Harnessing Science and Technology for America's Economic Future: National and Regional Priorities (1999)

The National Forum was organized in the context of an ascendant U.S. economy. The buoyant outlook of early 1998 stood in sharp contrast with the mood of a decade earlier, when key U.S. manufacturing industries, such as automobiles and semiconductors, seemed to be failing in the face of daunting international competitive challenges.¹³ Although many factors fostered the turnaround, including effective national monetary policies, strong support for entrepreneurship in U.S. institutions and culture, and more focused management of U.S. industrial firms, robust technological innovation clearly has been a central contributing factor. One of the tasks of the forum was to look to the future and ask whether the U.S. technological resurgence is sustainable over the next 10 to 20 years and from whence tomorrow's technology will come.

Innovation in two broad, science-based industrial sectors has contributed to U.S. innovative success in the 1990s. The first is information technology, including semiconductors, computers, software, communications equipment, and information technology services. The second is the complex of industries that feed new technology into health care, including biotechnology, pharmaceuticals, and medical devices. Among the 50 U.S. firms with the largest research and development (R&D) budgets in 1994, the 20 with the highest ratio of R&D spending to

sales were all in either the information or health care sectors. The increasing focus of private-sector R&D investments in these fields is a powerful illustration of their promise.

As pointed out in chapter 2, the process of harnessing science and technology for economic growth is complex and not adequately understood. Many economically important innovations are either imitative or represent incremental improvements of current practice. For example, through a series of individually minor incremental changes during the first 60 years after the introduction of insulin, impurities were reduced from 50,000 parts per million in 1930 to 1 part per million in 1980. Other innovations constitute important discontinuities in which radical changes usher in new categories of products or services, such as the introduction of magnetic resonance imaging. Although the discontinuities often flow from new science and technology, they also can result from the creative combination of already-available technologies, such as innovations in express delivery.

A key comparative strength of the United States has been the ability to initiate and rapidly exploit innovative discontinuities that stimulate economic growth, by transforming existing industries or giving birth to new ones. Several important changes appear to have occurred in U.S. innovation during the 1990s that will affect economic and industrial performance in the future. First, a recent analysis of U.S. patents issued to inventors from all over the world shows a dramatic increase in the reliance of inventions on recent science (Narin et al., 1997). The trend is especially pronounced for U.S. inventions in the medical and chemical fields. A large percentage of the scientific citations in recent patents resulted from work in universities and government laboratories.

A second trend concerns corporate research.¹⁴ In contrast with government, which funds research to advance national interests or the missions of particular agencies, companies fund research to gain proprietary advantage. Corporate research laboratories first emerged about a century ago and flourished during the post-World War II period. The corporate laboratories of companies such as Du

Pont, AT&T, IBM, and Xerox grew to become important sources of fundamental technologies. Deregulation and the rise of global competition have led companies to put more focus on short-term results. One result is that investments in research, particularly longer-term or speculative research, have come under increased pressure and scrutiny. In aggregate, as Table 3-1 shows, the level of basic and applied research in industry declined by 20 percent in real terms between 1991 and 1994 and, despite recent increases, had not regained the 1991 level by 1997 (NSF, 1998). The changes have been extensive among the companies that had been most prominent in their fundamental research capabilities, such as IBM and AT&T. Although some newer companies, such as Microsoft, are boosting their investments in longer-term research, many newer information technology companies appear to focus their research exclusively on Edison's Quadrant (EQ) and near-term product development.¹⁵ Other firms that require access to fundamental research—such as Intel, Motorola, and Texas Instruments—are pooling funds to support work in universities, but they perform relatively little in-house research.

The United States appears well positioned to profit by continuing to push incremental technological progress, but where will tomorrow's radical discontinuities come from? Important innovations increasingly are characterized by an extensive research base and an R&D environment in which institutional flexibility is tolerated and even encouraged. Information technology and biotechnology (including pharmaceuticals), the two most promising fields for the future, display those characteristics, but institutional relationships and funding trends differ between the two fields. In biotechnology, extensive collaboration between universities, start-up firms, and larger companies provides fertile ground for radical innovation; the current situation and trends in this field are encouraging. In information technology, however, there are grounds for concern. As noted earlier, firms are increasing their R&D spending but appear to focus more on short-term

results. It is unclear whether the products of emergent new institutional relationships in combination with diminished effort within research labs of large companies will suffice to sustain the flow of radical innovations in information technology.

An assessment of U.S. prospects for building the science and technology foundation for future economic growth must recognize global developments and their implications.¹⁶ As noted in Chapter 2, the U.S. R&D enterprise, although still by far the largest in the world, accounts for a much smaller share of the world total than it did in the 1960s. Despite the growth of scientific and technological capabilities outside the United States and the growth of international R&D interdependence through investments by multinational corporations and other mechanisms, national policies and innovation environments are still important. National governments still provide a substantial share of R&D funding in most countries, particularly in support of key institutions, such as research universities. Governments increasingly are pressed to deliver tangible economic benefits of R&D investments to citizens. International linkages have resulted in closer and more complex relationships between trade policy, regulation, technology policy, and competition policy.

The terms of domestic U.S. debate about the desirability of international R&D have undergone a number of shifts over the years. During the 1960s, increased investments in offshore R&D by U.S. companies raised concerns over loss of employment and other technological opportunities associated with the domestic performance of R&D. In the 1980s, as foreign investment in the United States grew rapidly, critics argued that foreign firms were creating mainly low-wage, low-skill employment and were not locating high-value-added activities, such as R&D. In the early 1990s, as the U.S. R&D investments of foreign companies grew, concerns were raised that these investments were a means of "cherry-picking" the fruits of government-funded R&D.

Although trends in the internationalization of R&D activities are not easy to track because of inadequate data, it is possible to formulate several generalizations based on existing information and analysis. First, some components of the innovation process, including product development and manufacturing, are much more international than are activities aimed at technology creation. Although multinational corporations still largely generate their inventions and basic technologies in their home countries, they are more likely to develop and manufacture technology-intensive products by using the capabilities of foreign subsidiaries and strategic alliance partners.

A second generalization that can be supported by available evidence is that international flows of R&D investment are attracted to national or regional economies that can nurture specific technology-based capabilities. Even as the nationality of firms is blurred, the innovation environment in specific nations or regions, including supporting policies, becomes more important.

How do U.S. science and technology policies compare with those of other countries? A detailed look at U.S. policies is provided in Chapter 4, but a summary comparison of basic elements of science and technology policies between the United States and other developed countries shows some important similarities and convergences. In the United Kingdom, France, Japan, and Germany, the share of R&D financed by government has declined, as it has in the United States. The share of R&D performed by government laboratories also has declined in developed countries except Germany, and the German trend reflects the influence of reunification.

At the same time, U.S. science and technology policies remain distinctive in several important respects. For example, the share of R&D aimed at defense needs remains considerably higher in the United States than in other countries of the Organization for Economic Cooperation and Development, although the U.S. defense-related share has been declining as well. Note also that a significant amount of R&D spending by the U.S. Department of Defense supports broadly applicable work in fields such as computer science, materials science, and engineering. Within civilian-oriented government R&D, the United States directs larger shares of its funding toward health and space-related research and a smaller share toward research aimed at general economic development than other developed countries.

U.S. policies for the future must recognize that scientific and technological excellence will be distributed broadly throughout the world and that cross-border flows of R&D investment and technology will increase. For the United States to remain an attractive platform for R&D and related investments by U.S. and foreign-based firms, continued strong public investment in the R&D infrastructure will be required. However, U.S. policies also must be based on a realistic conceptualization of the sources of economic benefit associated with innovation.

Rather than restricting foreign access to the results of publicly funded R&D, it might be more productive to focus on improving the domestic adoption and implementation of new technologies from both domestic and international sources.

Forum participants wondered whether the tendency of national governments to invest in research with identifiable payoffs will lead to underinvestment in research in Bohr's Quadrant (BQ), or work aimed at advancing fundamental knowledge.¹⁷ Some have argued that expanded international cooperation could help to leverage scarce resources (Government-University-Industry Research Roundtable, 1998a). Although international collaboration at the scientist-to-scientist level often works well, large projects that require extensive coordination between national governments have had mixed results. Still, the international space station and other projects illustrate that cost and other pressures will continue to provide governments with strong incentives to seek international cooperation for large-scale research. After the forum, steering committee members suggested that much work needs to be done by the United States and other countries to build an appropriate institutional framework for expanded cooperation in international science and engineering research. Although this is an important task for the federal government in coming years, the committee is cautious about how much can be expected in the near term.

Excellence in R&D is a necessary but insufficient condition for creating wealth through the development of science- and technology-based industries. Nations and organizations must also possess mechanisms and infrastructure needed to transform science and technology into products and services that are competitive in global markets. At the forum, participants focused on an aspect of the innovative infrastructure in which the United States appears to be enjoying considerable success—fostering an environment that encourages the formation and growth of science-and technology-based companies.¹⁸

It was noted that the launch of high-technology ventures has been concentrated in specific regions, with Silicon Valley in Northern California being the most notable. Silicon Valley's strong infrastructure for creating and sustaining high-technology businesses has developed over many years. In addition to the role of Stanford University as a source of talent and know-how, several serendipitous events have contributed to the growth of Silicon Valley. Perhaps transistor coinventor William Shockley's decision to move to Palo Alto was a key

determining event. Several managers of Shockley Semiconductor, including Gordon Moore and Robert Noyce, left to start Fairchild. Fairchild spawned a group of start-ups, including Intel. Xerox's Palo Alto Research Center was an important source of talent and ideas for software-related start-ups in the 1970s and 1980s.

Several elements of infrastructure are essential to Silicon Valley's continued success. First is access to scientific and technological talent. Second is access to businesses and professionals (e.g., lawyers, accountants, executive search firms) that cater to the needs of start-up companies, in areas such as equipment leasing, legal help, and so forth. Third is a business culture that encourages people to strike out on their own. Failure is not welcome but is tolerated. In fact, venture capitalists seem more willing to invest in someone who already has failed than in a first-time entrepreneur.

The final element, which deserves some detailed comment, is the availability of financing for science- and technology-based ventures. Start-up financing is available from several sources in several forms. Organized venture capital is only one of these capital sources. Of the total amount of venture capital disbursements, about $10 billion in 1997 (National Venture Capital Association, 1999), roughly 60 percent goes to high-technology companies. The rest goes to other private equity investments, such as shopping centers and other real estate projects. Of the venture capital that goes into high-technology, perhaps half or less goes to support technology development. This amount, about $3 billion in 1997, is not even as large as the $5 billion R&D budget of IBM.

Another source of financing for start-ups is corporate investment. For example, Adobe Systems started a venture capital fund several years ago aimed at allowing it to make superior returns on its cash reserves while providing a window on new technologies; the fund has been relatively successful so far. There are also "angels," wealthy individuals interested in investing in start-ups, who often have achieved success as entrepreneurs themselves. Angels constitute an important and growing source of financing for start-ups.

Note that the venture capital industry historically has been highly cyclical. The high returns of recent years have attracted more investment capital, so more money has been aimed at roughly the same number of potential ideas and entrepreneurs. If history is any guide, this ultimately will lead to lower returns and less money flowing into venture capital. Also, venture capital might be more or less available in different regions or for companies in different stages of development (e.g., seed capital versus capital for expansion).

Nevertheless, today's environment of relatively abundant capital has provided opportunities for other regions around the country to build infrastructure for supporting science- and technology-based start-ups patterned on Silicon Valley's success. The Boston area long has been a fertile region for start-ups, and other well-known areas of high-technology activity, such as Austin, Texas, and the Research Triangle region of North Carolina, have been building the necessary

infrastructure for many years. Other emerging high-technology regions include Seattle, Washington; Salt Lake City, Utah; Northern Virginia; and Southern California.

The health of start-up activities is only one of several factors that will determine the U.S. high-technology future; maintaining and enhancing the positive environment that exists today will be important as well. A number of forum participants mentioned the importance of avoiding actions that could damage today's strong infrastructure and incentives to launch science- and technology-based ventures.

Four specific critical issues were discussed extensively at the forum. The first is the availability of scientific and engineering talent to fuel the growth of start-ups. Several recent reports by government agencies and industry associations state that there is a severe shortage of high-technology workers, particularly in information-technology. One way to address the problem in the short term is to increase the number of visas available for foreign scientists and engineers to work in the United States. Over the longer term, it might be necessary to expand the pool of Americans capable of filling these jobs if the United States is to remain an attractive location for high-technology activities. This issue is explored further in Chapter 6.

The second issue is the effect of securities litigation. The stock price of high-technology companies tends to fluctuate a great deal, and the risk for investors can be high, particularly for undiversified investors over short periods. If business results fail to meet expectations and its stock price falls sharply, a company can be vulnerable to securities-fraud lawsuits by shareholders. Recent federal legislation is seen as upholding the rights of shareholders to bring class-action lawsuits for genuine fraud while limiting the scope of less meritorious suits. However, suits increasingly are being filed in state courts, and this has led many in the high-technology community to call for legislation that would establish national uniform standards for securities class actions.

The third issue is intellectual property protection. For small science-and technology-based start-up companies, intellectual property can be one of the primary corporate assets. Particularly in the software and healthcare fields, U.S. firms often lose out on revenue because of various forms of infringement on intellectual property rights (IPR). Because markets outside the United States will represent the lion's share of new growth opportunities in coming years, intellectual property-intensive businesses have an interest in steps that increase the effectiveness of IPR protection around the world.

Finally, concern has been expressed about abuses in the civil justice system that have raised the costs of doing business and created impediments to product innovation through the application of science and technology expertise. Frivolous lawsuits and excessive punitive damage awards have made many technology-oriented businesses less willing to undertake cutting-edge R&D for fear of being sued unjustly by product users. The problem is compounded by the prevalence of

so-called "junk science" testimony offered by experts engaged by litigants in these often complicated proceedings. Several participants in the forum remarked that product liability reform legislation such as that passed by the Congress but vetoed by the president in 1996 would relieve some of the concerns of those engaged in these enterprises and further free up the productive forces that have made the United States a leader in efforts to capitalize in the marketplace on our science and technology resources.