Those who work at the intersection of the academic and industrial research enterprises have heard the story about the US manufacturer that, having spent countless hours trying to negotiate a cooperative research agreement with a university, gave up in a huff and took its business to an institution abroad. Trouble is, no one can, or will, identify the parties involved.
Whether truth or urban legend, the persistence of the story points to difficulties that both sides agree continue to hamper industry-sponsored academic research. The strains persist at a time when universities are eager to supplement federal research funding, which has leveled off in recent years. Industry funding for research surpassed that of the federal government in 1980; now industry pays for two-thirds of all the R&D conducted in the US. But as a share of total research funding for universities, industry support has declined from its peak of 7% in 1999 to 5% in 2006, according to NSF statistics. In that year US companies spent just $2.4 billion of their $223.4 billion total R&D outlays at US institutions of higher education.
Most of the impediments to partnerships are related to the treatment of intellectual property—patents, trade secrets, and know-how. Universities are intent on protecting the intellectual property they have built up over decades of basic research financed overwhelmingly by federal agencies. Whether their IP negotiations end with an agreement or not, industry and academia agree that the process of negotiation takes far too long. A few companies are rumored to have given up talking in frustration and instead have worked out arrangements with foreign universities. No one interviewed for this article would provide a specific example of that.
Dow Chemical Co, however, is up-front about preferring to work with institutions abroad. Susan Butts, senior director of external science and technology programs at Dow, says foreign universities ranging from Cambridge to Tsinghua will routinely assign their intellectual property rights over to the industry sponsor. Rather than haggling over IP, foreign institutions and their faculties are typically more interested in the experience they gain working on real-world problems, the industry connections for their students, and the funding, she says. In a 2003 exercise, Dow, which spends between 1% and 3% of its $1 billion research budget at universities, examined the terms of the more than 100 sponsored research agreements it had in place with institutions around the world. The company found that in 69% of its US contracts, the university partner had title to the inventions that resulted. By contrast, industry either took sole title to inventions or shared title in 85% of its sponsored work abroad. Dow also found that its agreements with US institutions took an average of more than five months to negotiate. That compared with just a few weeks required for arrangements with universities abroad.
Bridging the gulf
In 2006 the National Academies formed the University–Industry Demonstration Partnership (UIDP), of which Butts is the current president. The 85-member organization’s initial project is the development of a software tool called Turbo-Negotiator, designed to identify both common interests and differences in expectations the two parties have about a research proposal that is being considered for funding. If a company is bringing a lot of IP to the table, for example, then it probably should be provided liberal rights to the resulting inventions. The program also would identify potential stumbling blocks and give some indication whether a timely agreement is likely to result.
Anthony Boccanfuso, the UIDP executive director, says industry’s ability to go overseas for their research needs is a recent phenomenon that has given the issue greater visibility.
Mark Crowell, associate vice chancellor for economic development and technology transfer at the University of North Carolina at Chapel Hill and a UIDP board member, says he has seen fewer than five deals fall through in the 21 years he’s been involved in the business. That’s despite the huge differences in the objectives of each side. The university is more driven by a desire to maintain its research program and academic mission than by making money, he says. But Crowell worries that UIDP’s industry membership, and hence its focus, is drawn predominantly from information technology and chemical companies, with little representation from the biotechnology and pharmaceutical sectors.
Bayh-Dole Act to blame?
Butts and others in industry point to the Bayh–Dole Act, a 28-year-old amendment to US patent law, as the root cause of the trouble. By granting universities and other nonprofit organizations the right to own the inventions that spring from their federally funded research, Bayh–Dole fueled a rapid growth in technology licensing, startup company formation, and other commercialization routes that are now commonplace on campuses. But some in industry argue that the 1980 statute also had the unpleasant side effect of impeding research partnerships. Though the act didn’t speak directly to industry–university collaborations, academic institutions have interpreted Bayh–Dole as giving them ownership of inventions that arise from sponsored research. Since the sponsored research will draw from a considerable foundation of expertise gained through taxpayer-funded research, universities argue that they are compelled by Bayh–Dole to retain their ownership of IP. Crowell points out that industrial partners can get all the IP rights they need, including the ability to cross-license, without taking ownership. Robert Gruetzmacher, director of technology commercialization at the DuPont company’s Center for Collaborative Research and Education, observes that Bayh–Dole directs universities to ensure that the technology will be used when licensed. To do so, he says, universities must retain ownership. DuPont spends between 1% and 3% of its $1.4 billion research budget at universities, including in sponsored programs and unrestricted grants.
Dow, for one, advocates small changes in Bayh–Dole and tax law to clarify the treatment of IP in sponsored research. But no one expects legislation in the near future—certainly not this year. Universities, which have come to see their IP portfolios as a revenue stream, have united against any alterations. The Association of American Universities, the Association of American Medical Colleges, and other higher-education groups say no fix is needed; they cite statistics showing that the law led to the creation of more than 5000 companies, 3641 new products, and 260 000 new jobs based on university research from 1980 to 2005.
Even so, few of the 230 or more universities that now have standalone technology transfer offices derive sufficient licensing revenues to be self-sustaining, according to the Association of University Technology Managers (AUTM). Few inventions turn into “home runs” in the mold of the fabled Cohen–Boyer gene-splicing patents that spawned the biotechnology industry and delivered $255 million in royalties for their inventors and their institutions, Stanford University and the University of California system. Most of the other big moneymaker licenses are drug related.
Technology powerhouses like UC, MIT, and Stanford have well-established royalty streams and patent portfolios. Charles Louis, vice chancellor for research at UC Riverside, told a congressional panel last summer that UC manages a portfolio of more than 7500 active patents, 80% of which have generated interest from the private and public sectors. Systemwide, UC licensing revenues totaled $193.5 million in 2006, while research expenditures reached $3 billion, according to AUTM.
Other showstoppers
To be sure, IP rights aren’t the only potential showstopper to cooperative research. Companies typically insist on delaying publication so they can comb the results for potentially patentable material, a restriction that runs counter to the academic researchers’ imperative to publish. And Internal Revenue Service regulations restricting the use of campus facilities that were financed with tax-free bonds led some institutions such as Case Western Reserve University and Arizona State University to erect separate facilities using taxable financing to house their industry-sponsored research.
Albert Johnson, a senior analyst in the science and technology division of Corning Inc, says that provided the requisite workforce is available, as is the case in India, China, and other Pacific Rim countries, it makes good business sense to locate a research operation in proximity to manufacturing facilities. Johnson describes Corning’s sponsored academic research as a “vanishingly small” 1% of its $500 million R&D budget; he says the cost of putting a sponsored project together and negotiating the terms may be greater than the value of the proposed research. But a company could have other motivations to support domestic academic institutions, including a desire to keep their scientists and engineers abreast of the latest developments in their fields.
Most sponsored research projects are small. Dow’s run as little as $10 000, says Butts, while DuPont’s range anywhere from $50 000 to $500 000 a year. Companies typically pay for some work by a graduate student or postdoc. A project is usually tied to a product or process and supplements work the company is doing internally. There are some exceptions—DuPont and MIT have an umbrella agreement that commits the company to sponsor $60 million in research over 10 years. With a master agreement in place covering IP and other issues, all the parties need to do is plug in the values and the topic of the individual projects, says Gruetzmacher. Even bigger exceptions are the 2007 agreement oil giant BP made with UC Berkeley, Lawrence Berkeley National Laboratory, and the University of Illinois to provide $500 million for research on new energy sources and Chevron Corp’s 2006 pact with UC Davis to fund up to $25 million in energy research.
Needs vary by industry
Jon Soderstrom, current AUTM president and manager of Yale University’s technology-transfer office, admits that most universities “probably have not been attuned to the needs” of information-technology and other industries whose products incorporate physical sciences technology. Because their research portfolios are skewed toward the biomedical arena—the National Institutes of Health alone provides about half of all nondefense federal R&D support—many universities, including Yale, have considerable experience working with biotechnology and pharmaceutical companies, according to Soderstrom and others.
The trouble is that the IT and semiconductor industries have a far different business model, in which product life cycles are short and rapid commercialization is the key. “A month for them is like a year to biotech,” says Soderstrom. Whereas a single patent covering a new chemical compound can often be the basis for a blockbuster drug, new IT products typically comprise many components, each incorporating many patents. Manufacturers of IT products typically need to bundle all the necessary patents, which are of little use in isolation, explains Butts. Companies trade and cross-license the portfolios, even to their competitors. They are also anxious to reach agreements quickly, since an improved technology may come along in a matter of months, and competitors may reverse-engineer their products. By contrast, a biotechnology or pharmaceutical company can afford lengthy negotiations during the years-long process of bringing a new drug to market. But they will want to ensure that they have an exclusive license from the university.
Multinational IT and semiconductor companies don’t need exclusive licenses, and they view royalties as “an accounting nightmare,” says Katharine Ku, director of Stanford’s office of technology licensing. They would rather pay a predictable annual fee to license a patent. Another big sticking point to sponsored research is the treatment of background IP. Ku says that Stanford, which has about 100 industry-sponsored projects going, is willing to grant a company rights to use the IP that was previously generated in the sponsored investigator’s laboratory but has balked at giving blanket permission to infringe on the inventions of other Stanford faculty. Companies in the IT sector, she explains, often fear that universities will sue for infringement and force them to stop production of chips or other items.
Ku believes that the days of easy terms from foreign universities are numbered. She says Japanese and Taiwanese universities are moving toward the US model of technology transfer, while Singapore is working at becoming Asia’s intellectual property capital. She’s not worried about companies sponsoring research offshore. “Our view is that they should go wherever they can get the best research,” she says.
Sources of research funding for US universities, hospitals, and research institutions, 1991–2006
Sources of research funding for US universities, hospitals, and research institutions, 1991–2006
