Laren Tolbert, Georgia Institute of Technology: One thing that was not mentioned about the so-called "home-run" patents, which were largely biomedical, and more specifically pharmaceutical, is that these areas produce a single compound. This has a very clear patent situation with a clearly identified profit margin, whereas many other patents represent an improvement on a process. The same may be said of software, although that is still an emerging field.
Should we separate the way we treat patents in these different fields to avoid a one-size-fits-all approach to intellectual property issues by making every innovation fit the biomedical model?
David Mowery, University of California, Berkeley: Public universities, and public institutions in general, often find themselves in a more complicated situation because the political overseers will tend to insist on accountability of various forms, which pushes people toward a uniform policy. So there will always be attention there for public universities. But by and large, it may make a lot of sense to stop looking at the entire world as though it's biotechnical or biomedical.
Christopher Hill, George Mason University: On top of that, software can be either patented or copyrighted or both, and the presumptive rights to copyright ownership can be different from the presumptive rights of ownership of the patent. Furthermore, it appears now that undergraduate students who work on software as part of a homework assignment that is later integrated into a software package may have a copyright interest. We are just beginning to thrash this out. So I don't think that anyone pretends that one size is going to fit all.
William Wakeham, Imperial College: The point about students is quite interesting. Upon registration, we require a student to assign any of their intellectual property to the college. This is not necessarily legally possible, but we do it anyway.
But there is a very big difference with the biomedical area because the patent situation is usually very different than in, say, mechanical engineering, where a particular device is created. This usually stands alone. However, in the biomedical area, the central patent has to be "decorated" with other things
to build a broad enough portfolio to be useful. We have some examples of instant successes, but mostly it has to be built up slowly. There is a lot more work in developing something in the biomedical area, but also there is a much bigger profit if one gets all the way to a patent. With software, once the code is written and professionalized, it's all ready to go.
David Schetter, University of California, Irvine: Dr. Mowery, is it not the case that, prior to the Bayh-Dole Act in 1980, the universities were not allowed to own the results of federally funded research, and that a primary impetus of Bayh-Dole was to create an incentive for universities to commercialize federally funded research? Second, if that is the basis of Bayh-Dole, do you have any data to determine whether or not it has achieved its goal?
David Mowery: Universities were, in fact, able to obtain patents on and license the results of federally funded research before Bayh-Dole under the terms of what were known as institutional patent agreements, which were negotiated by individual universities with individual federal funders of research. In fact, one of the pressures building up in the system before Bayh-Dole, notably in the University of California, was from what was then Health, Education and Welfare, overseer of the National Institutes of Health (NIH), to prevent the universities from negotiating exclusive licenses for the results of NIH funded research. So universities were already in the licensing game selectively, although not nearly as many and under much more complicated and perhaps less stable rules than what came about in 1980. There can be little doubt that Bayh-Dole has brought more universities into technology transfer and licensing, especially when you look at the number of universities that have established offices and the amount of patenting that has been conducted. Do we know that patenting of university-developed technology resulted in either a more rapid or a successful transition to commercialization? This is a much more difficult question to answer. And it is a question that will almost certainly vary among universities, among technologies, and a host of other things. It' s very difficult to measure because it is an experiment without a well-developed control group.
Christopher Hill: Another impetus for Bayh-Dole was the Department of Energy (DOE) Act of 1977, which had an explicit provision that was exactly contrary to Bayh-Dole that prohibited ownership of patents by any contractor or grantee. And because DOE was then so important, it had undercut the system you are talking about in a serious way.
Cheryl Fragiadakis, Lawrence Berkeley National Laboratory: One big contrast between the U.S. system and its history and what we just heard about what' s happening at Imperial College in the United Kingdom is the rather xenophobic approach that the U.S. government seems to overlay on research collaborations and licensing versus the very forthright desirability of multinational interactions in Europe. I see this as a limit to what is happening in both national laboratories and to some campuses that actually go by the law in the United States. Is the contrast as stark to you? Do you have any forecasts of what should be done or might happen in the future on this?
Christopher Hill: I did not hear anything in Professor Wakeham' s description of Imperial College that we do not do now at George Mason, a public university. The difference is that in Professor Wakeham' s presentation, "multinational corporation" was a euphemism for an American company. But we would have no problem doing a strategic alliance with an overseas corporation at this point. If we wanted to accept federal government money as part of the package, then we would have to fulfill some of the requirements in U.S. law written by Patrick Windham to deal with the xenophobic impulse.
Michael Marron, Office of Naval Research: This question is directed to Christopher Hill. You ended on a note of increased private investment in research as a rosy picture but you gave a negative picture of the role of government involved in these areas. When those of us in the government enterprise look at the studies done by the National Science Foundation (NSF) on how much the investment has grown in various sectors of the economy, we're taken aback by the inability to assess how much of that research investment is highly directed. And I'm trying to avoid the applied research versus basic research polarization, but rather highly directed research as opposed to the undirected broader base research which we in the Defense Department (DOD) refer to as 6.1 and 6.2 type of research for our technical base investment. We are faced with people saying, "Look at all this huge growth here. We don't really need you guys anymore." And we see a different story when we examine it on a case-by-case, industry-by-industry basis. I don't know whether it's appropriate, but I would like to hear your comments about how this affects this rosy picture that you see developing in the 1990s.
Christopher Hill: It is difficult to parse out what any of us mean when we talk about dollar expenditures for R&D. The official statistics on federal funding for R&D include 6.3, 6.4, 6.5, 6.6, and 6.7. However, a great deal of activity in 6.4 and 6.5 is not what universities would recognize as R&D. Four years ago the National Academies of Science and Engineering issued a report called Allocating Federal Funds for Science and Technology that advocated the use of a new construct that it called federal science and technology (FS&T), which is an attempt to correct the statistics to eliminate those things that go beyond R&D, such as testing and evaluation of new systems. By that measure, FS&T funding, instead of being $75 billion, is around $45 billion, and it has grown more slowly than R&D. On the industrial side, the same is true. The largest funders of industrial R&D are the car companies: General Motors, Ford, and Daimler-Chrysler. My guess is that a great proportion of General Motors' $5.5 billion or $6 billion R&D budget looks like DOD's 6.4 or 6.5 activity. It includes testing prototypes of new vehicles, which is not unlike what the universities would say is not R&D at DOD. It is not really "R&D" at General Motors either. But it is for tax purposes, and so it is reported as such to NSF. So we have to be careful about statistics for both industry and government R&D spending. Nevertheless, no matter how you measure it, the federal R&D investment during the 1990s has been flat, while the private investment has grown very rapidly.
Michael Marron: My point and your point—you seem to be agreeing with me at this stage—is that heavy growth is really in the highly directed arena and not so much in the broader seed region that we might otherwise argue. And so maybe the picture is a little bit askew here when we look at the role of the federal government in this investment partnership.
Christopher Hill: I am talking about the decline of the large, centrally funded industrial fundamental research labs. Many of them are gone.
David Mowery: You are right in the sense that, if you break this down into the discredited categories, industrially funded R&D during the 1990s is by far the fastest growing category, with more than 10 percent annual rates of increase since development, which is much more rapid than basic research, whatever that may be.
Christopher Hill: I would add that, in the information technology sector, a great deal of what generates new technology is not even counted by the firms who are doing this as R&D. In Northern Virginia, there
is a very dynamic, 1,500-firm information technology sector. Almost no one claims to be doing R&D, and yet it is a rapidly developing, changing industrial economy.
Fritz Kokesh, Massachusetts Institute of Technology: "Directed research" is an interesting term. To me it implies that there is a purpose in mind if it's successful. But the comments here assume that there is another feature to it that assumes that it's not frontier research. And the two aspects can be separate. To illustrate, if someone can see a purpose for something that's still occurring at the frontiers of science, then that should be considered every bit as basic as something that can't be seen.
Christopher Hill: I agree with you on that.
Thomas Manuel, Council for Chemical Research, Inc.: It's important to iterate some of the things that are implicit in some of the talks here that would refer to what I call a "social dimension" of collaboration. And the first one is that if you look at the agenda of this workshop, it's sliced into bilateral pieces. The fact is that most collaboration and the most fruitful and the greatest trend for the future is going to be tri- if not multilateral. This is implicit in Professor Wakeham's presentation.
Second, collaboration is a contact sport, and it's an iterative activity. So not only does one need patience in a particular relationship, one needs to try again and again and practice it. The third thing intersects directly with Dr. Hill's observation that everything fell apart in the late 1970s. The Council for Chemical Research attempts to continually address these types of needs and opportunities as it goes forward in many directions and as the picture changes. So there is a need for all of us to continue to work together in this area, seek ways to find new collaborations, and to not get fixed into any particular paradigm of the moment.