CONTACT: Stanford University News Service (650) 723-2558

U.S. taxpayers in a pickle without Cold War motivation for federal R&D

STANFORD -- There was a silver lining to the Cold War: The free world got cheaper and better products as a result of U.S. taxpayer support of research into technologies that might be used to fight the evil empire.

Now that the war is over, American economic competitiveness is often touted as the mission for federal research, but it lacks the political urgency of the old national security mission, say economists Roger Noll and Linda Cohen in the forthcoming September 1994 issue of Scientific American. Noll is the Morris M. Doyle Professor of Public Policy at Stanford University, and Cohen is on the faculty of the University of California-Irvine.

"Our conclusion is that the United States has not yet found a politically workable and economically attractive means of encouraging technological progress," Noll and Cohen say. As a result, federal government support of research and development can be expected to continue the decline that began in the late 1980s.

Research and development are the engines of economic growth, but who pays for it and who benefits from it are questions that remain problematic for U.S. politicians, say the authors, whose case studies of failed and successful science research projects were reported in their 1991 book, The Technology Pork Barrel.

"Competitiveness is not a politically powerful substitute for the cold war in forging a durable, bipartisan coalition for supporting R&D," they say. In addition, political problems in implementing competitiveness- based programs "are likely to undermine the economic performance of the programs. Eventually that will further reduce political support for the programs."

Japanese experience

The Japanese sustained a massive research effort without a Cold War enemy to motivate them, Noll and Cohen say, but the Japanese model won't work for the United States.

"The key lesson from the experience of Japan is that although policies providing extremely high financial incentives for private investment [in research] will produce rapid economic growth, the economic benefits of that growth will not be widely shared."

By keeping out foreign competition and facilitating the formation of domestic industrial cartels, the Japanese government "produced a system that ranks first in the world in its fraction of gross domestic product invested in R&D. It has a higher rate of sustained growth than any other advanced, industrial economy, and it has consistently low unemployment," Noll and Cohen say.

"The other side of Japan's remarkable performance is that the ordinary Japanese citizen has a standard of living far below that of citizens in other nations with approximately the same per capita gross domestic product. Although on average Japanese employees work substantially more hours per year than American workers, the real purchasing power of the average annual take-home pay in Japan is only about 75 percent of that in America - a noteworthy fact considering the real value of wages for average American workers has not increased for 20 years. "

Most Japanese are not pleased with their economic system, the authors say. "The recent political upheaval in Japan was sparked by scandals among members of the ruling party, but it has been brewing for more than a decade. It is rooted in the ordinary Japanese citizen's dissatisfaction with being unable to attain a standard of living roughly equal to that in North America and Western Europe."

In the United States, new government R&D programs encourage the creation of research consortia, including government-subsidized joint ventures between universities and U.S. companies.

Collaborative research can be beneficial if it expands the technological base of an industry but allows each company to develop its own products and production methods in competition with each other, Noll and Cohen say. It's just as likely, however, to encourage cartels, which leads to monopoly pricing - a threat to the average citizen's standard of living.

Federal funding declining

Adjusted for inflation, the government's R&D spending has fallen 7 percent since 1988, and is no longer keeping pace with the economy, Noll and Cohen say. The political will to continue R&D support is likely to be eroded further unless a more broadly appealing rationale can be articulated, they say.

The new emphasis on economic competitiveness has caused two changes in how R&D programs are devised and run. "One change is greater privatization of the selection and results of research projects. . . . The other change is increased collaboration among American firms and research organizations."

These range from the extreme case of the Advanced Technology Program of the National Institutes of Standards and Technology, which purposely avoids spending on any technology that might be of direct use to the government, to the Technology Reinvestment Program of the Department of Defense and industry-specific programs, such as Sematech for semiconductor manufacturing, the Flat Panel Display Program, the Clean Coal Technology Program, the National Aerospace Plane Consortium and the Clean Car Initiative.

The federally funded national laboratories are also seeking joint research projects with industry through cooperative research and development agreements.

In each program, projects are proposed and managed by the private participants, and they are to retain the property rights to whatever technology results. Proposals for projects are routinely exempt from the Freedom of Information Act, so only government officials can review them.

Why the taxpayers should support such programs is not evident from economic research, the authors say, even though there is strong evidence that heavy spending on R&D pays off in a nation's higher gross domestic product.

Economic value of R&D

The relationship between economic growth and technological progress has been "exhaustively" examined by distinguished economists, Noll and Cohen say. Among them are Nobel Laureate Robert Solow of the Massachusetts Institute of Technology, the late Edward Denison of the Brookings Institution, Moses Abramovitz of Stanford University, Edwin Mansfield of the University of Pennsylvania, Richard Nelson of Columbia University and F.M. Scherer and Zvi Griliches of Harvard University.

"The main conclusions from their work are that more than half of the historical growth in per capita income in the United States is attributable to advances in technology," Noll and Cohen say, "and that the total economic return on investment in R&D is several times as high as that for other forms of investment."

The consensus among economists has been that the average U.S. citizen has benefited from subsidizing R&D because, under the system put in place during the Cold War, "most of the benefits of innovation accrue not to innovators but to consumers through products that are better or less expensive, or both."

Because the benefits are broadly shared, individual companies may lack the financial incentives "to improve technologies as much as socially desirable." The government needs to provide either incentives or direct research and development investment.

Political liberals and conservatives have disagreed only on the specifics of how the government should increase investment, Noll and Cohen say. Conservatives prefer to strengthen innovators' intellectual property rights - legal protections such as patents and copyrights - that allow the innovating companies to make higher profits from their innovations. Liberals prefer to have taxpayers pay for R&D through targeted programs.

Both approaches have drawbacks, Noll and Cohen say. The conservative approach "creates higher profits through the establishment of monopolies, which are inefficient. Second, any form of protection for intellectual property limits the diffusion of research results" so that benefits from applications in other products and industries are limited.

The liberal approach gives broad access to innovations but requires "elaborate, costly and inflexible" monitoring of public research projects. "As a result, R&D done under federal contract is inherently more expensive and less effective than R&D done by an organization using its own funds." Government cost accounting and auditing requirements are so burdensome, they say, that many federal contractors separate their federal and private work so they can use cheaper methods to manage their private research work.

Open research system

In the past, a political consensus for federal R&D was achieved by using both liberal and conservative strategies. Business was expected to fund its own research in areas where government could easily create strong intellectual property rights, and government supported work in targeted areas considered to be in the national interest. Even much of the defense-related research was broadly disseminated and commercially adopted, such as with computers, microelectronics and telecommunications.

"Historically, the U.S. government has contributed to university research and encouraged the open and free dissemination of the results. That policy - both directly and through the movement of students to industry or other universities has offered significant benefits. Recent studies by [Edwin] Mansfield [of the University of Pennsylvania] show that the economic payoff from fundamental research is higher in the United States than in other nations."

Proponents of the new research consortia programs want to limit the benefits to American companies and not allow non-American citizens to have a free ride, but domestic industry research cartels could actually lead to decreased innovation, Cohen and Noll say.

The problem is that firms in competitive industries tend to focus on technology that they can easily protect with patents or limit to the members' use. Sematech, for example, worked on developing new machines that would give its members a manufacturing edge and could not be used by others.

R&D projects that do not limit the technology in this way are not likely to get political support from business, the authors say, "unless, as was the case in Japan, they are accompanied by a domestic production cartel and strong trade barriers."

"The industrywide centralization of applied R&D therefore confronts domestic consumers with a Hobson's choice. If the venture makes U.S. industry more productive than its foreign competitors, the domestic industry will retain most of the benefits of its expanded productivity by cartelizing the domestic market. If the venture fails to make U.S. industry more competitive, the domestic industry will lose market share to foreigners, leading to the imposition of import restrictions. Once again, a domestic cartel emerges, but in this case one that is inefficient as well as monopolistic."

"In either case, the main effect of centralized R&D is to transfer wealth to members of a domestic cartel, not to promote the economic welfare of most citizens."

An alternative is to subsidize separate competing proposals from the same industry, as has been done with the Advanced Technology Program and has been proposed for the Flat Panel Display Program. Such programs, however, are politically unpopular when they are successful.

"When a project becomes successful, outside firms perceive it as unfair on the grounds that the government is interfering with the success or failure of companies in the industry." A Department of Energy program to develop supercomputing technology was shelved, for example, when competitors of Cray Research, the government laboratory's private partner, complained.

"More often, such programs are not very successful," Cohen and Noll say.

"To some degree, failures should be expected because the outcome of an R&D effort is inherently unpredictable. Unfortunately, technical and economic failure seldom lead to the timely demise of a major project." The failure of a government-industry project to meet its early goals often creates more political pressure on the government to absorb an even bigger share of the cost, they say.

Economic research and historical experience indicate a combination of government policies is better than any one policy.

Government could take a "directing role" in subsidizing both fundamental research and research aimed at broadening the technology base and making the results widely available. It can best encourage applications research through use of differential tax treatment for research investment but otherwise take a hands-off stance.

While such a scheme would work economically, it has significant political liabilities, Noll and Cohen say.

"Consumers would benefit the most, but most of the political support for R&D comes from industry, not consumers. And because the entities that conduct technology-based research cannot usually keep the results for themselves, U.S. firms would not be the only beneficiaries."




This is an archived release.

This release is not available in any other form. Images mentioned in this release are not available online.
Stanford News Service has an extensive library of images, some of which may be available to you online. Direct your request by EMail to newslibrary@stanford.edu.