Capitalizing on New Needs and New Opportunities: Government-Industry Partnerships in Biotechnology and Information Technologies (2001)

The biotechnology revolution is likely to be emblematic of the 21^st century just as the information revolution has characterized the latter part of the 20^th century. The momentum of the information revolution is not spent, however, because computing and communications will continue to change our economy and society as electronic devices become smaller and more pervasive, and as information networks expand in capacity.

The biotechnology revolution, though presently in its early stages, will grow in importance, in part, because information technology (IT) will enable scientists to take full advantage of advances in areas such as genomics. At the same time, biological systems may provide useful models for advances in computing. Indeed, the information and biotechnology revolutions are not likely to continue as separate phenomena. Rather, biotechnology and computing will interact increasingly with each other and, in this way, contribute to the pace of advance in each field. Federal partnerships with other principal actors—namely, industry, universities, and laboratories—are likely to play a key role in the continued advance of both biotechnology and information technology.

The federal government’s role remains significant although non-federal entities increased their share of national funding for R&D from 60 to 74 percent between 1990 and 2000. Federal funding still supports a substantial component, 27 percent, of the nation’s total research expenditures.¹ Importantly, federal

expenditures constitute 49 percent of basic research spending. In addition, federal funding for research is often more stable and based on a longer time horizon than funding from other sources. Shifts in the composition of federal research support are therefore important in their own right and for the impact these shifts may have on the future development of biotechnology and computing—two of our most innovative industries and a source of substantial growth in the economy.

There are two main concerns. The first is the absolute amount of federal investment and the second is the allocation of the funds available. Concerning the first point, there is a growing concern that the United States is not investing enough and broadly enough in research and development. In the private sector, the demise of large industrial laboratories, such as IBM’s Yorktown facility and Bell Laboratories, has reduced the amount of basic research conducted by private companies.² Although private sector R&D has steadily increased in the United States in recent years, almost all of it has been product-oriented rather than geared towards basic research.³ In addition, the increase in corporate spending on research is concentrated in sectors such as the pharmaceutical industry and information technology.

The second area of concern regards the allocation of federal research funds, specifically, the unplanned shifts in the level of federal support within the U.S. public R&D portfolio. As highlighted in an earlier report by the STEP Board,⁴ the United States has experienced a largely unplanned shift in the allocation of public R&D (Figure 1).⁵ The end of the Cold War and a political consensus to reduce the federal budget deficit resulted in reductions in federal R&D funding in real terms. For example, the decline of the defense budget means that military support has fallen for research in physics, chemistry, mathematics, and most fields of engineering. The previous STEP study, noted above, showed that in 1997—even though the overall level of federal research spending was nearly the same as it had been in 1993—several agencies were spending substantially less

on research than in 1993: The Department of Defense had dropped 27.5 percent, the Department of the Interior was down by 13.3 percent, and the Department of Energy had declined by 5.6 percent.⁶ Declines in funding for the departments of Defense and Energy are significant because, traditionally, these agencies have provided the majority of federal funding for research in electrical engineering, mechanical engineering, materials engineering, physics, and computer science.⁷

After five years of stagnation, federal funding for R&D did recover in FY1998. In 1999, total expenditures were up 11.7 percent over the 1993 level. These changes were driven mainly by the increases to the NIH appropriations. Breakthroughs in biotechnology, and the promise of effective new medical treatments, have resulted in a substantial increase in funding for the NIH, which is slated for further increases by the current administration.⁸

Notwithstanding this change in overall R&D funding, the most recent analysis shows that even with the increase in federal research funding after 1997, the shift in the composition of federal support is largely unchanged. In 1999, the life sciences had 46 percent of federal funding for research, compared with 40 percent in 1993.⁹ This difference in funding trends between the physical sciences and engineering on the one hand and the life sciences on the other hand, is disturbing insofar as progress in one field seems to depend increasingly on progress in others. Growing imbalances in federal investment in research across disciplines may therefore have major consequences for our ability to exploit fully the existing public investments in the biomedical sciences and for continued United States leadership in biotechnology innovation and commercial applications.

The policy community has recently begun to recognize the implications of these differential trends in funding. Indeed, one of the early attempts to focus policy makers’ attention to address these concerns was the Committee’s October 1999 conference on Government-Industry Partnerships in Biotechnology and Information Technologies. This report summarizes the proceedings of that conference and includes several subsequent studies commissioned by the Committee for Government-Industry Partnerships for the Development of New Technologies. An overview of the conference and the papers are presented below. The Findings and Recommendations of the Committee, intended to address the issues raised in the conference and commissioned analyses, are presented in Part III of this volume.

The conference on Government-Industry Partnerships in Biotechnology and Computing took place at the National Academy of Sciences in Washington, D.C. Panel presentations and discussion centered on four related topics. These were (1) the nature and emerging trends of the federal research portfolio in biotechnology and information technology, (2) historical perspectives on federal support for these sectors, (3) new needs and emergent opportunities in biotechnology and information technology, and (4) how the United States might best meet these needs and capture potential benefits. Key points of speakers are summarized below:

Gordon Moore, Chairman Emeritus of Intel Corporation, in his review of the semiconductor industry, noted how a positive policy framework from the federal government has helped that industry to achieve its current position of competitive strength. This support, he noted, has evolved from policies that protected intellectual property rights to the support of industry research consortia to continuing partnerships between national laboratories and industry in emergent technologies.

In reviewing federal support to the biotechnology industry, Edward Penhoet of the University of California, observed that it has been mainly in the form of basic research support—much of it in university settings with funding from the National Institutes of Health. As biotechnology continues to mature, he noted, it needs greater multidisciplinary collaboration to reach its full capability.¹⁰ In turn, the nature of federal funding will have to evolve if it is to accommodate the changing nature of the industry.

Taking a historical perspective, Kenneth Flamm of the University of Texas noted that the federal government role in supporting information technology evolved from being the main customer to providing long-term basic research to supporting near-term targeted projects. This supporting role would continue to evolve, he predicted, even as the industry continued to mature. Somewhat surprisingly, Dr. Flamm reported that the level of public support for computing has declined in recent years.¹¹

In tracing the development of medical research and the biotechnology industry, Leon Rosenberg of Princeton University noted that universities play a critical research role and thus have to be accounted for in any government partnership framework. He further noted that medical and biotechnology research enjoys broad-based political support and is therefore well funded—a theme resonated by William Bonvillian of Senator Joseph Lieberman’s office. Indeed, both speakers concurred that biotechnology research is expected to become increasingly multidisciplinary. To adjust to this trend they suggested, the distribution of funding to support biomedical research will increasingly have to be more broad based—that is, across a variety of seemingly unrelated disciplines.

In exploring new needs and emergent opportunities in biotechnology, Marvin Cassman of the NIH noted that the marriage of biology and information technology will usher an incredible pace of change in the advance of knowledge. Using the data being generated in genomics and other fields, scientists will be able to understand biological systems more completely. To this end, skills in bioinformatics and related information technology will be increasingly called for. Yet, as Rita Colwell of the National Science Foundation observed, interdisciplinary training—needed for bioinformatics—is often limited by constraints posed by traditional departmental boundaries. As a result, she concluded, there is an inadequate supply of such scientists with the needed interdisciplinary skills.

Dr. Paula Stephan confirmed that there is great demand for information technology professionals, especially in the biotechnology industry, but that there are problems in meeting this demand. In explaining this supply problem, Dr. Stephan offered that universities respond more to research funding than to industrial demand. She also noted that traditional departmental structure in universities often inhibit interdisciplinary training. In addition, the operation of the market—where jobs in information technology traditionally pay more than do jobs in biology—also reduces incentives for information technology professionals to acquire training in biology.

Meanwhile, research in biotechnology holds promise for applications in a variety of areas. As publicized at the conference, crosscutting technologies—such as nanotechnology and functional genomics—are expected to yield useful applications ranging from tailor-made medication to biosensors that detect environmental dangers. In his luncheon address, Daniel Goldin of NASA outlined the enormous potential of biology and information technology, and explained how synergies between the two hold the key to future space exploration.

In order to capture the important complementarities between biological research and the physical sciences, policy shifts will be required to support and promote continued U.S. leadership in biotechnology and information technology. As Congressman Sherwood Boehlert pointed out, the question is not one of if, but more of how and at what levels government should support biotechnology and information technology research.

Indeed, as several speakers at the conference pointed out, the existing policy

framework has, in some respects, quite different impacts on the evolving needs of the information technology and biotechnology sectors. For example, Wesley Cohen of Carnegie Mellon University noted that appropriate patenting rules are especially important in the biotechnology sector since these are often the principle assets of a start-up venture.¹² By contrast, firms in the information technology sector rely more on trade secrecy and long lead times to protect intellectual property. Robert Blackburn of Chiron Corporation explained that the courts have been applying patent law doctrine, developed for the synthetic chemical industry, to biotechnology. However, unlike synthetic chemicals, which are marketable products, biotechnologies are research tools. This makes patents difficult to enforce, he noted, leading in some cases to friction between university and industry researchers. Thus, current applications of patent laws can sometimes act as a disincentive to drug development. In a similar vein, Maryann Feldman of Johns Hopkins University noted that existing patenting rules—as per the Bayh-Dole University and Small Business Patent Act—yield only modest benefits to universities in terms of licensing revenue while, in some cases, even discouraging university-industry collaborations.

Differences also exist in the nature of direct government support to the information technology and biotechnology sectors. While there is a long history of government-industry partnerships in information technology, government support of biotechnology has, to date, relied largely on funding for basic and clinical research, with industry concentrating its resources on the development of commercial applications, albeit under the supervision and review of the Food and Drug Administration.

Yet, as the gap between fundamental research and application narrows in biotechnology, and as the field becomes more multidisciplinary, the nature and distribution of government support—as well as the policy framework—will need to evolve. The conference presented two government-industry-university consortia—the Mouse Sequencing Consortium and the Nucleotide Polymorphism Consortium—as illustrations of possibilities in this sphere.

Following the conference, four research papers were commissioned to deepen the analysis of points raised at the conference. Two deal with macro trends—one on government-industry partnerships in information technology, and the other on the composition of the present portfolio of federally supported research. The remaining two take up complementary micro-level examinations of the rele-

vance of current patent policy to the biotechnology and information technology sectors, and the need to address critical requirements in bioinformatics. Together, the four papers provide empirical support and greater depth of analysis to key aspects of the conference deliberations.

The paper by Kenneth Flamm examines the history and current concerns in federal support and government-industry partnerships in U.S. computer research.¹³ He notes that large-scale R&D investment in computers and computer architecture has dropped in both absolute and relative terms over the 1990s. Further, he finds that longer-term investment in fundamental and basic research for computers is falling in comparison with historical levels.

This is cause for significant concern, argues Dr. Flamm, since computer technology, though pervasive, is by no means fully mature. Indeed, a growing body of economic literature links improvements in the technology industries with the recent healthy performance of the U.S. economy, and points out that continued technical advance in information technology yields substantial payoffs in biotechnology and other areas.¹⁴ To maintain the commitments of the industry and to ensure continued welfare-enhancing gains in productivity, the federal government should continue and expand its investments in the research that supports this critical industry.

In his analysis, Michael McGeary documents recent trends in federal funding of research related to health and information technology.¹⁵ Drawing on the recent assessment by the STEP Board,¹⁶ he describes the substantial shift in the composition of the federal research portfolio that has taken place over the 1990s. He notes that in 1999, the life sciences had 46 percent of federal funding for research, compared with 40 percent in 1993, while funding for physical science and engineering decreased from 37 to 31 percent of the federal research portfolio over the same period. Dr. McGeary’s paper raises the question of whether, consequently, “the federal research portfolio has become ‘imbalanced.’” While it is normal that the allocation of the nation’s research funding should evolve over time in response to new scientific opportunities, the sustained shifts in the distribution of this research portfolio allocation have important implications for future advances in both information technologies and biotechnology.

Wesley Cohen and John Walsh study patterns of publicly funded research and patent policy, and note their implications for industrial R&D in the pharmaceutical, biotechnology, and semiconductor and computer industries.¹⁷ In particular, they assess the impact of the Bayh-Dole patent and trademark act of 1984. This act seeks to exploit what was believed to be a stock of innovations in universities and other research centers by creating incentives to commercialize them. In particular, Bayh-Dole permits these organizations to obtain patent rights on federally sponsored research.

Based on surveys of R&D managers, Cohen and Walsh find that government support for research is important for the pharmaceutical, biotechnology, and semiconductor industries. Among these, however, patent protection is effective in stimulating research in the pharmaceutical and biotechnology industries, but less powerful in the semiconductor industry. This underscores the importance of understanding the differential effect of the current public policy framework on different industries.¹⁸ In the semiconductor industry, firms tend to protect their inventions through a combination of lead-time advantage, secrecy, and the exploitation of complementary capabilities rather than rely on patents.

Finally, Paula Stephan and Grant Black examine emerging opportunities and emerging gaps in bioinformatics.¹⁹ Bioinformatics, they note, is a highly interdisciplinary field, requiring understanding of mathematics, computer science, and biology. There is today a critical need for individuals with skills in bioinformatics to maintain the pace of research in biology.

Stephan and Black note that while the demand for individuals trained in bioinformatics is strong and professional salaries are high, too few students are being trained in this area. They suggest that the funding signals—which encourage faculty research, and hence the demand for doctoral and postdoctoral students—are inadequate. Moreover, the traditional organization and incentives of the relevant disciplines and university structures tend to discourage cooperation. Stephan and Black suggest, among other remedial measures, that federal research funds should be targeted to promote the institutional and pragmatic changes necessary to produce a greater supply of students trained in bioinformatics.

Despite some recent turbulence, the United States continues to enjoy, at this writing, an unprecedented period of economic growth even taking into account the recent cyclical downturn. This sustained growth, historically low unemployment, and—since 1995—substantial increases in productivity, combined with the great promise of new technologies, provides the conceptual basis for those who believe the U.S. is developing a “New Economy,” that is one with higher long term growth in productivity.

Despite the recent cyclical downturn, our nation’s scientists, engineers, and policymakers have reason to be proud of this economic performance. Yet, many of the new technologies that drive growth today are based on innovations that are a number of years, even decades, old. The Committee is concerned that recent downward trends in funding for basic research in the public and private sectors put at risk the investments that are necessary to nurture and sustain such growth.

It is for this reason that they believe that government policymakers and industrialists should devote more attention to the investments that must be made today to underpin future productivity and prosperity. Longer-term, sustained, and increasing support for a diversified R&D portfolio is key to sustained economic growth and increased national welfare. Effective government-industry partnerships are a key component of this productivity-enhancing policy framework. With this review of the opportunities and challenges posed by partnerships in the biotechnology and computing industries, the Committee hopes to encourage greater public understanding of the importance of the national research effort and of the need for well-conceived partnership programs to help bring the fruits of this research to all Americans.