositories, and retaining only enough of them to ensure reliable preservation over time, could create the opportunity for optimized preservation of and access to books and journals. Everyone gains under this scenario.

Several library and information technology leaders are developing models for such collective long-term management of titles—models that can potentially free up scarce funds for access to electronic resources. With the support of sympathetic, often cash-strapped administrators and some key funding agencies, pilot projects in the better management of so-called legacy collections through shared repositories are under way at the University of California, the Five Colleges of Massachusetts, and the Center for Research Libraries, among others. These efforts are being joined by the third-party provider JSTOR, which is building print back-up repositories to ensure the fail-safe replication of digital files in case of catastrophic loss—an action that will in turn increase trust in its digital services.1

With such an elegant solution nearly at hand, what is preventing a full-scale effort to build these repositories and enable local decision-makers to focus on access rather than preservation? The answer, it seems, is competition trumping collaboration. This leads to the second primary challenge: to effect these new approaches within a culture of often intense competition among research institutions. Just as universities and colleges compete for the best students, the best faculty, the best facilities, and the largest slice of the pie of external funds for research and teaching, so they also compete for the best provision of information. The library—and especially the size of the library collections—has served as a convenient, if sometimes misleading, symbol of excellence in this domain.

The reason is simple. Until recently, libraries have had to purchase and retain books and journals to provide local access to information. As a corollary, the measure of a library's ability to provide access has been the size or volume count of its print holdings. Some in the library community propose new ways of assessing the ability of a library to fulfill its mission to provide information in support of research and learning. This would take into account the activity of licensing (as opposed to purchasing) crucial e-resources, the growing portion of the acquisitions budget that goes to digital rather than print resources (with the digital portion now about 20 percent of that budget and growing rapidly), and the reality that both students and faculty want service now, no matter where that service comes from or who owns it. The Association of Research Libraries (ARL), a group that has been measuring its library members' performance for decades, is developing service-based metrics to complement and, in time perhaps, to supersede those based on volume count. And as the Digital Library Federation, a group of large research libraries with sizable collections of digitized content, is planning to aggregate that content to make it interoperable and thus more accessible, the group is working through the hard issues of how institutions share content (good for the user) and how they maintain a strong institutional identity online (good for the university).

How have the University of California (UC) and the Five Colleges of Massachusetts managed to work through similar issues? First, their cooperation stems not from altruism and a sense of good citizenship exclusively but also from the threat that if they do not cooperate, they will not be able to serve their faculty and students at the same level of standards they are committed to maintaining. Both are responding to real financial pressures. UC and the Five Colleges can implement solutions that work across several campuses because they already have a culture of cooperation. Their cooperative storage and collection-management programs are built on a history of cooperation on a variety of library, administrative, and teaching initiatives over decades.

As to be expected, the hardest part of such collaboration is to cede control from the individual to the group. But as Daniel Greenstein, of UC, points out, what matters is not ownership of research materials but instead governance over how the materials are deployed on campuses. Libraries of the future, he and others predict, must focus on competition based not on the size of collections but on the service environments that individual campuses craft for their faculty and students—a much harder task, of course. In addition, collaboration requires extensive and time-consuming consultation with key stakeholders, from faculty to college/university presidents. But we cannot wait for consensus to emerge. By then it will be too late.

At a minimum, we can agree that creating a handful of repositories throughout the United States to serve as a home to low-use materials makes sense. In the end, the efficacy of these systems will rely on trust among those institutions that, in the analog world, have been rewarded for friendly competition.

On the Internet, no one knows how big your library is. People know only how well your library delivers content.


1. A study published by the Council on Library and Information Resources in June 2003—Bernard F. Reilly Jr. and Barbara DesRosiers, Developing Print Repositories: Models for Shared Preservation and Access, http://www.clir.org/pubs/abstract/pub117abst.html (accessed November 25, 2003)—describes in detail these and other efforts, as well as outlines what remains to be done to effect a full system of print repositories that would ensure the persistence of these information assets. See also http://www.ucop.edu/cmi/welcome.html (the University of California); Willis E. Bridegam, A Collaborative Approach to Collection Storage: The Five-College Library Depository (Washington, D.C.: CLIR, 2001), http://www.clir.org/pubs/abstract/pub97abst.html (the Five Colleges of Massachusetts); and http://www.crl.edu/content.asp?l1=4&l2=19&l3=35 (the Center for Research Libraries) (all URLs accessed November 25, 2003).

Abby Smith is Director of Programs, Council on Library and Information Resources.