AEI Press  (Washington) Publication Date: January 1997 Paperback ISBN: 0844740152 Hardcover ISBN: 0844740144

Science for the Twenty-first Century: The Bush Report Revisited
Edited by Claude E. Barfield

Vannevar Bush's report, Science: The Endless Frontier, has been the touchstone for all discussions of U.S. science policy since its release in 1945. The five essays in Science for the Twenty-first Century reexamine the Bush Report and its relevance to the challenges facing contemporary science policy.

The editor, Claude E. Barfield, is a resident scholar at AEI. He has also edited, with Bruce L. R. Smith of the Brookings Institution, The Future of Biomedical Research, which AEI Press will publish in December.

Vannevar Bush's seminal 1945 report for President Franklin Roosevelt, Science: The Endless Frontier, is justly regarded as the single most influential document shaping U.S. science and technology over the past half-century. The two purposes of this collection of essays are, first, to evaluate Bush's underlying assumptions and recommendations in light of current challenges to U.S. science and technology policy and, second, to suggest areas where the report remains valid or prescient, areas where its vision is flawed, and areas where new conditions and circumstances have overtaken it.

To accomplish those tasks, AEI recruited a diverse group of economists. Three of the authors, Richard R. Nelson of Columbia University, David C. Mowery of the University of California, Berkeley, and Murray Weidenbaum of Washington University have written extensively on issues related to research and development; the fourth, Raymond Vernon of Harvard University, has long studied the relationship between technological comparative advantage and international trade; and the fifth author, William A. Niskanen of the Cato Institute, was chosen to bring a fresh perspective from an economist who has not specialized in science policy issues. Not unexpectedly, the results contain points of consensus as well as disagreements over the continuing legacy of Vannevar Bush.

The Linear Model

From our contemporary perspective, the most fundamental flaw in the Bush Report is its simplistic view of how innovations occur and the role of research in that process. According to that view, the innovation process is a relatively straight line with a clearly delineated progression from basic research to applied research, demonstration, and full-scale development. Richard Nelson explains in his essay that the dynamics of the innovation process vary greatly by industry and that the relationships among companies, universities, and public institutions are complex, as is the progress of scientific knowledge and technology.

Nelson observes that Bush, in his zeal to protect research from political intervention, adopted an overly narrow description of basic research. Bush argued that "basic research is performed without thought of practical ends," but as Nelson shows, a good deal of scientific research has been clearly targeted at a set of particular applications. Nelson notes that "it is striking how the most successful industries in international competition have been drawing on exactly those areas of technology in which substantial targeted basic research has occurred."

Structure and Mechanisms for Research Support

The Bush Report envisioned a federal agency that would oversee all federal scientific research and use a system of large-scale grants to universities as the basic means of government support.

In his essay, David Mowery explores a number of ways in which the postwar history of scientific institutions has diverged from Bush's plan. In the first place, although the National Science Foundation was created in 1950, it never became either the central policy-making body for science or the largest funder of science projects. Since the 1950s, the Defense Department, the National Institutes of Health, the National Aeronautics and Space Administration (NASA), and the Department of Energy have all outspent NSF in funding basic science. Even today, although NSF is the most important patron of university research, it still accounts for only about 15 percent of all federal funding for basic research.

Moreover, rather than depending on large institutional grants to universities and allowing them to set internal basic research priorities (as envisioned by Bush), the hallmark of the postwar science grant system has been project grants to individuals and to teams of investigators. From today's vantage point, Mowery sees certain virtues in Bush's original recommendation. Mowery argues, for instance, that the lack of a powerful central agency to develop science policy and administration has meant that "postwar science policy has had a vacuum at its center, with little or no comprehensive review and assessment of the allocation of the federal basic research budget." This has resulted in "little apparent logic" to the current allocation.

The Private Sector and Innovation

Although the Bush Report refers in passing to the role of the private sector in promoting scientific advance and innovation, Murray Weidenbaum notes in his essay that the report "gave the private sector short shrift." That has become an increasingly grave omission, for while federal support dominated U.S. spending on research and development until the mid-1960s, today private sector support for R&D constitutes two-thirds of the U.S. total.

Given the central role of the private sector in the innovation process--even when the federal government supplies most of the funds--Weidenbaum states that Bush erred in not paying greater attention to broader governmental policies, such as education, tax, and regulatory policies, which exert a strong influence on the private development of technology in the United States. Weidenbaum gives the example of the "growing general recognition of the need to reform the government's revenue structure to encourage the saving and investment needed to commercialize the new technology resulting from scientific advance."

Of equal importance, in Weidenbaum's view, are the negative consequences of misguided regulatory policy. Weidenbaum cites specific regulatory barriers to innovation such as the rigid permit process established by the Clean Air Act; the Federal Aviation Administration's certification process, which inhibits new technology; and the continuing application of export controls on supercomputers, which undercuts U.S. companies in competition with companies from other nations.

Globalization and Multinational Enterprises

Raymond Vernon explores the implications for national R&D systems of emerging multinational enterprises and of the increasingly complex world of borderless strategic alliances. He notes that "implicit in the Bush Report is the idea that U.S.-based scientists would generate research in U.S. laboratories for use on U.S. territory by U.S.-owned firms." The reality has become quite different over the past four decades, as U.S. multinationals have blanketed the world with subsidiaries and foreign-based multinationals have established subsidiaries in the United States. By the 1990s, a third of the profits of U.S.-based multinationals came from foreign subsidiaries; and conversely, by the early 1990s foreign enterprises were spending almost $15 billion on U.S.-based R&D and were employing over 100,000 researchers in the United States.

Vernon argues that these new phenomena pose great challenges for national science and technology policies, not the least being a defense of national expenditures for R&D when the results will spill out over the entire world. "There is no practical way in which the fruits of this support can be channeled predictably to the United States," he asserts. Still, Vernon argues against policies that would attempt to dam up spillovers or to limit foreign investment. He rejects those as "quaintly futile gestures."

While stating that he has no "panaceas" for these dilemmas, Vernon concludes that "for a nation whose strength appears to lie in assimilating and integrating the diffuse efforts of others, the prospect of maintaining a relatively open climate for the exchange of technology ought not to be very daunting."

What We Don't Know

William Niskanen performs the role of skeptic in this volume, raising fundamental questions about the received wisdom from the Bush Report and about key premises of current U.S. science policy. He writes that while a number of economists have made their reputations by estimating the role technology plays in productivity and economic growth--about 50 percent of economic growth and 80 percent of productivity, by some calculations--"technology is one of economists' favorite code words for what they do not understand. All of these estimates of the effects of technology are residuals, estimates of the percentage of economic growth that economists cannot explain by the measured increase in conventional inputs."

Niskanen does not flatly oppose all government support of basic research. He concedes that private incentives "are probably not sufficient to induce the optimal level of R&D," but he finds that the current system of direct subsidy through bureaucratic decision making is flawed and inefficient. Niskanen concludes that the best policy for increasing private investment in R&D would involve the tax code, specifically, a tax credit for R&D expenditures. He would also make the credit refundable to aid start-up firms with no current tax liabilities. For universities, he would institute a system of matching grants from private firms and foundations.

Niskanen concludes by saying, "The outcome of these proposed changes in the public instruments to support R&D may be either a higher or lower total level of support: I do not know or much care. My concern is to make government support of science and technology more than a raid on the Treasury."