university mission as well. Students expect to have ready access to the full panoply of digital content, data included. Faculty need an increasingly sophisticated work environment and a fair amount of IT support. Faculty also need bandwidth, systems support, and library resources delivered to the desktop; all of which require institutional investment—as do new labs, and redesigned, updated existing labs, and new buildings in which to put big science.
Research universities also need academic disciplines to consider how new research methods should be calibrated in the context of promotion and tenure decisions. There was an interesting conversation at MIT not too long ago about how faculty members can subject their work to the scrutiny of peer review when the scholar is working in an environment that is entirely electronic and on the Web.
Finally, IT in the legal, regulatory, and compliance context has caused a distinct rise in the cost of doing business for research universities. For example, libraries now license databases. The MIT libraries are probably not unusual in dedicating two full-time people to license agreement negotiation. Likewise, MIT's information systems department has a staff of lawyers who negotiate software licenses for the university. And there is now a staff member whose job it is to respond to calls relating to the provisions of the Digital Millennium Copyright Act (DMCA). Bear in mind that people can call up and demand that you take content down, and the bias is in their favor. Research activities also incur new expenses. Vice presidents of research work hard at protecting the scientists and students in their institutions.
These new costs of doing business are all total overhead. They are defensive, and they are dead-weight costs to the university. There has been some question earlier in this symposium about whether the DMCA costs money. The answer is yes, the new legal environment costs a great deal of money in dead-weight overhead.
As research universities have confronted the reality of rising dead-weight costs and diminishing flexibility, in an increasing Draconian intellectual property (IP) environment, they have become increasingly aware of the importance of advocating for and supporting openness. The best students will be diverse in their nationalities and religions. These students need access, as we have heard from Harlan Onsrud, 1 to high-quality information, and they should not have to choose between buying lunch, buying a dataset, or buying an article. Faculty should be able to teach the best way they know how, without the requirement to plow through endless permissions and approval processes to obtain the ability to use in their courses information that is now (by default) protected by IP regimes.
You have already heard a great deal about the importance of openness in research and about the need for work to have visibility and impact. I want to second Harlan Onsrud's comments on subscriptions. 2 Because as a practical matter, if the size of a subscription base is reduced to 50 institutions, and the license terms of digital access to that database prohibit interlibrary loan, then one really has to ask the question as to whether publishing in that particular journal does indeed provide appropriate visibility and impact.
One might reasonably conclude from these remarks that research universities, and the home that they provide for many scientists and engineers, are in deep trouble. Imagine for a moment that universities are not focused on teaching students, are not conducting not-for-profit research, but rather are engaged in some other enterprise. Imagine a business trying to operate under the various constraints and uncertainties that apply to research universities. Imagine that MIT was not MIT, but rather a major metropolitan newspaper publisher. The business would be in a situation where its staff authors were writing material on the premises and then sending that material to an outside third party. This newspaper would then have to buy back the work of their staff authors at an arbitrary price so that it could be used in the publication of the newspaper. Imagine, worse yet, that the third parties were near monopolies. Imagine, moreover, that the newswire services it dealt with chose to license content to it under arbitrary and unilateral terms, so that it had no control over the data stream. Then imagine that compliance with all the IP requirements that surround the content that goes into the newspaper had become unrecoverable and unsustainable. This is the situation that universities find themselves in today.
We believe, at least at MIT, that new educational and information data management strategies are required. There are two initiatives under way at MIT, which reflect a market response based on a commitment to openness.
1See Chapter 25 of these Proceedings, “Emerging Models for Maintaining Scientific Data in the Public Domain,” by Harlan Onsrud.