The Back Page

Exploring the Connection Between Physics, Technology, and Economic Growth

by J.C. Tsang, H.D.I. Abarbanel, G.A.N. Connell, L.J. Lanzerotti, and J.D. Sullivan, members of the APS Panel on Public Affairs (POPA)

J.C. Tsang

How does scientific knowledge provide economic value in today's environment. Many of us "think" that this question is trivial. It isn't. Public appreciation of the value of scientific knowledge and physics is critical for continued support for our work. Understanding how concerned non-scientists now think about the economic benefits of such support is fundamental to responding to today's pressures on physics research and education. A detailed understanding of the practical value of physics knowledge in the contemporary economy and the ability to communicate that understanding is essential in justifying future investments in R&D.

Such efforts are necessary because financial support for R&D in the US is substantial. If this support is taken for granted by scientists, it can be significantly reduced or reallocated in ways that we may believe are unproductive. About $215 billion will be spent on R&D by the public and private sectors this year. Throughout the 1990s over 2.5% of the US GDP was spent annually on R&D. Total US R&D expenditures since the late 1970s have doubled. This growth has been driven by industry, which now provides over 65% of all US funding for R&D, reversing past patterns, where about 60% of all US R&D spending was supplied federally.

The $63 billion federal spending on R&D in 1998 is an impressive part of total discretionary federal spending. Even the $15 billion spent annually by Washington on "basic research" is a measurable fraction of discretionary spending. In contrast, total federal funding for the arts and humanities before the wave of interest in deficit reductions was only $350 million. Scientists should ask why they have been able to attract so many more federal dollars. Physicists should also consider the growth in private sector support for R&D over the last two decades and understand the opportunities it presents.

As physicists, we believe that physics education and research benefit society in numerous intangible and tangible ways. The intangible contributions of physics include the satisfaction of the human instinct to comprehend the world in which we live through a fundamental understanding of the physical universe. The contributions of modern physics to many different parts of contemporary life make an unchallengeable case for how physics has enriched the nation. While we may not have built the things we see around us, we know they are based on our ideas. For many of us, the intellectual beauty and anecdotal connections justify the national investment in our educations and work.

While we may wish to justify physics on intangible grounds, we must recognize that more tangible rationales have always been critical to our fellow citizens. There is general agreement that scientific discoveries based on research are often necessary, but almost never sufficient for new technologies and economic growth. While many new products and processes originate in scientific discoveries, the scientific discovery is often not the most "difficult" of the steps leading out of the lab. It is almost never a sufficient step for the civilian marketplace. The "I didn't actually build it, but it was based on my idea" rationale now produces scorn rather than respect.

Two subjects are at the heart of the POPA bibliographic essay on research, technology and economic growth. One is whether the existing organization of research in the US, and the system for its support, provides satisfactory economic returns given the current level of investment in R&D. The second is the effort to quantify the economic benefits of scientific research.

The debate over the first question challenges the traditional distinctions between basic and applied research and the importance of basic research as the major driver of innovation. It suggests that the systems and organizations that were very successful after WWII may require significant modifications to meet today's challenges. This can severely impact support of physics research as we know it.

With respect to the second, there is wide acceptance that research is necessary for economic growth. However, there is not a consensus on how the economic benefits of research can be quantified. The intellectual infrastructure connecting levels of research support to quantitative increases in the GDP does not exist. In contrast, analysts believe that we quantitatively understand the economic consequences of many other investments. Since budgets are quantitative instruments, it is an advantage to be able to relate quantitative results to quantitative inputs. We must take seriously the efforts of economists and policy analysts to quantify the economic contributions of R&D.

The Science-Technology Connection
Our current institutions and policies are solutions to yesterday's challenges and problems. Since WWII, some of these challenges and problems have changed. Occasionally, this can require new solutions. Studies of the role of science and technology in innovation improve our understanding of the contribution of research to economic growth. Understanding in this area traditionally occurs through complex case studies (data) and the creation of simple models (theory) which try to summarize these studies. The models inspire controversy both with respect to internal consistency, and whether they describe reality. These controversies are important because alternate models provide different conceptual frameworks by which policy makers, scientists and engineers view the complex interaction between science, technology, and economic growth.

Many developments in the first half of the 20th century showed the limits of "uneducated, practical" inventors and the need for highly educated innovators capable of understanding and performing scientific and technical studies. This produced a model of innovation, where technical progress began with basic research and "basic research is the pacemaker of technological improvement." Projects producing revolutionary innovations were theorized to begin with basic research, pass through applied research and end in development. This so called "linear model" provides an appealing rationale for the support of basic research. Its underlying assumptions had significant influence on R&D practices and policies in the US for many years after the end of WWII.

Today, the "linear" model has been abandoned in many policy circles. In the 1970s and 1980s, concern about US performance in the global economy spurred increased interest in the connections between scientific knowledge and economic growth. Reassessments of the "linear model" led to more complicated models of the innovation process, and the role of research in it. These models draw on what are perceived to be unique features of today's economy. These include emphasis on time to market, extreme competitiveness, globalization of design, and manufacturing expertise, etc. All emphasize the importance of applied research, and feedback from users, customers, etc. in driving technological innovation. The rapid, continuous evolution of products described by "Moore's Law" is seen as the major way in which research can contribute to economic growth. Today's understanding of how we now view the connection between scientific knowledge and economic innovation has a critical impact on how support for research is currently justified and structured.

Valuation of R&D
A satisfactory quantitative understanding of how research pays off in the marketplace would allow the direct comparison of the returns on investments in research and other kinds of investments. For physicists who know the history of the transistor, the semiconductor industry seems a splendid example of how research can have enormous economic payoffs. But within the current methodologies of modern economics, there is no generally accepted numerical or analytic model of the contributions of scientific research to quantitative economic growth. The controversies in this field are intellectually challenging and well worth reading.

This problem is closely related to the management of research in business. The dominant methodology of the past has been the calculation of the net present value (NPV) of a project. This technique was designed for investments such as building factories and buying airliners, but is widely criticized as a blunt tool more likely to provide the wrong answer than good guidance for R&D policy.

The emerging field of the quantitative assessment of risk has studied this problem in the last decade and offers a different framework for analysis. The new approach stresses that current research investments provide information allowing for the better use of future investments. From this viewpoint, a research investment is analogous to a financial option. A research investment is viewed as the cost of keeping such an option alive until the decision to go ahead with or cancel the project can be made with greater certainty about the outcome based on the results of the research. Such systematic approaches to investing in research raise challenging questions about how different types of research can be compared, how strategic directions against which the option value of a proposed research activity can be assessed quantitatively are arrived at, and how metrics to assess the impact of the strategic options can be defined.

Conclusion
Anecdotal examples of the value of R&D are basic to any consensus on the value of physics research. They do not, however, help individuals who are trying to understand the additional benefits of a 7% annual increase in federal basic research funding as compared to, for example, a smaller annual increase in research funding, and new tax credits for corporate R&D or increased funding for Head Start, etc. Even in an era of budget surpluses, choices like these will have to be made. A better understanding of how R&D contribute to our economy, and the development of metrics that quantify these contributions would make trivial the justification of R&D to our fellow citizens.

For the present, however, physicists must be realistic about the kinds of arguments we can present. We can clearly point to where we have had great intellectual impact but must recognize that our economic impact has generally come as part of a community of scientists, engineers, and technically minded individuals. If we work together with our professional colleagues, accept that many of the choices involved in the crafting of science and technology policy are political choices, and are prepared to engage in the political process by which policy choices are made, we now know enough about the economic impact of physics research and education today to feel comfortable, though not complacent, about the economic future of our profession.

The above is a summary of a recent bibliographic essay produced by the Science, Technology and Economic Growth Committee of the APS Panel on Public Affairs (POPA). The full report, with complete citations of source material can be found on-line at www.aps.org.