Linking Trade and Technology Policies: An International Comparison of the Policies of Industrialized Nations (1992)

DAVID C. MOWERY

No observer of recent developments in the trade and investment relationships among the industrial and industrializing nations of the world can fail to be struck by the extent to which trade and investment flows, and the foreign and domestic public policies affecting them, now influence the technology development and investment decisions of U.S. firms. International and domestic collaborative ventures; R&D subsidies; technical standards; intellectual property protection; foreign investment decisions; ''dual-use" technology development, transfer, and support; technology transfer and "offsets;" and access to offshore sources of technological and scientific research, to name only a few issues, now are linked with trade policy and trade negotiations (both bilateral and multilateral) in complex ways. As this list suggests, technology-related issues now occupy a much more prominent place on the trade policy agenda of the United States and other industrial and industrializing economies.¹

This paper surveys several aspects of the challenges to domestic technology policy created by greater international economic and technological interdependence. Technology policy is difficult to define with great precision, since the innovative performance of an economy is affected by so many policies and influences. For purposes of this paper, it is defined to be the set of public policies that influence the creation, commercialization, and adoption of new technologies within an economy. This survey focuses primarily on the U.S. response to the linkage of trade and technology policies, although I make passing references to similar or contrasting responses in other industrial economies. There are two reasons for this focus: (1) For a number of reasons, these challenges have arisen with greater starkness and suddenness in the United States than in either the European Community

(EC) or Japan; and (2) Because of its large presence within the global scientific and technological (S&T) communities, as well as the size of the U.S. presence in world trade flows, U.S. responses to these challenges will have profound implications for the future of the global S&T and trading systems.

The growth in international trade and investment within the U.S. economy have created at least six issues for technology policy, many of which have also received considerable attention within the EC and Japan:

1. Equality of access to research facilities and results among industrial economies with contrasting domestic systems of research and industrial governance (the finance, ownership, and oversight of corporate organizations).

2. Developing technology policies that promote domestic economic welfare with minimally disruptive effects on trade and investment flows.

3. Improving domestic coordination of the formulation and implementation of trade and technology policies.

4. Adjusting to change in the technological and economic relationships between military and civil applications of "dual-use" technologies.

5. Adjusting to higher levels of foreign ownership of domestic technological assets (high-technology firms, R&D installations, etc.).

6. Adjusting to a "multipolar" world of sources of scientific and technological advances and inputs.

This list is not exhaustive, nor are all of the issues on it equally urgent in all of the industrial economies. Nevertheless, it frames an agenda for discussion at this symposium, and highlights many of the key issues in the U.S. domestic debate.

U.S. technology "policy" is the outcome of a series of loosely coordinated and often inconsistent decisions made in a broad array of federal agencies and policy areas. The most important characteristic of federal science and technology policy is that it has arisen out of the decisions and effects of policies (including procurement) designed to further the varied missions of numerous individual federal agencies, rather than being designed with any comprehensive economic strategy in mind.

The key elements of this system, many of which are unique among the industrial nations, arose during and after World War II. The loose coordination and weak central oversight characteristic of federal science and technology policy are attributable in part to the failure of Congress and the executive branch to agree on the structure and powers of a federal "science agency" to replace the wartime Committee on Medical Research and the Office of Scientific Research and Development, consistent with the recom-

mendations of Vannevar Bush's famous report, Science: The Endless Frontier. The National Science Foundation (NSF) was established only in 1950, well after the military services, the Atomic Energy Commission, and the National Institutes of Health had begun their own ambitious intramural and extramural science and technology research programs. The NSF budget was and is dwarfed by the research budgets of these other entities. No single agency, office, or committee in either the executive or the legislative branch of the federal government reviews the allocation, costs, and benefits of the entire federal R&D budget (including military R&D) on a regular basis. The interaction between science and technology policies and other policy areas (e.g., antitrust and trade policies) also are not reviewed on any but a sporadic basis.

Other key characteristics of postwar U.S. technology policy and the U.S. innovation system include the following:

1. Dominance of the large federal R&D budget by defense and related agencies: Throughout the postwar period, federal funds have accounted for a large share (45–60%) of total national R&D spending. The federal R&D budget has in turn been dominated by defense-related expenditures, the effects of which on commercial technologies and the firms marketing them were accentuated during the 1950s and 1960s by large-scale military procurement of components and systems in such areas as computers and microelectronics. In many instances, the combined effects of "spillovers" and military procurement supported the development of commercial applications.

2. The prominent role of U.S. universities as performers of research, especially basic research: A large share of the basic research performed in the U.S. economy, more than 50 percent in recent years, is carried out within universities (including federally funded R&D centers, FFRDCs). U.S. universities may well account for a larger share of total national R&D performance than is true of many Western European nations. Comparative statistics suggest that Japanese universities account for a larger share of that nation's total R&D investment (see National Research Council, 1989b), but the quality of these data and the quality of much Japanese university research suggest that the contribution of Japanese universities to basic knowledge is modest in many areas.

3. The role of new firms as agents for the commercialization of new technologies: Another important and unique element of the postwar U.S. research system is the prominent role of new firms as agents of technology commercialization in such technologies as computers, semiconductors, and biotechnology. Although in numerous instances, these firms commercialized technologies that drew on research performed within larger firms, universities, or government laboratories, the fact remains that their role in

this economy was significantly more important than in the postwar Japanese or Western European economies.

4. Minimal assistance for industrial technology adoption: With the significant exception of agriculture, the large federal investment in military and civilian research and technology development devoted virtually no attention or resources to support for the adoption of new technologies. In this respect, as Ergas (1987) has pointed out, postwar U.S. policy resembles that of France and the United Kingdom, both of which supported large defense-related R&D budgets, and contrasts with those of Germany, Sweden, and Japan, where a larger investment of public resources was directed to the support of technology adoption.

Postwar U.S. technology policy relied heavily on federal funding of scientific research, especially basic research, and on federal funding of development or applied research in defense-related areas. This de facto technology policy was further differentiated from those of many other industrial economies by the important role of federal procurement, primarily for defense purposes, and a tough antitrust policy during much of the postwar period. Both of these features contributed to the importance of small startup firms in technology commercialization. The U.S. research system exhibited enormous diversity in funding sources, performers, and objectives, and arguably linked scientific research more closely to postgraduate education than was true of other industrial economies during this period. The structure of the U.S. R&D system, like that of the R&D systems in other industrial economies, also was heavily affected by the structure of U.S. financial markets, firm structure and ownership, and corporate management and oversight.

Current challenges to U.S. technology policy are driven by the growing role of international flows of trade, capital, and technology to and from this economy, as well as by other developments. These developments have intensified international competitive pressures on U.S. firms and arguably have reduced the economic payoffs to U.S. firms and citizens from the large federal investment in R&D. Since little of the postwar federal R&D investment was motivated by economic objectives, this result should not be surprising. Nonetheless, growing demands for improvements in U.S. living standards and international competitiveness mean that the value of the federal R&D investment increasingly is being measured in economic terms.

One of the most important of the "other developments" mentioned above is the declining contributions of military-funded R&D and military procurement to civilian applications in many (not all) technologies. The celebrated

examples of semiconductors, computers, jet engines, and airframes in the 1950s and 1960s, technologies in which military R&D and procurement yielded important civilian applications, have few contemporary counterparts. Military procurement has declined as a share of total demand in many of these industries, and technologies and applications increasingly flow from civilian to military applications.

This change or reversal in the flow of technological spillovers also has an important economic component. Many U.S. defense suppliers of high-technology components and systems now are economically dependent on their fortunes in the civilian market.² This shift in the economic relationship between military and civil technologies contributed to the decision of the Defense Advanced Research Projects Agency (DARPA) to support the Sematech (Semiconductor Manufacturing Technology) and National Center for Manufacturing Sciences (NCMS) initiatives. Both of these research consortia, in which Pentagon research funds supplement contributions from private industry, focus on the development or improvement of civilian, rather than military, technologies and manufacturing processes.

Postwar federal science and technology policies were not designed primarily to aid the competitive capabilities of U.S. firms in civilian technologies.³ Nevertheless, U.S. firms benefited from the large federal R&D investment, because of the relatively slow pace with which the results of this investment moved across international boundaries and because many foreign firms were not well-equipped to apply these results quickly to commercial applications. Neither of these conditions now applies—scientific and technological knowledge move more quickly within the international economy, and foreign firms have dramatically improved their ability to apply advanced scientific or technological knowledge. In a number of industries, U.S. firms no longer dominate industrial practice or technological performance. As R&D costs and risks rise, access to foreign markets is increasingly important to the viability of many U.S. high-technology firms, even as these firms in many instances now derive a larger share of their components or advanced subassemblies from foreign sources.

Clearly, U.S. technology or trade policies that lead other industrial nations to retaliate by restricting access to their markets may be harmful to U.S. high-technology firms. Existing formal and informal restrictions on U.S. firms' access to foreign markets are equally harmful. As the tariff-based market access barriers of the earlier postwar period have been reduced through successive rounds of multilateral negotiations, they have been replaced by ''nontariff barriers," which greatly complicate trade policy-making and link it tightly to technology policy. Nontariff barriers are numerous, difficult to define and measure, and often involve instruments of domestic technology policy. The redefinition of the trade policy agenda during the past 20 years has propelled intellectual property regimes, anti-

trust policy, technical standards, regional development policies, and a broad array of other policies to a central position in bilateral and multilateral trade negotiations.

The importance of international trade and capital flows has expanded greatly in the U.S. economy during the past 25 years. The share of imports and exports within U.S. gross national product has doubled since 1965, a larger increase than either Japan or Western European economies have experienced during this period. More recently, foreign direct investment in this economy also has grown rapidly, although foreign investment in the U.S. economy remains lower than foreign investment in most Western European economies.⁴ The speed with which the U.S. economy has "internationalized" is central to understanding the U.S. domestic debate over technology and trade policies.

Along with a relatively open market for imports and foreign investment (albeit one that has faced significant demands for protection in recent years), the United States maintains a relatively open research system. The structural characteristics of the U.S. R&D system, with its high labor mobility, heavy reliance on university research for basic science and training, and the importance of small firms for technology commercialization, mean that access by foreign firms to U.S. scientific and technological advances is relatively easy. The R&D and industrial governance systems of many other industrial economies, however, differ considerably. The relative importance within many foreign R&D systems of "open" and "closed" research institutions, respectively universities and private firms, contrasts with that of the United States. Moreover, the institutions of industrial finance and governance in many industrial economies may make it difficult for U.S. or other foreign firms to gain access to industrial technologies or innovations through acquisitions of firms or intellectual property. As international economic integration advances, competition among firms increasingly takes the form of competition among different systems for the organization of research, finance, or industry, and these structural differences loom much larger.⁵

Much of this foreign challenge to U.S. economic and technological hegemony, of course, reflects the postwar reconstruction of foreign economies and technological infrastructure that was a central goal of postwar U.S. foreign policy. For much of the 1900–1940 period, one during which the U.S. economy exhibited strong growth in income and productivity, few observers would have characterized U.S. firms as either technologically or scientifically dominant. Nevertheless, the passing of the postwar technological hegemony of U.S. firms has occurred quickly. In many industries, U.S. firms now are first among equals in technological capabilities (or in others, well behind the state of the art), rather than dominant, even as the economic returns from the postwar U.S. R&D system are increasingly available to U.S. and foreign firms alike.

Moreover, for many U.S. firms, access to foreign science and technology is increasingly important to their competitive future. This factor, as well as the need to gain more rapid access to foreign markets and capital, and the political obstacles to market access or investment in many industries, has led U.S., Western European, and Japanese firms to broaden the international reach of their R&D operations.⁶ U.S. firms, for example, have formed strategic "alliances" with foreign firms that engage in joint development, manufacture, or marketing of high-technology products. These alliances, which have grown rapidly in number during the past 20 years, now span a wide range of nations and industries, including such high-technology sectors as semiconductors and commercial aircraft, but they have appeared as well in automobiles and steel.⁷ Similar alliances have grown between firms within the EC, in many instances with financial support and encouragement from EC technology programs. In many industries, joint ventures between U.S. and foreign firms have expanded simultaneously with growth in domestic research collaboration among U.S. firms and between U.S. firms and universities. These alliances are contributing to the accelerating pace of international technology transfer and to the internationalization of components sourcing, and these trends are in turn intensifying economic and technological interdependence between U.S. and foreign firms.

Although debate over appropriate responses to these new circumstances continues within the U.S. government and private sector, the 1980s witnessed important new developments in U.S. trade and technology policies. Along with other governments, the Reagan administration initiated several policies intended to capture more of the returns of publicly financed research for the U.S. economy. Trade policy became far more salient and politicized during the Reagan administration, as congressional involvement increased significantly. These actions tightened the interdependence of U.S. trade and technology policies, although they did not always result in consistency between trade and technology policies.

The Reagan administration entered the White House in 1981 with a pledge to remove the federal government from a major role in the commercialization of new technologies. In this view, the appropriate federal role in civilian technology development was limited to funding of basic research, commercialization of which was best handled by the market.⁸ The contrast between its 1981 posture and the administration's 1987 response to the demonstration of the phenomenon of high-temperature superconductivity or the formation of Sematech (the Semiconductor Manufacturing Technology consortium), is dramatic. In these instances, as well as in the National Science Foundation programs for university-industry cooperation,

Engineering and Science Research Centers, and other initiatives, the Reagan administration, with strong bipartisan support, proposed or implemented policies designed to increase the national economic returns to the large federal investment in basic research. Many of these initiatives have been continued by the Bush administration, and Congress has if anything pressed for more far-reaching action.

These developments represented a considerable shift from the rhetoric of 1981, and changed the historic posture of federal policies toward commercial technology development (outside of agriculture). Previous federal initiatives for commercialization generally were modest but, where significant, supported the commercial development of technologies for which market mechanisms and incentives were deemed to be lacking or insufficient. Examples include the commercial supersonic transport, coal liquefaction and synthetic fuels, "Project Breakthrough" in residential housing construction, and the liquid metal fast breeder nuclear reactor. These initiatives of the 1980s, however, indicate some movement in U.S. technology policy to a posture that more closely resembles that of the EC or Japan.

In contrast to the federal commercial technology programs of the 1960s and 1970s, the technology policy initiatives of the 1980s and those of the foreseeable future are designed to accelerate the commercial development of basic research advances for which the private returns are likely to be high, but only in the absence of faster commercialization by foreign firms. As a result, some of these initiatives have attempted to restrict the transfer of research results to foreign enterprises. Efforts were made, for example, to control foreign access to publicly funded research in high-temperature superconductivity in 1987 and 1988. The Sematech and NCMS consortia currently exclude foreign firms, and transfer of NCMS-developed technologies by member firms to their foreign subsidiaries is subject to restrictions.

The EC also inaugurated a series of regional "strategic technology" programs during the 1980s. In many cases, these programs replaced or supplemented the technology policies of member states (primarily France and the United Kingdom) that had relied on "national champions" during the 1960s and 1970s, often attempting to use defense and other public sector procurement policies to bolster these domestic champions.⁹ The EC programs of the 1980s, including ESPRIT and others, as well as the EUREKA program, invested large sums of public funds in "precommercial" research in information technologies, microelectronics, etc. With some important exceptions (IBM Europe has been allowed to participate in parts of ESPRIT and EUREKA), these programs have not welcomed foreign firms.¹⁰

Both the U.S. and European initiatives, of course, were in part responses to the perceived successes of Japanese technology development programs, supported by public and private funds, that produced some successes during the 1960s and 1970s in such technologies as very large-scale integration.

These programs also excluded foreign firms, including Japanese affiliates of non-Japanese multinationals. At least some more recent Japanese research and technology development programs (e.g., in high-temperature superconductivity), however, have been open to foreign participation.

Other U.S. technology policy initiatives during the 1980s strengthened intellectual property protection and reduced antitrust restrictions on collaboration in research. By one count, since 1983 the Congress has passed 14 laws (including the establishment of the Court of Appeals for the Federal Circuit in 1982) increasing protection for intellectual property owners. Improved international protection for intellectual property has also been a central goal of U.S. trade policy in bilateral negotiations, including the use or threat of Section 301, and in the multilateral Uruguay Round of trade negotiations. The 1984 National Cooperative Research Act, another U.S. response to the success of Japanese technology and industrial policies, reduced the antitrust penalties for collaboration among firms in precommercial research. In 1990 the House of Representatives passed a bill that reduced antitrust penalties against consortia that engage in production (the bill excludes from its coverage consortia involving firms with more than 30 percent of their equity owned by foreigners).¹¹

Just as U.S. trade policy frequently has been called on to redress the problems created by unbalanced macroeconomic policy,¹² the lack of a coherent U.S. technology policy has placed enormous demands on trade policy. The 1980s witnessed the occasional resort by U.S. policymakers, believers all in free markets, to trade policy to compensate for the absence of a politically unacceptable, "interventionist" U.S. technology policy. The proposal by the Massachusetts Institute of Technology to purchase a supercomputer from the U.S.-based joint venture involving Honeywell and Nippon Electric Company of Japan is an example of this phenomenon (the fact that U.S. policymakers treated the products of this U.S.-based joint venture as Japanese in origin also points out the complexity of determining the national origin of products in a technologically interdependent world).¹³ Threatened by the Department of Commerce with an investigation of dumping in supercomputers, MIT elected to postpone the procurement, instead seeking financial support from the National Science Foundation for a supercomputer research center that would involve U.S. firms and U.S.-based technology. MIT Provost John M. Deutch stated that "it became clear important elements of the federal government would prefer to see MIT acquire a supercomputer based on U.S. technology. Since the federal government would ultimately bear nearly all the costs of the machine through research grants to MIT, the preferences of the U.S. government must be seriously addressed" (Massachusetts Institute of Technology [1987]; See also Putka [1987] and Sanger [1987]).

The discovery of the technology policy potential of antidumping policy

is not confined to the United States. During the late 1980s and 1990s, the Commission of the European Communities has resorted to increasingly elastic and creative definitions of dumping as a means of inducing foreign-owned firms to locate more of their high-value-added manufacturing activities and R&D within Western Europe (Economist, September 9, 1989).

Although the increased concern of a number of recent U.S. technology policy initiatives with commercial development of the fruits of basic research investments arguably is a positive development, the mercantilistic flavor of many of them may have unfortunate consequences for U.S. firms. Proposals to restrict scientific and technological cooperation at the water's edge fly in the face of the growing interdependence of national R&D systems. To the extent that U.S. policymakers design technology policies that ignore the growing interdependence of U.S. and foreign scientific and technological research, both U.S. and foreign technological development will be hampered. Similarly, efforts by Western European or Japanese policymakers to restrict foreign access to publicly funded technology projects may have serious negative consequences for their domestic firms.

Proposals to restrict access to U.S. research facilities and findings also overlook the historic futility and ineffectiveness of such restrictions. Indeed, the reality of global technological interdependence is well illustrated by the recent decision of Texas Instruments, a major participant in Sematech, to enter a technology-sharing joint venture with Hitachi of Japan, presumably one of the major technological threats to the firms participating in Sematech.¹⁴ Efforts to impose strict limitations on international transfer or foreign participation attempt to deny the reality of this interdependence.

Restrictions on foreign participation in U.S.-based research consortia that involve significant public funds (in many cases from state, as well as federal, sources), will also complicate U.S. efforts to gain access for U.S. firms to similar consortia in other industrial economies (needless to say, this observation also applies to governments and firms in Western Europe and Japan). Paradoxically, the efforts of the U.S. and foreign governments to establish "closed" national technology development programs are contributing to the development of international alliances among large global firms that partially frustrate the aims of the "national" technology development projects.¹⁵

The recent Pentagon initiatives for the support of commercial technology development that have been spawned by the change in the civil-military technology relationship also pose a dilemma for U.S. trade policy. These programs undermine the basis for U.S. opposition to large-scale foreign

technology development subsidies such as those of the European Airbus program. Admittedly, Airbus (estimated to have consumed $12–15 billion in public subsidies since its foundation) is targeted more precisely on a specific commercial technology (indeed, on a set of commercial aircraft designs) than recent Pentagon programs, and Airbus subsidies support production as well as development.

The fact nevertheless remains that the Airbus program is driven in part by the desire of the participant governments to maintain military aerospace industries by supporting the participation of their national aircraft firms in a major commercial project. This justification is similar to one of the reasons for DARPA support of Sematech and NCMS. The development of similar U.S. programs places this nation's trade policymakers on a very slippery slope. If the difference between U.S. and foreign technology-subsidy programs becomes one of degree, rather than kind, the limits to foreign abuse of subsidies that are imposed by U.S. opposition and persuasion are likely to be eroded still further.

These recent U.S. technology policy initiatives also fail to address a fundamental competitive weakness. Both stronger protection for intellectual property and the relaxation of antitrust restrictions on collaboration in research or in production tend to favor the creation of new commercial technologies. Yet one of the most serious competitive deficiencies of U.S. firms is their slow adoption of new manufacturing technologies, an area in which U.S. firms appear to lag behind their Japanese, German, or Swedish counterparts. Policies supporting commercial technology creation may conflict with those aiding adoption.¹⁶

Further relaxation of U.S. antitrust policy, for example, could allow domestic consortia to acquire market power in the initial production of innovative technologies and thus discourage rapid adoption. This possibility is unlikely if U.S. markets remain open to imports. Nevertheless, protection of the U.S. domestic market from ''unfairly traded" imports combined with antitrust exemption for a domestic production consortium (e.g., in high-definition television) could create a high-cost, noncompetitive domestic producer and severely impede the adoption of its technologically sophisticated products within the U.S. economy.¹⁷ Similarly, intellectual property protection often represents a compromise between the interests of innovators and those of potential adopters.

A similar tension is present in the EC debate over policies to bolster the Western European electronics and information technology industries. The current strategy, which combines large-scale R&D subsidies with formal and informal restrictions on access to the Western European market by foreign producers, has raised the prices and lowered the quality of the information technology and microelectronics products available to Western European firms and consumers, slowing adoption.¹⁸ Throughout the industrial

and industrializing economies, policymakers should recognize that while the domestic production of advanced technologies yields important economic benefits, policies that restrict or penalize domestic adoption of these technologies deny an economy the benefits of their application in other industrial sectors. At least some recent U.S. and EC policy initiatives, inspired by the example of postwar Japanese industrial and technology policies, appear to overlook one of the central goals of these Japanese policies: encouragement for domestic technology diffusion.

Within the U.S. government, science and technology policy oversight and coordination within and between Congress and the executive branch are unequal to the task of reviewing strategic technology policy initiatives and coordinating these initiatives with the policy agenda in international trade negotiations. For example, despite the popularity of this claim by interested parties, the executive branch and Congress are not well equipped to evaluate the arguments that one domestic industry or another is "economically strategic" for U.S. living standards, competitiveness, and economic growth. Lists of "critical technologies" proliferate, but few of these lists' authors have analyzed the connection between their critical technologies and U.S. R&D policy, trade policy, or industrial structure. Agreement within the U.S. government on the U.S. position in the Uruguay Round negotiations on subsidies has been difficult to achieve, in large part because of the lack of a consensus among U.S. policymakers and private firms on the appropriate limits for R&D subsidies. During the planning of the Uruguay Round, U.S. government suggestions that a ''high-technology" negotiating group be organized also led to naught because the U.S. government could not reach consensus on the mandate for such a negotiating group. The list goes on.

Among other things, as was noted above, the development of a more comprehensive technology policy that included as one of its goals the improvement of U.S. competitive performance could take some of the pressure off of trade policymakers to compensate for a current de facto technology policy that (exaggerating only slightly) in some areas is perverse and in others nonexistent. Nevertheless, technology policy initiatives have important implications for U.S. trade policymakers, as in the case of Sematech, and for policymakers in the EC, who must address the consequences of recent antidumping actions and those of restrictions on foreign participation in EC R&D programs.

Policymakers in the EC, the United States, and Japan are well advised to tilt in favor of technology policy instruments that are less trade-distorting. Some forms of assistance for technology adoption (e.g., industrial exten-

sion, equipment leasing, cooperative research, or small-scale demonstration projects), for example, are likely to prove less disruptive to the multilateral trading system than targeted subsidies or government procurement policies that exclude foreign firms. Other instruments of technology policy that may be less trade-distorting include tax expenditures and other support for university graduate education and for on-the-job training of production and professional employees.

Efforts to restrict technology transfer from U.S. universities or firms to foreign entities may prove more harmful to the United States in its current position as "first among equals" than would have been the case two decades earlier, during the period of U.S. dominance. U.S. firms now are in a position to gain more from balanced international exchange of technologies and research than at any time in the postwar period. Restricting foreign access to U.S. research is likely to impair U.S. firms' access to foreign technology, ultimately eroding, rather than improving, their competitiveness.

Equality or reciprocity of access nevertheless remains an important issue for the trade and technology policy agenda. This issue involves international differences in systems of finance and corporate governance, as well as differences in the structure of national R&D and innovation systems, and therefore is a difficult problem to resolve. The "structural" origins of this problem also mean that it looms largest on the U.S.-Japan negotiating agenda. The following discussion can only sketch out some issues for further consideration.

As was noted earlier, foreign acquisitions of U.S. firms, specifically small, high-technology firms, are frequently cited as an important source of asymmetry in technology access, since such firms have few analogues in the Japanese or Western European economies. But the view that foreign acquisitions of U.S. start-ups result in a one-way "technology drain" to the foreign buyers may well be based on an unrealistic view of the nature of technological assets and the characteristics of the technology transfer process.

The key technological assets in many high-technology start-up firms are rarely embodied in patent or license agreements, which often convey limited coverage or control of these assets. Instead, the critical knowledge is often "tacit" (e.g., not codified in blueprints or other documents), consisting of know-how and other less easily transferred forms, and this knowledge is embodied in the firm's employees. In the vast majority of acquisitions of U.S. firms by foreign enterprises, these critical human assets do not leave the United States following the acquisition—instead, they frequently leave the firm, transferring their skills and know-how to other U.S. firms. As such, the putative "drain" of U.S. technology through foreign acquisitions of high-technology firms may be exaggerated.¹⁹ Nevertheless, the serious impediments to U.S. acquisition of firms in other industrial economies,

particularly in Japan, are not exaggerated, and will continue to create serious tensions until they are reduced or removed. Formal or informal restrictions on foreign investment in the United States, however, are difficult to square with the stated position of the United States in the Uruguay Round and elsewhere that restrictions on foreign investment should be removed.

A second important issue in reciprocal access to research concerns the role of U.S. universities, which play a more prominent role in the United States and global scientific enterprise than do universities in some other industrial economies.²⁰ Foreign access to U.S. university research raises two separable issues. The first concerns access by foreign firms to U.S. research that has been funded largely or partially by U.S. public funds, for example, through research grants or overhead payments to the academic institution. To the extent that the establishment of cooperative research agreements between U.S. universities and foreign firms allows these firms to gain access to research results without repaying these subsidies, they may be able to "free-ride." Just as U.S. state universities charge higher tuition to nonresident students, however, U.S. universities may be well advised to consider assessing higher overhead charges on foreign firms with whom they negotiate research agreements, in recognition of the substantial contribution of public funds to their research prominence.

The establishment of a multitiered fee structure for university-industry research agreements, however, does not resolve the perceptions of inequity that arise from the fact that the U.S. research system, in which universities play a prominent role and welcome foreign visitors and industrial research agreements, is in many respects a more accessible system. The structural differences between the U.S. and foreign research systems are such that a strict requirement of reciprocity in access to research facilities is either worthless or infeasible. Assurances by the Japanese government of complete access to Japanese universities, for example, may be of limited interest to U.S. firms, in view of the modest amount of world-class research performed in Japanese universities. A "results-oriented" reciprocity requirement that Japanese firms open their industrial research facilities to foreign researchers could impose a similar requirement on U.S. firms, and is scarcely likely to elicit the support of U.S. firms.

The structure of the U.S., Japanese, and Western European research systems may be converging in some respects, as Japanese quasi-public or university-based research institutes become more important performers of high-quality basic research (and as the level and the "spillovers" associated with U.S. military R&D spending decline). If this convergent trend is significant, access to relatively "open" Japanese research institutions may become more attractive to informed U.S. and European firms. This process of institutional change and convergence nevertheless is likely to move so slowly that this

issue of reciprocal access will remain very difficult for the foreseeable future.

U.S. technology policies can be criticized for their inconsistency with trade policy, their perverse effects on U.S. competitiveness, and their failure to take into account the realities of international technological interdependence. Nevertheless, the current debate in the United States will be powerfully influenced by evidence that foreign research establishments are being made accessible to U.S. firms and individuals and that markets for U.S. exports are being liberalized, and by agreements in the Uruguay Round that allow U.S. exporters to promote the benefits of multilateral liberalization to an increasingly skeptical domestic polity and Congress. The challenges of trade for technology policy are challenges that must be addressed by the U.S. and foreign governments alike.

Barfield, C. E. 1982. Science Policy from Ford to Reagan. Washington, D.C.: American Enterprise Institute.

David, P.A. 1986. Technology diffusion, public policy, and industrial competitiveness. Pp. 373-391 in The Positive Sum Strategy, R. Landau and N. Rosenberg, eds. Washington, D.C.: National Academy Press.

Destler, P. I. 1986. American Trade Politics: System Stress. Washington, D.C.: Institute for International Economics.

Economist. September 9, 1989. A gun that needs to get knotted. 82-85.

Ergas, H. 1987. Does technology policy matter?" Pp. 191-245 in Technology in the Global Economy, B. Guile and H. Brooks, eds. Washington, D.C.: National Academy Press.

Financial Times. March 27, 1991. Europe and electronics. 16.

Graham, E., and P. Krugman. 1989. Foreign Direct Investment in the United States. Washington, D.C.: Institute for International Economics.

Hayes, T. C. 1988. Developing a computer chip for the 90's. New York Times, December 23, 1988, C6.

Hudson, R.L. 1991. In Europe, electronics industry sends protectionist message to ICL. Wall Street Journal, March 27, 1991, B2.

Massachusetts Institute of Technology. 1987. MIT exploring linked supercomputer center. MIT News Office. November 5, 1987, p. 2.

Mowery, D.C., ed. 1988. International Collaborative Ventures in U.S. Manufacturing. Cambridge, Mass.: Ballinger.

National Research Council. 1989a. Learning the R&D System. Office of Japan Affairs. Washington, D.C.: National Academy Press.

National Research Council. 1989b. The Working Environment for Research in U.S. and Japanese Universities. Office of Japan Affairs. Washington, D.C.: National Academy Press.

National Research Council. 1991. Europe 1992: The Implications of Market Integration for R&D-Intensive Firms. Academy Industry Program and Office of International Affairs. Washington, D.C.: National Academy Press.

Nelson, R. R. 1984. High-Technology Policies: A Five Nation Comparison. Washington, D.C.: American Enterprise Institute.

Nelson, R. R. 1990. U.S. international competitiveness: Where did it come from and where did it go? Research Policy 19:117-132.

Ostry, S. 1990. Governments and Corporations in an Interdependent World. Washington, D.C.: Council on Foreign Relations.

Panel on the Impact of National Security Controls on International Technology Transfer. 1987. Balancing the National Interest: U.S. National Security Export Controls and Global Economic Competition. Committee on Science, Engineering, and Public Policy. Washington, D.C.: National Academy Press.

Pollack, A. 1990. Sony to share data at U.S. plant. New York Times, February 21, 1990, C1.

Putka, G. 1987. MIT cancels supercomputer plan, citing U.S. pressure to reject Japanese bids. Wall Street Journal, November 6, 1987.

Sanger, D. E. 1987. M.I.T., pressed by U.S., won't buy computer. New York Times, November 6, 1987.