Washington, D.C.
October 18-20, 1995
Peter Lyman, University Librarian
University of California, Berkeley
On my first day as University Librarian at Berkeley, I got a call from Mike Keller, University Librarian at Stanford, who said, “Let’s merge.” “What?” “Well, we always talk about cooperation,” he said, “and when we talk about cooperation we tend to talk about things at the margin. Let’s merge, put everything on the table. Let’s assume we can work cooperatively in every area as a matter of principle, and discover whether there are actually problems we can’t solve together.” In retrospect, a year later, I think he was exactly right.
If we are going to solve our individual problems (if, for example, we must create a new system for the electronic publication of scholarly works, as was suggested earlier this morning), new modes of cooperation will be required, going far beyond anything we have done before. As a profession, we have a history of successful collective problem solving and inter-institutional cooperation, but budget problems and the promise of network technologies as a communications medium are forcing us to explore new kinds of partnerships that imply genuine interdependencies. To begin to explore the design of successful partnerships I will briefly explore four questions which might serve as a framework for our discussions: First, how do we measure success? Second, how are partnerships established, and grow? Third, what will the ARL/AAU Latin American demonstration project teach us about library cooperation? And, lastly, how might the proposed changes in the copyright law affect cooperation among libraries?
When we talk about shared resources, there are, of course, many levels of problems, but I think the fundamental problem may concern our values. The value problem that perplexes me the most is understanding how we measure the quality of libraries. Are cooperation and shared resources, or, more precisely, cooperation when it means shared resources, ultimately consistent with the way we measure the quality of research libraries?
The reason I raise that question is that, when we look at the quality of research libraries as measured by the ARL Statistics, and particularly as they are interpreted by our faculty and administration when they read The Chronicle of Higher Education, the emphasis is on input measures of quality. How much do we spend? How much do we buy? If that is our measure of quality, then sharing resources diminishes our quality when it is measured by input. While we are cooperative, we are also competitive, and our institutions compete for faculty, grants and students. Universities are ranked by, among other things, the quality of the library (the size of the collections, number of staff and so forth). So, if the winners are the ones that own the most, what is the incentive for sharing resources?
If we are really going to talk seriously about mergers, about shared resources and interdependency, we have to have a different way of measuring the quality of libraries. I suspect the new criteria would have to be output or performance measures. Performance measures would have to include quantitative answers to questions like: What information do the faculty and students need, and how well does the library satisfy those information needs? Most fundamentally, performance measures would evaluate the library by measuring its contribution to the productivity and the creativity of the faculty and students. It would not necessarily be the size of the collection alone, but the difference that information resources makes to our clients. Thus, our first problem is to make sure our organizational strategy is consistent with our values. This is not to criticize the ARL Statistics, which have the function of measuring our value to each other as potential collaborators, but they are not sufficient to measure quality. I would suggest that the first step towards sharing library resources is to define quantitative measures of library quality based upon performance.
Organizational theory and the business literature, especially the work of Rosabeth Kanter, has much to teach us about institutional cooperation. Her first book was on utopian communities (and maybe that’s relevant to this discussion, too), but she later wrote a wonderful book called Men and Women of the Corporation, one of the first books on sex equity and equal opportunity, and recently she has written a series of articles on how partnerships grow. In “Collaborative Advantage: The Art of Alliances” (The Harvard Business Review, July/August 1994, v.72 n.4, pp. 94-108) Kanter defines three variables that I think are quite useful in our thinking about partnership and cooperation: the role of personal relationships, seeking mutual benefits, and accepting differences.
Kanter’s fundamental argument is that organizational agreements start with personal relationships, and require strong personal commitments from the organizations’ leaders. Both the reorganization of work and conflict resolution almost always require personal commitment by top leaders, and even negotiations must be based upon a strong personal ties.
Seemingly, this emphasis upon personal relationships is in tension with network technology, which we are contemplating as the means of cooperation because it can span vast geographical distances. When we think about cooperation we tend to think about something that scales at least on a national basis, and it looks as if computer networks may also be a medium for global cooperation. And yet, if you look at the literature on network communication, for example, the use of e-mail for decision-making in corporations (for example, the work of Lee Sproull and Sara Kiesler at Carnegie Mellon, Computing and Change on Campus, Cambridge University Press, 1987), it turns out e-mail is a very poor medium for group decisions and conflict resolution. This may be something every manager must learn by experience: e-mail is only for messages, not for deliberation. When you get an angry message, you have to say, “Please come see me.” The exception to this rule is found among corporations or groups that share a very strong culture, or in which strong personal face-to-face relationships precede e-mail decision making.
So when we think about information technology as a medium for cooperation and shared resources in the national information network, it is important not to underestimate the importance of personal relationships and commitments. The fundamental structure of the network is social, not technological, and we are only beginning to think about designing the social organization of the network as carefully as the technology.
Kanter’s second key point is mutual benefit. Yet, benefits cannot be measured until costs are understood, as was illustrated in this morning’s discussion of the Journal Storage (JSTOR) project. In this respect, one of the weaknesses of libraries may be that we have used the word “free” to describe our services, when we meant “subsidized.” Economists have suggested that the inflation rate for library materials will not moderate until our users understand the economic implications of their information choices. By extension, we must fully understand costs before any partnership commitments can be made. The second thing we need to do in designing partnerships, then, is to define models and collect data about the costs of library services and collections.
Kanter’s third point is that partners must understand that they are different, and accept those differences. I am very aware of the remarkable differences between the institutions that we lump together under the name “research libraries,” having just moved from a private university to a public university. This is only an example, not a model, since my sample is skewed: I moved from USC, one of the most entrepreneurial and management oriented of all universities, to the University of California, which is governed by a civil service style bureaucracy. In retrospect, I remember negotiating a contract to establish document delivery and high speed delivery of monographs to USC from UCLA’s great collections. At the end of this negotiation I said to Gloria Werner, UCLA’s University Librarian, “You know, you really understand the responsibility of librarians to share information.” She replied, “Well, actually, Peter, it was about parking. We need back the parking spaces that USC faculty and students use when they visit the UCLA Library, [in order] to meet the Air Quality Management District quotas.” (If I’d only known that earlier!) The differences between institutions, in mission and motivation, are potentially very serious obstacles to partnerships if they are not recognized. I have noticed three kinds of differences between public and private universities that might serve to make the more general point.
First, differences in funding sources might imply differences in management style. Private universities are often management oriented environments in which planning is important, but not as important as having great ideas; in entrepreneurial environments ideas attract resources. But public Universities are more likely to be governed by civil service rules which emphasize process and consultation, are risk averse rather than entrepreneurial, and are organized around rules and procedures rather than problem-solving.
Secondly, while private universities have relatively sharp boundaries, public institutions often have diffuse boundaries, are woven together with other parts of the public sector. As a librarian at UC Berkeley, I participate in a series of partnerships with the other campuses of the University of California. I must think about how my actions as an independent agent working with the Latin American project interacts with my responsibility to my colleagues within the University of California, and also how any agreement will affect the California State Universities and the public libraries in California for which Berkeley is “the collection of last resort” in the state master plan. In an extreme case, these differences might even constitute fundamental differences in mission, which would make cooperation between public and private research libraries very complex indeed.
The third difference, one that really is potentially very important when identifying costs and benefits, is the difference in scale and the degree of specialization between institutions. When using performance measures, it’s not the ARL rankings that count, it’s the ratio of resources to users. Basically, although there are a number of public research libraries in the top 20 list of ARL statistics, if you look at the ratio of books or staff to Ph.D. programs (or number of faculty and students), the publics tend to fall behind the privates, because the scale of use of collections and services is often much greater.
These examples are only intended to illustrate the general point that every research library is different, each reflecting its own institutional history and mission. In building partnerships we will have to overcome real cultural and organizational differences, and the first step will be to recognize them. This is also why leadership and commitment are essential resources.
The Association of American Universities Task Force’s report on Acquisition and Distribution of Foreign Language and Area Studies Materials (1994) suggested that the problem of the relatively high cost and low use of area studies collections is one of the most important library problems, and one which might be solved by inter-institutional cooperation. It is, of course, important to recognize that library costs differ radically across different fields. To take an example, when we began to look at cost/use ratios, we discovered that some of our most expensive journals, in engineering and computer science, are reasonably cost effective when measured on a per use basis. Latin American collections are a very high priority for us in California, but we need a more cost-effective model of how to build these collections, given that they are essentially archival collections, from the point of view of cost per use.
The Latin American project, funded by the Mellon Foundation, is the second phase of the ARL Latin American demonstration study to learn about institutional cooperation, to explore ways to connect Latin American collections at Berkeley, Florida, Stanford, Texas, and Yale (see http://www.arl.org/collect/grp/index.html ). The first phase, as you know, has concentrated on building a data base and a strategy for distributed information resources. That’s just now coming into operation at a level that can be evaluated, and is an extension of known methodologies for institutional cooperation through building shared information systems. In the next phase, we are beginning to look at a second question: Assuming that the idea of a distributed, shared resource with a common data base is robust (and I do think that will be the case), how will it make a difference to the operations of our differing institutions, that is, to the quality of our performance? In the business literature it is axiomatic that technology only lowers costs when work is reorganized to optimize efficiency, otherwise it is more likely to add quality than increase productivity. Thus each institution will focus on how its own internal organizations must change in order to take advantage of a shared resource, and, given the discussion on differences above, it is unlikely that there will be one single answer.
At Berkeley we would like to focus upon lowering the opportunity costs for the faculty and the students in gaining access to remote information resources, because one of the reasons people prefer to browse in stacks is that the opportunity costs are very low (and the infrastructure costs, being defined as free, are invisible). Our question is: What service level would we have to reach to make faculty and students indifferent to whether we own the collection or whether we are providing access to remote information resources? If we want to lower user opportunity costs, we must define our services through our customer’s perceptions of their needs, and develop the technologies to accomplish our service goals. In practice, this requires that we rethink interlibrary services and document delivery costs. A more fundamental alternative would be to develop an electronic publishing strategy, changing the structure of scholarly communication by linking Latin American sources to research libraries by network.
Finally, copyright policy may become a significant constraint on inter-library cooperation. The Commerce Department’s Working Group on Intellectual Property Rights has now produced its White Paper, entitled Intellectual Property and the National Information Infrastructure (September 1995). I’m extremely concerned about the direction of copyright policy for several reasons.
As both a political scientist and as a librarian, I’m very concerned that a public policy document on such an important question would be written in a technical language that is a significant barrier to public understanding and discussion. The White Paper is deliberately written as a legal brief, the rationale being that it is only a technical update on established legal understandings, an extension to accommodate new information technologies. In fact, the White Paper implies fundamental changes in copyright law, and these changes are partially concealed by the rhetorical structure of legal discourse (all of the citations in the paper are to statutes or litigation).
I’m particularly concerned about the economic impact of the newly defined “transmission right,” that the transmission of copyrighted information by network will be defined as making an illegal copy. Transmission rights potentially add substantial costs to the some of the most exciting technologies for access in higher education, particularly distance education and shared digital collections. If a copyright charge is added every time something is transmitted, the costs of using technology for shared information resources are going to be driven up dramatically.
The argument about transmission rights is consistent with the assignment of intellectual property issues to the Commerce Department; tacitly, the White Paper assumes that the market is the sole mechanism for achieving the public interest in access to information. In itself, the idea of transmission rights might be an effective way to help create a stable marketplace for electronic publication. Ominously, however, the White Paper is virtually silent on the issue of Fair Use in digital environments, deferring consideration of Fair Use until later. Transmission rights combined with silence about Fair Use, and a recommendation that licensing be the primary mode of access to information, is imbalanced public policy, designed to solve the problems of publishers without the concern for public education which is traditional in copyright policy. This imbalance is best illustrated by the White Paper’s recommendation that children be taught that information is private property, but not a word is to be said to children about information as an educational resource. It is curious that NII policy emphasizes access in a technological sense, recommending subsidized connectivity to public institutions like schools and libraries, at the same time that the White Paper is creating new economic barriers to access for educational purposes.
The White Paper is not law yet, but I think that the vital interests of public education are at stake in this discussion, and particularly in the silence about the status of Fair Use in digital environments; without Fair Use, and with copyright extended to include transmission rights, institutional cooperation will become more expensive. (In fairness, the White Paper does not entirely see the marketplace as a substitute for libraries, for it reserves special rights to libraries in the area of preservation and digital copies.)
In a sense it is correct that the White Paper is only incremental policy making, for very little is known about the future development and use of the network at this point. It is probable that in time these new policies will be made moot by technology and social forces; in each era of technological innovation traditional economic interests have tried to protect their power through political influence, but history tends to treat them like the tides treated King Canute. In this sense, perhaps the greatest danger is cultural, not legal, namely that in this brave new world libraries will begin to establish market relationships between themselves and their users, becoming distributors for the publishers, and, as the Commerce Department recommends, begin to negotiate commercial licenses which substitute for Fair Use and cooperative inter-institutional agreements with other libraries.