Close this search box.
Close this search box.

The Future of the Library in the Research University

Paul N. Courant *

(Paul N. Courant is University Librarian and Dean of Libraries, Harold T. Shapiro Collegiate Professor of Public Policy, and Arthur F. Thurnau Professor of Economics and of Information at the University of Michigan).

I have been asked to write about the role and functions of research libraries in the 21st century, informed by three perspectives that combine quite naturally for me: those of a former provost, an economist, and a library director. The principle that unites these perspectives, at least in my mind, is the economic idea of a “public good.” Where economists may be fairly or unfairly thought of as bean counters who care only about the bottom line, the field of public economics, which I have been practicing for several decades, is largely concerned with the production of goods and services whose value is best realized through collective action, rather than in the marketplace. Universities, the practice of scholarship, and libraries all fall within this category.

The key feature of public goods is that their consumption is nonrival, meaning that the cost of adding users is zero. National defense, the brightness of the sun, the view of a distant mountain range, access to a catalog record available on WorldCat, and information generally are public goods in this sense. Markets are not effective at producing the optimal amount of such goods, because it is inefficient to exclude people if the cost of adding them is zero, but charging a price of zero will not cover the fixed costs of producing the good. Thus, production of public goods is generally left to public institutions, and the mechanisms for determining the best amount to produce are political as well as economic.

The fact that markets are not good at producing public goods, including those produced by the university library, does not imply that economic considerations of cost, technology, and demand are irrelevant. On the contrary, where markets cannot be relied upon to produce things of value, the job of determining how and how much to produce is that much more difficult; this is the subject matter of public economics. Our task requires knowledge of the library’s purposes, the technologies that can be deployed to accomplish those purposes, and the cost and effectiveness of deploying alternative technologies. Recent and foreseeable changes in information technology are especially interesting in this context because they make possible new and different ways of organizing, finding, and publishing (in the broad sense of making public) scholarship both old and new. Whoever are the actors (and most faculty, provosts, and librarians may well lead long and productive lives without developing more than a nodding acquaintance with any economist), the problems that the research library faces in responding to changes in information technology are very much within the purview of economics.

The provost’s job is to articulate the demand for the library’s collections and services. The provost must ensure that the library is delivering value for the institution in which the library sits, and must continually assert the primacy of scholarship and academic work, including teaching, in the library’s mission. Thus, the provost identifies the library’s objectives in the context of all of the university’s missions.1 The provost also helps define and implement the business model(s) that sustain(s) the library, especially the revenue side. The provost is similarly responsible for the efficient production of many other public goods within the university.

Although I cannot speak for provosts in general, my experience as the provost of a major research university persuades me that the quality (both academic and in the marketplace) of undergraduate education is vital to the continued success of the great research universities. This observation has important consequences for how we think about the future of research libraries that are embedded in universities. Thus, although the assigned topic of this essay is the future of research libraries, my subject will be both narrower and broader—namely, libraries in research universities. Such libraries, with their universities, cannot limit their purview to research, although I will argue that their signal contribution to undergraduate education is the teaching of scholarly methods.

The library director—the position that I currently hold but know least well of the three—helps the provost determine the library’s missions and the mechanisms that the library can employ to greatest effect in service of those missions. (Of course, the director also does a fair amount of the heavy lifting in seeing to it that the work is done.) The director brings to bear the expertise of the librarians as well as perspectives on both scholarship and education that derive from the rich interactions among the library, faculty, and students. Libraries and librarians know a great deal that is crucial to the effective functioning of academic institutions. When things are working well, this knowledge is shared at many levels, including the reference desk (or the reference IM site) and conversations between subject specialists and faculties. It is the director’s job to ensure that the expertise of librarians and other library staff is developed and deployed to maximum effect. As the technology and associated social and political structures that affect academic work and research libraries continue to change, a vital part of the library’s role is to keep up with the changes and to develop and deliver an array of tools, services, and ideas that help students and faculty do their work.

With public goods as well as private ones, much of what is at stake can be reduced to demand and supply. In the public goods case, the determination of demand is the more difficult problem. Taking into consideration technology, publishing, the rights environment, the state of current collections, and the cost of adding to those collections and maintaining them in any number of possible configurations, what do we want the library to do?

The Library as Intermediate Good

The library provides essential infrastructure—largely in the form of reliable and well-documented access to prior scholarship, data, the cultural record, and other research materials—that is necessary to the effective practice of scholarship. It is worth noting that this description of the mission of the library is robust with respect to history, organization, and technical change. Providing the infrastructure of scholarly work was the library’s mission before the invention of the printing press, and continues to be the library’s mission in a world where making public materials that are used in scholarship is accomplished in myriad media, many of them digital (making content easy to copy and to transmit, at least technically).

The library’s value is derived from the requirement that scholars, teachers, and students have easy, rapid, reliable, and documented access to the rich set of materials that constitute the scholarly and cultural record. If some other institution were to provide the same or essentially similar access to those materials, the university library would have no clear mission. What matters is that the academic work be done, and be done well. The library’s future depends on its effectiveness in delivering materials and expertise requisite for excellent academic work.

Libraries should also be efficient, in the sense that they should deliver the services they provide at as low a cost as possible. Of course, for many services the library provides, the lowest possible cost is not low. (The same applies to much in the research university.) Work that requires individual attention by highly skilled academics does not, and cannot, come cheap. Academics and academic librarians will sometimes argue that our calling is so exalted that it would be wrong even to think about cost.2 This is nonsense. Precisely because academic work is so valuable, we should attend to cost all the time. If we do not, we will waste resources that we could have been using to advance scholarly work. It is not helpful to assert that the library should be great no matter what the cost.

Harold Shapiro once wrote that a competent faculty and administration will always see important things to do whose costs greatly exceed the resources available.3 If research universities and their libraries were ever at a loss for things to spend more money on, we would either have (a) solved all the world’s problems or (b) used up all our ideas, and hence should give way to someone else. Neither circumstance looks likely. I am not saying that the university library should be a profit-oriented business, or that we should count beans and base our decisions solely on the sorts of things that accountants or librarians are good at counting. But I am saying that we should always be looking for ways to do what we do more cheaply as well as better, because if we can do some things more cheaply (holding quality constant) it enables us to do everything better, including improving quality.

Some Economics of University Libraries and Academic Publishing

Historically, the university library has met the definition of a public good with respect to its own campus. For the population associated with the university, and geographically nearby, the collections could be made almost freely accessible with little or no effect of adding to the population. Going back to the definition of public goods developed above, use of library materials on campus is very close to being nonrival. (Textbooks and assigned work in large courses are the exception that proves the rule, here, and with respect to those materials, the academic library acts very much like a public library.) In the larger economy of higher education, however, the quality of the library was, and still is, of significant consequence, because the better the library, the higher the quality of faculty and students that can be attracted to the university, and the higher the quality of research that can be undertaken there. Both reputation and economic resources depend on academic quality, and the library is a vital intermediate good in any university’s ability to produce academic quality.

When almost all resources were in print, the economist, the provost, and the librarian were in sync, although the librarian might have been a little less concerned with cost than the others. By improving the library, one could improve the competitive position of the particular university relative to other universities. Distinguished departments were built around distinguished collections. An excellent collection at some distance was not a substitute for an excellent collection at one’s home institution. Hence, the payoff to investing in excellent collections at home was clear to all. To be sure, Interlibrary Loan makes it possible for scholars who are not proximate to excellent collections to do their work, but there is a significant difference between having essentially instantaneous access and access that is removed in both time and space. Thus, marketlike competition among the great research institutions provided a mechanism for supporting a system of superb research libraries.

The world of print also used a complementary set of technologies that supported the publishing industry. Printing, paper, and binding are expensive. Making print copies is expensive. There are substantial scale economies with respect to the size of a print run. Before undertaking the investment required to publish several thousand copies of a monograph (the high end for scholarly work even in the good old days), the publisher would want assurances that there was a market for the work. So the publisher would take steps to have it carefully vetted by experts in the field—the very sorts of people who would buy it and ask their libraries to buy it—and edited by people who were good at making the product attractive and usable, the better to sell more copies.

The interests of universities and academic publishers meshed well. To be sure, the publishers would have been even happier without the doctrine of first sale, but basically, they produced for our libraries, using the talents of our faculty, and we were able to take their editorial behavior as a strong indicator of the academic value of the work. Articles in refereed journals or monographs of established presses were just the ones we wanted, and libraries bought them, took care of them, and made them (approximately) permanently available to our campuses. If we had lots of them, we could support scholarship of great breadth and depth, so we have the great universities attached to their great libraries, and tenure cases being decided based on the peer review of the people who created and used the works that filled those libraries. As a bonus, because most works were collected by several libraries, copies could be compared, albeit at some expense, for authenticity and consistency.

In the digital world, the technological underpinnings of this economy—expensive print and expensive distribution (hence an advantage to local access, replicated across institutions) disappear. With digital production of text and images, making copies of things is essentially free (in one form) and getting cheap (print-on-demand), even in the traditional book. It’s still very expensive to produce a beautiful book, but now it’s quite cheap to produce a pretty nice book, and essentially free to distribute a usable digital file, and for many purposes the cheap and free alternatives perform adequately or better than the costly options. So, much of the academic work of the present and the future exhibits the character of a public good—once produced (and, of course, the initial production is still not cheap, but those costs are borne almost entirely by academic institutions and granting agencies) the work can be distributed to all who wish to see it at essentially no cost. It is possible to exclude users, of course, and under current law and custom, exclusion is straightforward. But exclusion under current technology is plainly inefficient and difficult to sustain, especially when authors, at least in many fields, cheerfully post their stuff on the Web where anyone can find it, legally or not.

Publishing, academic libraries, and the particulars of peer review were all developed in a world where printing, copying, and distributing were expensive. We no longer live in that world. How then, should we configure the library—and publishing and peer review—to take full advantage of the change?

The Future of Libraries in Research Universities

Research universities continue to require easy and quick access to reliable and replicable scholarly resources. Essentially all of the scientific journal literature is now distributed electronically, and most of the rest of the journal literature will follow shortly. Digital distribution has been relatively unimportant for monographs, but they, too, are basically born digital, and in the fairly near future one expects that the technical problems of e-books will be solved, although I would not be surprised if inexpensive print-on-demand is an important part of the solution for quite some time. Meanwhile, many great print collections are being digitized. It is likely that almost all of the scholarly literature will be available in digital form, at least somewhere, within the next 10 years or so.

The implications for the cost of library storage are potentially huge. One can imagine that new work will take up relatively little space, and that a substantial part of the existing monographic literature can be moved off site and replaced for most purposes with electronic files. (This will require that reasonable arrangements can be made with holders of rights who can be found easily, with reasonable statutory arrangements made for the cases where rights holders are hard to find. There would be enormous gains relative to the status quo for both users and rights holders, so the economist in me believes that such arrangements should be relatively easy to make.) Thus the library’s ever-growing claim on space would attenuate, affording, among other things, the opportunity to provide both services and improved access to scholarly materials that are currently stored securely but inaccessibly.

Before most humanities faculty—and, hence, their libraries and provosts—would be willing to substitute electronic copies of print works for the originals, they would have to be assured permanent access to the originals. This could most effectively be accomplished through a small number of print repositories, with very good security, climate control, and the like, with costs and access shared across a network of libraries. I expect to see an interesting tension for provosts, librarians, and other academic leaders as they begin to move in this direction. On the one hand, the availability of shared and sharable depositories will tend to erode the competitive advantage that the most prestigious libraries derive from the size of their local collections. On the other hand, it is precisely those libraries that will stand to save the most from using collective repositories and that have the most special collections materials that could be used more effectively in liberated local space. Moreover, the most prestigious institutions will have the greatest demand for developing new tools and services that exploit information technology. If the library continues to be the source of local expertise and innovation in both services and collection development, investments in the library will continue to generate competitive advantage for their institutions, although some collection development, notably the collection of audio and video clips, Web pages, Flickr sites, and the like, would benefit greatly from coordinated strategies across institutions.

Bibliographic reliability is much more difficult to guarantee in a digital world than in the world of print, and will require a set of social institutions that can identify and assure the stability of copies of record. Portico and LOCKSS have developed mechanisms to deal with this for a subset of the journal literature. It remains to be seen if their efforts are sufficiently comprehensive and trusted. I believe that the leadership of the great universities will have to create a collective institution whose job it will be to assure the “bibliographic” integrity of digital scholarly materials. I can’t imagine anyone but librarians in charge of these institutions if they are to succeed. The level of interuniversity cooperation required for this to work is unprecedented, as is the level of cooperation required for the shared print repositories outlined in the previous paragraph. As is often the case, the technical requirements are easier to attain than the social and institutional arrangements that are necessary to take full advantage of technical change.

The library will succeed (because it will have plenty of valuable work to do) if it continues to be the locus of expertise and innovation regarding scholarly information, how to find it, and how to use it.

Near the beginning of this essay, I suggested that effective undergraduate education would be essential to the success of research universities, and that the teaching of scholarly method is the most important aspect of undergraduate education.4 Our students must learn how to make judgments about the quality of information that they use. I do not presume that things in libraries are “good” and that things on the open Web are not good. Rather, I argue that it is essential that students be able to check on facts and assertions using reliably replicable sources. Libraries provide the infrastructure for this kind of teaching just as they do for scholarship, as the methods involved are the same in both. Success here will require two things: (1) coordinated curricula in which librarians and faculty demand the engaged use of library materials and library expertise; and (2) the ability to search library collections with something like the same ease and efficacy with which one can search the open Web. I do not suggest that we compromise our standards, but that we spend significant resources in making our resources and methods easily available to the world. How better for research libraries to advance the public good?


* I am grateful to Matthew Nielsen for excellent research assistance and even better conversations and comments.

1 Eleanor Jo Rodger makes a brief and persuasive argument that in order to succeed, libraries must deliver value to their “host systems.” Adapting her nomenclature, in this essay the university is by far the most important part of the host system, and the provost is its agent. See Rodger, E. J. 2007. What’s a Library Worth? American Libraries 38(8): 58-60.

2 Consider, for example, the assertion that “cataloging is a public good which should be supported regardless of economic concerns.” (Fallgren, N. J. 2007. Brief meeting summary: May 9, 2007—Structures and standards for bibliographic data. Accessed December 3, 2007, from

3 Shapiro, Harold T. 1987. Tradition and Change: Perspectives on Education and Public Policy. Ann Arbor: University of Michigan Press, p. 139.

4 For an extended version of this argument, see Courant, P. N. (forthcoming), “Scholarship: The Wave of the Future in the Digital Age,” in Richard N. Katz, ed., The Tower and the Cloud: The Co-Evolution of Higher Education and the Web (in press).

Skip to content