National Science and Technology Strategies in a Global Context: Report of an International Symposium (1998)

__ From the presentations made at the symposium, it appears that a number of countries and regions are moving in similar directions with regard to national science and technology strategies. Harnessing science and technology for economic growth is often a key priority. At the same time, maintaining a long-term view for government investment priorities can be difficult, due to fiscal constraints. Players in the science and technology (S&T) system are challenged to do more with less, and to demonstrate the concrete benefits of investment in research and development (R&D). Developing effective mechanisms for cross-sectoral cooperation in R&D, and for linking the S&T community with aspirations of the public and the political leadership, are pressing tasks in many countries.

__ Nations and regions are converging from different directions in terms of defining public and private roles, and in setting appropriate time horizons. In Korea, for example, the government has played an important supportive role in providing technical infrastructure for industry. As Korean industry has become more globally competitive, the government role is shifting toward ensuring sufficient funding of applied R&D with longer time horizons, particularly in partnership with industry. In Mexico, national leaders recognize the need for expanded government-industry-university cooperation in order to strengthen the national S&T infrastructure needed for Mexican industry to compete in global markets.

__ Participants in the workshop recognized that international cooperation is a significant element in their various national strategies, but its importance and role differ according to the country. In countries that have pursued a focused and coherent national S&T strategy, such as Finland and Korea, developing the international component of that strategy appears to be straightforward. For some other countries struggling with fundamental domestic issues and debates, such as the United States and Mexico, there appears to be less policy focus on international cooperation, even as individual companies and other institutions increase their international linkages.

For the purpose of this activity, a "national science and technology strategy" may be defined as the goals, time horizons, performers, funders, and funding criteria that characterize a nation's science and technology investment portfolio. The roles of government, industry, and universities, and their relationships, are a particular focus.

An examination of international trends in science and technology policies and

performance reveals several important developments.² One long-term trend is a decline in the U.S. share of world R&D spending to an estimated 32.3 percent in 1994, down from about half the world total in the early 1970s. This decline in U.S. dominance corresponds with the growth of science and technology capability outside of the United States. A second long-term global trend that has important implications for science and technology is the movement toward more open, market-oriented economic policies.

One hypothesis that was tested and discussed at the conference is the significant movement toward convergence in national science and technology strategies. The elements of this emerging modal strategy might include reliance on the private sector for 60–75 percent of total R&D spending, a public spending focus on science and technology infrastructure needed for improved productivity and economic growth, and emphasis on building more effective collaboration between the government, university, and industry sectors.

Countries are approaching this model from different directions.³ For countries such as the United States and France, where the national government has traditionally provided about half or more of R&D funding, this has involved constraints on overall government spending and a relative decline in national security-related spending. For Korea and a number of the emerging Asian economics, where R&D has been highly private-sector driven, this has involved increased public sector funding and focus on improving key elements of the infrastructure, such as graduate science and engineering education. Improved inter-sectoral linkages and learning from foreign approaches are being pursued by many countries. A final global trend is the growing international flow of R&D resources. Education and training of foreign science and engineering students in the United States is a long-standing trend. In addition, multinational corporations are playing an expanding role in the globalization of science and technology through overseas R&D labs and related activities.

Mexico can be characterized as an industrializing country whose companies are not generally technology-based. We have experienced adjustment pains in moving toward a more open market system. Recent milestones include implementation

of the North American Free Trade Agreement (NAFTA) in 1993, and entering the Organization for Economic Cooperation and Development (OECD) in 1994.

Several elements of the social and policy environment in Mexico have a major impact on science and technology strategies. First, scientific and technological capabilities in the public and private sectors are heavily centralized geographically around Mexico City. Second, there is a strong contrast between the awareness of a small number of policy makers and middle class professionals concerning the importance of science and technology, and that of the large rural population living under marginal conditions. Third, domestic industry is dominated by small business, most of which is not S&T-based. Finally, Mexico's demographic distribution has a large fraction of very young people demanding education and new jobs. Table 2-1 shows a comparison of Mexico and the United States along some important dimensions of S&T policy.

Scientific research in Mexico is a very young endeavor. The first research centers were created during the 1930s in most disciplines. Institutionalized scientific activity following international standards was only consolidated in the 1950s when the National University (UNAM) moved to its new campus south of Mexico City. The Metropolitan University and the Research Center for Advanced Studies were established during the 1960s. Along with the National University, they account for over half of the basic research performed in Mexico today. Large universities account for about 19 percent of government investment in R&D.

In 1971, the National Council for Science and Technology (CONACYT) was created as a federal agency under the Ministry of Education charged with the establishment of national science S&T policies. Over the years, CONACYT has provided more than 90,000 scholarships for graduate studies in Mexico and abroad, and also runs a grant program for basic science projects.

A great deal of Mexico's R&D is performed at government facilities administered by CONACYT and other ministries. The CONACYT structure consists of 27 research centers in the natural sciences, social sciences and humanities, and technology development and services. Altogether, they absorb 28 percent of the federal investment in S&T. They are tasked with allocating resources based on objective criteria, increasing the quality of basic research, and encouraging innovation by promoting links with industry.

Federal investment in R&D activities is currently 0.38 percent of the GDP and the present administration has committed itself to reach 0.7 percent by the year 2000.

Historically, industry has enjoyed a privileged position as industrial development was defined by an import-substitution model. Mexican industry was a closed system, protected by strict import laws and trade barriers. A number of the largest enterprises have been nationally owned. Although this system encouraged the growth of domestic-owned industry, incentives to increase efficiency or innovate were not strong. With little need to exploit proprietary technologies, Mexican industry was slow to develop R&D organizations. Any new technological process required by local industry was acquired from foreign companies, often embedded in imported machinery and equipment. Multinational subsidiaries consulted their home offices for technological assistance.

The current industrial structure of Mexico is well described by a recent report of the OECD.⁴ On the one hand, there are a limited number of big companies operating in the mining, steel, cement, glass, petrochemicals, and metallurgical industry sectors; some are subsidiaries of big foreign corporations. On the other hand, there are many small companies with local markets. In between, there are some medium-sized enterprises and the so-called maquiladoras, wholly owned manufacturing subsidiaries of U.S. and other international companies, which are mainly located in the border regions.

Mexico began opening its markets to global competition during the 1980s, a process accelerated by NAFTA and entry into OECD. Domestic industry is now exposed to competition both at home and abroad. However, only a dozen or so large companies have their own R&D organizations. Mexican industry now faces

the urgent challenge of developing innovative capabilities, producing highly trained personnel, increasing investment and developing synergistic linkages with the academic centers.

Up to now, the Ministry of Trade and Industry has not been a major player in funding S&T activities. There is little coordination between CONACYT programs and this Ministry. The creation of a strong independent technology agency, such as the Science and Technology Agency in Japan, might work to fill this gap and build effective links between academic science and industry. This seems to be a fertile time to undertake new initiatives due to the recent privatization of state monopolies and the inflow of foreign investment. A new law on intellectual property was established in 1991, and the creation of the Mexican Institute of Industrial Property has favored imports, including technology. As a result, more intense technology-based competition is already evident in industries like telecommunications.

In 1989, the Presidential Science Advisory Council was created. Its members are the recipients of the National Science and Humanities Award. The Council is charged with producing policy recommendations to the President. Although the number of permanent members (more than 60) appears too large for effective action and coordination, in 1996 the Council signed an agreement with the Mexican Academy of Sciences and CONACYT to undertake several projects, including production of an inventory of advanced, applied science laboratories and personnel.

The Mexican Academy of Sciences has had a long tradition in conducting self-defined, independent studies in S&T policy. The Academy has collaborated closely with the National Academy of Sciences of the United States in recent years on complex, regional and global problems that require a multi-disciplinary approach. In 1995, the Academy published reports on the water supply from the Mexico City Metropolitan Area aquifers, the industrial requirements for technological modernization, and challenges and opportunities of science in Mexico.

During the last 15 years, several initiatives have been pursued to promote Mexican S&T. However, the country still faces challenges in educating the public on the importance of S&T and, more broadly, of knowledge itself as the most important source of wealth in the coming years.

In developing this plan, the government recognized that it faces important problems. These include the small scale of graduate programs, the relatively low proportion of graduate students (1.9 percent) enrolled in natural sciences programs, the relatively low proportion of faculty with graduate degrees, the concentration of basic and applied research in Mexico City (more than 55 percent of scientists and engineers work in the metropolitan area) and the lack of strong

linkages between universities and industry. In addition, government agencies in charge of R&D activities are not well coordinated.

The five-year plan has set some general policies to alleviate these problems. In addition to steps aimed at enhancing support for advanced research and attracting young people to scientific careers, this plan devotes considerable attention to international collaboration. Since a large number of Mexican scientists were trained abroad, there are strong personal links with foreign researchers and institutions. Indeed, access to large facilities in the United States and Europe has helped the Mexican S&T community to stay in close contact with current international problems and quality standards. The plan, therefore, aims at catalyzing these collaborations by creating new multinational programs for research. One positive goal is to promote international meetings that would focus on how to articulate university-industry links.

One of the main instruments that CONACYT and other federal funding agencies have put into effect is the financing of research projects through grants approved by peer review committees. Since 1989, CONACYT established that there should be no priorities in scientific or technological areas regarding research grant allotment. The main criterion for approval should be the quality of the proposal as judged by the committees, the rationale being that since the R&D community is still very small, Mexico needs top grade scientific and technological research work in all fields of knowledge. This decision was aimed at consolidating the strong, productive research groups in the country. Furthermore, the Support Program for Science in Mexico was also established with a loan from the World Bank of U.S. $500 million in 1993. Its main objectives are the replacement of obsolete equipment in recognized research laboratories and the enlisting of former Soviet Union scientists and engineers in research centers in Mexico.

These initiatives have had a positive impact, but also contribute to a lack of focus on several problems that require urgent attention, such as the need to place a higher priority on areas of knowledge that are weakly developed in the country such as computer science, communication technologies, and agriculture. Also, there are no explicit programs to advance entrepreneurial initiatives within the public higher education system. In fact, academic policies and SNI (see below) regulations act to discourage scientists and engineers from working with industry.

On the other hand, CONACYT and the public higher education system are doing a great deal in the area of S&T capacity building, particularly in human resources. Mexico is committed to doubling the R&D work force by the year 2000 (from the base level in 1993). There are more than 24,000 students receiving graduate scholarships both in Mexico and abroad. Important steps are being made to assess the quality of graduate departments. At UNAM, graduate teaching and

research in science and engineering was reviewed by mixed panels of experts from the Mexican Academy of Sciences and the U.S. National Academy of Sciences.

The National System of Researchers (SNI) is an imaginative program created in 1984. It was originally aimed at awarding fellowships to productive, full-time members of the S&T community, including social sciences and humanities. However, the economic crisis of the 1980s had a negative impact on faculty income and SNI became, in fact, a significant portion of the scientist's salary. Researchers qualify for the fellowships on the basis of scientific productivity and their training of students. The amount they receive is nontaxable, indexed to multiples of official minimum salaries, and graded according to level. To stimulate the decentralization of research activities, compensation at each level is higher for active scholars working outside Mexico City. SNI applications are evaluated by peer review committees whose members are distinguished Mexican scientists.

A useful byproduct of SNI is that it provides extremely valuable information on the state of science in Mexico. The information provided by individuals has produced a database that helps in monitoring the situation of Mexican research. There are currently 5,969 researchers in SNI and it is estimated that the country has around 22,625 scientists and researchers hired by universities, research institutions and industry.

In another example, over 20 years ago UNAM initiated a long-term project aimed at opening new research laboratories in different states. The most important projects were the acquisition of two research oceanographic vessels, one each for the Atlantic and Pacific, and the construction of the two-meter telescope in San Pedro Mártir, Baja California. This policy has been quite successful due not only to the benefits of bringing superior scientists and facilities to an otherwise average provincial campus, but because of closer contacts forged with local industry. For example, the technicians and engineers involved in the construction of the telescope's automatic control created a sophisticated instrumentation group that has grown and matured. Many technicians, seeking better salaries, have joined industrial firms throughout Mexico in what seems to be a crucial—though indirect—link between universities and private industry. An important future task will be to build on the insights of this model.

In conclusion, it is fair to say that in Mexico, the three legs of the government-university-industry triangle are evolving in positive directions and with more solid foundations than in the past. The main task for the future will be to make these legs more mutually reinforcing. Government needs to establish better coordination among its agencies and increase its investment in S&T, academic science needs to become more entrepreneurial and responsive to market demands, and industry has to face in a realistic way the need to acquire new knowledge to promote technological innovation. In addition, there is one obstacle that has to be overcome as a condition sine qua non. A healthy policy cannot be formulated without greater recognition of the long-term view.

Mexican S&T policies have traditionally taken the form of short-term projects during successive federal administrations. Decision makers in federal agencies want to obtain tangible results from their planning efforts before the expiration of their terms. Long-term projects are viewed as risky since there is no pledge that the next administration will support such programs and society in Mexico does not share a culture that would allow heavy public investment with acceptance that returns will only be seen over a 20- or 30-year period. Public universities are an exception: their autonomy to administer themselves allows them to plan for longer periods. This fortunate situation, combined with a closer collaboration or partnership with industry, represents Mexico's best hope for sustainable economic progress in the future.

As in other countries, Korean S&T policy has been influenced by economic, social, and political factors. A strong national consensus that economic development and industrialization would be the primary national goal has been the major reason for this coherence, although it has also meant that S&T policy is subordinate to economic policy. Korean S&T policy has gone through several phases over the past 35 years in response to changes in national development objectives. Per capita income has increased from $87 in 1962 to $10,076 in 1995, while real gross national product (GNP) has increased at an annual rate of about 9 percent per year over that time period. Korea is now the eleventh largest economy in the world, and is among the top five countries in the global production of electronics, semiconductors, ships, petrochemicals, and textiles. Continuous changes in industrial structure have made this impressive performance possible, and science and technology have made important contributions.

As rapid economic growth was achieved in the 1960s, policy makers recognized S&T as an essential element and took steps to build infrastructure. Through the 1960s and 1970s, the main focus was on building the technological capability needed to succeed in export oriented industries such as chemicals, heavy machinery, and shipbuilding. Since most of this technology was available from abroad, policies were aimed at providing the foundations for adopting and utilizing foreign technology. In the 1980s, the need to encourage domestic R&D activities was recognized, and several new national programs were established.

Korea's spending on R&D as a percentage of GNP increased from 0.81 percent in 1981 to 2.71 percent in 1995, the latter being a level comparable to other OECD members. This increase is particularly remarkable in light of the rapid growth in GNP occurring at the same time. Industry is mainly responsible for the increase. Korean industry supports 81 percent of the overall R&D effort. The number of researchers has also expanded dramatically over the past decade and a half. Korea has established several new institutions for advanced science and engineering education, but about two-fifths of Korean science and engineering doc-

torates are still earned abroad, primarily in the United States. Despite the rapid increase in R&D activity, the large deficit in technology trade (payments for licenses and royalties) shows that Korea still depends a great deal of foreign technology. The scope of basic research is still modest.

The Ministry of Science and Technology (MOST) was established in 1967 and has responsibility for formulating S&T policies and basic plans as part of a five-year economic and social development planning cycle. MOST is also responsible for coordinating the activities of the other R&D supporting and performing agencies. MOST does technology forecasting, supports research at government research institutes, universities and industry, and is responsible for "big science" and nuclear technology.

The Presidential Council on Science and Technology (PCST) and the National Science and Technology Council (NSTC) are two mechanisms utilized in the coordination process. PCST was created in 1991 to advise the president on developing blueprints for basic S&T policy, on creating networks and institutions to promote S&T, and on formulating national policies to utilize S&T to promote competitiveness. NSTC was created in 1973 as a cabinet-level body that monitors agency S&T plans and coordinates their implementation. NSTC currently has 18 members, including the prime minister, the minister of the Ministry of Finance and Economy (MOFE), officials from S&T agencies, and four private sector experts. MOFE plays a key role in the S&T budgeting process.

S&T policy in Korea operates under several laws dealing with promotion of various aspects of R&D. One notable recent law is the Special Law for Science and Technology, enacted in March 1997, which calls for the Korean government to push for a total R&D funding level of 5 percent of GNP by the early twenty-first century.

The Korean Advanced Institute of Science and Technology (KAIST) is the only university not controlled by the Ministry of Education. Established in 1971 as the Korean Advanced Institute of Science, the institute was established to fill the increased demand for scientists and engineers with advanced training. As of 1994, almost 2,500 students were enrolled in the bachelor's program, over 1,200 in the master's program, and almost 2,000 in the Ph.D. program. In general, Korean universities face challenges in maintaining top-flight research due to tight funding. The Engineering Research Center/Scientific Research Center (ERC/SRC) program was established to promote research in universities. There are now 21 ERCs and 17 SRCs.

The Korean Institute of Science and Technology (KIST) is a government institute set up in 1966 to meet industrial needs across the broad spectrum of applied technologies. KIST conducts feasibility studies, provides technical services for small-and medium-sized firms, and performs pilot-plant scale engineering studies. Now that Korean industry has become much stronger technologically, the role of government institutes is being reappraised. Rather than basing funding on lump-

sum budgets, institutes are moving to a project-based system. Eventually, it is expected that funding will move to a competitive grant system.

If Korea's S&T strategies of the 1960s and 1970s were mainly expressed in the formation of new institutions such as MOST, KIST and KAIST, strategies of the 1980s and 1990s are mainly expressed through a series of national programs. National R&D programs are public-private partnerships aimed at catalyzing industrial restructuring and movement into higher value-added industries through domestic innovation.

The National R&D program was established by MOST in 1982. The Highly Advanced National (HAN) Project was implemented in 1992 as an inter-ministerial program and a significant scaling up of national R&D efforts. The goal is to strengthen capabilities in strategic industrial technologies, both specific product technologies in areas where Korea can potentially capture competitive advantage in global markets, and fundamental core technologies necessary for the broader economy and improving the quality of life.

The government also has a number of tax and other incentives for private R&D investment. As illustrated by the shift in the share of Korean R&D financed by the private sector, from 30 percent in the early 1980s to about 80 percent today, industry plays the leading role in transforming science and technology into economic growth. Industrial R&D is highly concentrated, with the top five companies accounting for about 35 percent of the total.

In industries such as nuclear power generation and microelectronics, Korean industry has been very successful in pursuing a "middle technology" strategy. In these industries, a product niche was identified in which the technical requirements were appropriate for Korea (sophisticated but proven technologies where the fundamentals could be acquired from abroad and quickly mastered) and where Korean firms could quickly develop products that delivered value to a significant range of global customers. The leading position of Korean firms in the semiconductor memory business established in the late 1980s has led to additional opportunities in flat panel displays, which have a similar technological foundation. In nuclear power generation, Korea's focus on lower-cost reactors has opened market opportunities in a wide range of developing countries.

The focus of Korea's national S&T strategy in recent years has been to build on its recent successes and to lay the foundations for the capability needed to compete in the growth industries of the next century. Rapid growth in R&D spending, new programs, and institutions for advanced research and education, and efforts to forge new international linkages have been prominent.

It will be some time before most of these aggressive, long-term oriented initiatives can be accurately appraised. In the meantime, Korea's efforts to play a

greater role in international S&T are worth noting. One example is Korea's active participation and leadership in the S&T activities organized under APEC. As in Korea's overall S&T strategy, the private sector is playing a leading role in the globalization process. Leading Korean companies now have extensive manufacturing infrastructures, particularly in Southeast Asia and North America, but increasingly in China as well.

To begin with the big picture, it is clear that the policy environment for formulating national strategies and programs in science and technology was dominated by considerations related to the Cold War through the 1980s and even into the 1990s. Since that time, there has been a remarkable concentration of focus on economic growth as a policy priority, both in the United States and in a number of other countries around the world. Economic growth is a focus not only because increasing the standard of living is good in itself, but also because a strong economy gives countries the maximum flexibility to pursue an array of both foreign and domestic policy options.

A second important trend is that the global economic and competitive environment has changed enormously over the post-World War II period. During the 1950s and 1960s, U.S. companies were dominant and therefore often had an advantage in benefiting from new technology. During the 1970s and 1980s observers came to describe the competitive situation as a horse race among companies based in the developed "triad" of the United States, Europe, and Japan. Particularly at the end of the 1980s, Japan appeared to be in the lead in this race, with Europe second and the United States starting to trail badly. Today, this horse race looks more like a steeplechase with big jumps in it. In the 1990s the order of the horses has changed somewhat with the United States doing better in a whole series of industries. In addition, other "horses" have joined the race, such as Korea and Southeast Asia.

From the point of view of the U.S. government, we have traditionally focused research and development funding on defense and space, energy, and health, which are basic missions of the government. While funding basic research in these areas leads eventually to commercialization of new technology, it is a relatively slow, often haphazard process. What happened in the 1970s and 1980s is that other countries began to develop the infrastructure to absorb and utilize our basic research as well. Increasingly, it was no longer sufficient to do basic research in the United States in order to get a unique national advantage.

Also during this period other nations developed the capacity to rapidly commercialize technology and survive in a climate characterized by shorter life cycles for technology. This led to a policy climate in the United States favorable to efforts aimed at facilitating rapid commercialization of new technology. The differences

in the specific rationales for programs like the Advanced Technology Program or the Technology Reinvestment Project are not as important as the common motive of bringing people together to speed up the commercialization of research.

The statement of U.S. technology policy entitled Technology in the National Interest, which my office has worked on with other parts of the government, sets this context very well. Countries around the world are focused on economic growth, advances in technology are the single most important factor in creating growth, and capital will seek the best returns globally. The United States wants to attract the engines of economic growth to locate here and also promote growth in domestic industries by investing in assets that will stay here. These assets include trained and educated people, infrastructure, and a positive business climate that encourages people to develop businesses and to innovate in the United States.

We are all pleased about this policy statement, but when we look around we see that this is the same policy almost every other country in the world has. There has been an almost complete convergence. To illustrate, look at the number of countries that are in the World Trade Organization (WTO) and more importantly, look at the list of countries who are applying for membership. It includes all the former Communist bloc countries of Eastern Europe and others in the developing world. This is a real victory for the Western capitalist system.

The difficulty is that the position of the United States has changed. In the early 1950s we produced 40 percent of the world's gross domestic product (GDP). By 1970 the U.S. share had declined to less than one quarter of the world's GDP, about where it is today. The shift is even more dramatic with regard to R&D. In the 1950s and 1960s, the United States performed two to three times as much R&D as the rest of the world put together. Today, the rest of the world is doing about twice as much R&D as we are. This makes a dramatic difference when we contemplate policies that pretend to serve the U.S. interests by restricting the flow of ideas. This may work when we have all the ideas, but definitely will not when the rest of the world has good ideas too.

Most importantly, what we believe is happening, in a fairly crude simplification, is that as GDP per capita rises, countries have to spend a larger share of GDP on R&D to create value added jobs. For example, if average income is only $5,000 per year, creating new jobs does not require much new technology. If, however, you want to create jobs paying $30,000 or $25,000 a year then you have to invent something significant. What we are probably seeing is a progression of countries with growing per capita GDP that are expanding R&D. The technology policies and the programs countries adopt reflect to a large extent where they are in this process.

There are some anomalies. For example, the Republic of Korea is clearly spending what would appear to be a significantly larger portion of GDP on R&D than other countries with similar GDP per capita. Although some of that is military, I believe that it represents a deliberate attempt to move Korea dramatically

and dynamically up the ladder. This may represent a change in the rules. One possible worry for the United States is that we are not spending a high proportion of GDP on R&D compared with other countries with similar GDP per capita, and we look even worse when we take out military R&D.

So what does this mean for countries? To a first approximation, it appears that as countries move up this ladder they move through a first phase in which they develop the infrastructure that allows multinationals to operate. In the next phase, domestic policies shift toward helping domestic industries acquire technology. We have seen this in Korea for some time; we see it now in China and many other countries moving up the ladder.

But what happens, we think, is that by the time a country reaches a very high GDP per capita, say $15,000–20,000 dollars per year, it cannot compete as a low-cost manufacturing source with lower income countries. It really needs to invent something different. This is the point where you see a full fledged set of research and development initiatives and infrastructure, such as advanced research universities. And from my recent visit it appears that Korea is making a transition to this third phase well ahead of where it might have been expected to from its $10,000–11,000 GDP per capita. This includes efforts to bring Korean scientists and engineers trained in America back to Korea.

What we have seen in the United States from about 1980 is the emergence of a whole series of partnerships, including legislation to allow research and technology results developed in the federal government to flow to universities and ultimately to industry. The government began to realize it could take steps to make it easier for ideas to flow, in areas like antitrust exemption, funding collaborative research, and so forth.

As a result, we are seeing a paradigm shift. It used to be that government was primarily a customer for research, and effectiveness was evaluated by whether the agency funding the work achieved its mission. Industry's role was to act as a contractor to that end. We are increasingly observing a new paradigm emerge in which more attention is being paid to whether we create jobs in the civilian sector, whether anybody made any money, and whether we actually moved America forward along our mission for economic growth. Just as the role of the Department of Agriculture is not to make the agency rich but to help farmers, a similar attitude is spreading into the Department of Commerce. By interviewing the customers for these programs we learn that they work better precisely to the extent that the government listens to what industry says. The major message is that the government role should be to facilitate information transfer and links between universities, the entrepreneurial sector and larger enterprises.

The historical trends in public and private R&D spending are not as widely appreciated as they should be. U.S. government spending on R&D peaked at about 2.2 percent of GDP in 1966, due mostly to defense and the space race. Since that time it has declined almost continuously, with a small uptick during the build-

up in defense spending by the Reagan Administration. But over the last decade government R&D as percentage of GDP has fallen uniformly. Intriguingly, the private sector share has grown more or less continuously, so that now it stands at about 1.5 percent of the GDP compared with the government's 1 percent. So industry has become the dominant source.

Also, R&D as a percentage of sales by U.S. companies started shooting up in the 1980s from its historic level of just under two percent. This is very significant. What has happened is the increase is concentrated in two areas: information/communications/electronics, and pharmaceuticals/biotechnology. What we have is a transformation in the industrial research base of the United States that seems to have gone relatively unnoticed. Our models of the industrial research process which are based on the chemical and manufacturing industries represent an increasingly small part of what is going on. This shift has enormous implications. For example, we have continuing shortages of skilled people in information technology related areas. Companies are going to India and elsewhere to fill these jobs. Going forward, the players in the U.S. R&D enterprise, including universities, will need to be more agile in this dynamic environment.