Capitalizing on Investments in Science and Technology (1999)

COSEPUP believes that our nation is well positioned to sharpen its ability to perform and capitalize on research. We have emerged from a defense-oriented era of superpower tensions into a more fluid and flexible environment in which ideas, people, capital, and goods flow more freely among nations. New technologies and new institutions, most notably small- and medium-size firms, are setting a rapid pace of innovation. Newly competitive nations are entering the global marketplace. Traditional institutions, especially universities and government agencies, are testing new policies and partnerships that will allow them to adapt to this fast-paced and open environment.

Finding 1: Capitalization on science and technology is a major national strength, although there is much room for improvement.

Capitalization appears to be quite healthy in the United States today, delivering significant benefits to the nation. Nonetheless, COSEPUP believes that there are many opportunities in every sector to improve the capitalization process. As outlined in various parts of the report, the United States has weaknesses, and complacency could lead to a decline in its strengths.

This finding contrasts with the situation a few years ago, when several U.S. industries faced serious challenges in global markets (Dertouzos et al., 1989). Foreign based companies used superior product development and manufacturing to surge forward in high-technology industries pioneered in the United States. Pointing to continued U.S. strength in basic research, some observers were concerned that the United States was losing the ability to capitalize on its research investments, while foreign countries were reaping the lion's share of benefits (Prestowitz, 1988).

As this is written, the situation looks quite different because of two important shifts. First, many established U.S. companies and industries have improved their performance in product development, manufacturing, and marketing (see Chapter 2 and STEP, 1999). Second, a wave of new industries and companies has arisen in the United States, many of them with clear and direct links to public and private research efforts initiated several decades ago (such as the Internet and life sciences

examples cited earlier). Both of these trends have benefited from, and contributed to, a favorable macroeconomic environment.

The discussion in Chapter 3 illustrates the complexity of the emerging challenges and reinforces the caution expressed in Chapter 2, against formulating policies based on overly simple models of innovation.

Finding 2: The key elements contributing to effective capitalization are

• strong, stable funding for a portfolio of research investments that is diverse in terms of funders, performers, time horizons, and motivations;
• a favorable environment for capitalizing, characterized by a strong incentive structure for investors, competition in the market, and free movement of ideas and people between institutions;
• a skilled, flexible science and engineering human resource base that allows the United States to maintain research at the cutting edge and to capitalize effectively;
• mechanisms for research and capitalization that support cooperation between academic, industry, and government sectors.

These elements increasingly interact with each other. The key challenge and task for the science, engineering, policy, and business communities will be to continue to innovate so that the elements underlying capitalization are strengthened in the face of changing circumstances. The remainder of this chapter deals with that challenge.

The rationales and mechanisms by which our institutions support research will be centrally important in the twenty-first century. The essential seedbed for capitalization is a diverse portfolio of research programs, both long-term and short-term, across the spectrum of major fields. An effective research policy can provide continuing, long-lasting benefits to society in the form of new insights and products and an open intellectual environment in which future generations of scientists and engineers are educated. Research investments create several of the key ingredients needed for capitalization, such as the science and engineering human resource base that transforms science and technology into practical benefits through entrepreneurship and other mechanisms.

A central role of the federal government is to monitor and assess the national science and technology investment portfolio to ensure that U.S. scientists and engineers work at the forefront of all major fields and attain clear leadership in fields deemed essential to national objectives. The federal government must serve as the "funder of last resort" to support research and capitalization efforts in fields of national importance that are not able to secure funding from other sources. This task is increasingly important in an environment where industry is the predominant funder of research and development (R&D), and both federal and industrial

research funders face pressure to support work that delivers measurable, short-term results.

For example, the federal government has a unique responsibility to develop and maintain the infrastructure and technology that support modern research. For a nation to be a world leader, its scientists and engineers must have access to state of-the-art facilities. Many of these facilities are too expensive for a single institution or even industry to support. In the case of materials research, for example, facilities and equipment in several foreign universities now outclass those at most universities in the United States. Of particular concern is the need for modern equipment for materials synthesis and processing, where the United States lags Europe and Japan (COSEPUP, 1998, p. 34).

According to workshop discussions during this study and to a recent expert panel report to the President (See Box 3-1 and President's Information Technology Advisory Committee, 1998), long-term research to generate future innovations in information technology is inadequate. In the area of applying research on cognition and learning to address the nation's educational challenges, there is significant, difficult work to be done (Appendix A). The federal government will need to recognize and respond to such funding gaps and needs.

Recognizing and responding to emerging funding needs will require new tools for policy makers. Mechanisms for monitoring and assessing the national science and technology investment portfolio are emerging, but developing these tools is a task that will require additional study and experimentation. For example, science and engineering might benefit from a continuing, regular program of assessment, or "benchmarking," for individual fields, such as those recently conducted by COSEPUP. The purpose of these assessments would be to help funders and policy makers determine appropriate levels of funding, not to set milestones or predict outcomes.

The science and engineering communities can assist in the development of monitoring and assessment mechanisms. Researchers are encouraged to determine appropriate tools to assess their own particular fields. It is tempting to try to apply universal methods of assessing the return on research investments. Some forms of research, particularly those in which a certain outcome is expected, lend themselves to quantitative evaluation. Other forms, notably internally driven, long-term research, are not assessed easily by metrics or milestones because their specific outcomes and rate of progress cannot be known in advance. The search for new knowledge also contributes to goals other than those prompting the initial research.

In recent decades, a central mission of many graduate science and some engineering programs has been to prepare the future professoriate. In accord with this mission, most U.S. graduate schools impart a solid grasp of the principles and practice of research, and of the intellectual openness of universities. However, the majority of Ph.D.s in science and engineering enter employment positions outside

the academic research community—in industry, government, and teaching—where job demands and cultures may differ appreciably from those of academic research.

An important influence on how graduate students are prepared for employment is the type of funding mechanism they receive, such as fellowships, traineeships, and research assistantships. The proportion of these mechanisms has varied over time, but less in response to careful planning than to political or economic imperatives.

COSEPUP (1995) suggested in an earlier report that certain forms of financial support might allow some graduate students to gain greater flexibility in making educational choices, which could in turn allow them to select from a broader range of options and to adapt their preparation to a variety of careers. One model is the National Institutes of Health program grant, and another is the training grant recently introduced on a small scale by the National Science Foundation (NSF). Individual universities are developing their own approaches (Box 4-1). Yet little is known about how different types of grants may alter student or faculty behavior.

During a joint COSEPUP-National Academy of Engineering workshop on the Role of Human Capital in Capitalizing on Research, experts in the computer networking field stated that commercial growth is so strong that students inclined to work in industry have excellent job prospects with a bachelor's or master's degree. At the same time, the long-term academic research being undertaken in some relevant fields is becoming more specialized and "mathematical," so that students at the Ph.D. level have less opportunity to work on systems-oriented

problems (see Box 3-1). The incentives for both industry and academia to do the systems-oriented research and education needed for future growth in the networking area are apparently weak. How can we ensure that our research and education investments create the necessary human capital for interdisciplinary fields in which commercial opportunities are expanding rapidly?

One member of the working group suggested that small-scale, randomized experiments could help to answer these questions (Romer, 1998). For example, some students might receive funding via portable project fellowships, whereas others might receive more traditional fellowships or research assistantships. Over time, several questions could be asked: How would students' career choices vary? Would universities respond to incentives in the form of potential tuition dollars rather than research dollars?

In the same way, one could conduct an experiment whereby principal investigators at certain institutions would apply for program grants, while investigators at other institutions would apply for traditional research grants. Would the nature of the research vary at the two groups of institutions? Would faculty promotion criteria change? What about faculty practices of teaching or mentoring? Admittedly, a controlled, randomized experiment in this area would be difficult to implement. Still, it will be important for policy makers and the science and engineering community at large to design programs in ways that the results can be evaluated.

It is desirable, by whatever mechanisms, to increase the attractiveness of careers in science and engineering. Goals that have been proposed include bringing additional real-world and teamwork experiences to the classroom, creating more industrial internships, producing more interesting courses (especially at the introductory level), and stabilizing the levels and consistency of funding policy.

Graduate students achieve richer educational experiences and greater employment opportunities through more experience in industrial labs. NSF's Grant Opportunities for Academic Liaison with Industry (GOALI) program supports university-industry linkages. Some institutions, including Massachusetts Institute of Technology, the University of Michigan, and Lehigh University, offer industrial research opportunities. From industry's point of view, well-prepared students are essential. "We believe that if we hire the right people, products and profits will follow," said a semiconductor industry executive. "If we didn't have human capital, we wouldn't exist" (Wollesen, 1998).

The interchange of ideas and people at university-industry-government interfaces is a key to capitalizing on research. In particular, university-industry collaborations that transcend disciplinary barriers and focus on real-world problems bring many benefits, such as exposing students to the industrial environment and culture and allowing industry access to cutting-edge research. In most cases, project initiation and technology transfer decisions should be made by the private sector; the internal effort and skill of firms are the essential ingredients of innovation. The

government should play a catalytic and funding role in partnerships selected for their potential to capitalize on research (Nelson, 1993, p. 510).

Many large corporations now depend heavily on external sources of research. For example, DuPont's partnership with researchers at the University of North Carolina has led to new olefin polymers that may open a multimillion-dollar business. IBM, Toshiba, and Siemens are collaborating to produce 256-megabit memory chips. Even as they compete in the marketplace, companies must remove more barriers in order to maximize R&D efforts. New understandings from universities can improve a company's ability to improve products and exploit new opportunities. To support this process, companies once relied on their own expertise; now they maintain outside partnerships for this purpose. Of course, partnerships are difficult to form and manage effectively; many fail to live up to expectations.

In a true collaboration, both partners find areas of mutual interest, benefit from synergies of ideas, and share results equitably. Benefits can include shorter development times, better products, lower risks, and lower costs. However, not all areas can benefit from partnerships, and there is a danger that some forms of collaboration between university and industry could create conflict with the basic educational purpose of the university. Harvey Brooks (1993) proposes "buffer institutions" at, but not quite of, universities that would pursue these agendas.

Many universities are struggling to align internal policies, especially those regarding intellectual property rights, with industrial partnerships. One common formula is to grant the industrial sponsor rights of first refusal to an exclusive license; partners delay royalty discussions until they actually make a patentable discovery (Council on Competitiveness, 1996).

In some fields, potential partners continue to be isolated by cultural barriers. Several participants in a workshop on piezoelectric ceramics mentioned that U.S. ability to capitalize on research in this field would be improved if students and academic researchers had a better understanding of the potential applications of their work (Freiman, 1996).

Innovative partnerships are being tried by state governments. The Minnesota Technology Partnership Fund seeks to stimulate relationships between small companies and postsecondary institutions. Its objective is technology transfer—to increase the access of small, technology-oriented companies to academic resources. A company is invited to apply jointly with an academic partner to fund R&D that is designed to lead to near-term commercialization. One rationale is that economic activity can be stimulated by state incentives and the presence of a major research university.

State governments have improved their ability to help manage programs. The U.S. Innovation Partnership, which links federal research and innovation policy making to states through the National Governors' Association, provides an important new mechanism. State governments are increasingly responsible for delivery of technology and training services and, more generally, for technology diffusion

and utilization. They are also better situated than the federal government to integrate training and education in the local setting.

For the nation to realize returns on its investments in science and technology, a favorable environment for capitalization is necessary. The discussions in Chapters 2 and 3 illustrate the importance of a strong environment, and how it interacts with the other capitalization ingredients identified by the working group. For example, the favorable environment in the United States for commercializing technology through the formation of new firms has accelerated capitalization in areas such as the Internet, monoclonal antibodies, and other areas of biotechnology. For the most part, the U.S. antitrust environment has encouraged innovation and the free flow of information about key innovations. Without a favorable environment, realizing returns on investments in cutting-edge research and the creation of superior science and engineering human capital would take longer or would not occur.

Although the working group did not uncover any general concerns in this area that require corrective action, conditions and perceptions can change quickly. As pointed out in Chapter 3, the financing environment for new science- and technology-based firms historically has exhibited wide swings. Shifts in the investment environment can slow innovation in other ways as well. Some years ago, it was asserted that Japan-based companies enjoyed an advantage over others in pursuing long-term innovation strategies because of their ability to access low-cost, patient capital (Prestowitz, 1988). Now it is apparent that the efficiency of capital deployment is critical as well (Lahart, 1998). Although recent U.S. economic performance has been excellent, the low U.S. savings rate and short time horizons for investment could reassert themselves as U.S. weaknesses in the future (NRC, 1999).

Other aspects of the capitalization environment are changing, and will undoubtedly require adjustments and adaptations in the future. For example, the growth of new high-technology industries, particularly computing and information technology, is posing challenges to the enforcement of competition and antitrust policies. Differences in national systems for trade, investment, and industrial development still cause international frictions (Hamburg Institute for Economic Research et al., 1996). Other issues, such as product liability, have been mentioned as barriers to innovation in particular industries (Hunziker and Jones, 1994).

In short, it will not be enough for the United States to ensure a strong, diverse portfolio of science and technology investments, strengthen science and engineering human resources, and facilitate cooperation between sectors. Policy makers and the science and engineering enterprise must continue to recognize the importance of the capitalizing environment, and help to maintain and improve it.