Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 74
U.S.-Japan Strategic Alliances in the Semiconductor Industry: Technology Transfer, Competition, and Public Policy 10 Conclusions We believe that the alliance boom is clearly going to continue into the foreseeable future—until the computer and semiconductor industries pass their peaks of maturity and enter the stage of downward decline; even then, incentives to form alliances will remain strong. As long as the underlying forces identified in Chapters 3–5 are unchanged, the growth of American-Japanese alliances will not slow. Indeed, as foreign markets expand, commercial pressures to form alliances with companies in many different foreign countries will arise. The world semiconductor industry will see more alliances concluded between American and European firms, between Asian and American companies, and between European and Japanese corporations. In addition to the usual bilateral alliances, the number of trilateral and multilateral tie-ups is bound to increase. As sites of the two largest semiconductor industries, linkages between the United States and Japan will probably continue to constitute the bulk of international marriages. Alliances between U.S. and Asian companies will extend beyond the United States and Japan to include new players in Singapore, Taiwan, and Korea, which now have significant capabilities in semiconductor production. Just as the types of alliance have become more varied over time, so too will the number and nationalities of corporate participants. Strategic alliances will therefore continue to function as the most important mechanism for technology transfer in the semiconductor industry well into the next century.
OCR for page 75
U.S.-Japan Strategic Alliances in the Semiconductor Industry: Technology Transfer, Competition, and Public Policy EXAMPLES OF U.S.-JAPAN ALLIANCES: ASSESSING COSTS AND BENEFITS To assess the costs and benefits of alliances, it is necessary to look in some detail at specific examples. A number of examples of U.S.-Japan semiconductor alliances are treated in this report. In Appendix A, three cases are described in detail: Motorola-Toshiba, Sun Microsystems-Fujitsu, and Kubota's series of alliances with U.S. semiconductor and computer companies. Other examples are described less extensively in the section on asymmetrical pairings (Power Integrations-Matsushita, TriQuint's difficulty in forming a satisfactory alliance, and nCHIP-Sumitomo Metal and Mining) or symmetrical pairings (LSI Logic-Kawasaki Steel). Finally, in Chapter 8, examples of alliances are used to illustrate the trouble U.S. start-up companies in semiconductor equipment and materials have in attracting capital and the implications of computer industry trends for the semiconductor industry. Together, the cases and examples are representative of the broad range of partnering firms, motivations, and mechanisms that characterize U.S.-Japan semiconductor alliances. Motorola-Toshiba is a large U.S. company-large Japanese company alliance that includes a variety of mechanisms, such as licensing, a manufacturing joint venture, joint product development, and marketing cooperation. Sun-Fujitsu is a small (at the time the alliance was formed) U.S. company-large Japanese semiconductor company linkage that centers on a supplier-manufacturer relationship comprising licensing, consigned product development, and manufacturing foundry aspects. The Kubota case illustrates the variety of mechanisms that a large Japanese company can use to partner with small U.S. firms in support of an effort to enter the semiconductor and information industries laterally. The examples of asymmetrical alliances describe the considerations of U.S. start-up companies in forming alliances with large Japanese companies. Gaps in the representativeness of these examples are generally in areas that are important but have a limited number of examples. One fairly new type of alliance that is not included is licensing of Japanese product technology to U.S. firms. NEC and Mitsubishi Electric have licensed products to AT&T, but this pattern is still unusual. A large-scale collaborative product development effort is another type of alliance not covered in detail here, but Texas Instruments-Hitachi is perhaps the only alliance that fits this category.38 Large government-sponsored R&D consortia, such as Japan's Very Large Scale Integration Project and SEMATECH in the United States, 38 In liquid crystal displays, a type of semiconductor for all practical purposes, the DTI joint venture between IBM and Toshiba might also fit into this category. As a manufacturing joint venture, DTI is perhaps more complex than Texas Instruments-Hitachi.
OCR for page 76
U.S.-Japan Strategic Alliances in the Semiconductor Industry: Technology Transfer, Competition, and Public Policy are not discussed because international participation has been rare until now. However, U.S.-Japan collaboration through this mechanism is a possibility for the future, because the Japanese government has been opening consortia to foreign participation and is currently planning to launch several new semiconductor-related projects. A closer look at alliances that have failed would also have been useful. The historical background contained in the Motorola-Toshiba case study does describe the dead ends that Motorola has encountered, but a detailed treatment of the Motorola-Hitachi or Intel-NEC alliances and the reasons they ended in court might be illuminating in light of the fact that U.S. companies such as Digital Equipment, Hewlett-Packard, and Sun are bringing a new generation of RISC microprocessors to market and licensing them to a variety of Japanese and U.S. companies. It is generally quite difficult to get access to information about U.S.-Japan alliances that have failed, particularly those that are currently under litigation. What do these examples tell us about the costs and benefits to U.S. companies and industry of alliances with Japanese firms? Is the calculus different today than it was during the period prior to the 1980s, when American firms traded short-term income flows for technology that enabled the Japanese to build a world-class semiconductor industry? It is safe to say that the calculus is much more complicated today. A review of the experiences contained in our examples suggests conclusions about the overall costs and benefits from the U.S. perspective. Costs Costs of U.S.-Japan alliances can be divided into those incurred by the company that forms the alliance and those incurred by industry, including the upstream and downstream industries for which semiconductor companies are important customers and suppliers, as well as semiconductor manufacturers themselves. For individual companies, perhaps the biggest potential cost is creating a formidable competitor through technology transfer. For U.S. industry as a whole, the loss of semiconductor manufacturing capability and infrastructure resulting from a pattern of alliances may be the most serious potential cost. Transferring Enabling Technology: Perhaps the most important technical area of the semiconductor industry in which Japanese firms are still behind the United States is the design of software codes that are embodied in advanced microprocessors. Possession of this capability would allow Japanese companies to challenge U.S. industry in its most important stronghold, and would create new competitors for a wide range of U.S. semiconductor and systems companies. The Sun-Fujitsu alliance, the Motorola-Toshiba alliance, and the Kubota-MIPS alliance involved the transfer
OCR for page 77
U.S.-Japan Strategic Alliances in the Semiconductor Industry: Technology Transfer, Competition, and Public Policy of microprocessor technology. In these cases it has been a number of years since the transfer took place. The Japanese partners have yet to forge ahead and become world leaders in microprocessors. It appears that in general, Japanese firms are still working to acquire the necessary software skill base. As described in Chapter 8, microprocessing unit markets are affected by the complex interaction of compatibility, the need to have a base of applications software that can be run on an MPU, and the possibility of technical discontinuities that can quickly change the shape of particular markets (RISC). Japan's underdeveloped personal computer market is an obvious disadvantage to acquiring a critical mass of skills. The creation of competition in this area appears to be only a potential cost for U.S. companies today. However, industry trends toward open systems may reduce the importance of proprietary MPU architectures (see discussion of downstream trends in Chapter 8). If this is the case, then the lack of capability to design superior MPUs would be less of a hurdle for Japanese companies seeking to compete in systems markets, particularly if they control the manufacture of critical components and possess sufficient marketing capability. The MPUs licensed by Sun and MIPS to Fujitsu and Kubota have allowed the Japanese partners to build manufacturing and marketing capability in advanced work-stations and supercomputers. Thus far, the critical importance of MPU designs has enabled the U.S. partners to ''manage" these relationships, but there is no guarantee that Japanese firms will fail to acquire the skills necessary to design MPUs or that MPUs will not decline in importance—to the benefit of Japanese companies—in the future. Transferring Incremental Technology: The costs incurred from transferring incremental technology are less severe than those incurred by transferring enabling technology. There are no examples in this report of such costs having actually been incurred. Kubota's investment in C-Cube Micro-systems and their joint development of video compression technology is, however, an example of technology transfer in an area that may have large potential in the long-term, potential that Kubota is now positioned to exploit. From the U.S. standpoint, transferring the capability to develop emerging markets is an opportunity cost, yet assessing the risk in advance is inherently difficult. The actual cost depends on the technology or product achieving greater market success than predicted at the time it was licensed or acquired. Some American critics of U.S.-Japan semiconductor alliances say that they are partly responsible for progress made by Japanese companies in markets such as gate arrays and ASICs. Perhaps the most notable example from the past that fits this general category is liquid crystal display (LCD) technology, which was developed in the United States but which Japanese companies continue to develop and commercialize. The potential market for displays in all sorts of portable information processing and consumer
OCR for page 78
U.S.-Japan Strategic Alliances in the Semiconductor Industry: Technology Transfer, Competition, and Public Policy electronics products is enormous. The Power Integrations case is a good example of how it is increasingly possible for U.S. companies to control their technology and avoid this risk, at least in areas where the potential for a technology can be reasonably gauged. Because it has a good patent position and the courts are more inclined to protect intellectual property rights than they have been in the past, Power Integrations was able to structure an alliance in which it receives a royalty from Matsushita for internal use of the power management technology. Matsushita may be able to develop improvements on this technology that it could then utilize, but Power Integrations can keep Matsushita from selling these improvements outside (or license and incorporate them into the product) as long as the original patents remain in force.39 Power Integrations takes some risk, but has a better chance of being able to sustain product development and stay in the race for incremental improvements of its technology. Low Return on Resources Expended: Low return on expenditures is a short-and long-term cost that has been incurred by U.S. companies trying to break into the Japanese market. Despite the long-term organizational learning benefits described below, the Motorola-Alps joint venture is a clear example of costs incurred in attempting to break into the Japanese market. In that case, Motorola attempted to use an alliance to enter the market and avoid enabling a long-term competitor or incurring the technology transfer-related costs outlined above. Motorola finally had to give up technology to Toshiba in order to gain significant access to the Japanese market, and it has tried to structure the relationship to minimize the risk of long-term costs. The difficulty of selling in Japan has imposed long-term costs on the American semiconductor industry and may still discourage U.S. firms from investing in Japanese marketing and the technology development capability that will be necessary for long-term growth and survival. Unsuccessful Licensing Alliances: Unsuccessful licensing alliances occur when a Japanese company uses the technology in a way that makes the American company believe the agreement is being violated. Even apart from the issue of fault in a legal sense, it is safe to say that the licensing firm in this situation regrets granting a license and typically sues. This is a short-term technology risk as opposed to the first three costs because licensing agreements that have succeeded over several years are less likely to turn acrimonious. The Motorola-Hitachi case is referred to in this report, but there is no detailed treatment of this type of case here. Two additional points should be made: (1) This risk can be incurred in licensing to Amer- 39 Significant differences in the patent systems and corporate approaches to intellectual property protection in the United States and Japan (such as patent "blanketing" in Japan) must be kept in mind constantly by small U.S. companies with limited resources to expend in litigation.
OCR for page 79
U.S.-Japan Strategic Alliances in the Semiconductor Industry: Technology Transfer, Competition, and Public Policy ican as well as Japanese companies; U.S.-U.S. suits over microprocessors currently appear to be more common than U.S.-Japanese suits. (2) From the perspective of an American company, being in a position to license and then bring suit is preferable to not being able to enforce intellectual property protection at all, as in the case of the Texas Instruments basic patent for integrated circuits which was not granted in Japan until many years after the application. Semiconductor Specific Opportunity Costs: There are long-term opportunity costs other than those related to particular products, which are covered in item 2. These costs can be assessed in advance, to some extent, and are incurred by individual companies and by U.S. industry as a whole. For example, when a fabless start-up decides to consign manufacturing to Japanese companies rather than build internal manufacturing capability, it passes up experience in making devices that could bring greater technical independence and could facilitate entry into other product lines. At an industry level, lost demand for semiconductor equipment and all the inputs related to process technology may result in the disappearance of U.S. manufacturing infrastructure if this is a general trend. For a multinational corporation like Motorola, assessing the benefits and costs from corporate and industry standpoints is more complex. For the company, building process-intensive devices in Japan with Toshiba means that it gains some of the benefit of that experience for use in U.S. operations. This experience and increased access to the Japanese market put Motorola in a better position to survive as an integrated company, which will support U.S. managerial and technical jobs. If Motorola takes purposeful action to use U.S.-based equipment makers in its Japanese fabrication plants, it may largely avoid the negative implications for the U.S. manufacturing infrastructure of producing advanced devices in Japan. Foregone Synergies: There are also opportunity costs that are more difficult to quantify because the impact goes beyond semiconductors. By definition, these costs are incurred by systems companies and industries. There are no clear-cut examples in this report, but one might speculate that if Sun had built a semiconductor manufacturing facility, it would have the incentive and the capability to develop components besides microprocessors that add value to its systems, which would constitute a competitive advantage in the future. Just as it is difficult at this point to assess the long-term benefits of Japanese capability built through linkages, it is difficult to assess the costs of U.S. manufacturing capability foregone through alliances. We can clearly see the long-term costs incurred by the U.S. economy and by upstream industries such as semiconductors because of the American exit from much consumer electronics manufacturing during the 1970s. If information processing systems take on more consumer electronics characteristics, as
OCR for page 80
U.S.-Japan Strategic Alliances in the Semiconductor Industry: Technology Transfer, Competition, and Public Policy discussed in Chapter 8 under downstream trends, the costs of consigning semiconductor manufacturing will likewise come into better focus. Technical Dependence: Technical dependence is related to foregone synergies, but represents an extreme case. Reliance on a supplier or small group of suppliers for a critical component incurs the risk of supply cutoff or price gouging. Again, it is systems companies like Sun that are mainly vulnerable. Fujitsu did a great deal of custom design work on the SPARC chip, but in that specific case, Sun added the most significant value to the chip, so adequate IPR protection and competition among licensing partners safeguard Sun. In the short term, American systems makers are more vulnerable in areas such as LCDs and DRAMs where U.S. manufacturers are weak. In the long-term, concentration of the most advanced manufacturing capability in Japan might give Japanese foundries more bargaining power visà vis companies like Sun. Lost Political Independence: Companies involved in alliances may find themselves constrained in terms of political action. A hypothetical example would be one in which a U.S. company involved in a U.S.-Japan alliance takes a position on policy matters that it would not otherwise take because of a desire to maintain good relations with the Japanese partner. Balancing the interests that relate to the alliance and those that relate to U.S. industrial and national well being may be problematic in some cases. Another aspect of this issue is not illustrated in any of the case studies, but would impact systems makers. To the extent that semiconductor alliances consign manufacturing to Japan, there is a risk that technical dependence as outlined in item 7 would serve to constrain the political positions taken by U.S. Companies. Benefits The benefits that accrue to U.S. companies and industry as a result of linkages with Japanese companies can be largely characterized as access to resources and capabilities that allow American companies to bring products to market more quickly and effectively. Capital for Survival: Some Kubota alliances may have ensured the survival of a small, financially weak U.S. partner. It is difficult to determine whether companies would have survived even without Japanese backing, and the Stardent bankruptcy indicates that even the deep pockets of a Japanese partner do not always prevent failure. In many cases, a U.S. start-up firm with superior technology can survive without a Japanese alliance. In the case of a U.S. start-up that would not have survived without Japanese backing, whose products do not find a significant market, benefits to U.S. industry and the economy are likely to be minimal. It may be that "capital
OCR for page 81
U.S.-Japan Strategic Alliances in the Semiconductor Industry: Technology Transfer, Competition, and Public Policy for survival" benefits typically accrue to the stockholders and management of the U.S. firm. Leverage of Resources for Development and Growth: Sun-Fujitsu is a perfect illustration of the category of benefits that can accrue to American firms when an alliance is well structured. Sun was able to bring its work-station to market faster as a result of linking with Fujitsu. In this instance, the timing and access to manufacturing capability were critical, and Sun essentially leveraged Fujitsu's capabilities to build a leading position in a rapidly growing market. Such success is unusual, and requires superior technology and skills from both the American and the Japanese partners. Linkages between fabless U.S. companies and Japanese foundries yield these benefits when they are successful. In the short term, access to these resources (including investment capital and foundry services) allows the U.S. partner to build a company. This also means that U.S. entrepreneurs and engineers continue to focus on opportunities in semiconductor design, to the benefit of the engineering and chip design infrastructure in the United States. The long-term viability and contribution of design companies that do not manufacture are more difficult to assess. Leverage of Investment Resources: Linking with Japanese partners sometimes helps larger U.S. companies leverage investment in fabrication facilities. This is largely a short-term benefit. The LSI Logic-Kawasaki Steel alliance falls into this category. For the U.S. company, the benefits of access to a world-class fabrication facility at a cost lower than it would have paid on its own are considerable. From industry and national perspectives, the benefits are less clear. The increased viability of the U.S. firm is one benefit. When the fabrication line is built in Japan, a U.S. partner may be able to contribute toward maintaining the U.S. manufacturing infrastructure by linking with U.S.-based equipment manufacturers. Leverage of Technical Resources: American partners also benefit when Japanese partners bring technical resources to the alliance. The design work that Fujitsu performed for Sun and the DRAM process know-how Motorola receives from Toshiba illustrate this benefit, but the Hitachi-Texas Instruments alliance may be the best illustration. For the 16 megabit DRAM the firms exchanged designs, and for the 64 megabit generation they have agreed to develop a common design. To some extent the leveraging of technical and investment resources is closely related; in the Sun-Fujitsu case these benefits overlapped. These benefits are short-term in nature, but if an alliance is managed properly and the U.S. partner maintains its technical and other capabilities, continuation of the alliance and long-term benefits are possible. Access to the Japanese Market: A short-term benefit can be turned into a long-term plus through the sustained effort of the U.S. partner to produce superior technology and to develop the capability to market in
OCR for page 82
U.S.-Japan Strategic Alliances in the Semiconductor Industry: Technology Transfer, Competition, and Public Policy Japan. Sun gained better access to the Japanese workstation market by linking with Fujitsu. Motorola has gained better access to the semiconductor market by linking with Toshiba. The Power Integrations-Matsushita and nCHIP-Sumitomo Metals alliances show promise for delivering market access to the U.S. partner on better terms than would have been possible 10 years ago. These alliances demonstrate that it is often necessary to trade technology for access, but a presence in the Japanese market is increasingly essential for long-term survival and growth in the semiconductor industry and in the upstream and downstream industries to which it is linked. Freedom to Focus Resources on High-Return Activities: The Sun-Fujitsu case illustrates another benefit: rather than invest in semiconductor manufacturing, Sun was able to focus on building marketing resources and on designing the next-generation SPARC chip, which will be on sale soon. The opportunity costs and risks associated with not manufacturing, which are covered above, must be balanced against this concrete, short-and medium-term benefit. Organizational Learning: Learning accrues to U.S. companies that use their alliances to build a capability for extracting benefits from their technical, marketing, and manufacturing activities in Japan, whether or not these activities specifically involve alliances. Motorola is a good example. The experience that Motorola has gained at considerable expense through its alliances and other long-term efforts can be utilized in the service of long-term strategic objectives. The benefits of possessing this capability can extend beyond the semiconductor industry. For example, Motorola's telecommunications equipment business is affected by government regulation, and knowledge of Japanese government operations and key officials will be useful to the company in advancing its interests in Japan's regulatory and policy spheres. Evaluating these costs and benefits in individual cases is an extremely complex undertaking, but several generalizations can be drawn from the examples covered in this report. From the U.S. perspective, the costs of U.S.-Japan semiconductor alliances are largely—but not exclusively—long-term, potential, and difficult to quantify. Further, if the most serious costs were incurred, they would fall on actors and interests external to the company forming the alliance and would be borne in a manner that is impossible to predict. By contrast, the benefits from U.S.-Japan alliances are often more immediate, concrete, easy to quantify, and directly appropriable by the U.S. company that forms the alliance. American-Japanese alliances, like other business relationships, generally involve both risks and benefits, and require continuous adaptation to new circumstances. Over time, the "terms of trade" in semiconductor alliances appear to be improving for U.S. participants. It is now possible to gain valuable resourc-
OCR for page 83
U.S.-Japan Strategic Alliances in the Semiconductor Industry: Technology Transfer, Competition, and Public Policy es, such as investment capital, access to the Japanese market, and manufacturing services and know-how. The key factors that contribute to greater bargaining power for U.S. companies are (1) competition among larger firms seeking an alliance and (2) strong intellectual property protection. In making U.S. investments, Japanese companies have consciously steered away from the strategy of leveraged buyouts and hostile takeovers. Self-interest and the near absence of hostile takeovers in corporate Japan explain this circumspect approach; engaging in leveraged buyouts and hostile takeovers would surely ignite outrage in the United States. Although the record of the past illustrates significant costs associated with U.S.-Japan strategic alliances, the context is changing and the examples in this report illustrate that in some cases, tangible benefits accrue to both sides. It will be important for U.S. policymakers, the American semiconductor industry, and other analysts to continue to monitor the mechanisms and impacts of U.S.-Japan strategic alliances, to assimilate experience, and to consider changes in public policy and corporate strategy where appropriate. The U.S. government and industry need to create an environment in which American firms have sufficient bargaining power in forming alliances, and the necessary information base must be developed to take full advantage of the possible benefits. SEMICONDUCTORS AS A STRATEGIC INDUSTRY Strategic alliances constitute an important—perhaps the most important—mechanism for semiconductur technology transfer between the United States and Japan. Although these alliances are normally seen as instruments of technology transmission, they also reflect deep-seated forces at work within the two economies. We believe that it is unrealistic to expect alliances to correct structural deficiencies in America's own manufacturing infrastructure (even though U.S. companies ought to insist on a reverse flow of manufacturing know-how from Japan to strengthen the U.S. infrastructure); nor is it realistic to expect alliances to change the short-term time horizons of U.S. management. Strategic alliances are manifestations of underlying conditions, not the root cause of what ails the U.S. semiconductor industry. In part, the erosion of U.S. manufacturing capabilities is a reflection of changing comparative advantage and the attractiveness of shifting into the higher value-added segments of the semiconductor industry—specifically into software, microprocessors, and customized or semicustomized areas. Whether it makes long-term sense for the U.S. semiconductor industry as a whole to abandon mass manufacturing in stepping up the ladder of value added is cause for concern. Can commodity markets be abandoned without
OCR for page 84
U.S.-Japan Strategic Alliances in the Semiconductor Industry: Technology Transfer, Competition, and Public Policy losing control of process technology, important revenue streams, key components, and strategic flexibility for downstream applications? If the history of the U.S. automobile and machine tool industries offers any lesson, it is that the costs and dangers of neglecting manufacturing must be taken seriously. Although focusing on improving the outcomes of strategic alliances will not, by itself, provide a panacea for the problems of U.S. industry, this study suggests the need for a conscious effort to structure the alliances to ensure reciprocal and long-term benefits to the United States. This is primarily a challenge for the private sector, particularly U.S. companies, but in a number of ways government policy could have a positive impact. As a prerequisite, it is necessary to have a clear sense of what alliances bring benefits to the United States and, therefore, should be encouraged. Alliances are growing more complex, but this study has suggested some essential elements of advantageous alliances (see Table 13). Generally speaking, these alliances require ongoing interaction, considerable investment of resources, more costs and risks, and management challenges, but the rewards are also greater. The worst alliances, from a U.S. perspective, have been one-shot transactions that mortgage the company's and America's future by selling its more valuable, state-of-the-art technologies without bringing a reverse flow TABLE 13 Elements of Advantageous Alliances Require, as a prerequisite, thoroughgoing analysis and careful preparation—knowledge of the strengths and weaknesses of the partner and of the value of technology and other assets, as well as a clear set of objectives; provide mechanisms for accessing Japanese technology, particularly manufacturing technology, and for strengthening U.S.-based production; open opportunities for U.S. firms to sell their products in Japan; provide intellectual property protection for key technologies; balance the complementary assets and resources of the Japanese and U.S. partners; provide mechanisms for flexible adaptation and periodic review of the immediate impacts and possible longer-term implications; and bring together company interests and those of the nation (by enhancing value-added production and technology diffusion through training in advanced engineering, and constructive relationships with U.S. universities, R&D consortia, and other firms in the industry).
OCR for page 85
U.S.-Japan Strategic Alliances in the Semiconductor Industry: Technology Transfer, Competition, and Public Policy of benefits in return (through technology transfer to the United States or the enhancement of a U.S. company's market position in Japan). The working group believes that the semiconductor industry, a vital upstream segment of the crucial information industries, is a ''strategic industry" essential to the nation's well being. The reasons and criteria for labeling semiconductors a strategic industry, though seldom spelled out, are (1) the importance of semiconductor components for superior performance in military hardware; (2) the centrality of semiconductor technology for achieving breakthroughs in computers and information-based high technology; (3) the broad range and versatility of semiconductor applications for large end user industries, including automobiles and the service sector; and (4) the pivotal position of the semiconductor industry in the system of institutions and practices underlying America's capacity to innovate—employment, manufacturing infrastructure, R&D, graduate training, and so forth. The case for semiconductors rests on nothing less than their importance in maintaining America's capacity to innovate and commercialize technology—and all that this implies for military, industrial, and technological leadership. Strategic importance is time bound. No designation is immutable, and in the case of semiconductors, the leeway to undergo continuous technological change, permitting major new product breakthroughs in computers, telecommunications, and consumer electronics, is not going to last forever. If and when the technological and commercial growth trajectories level off or fall, the rationale for calling semiconductors "strategic" will disappear. For the moment, however, and into the foreseeable future, the case is compelling. Because the semiconductor industry is strategic, a number of imperatives should be considered by the industry, government, and universities. It is essential that the United States maintain a strong, efficient, and world-competitive semiconductor industry. To remain a world-class competitor, the U.S. semiconductor industry needs to maintain the full complement of capabilities, including leading-edge R&D, fabrication, equipment making, manufacturing, some testing and assembly, marketing, and servicing. American producers must not abandon key product markets with respect to technology (such as mask alignment) or revenue flow (e.g., DRAMs and microprocessors). Since volume production and sales across a number of product markets will separate the front-runners from the rest of the pack, the U.S. semiconductor industry cannot afford to rely exclusively on the computer industry to drive its growth; it must also ride the wave of growth in consumer
OCR for page 86
U.S.-Japan Strategic Alliances in the Semiconductor Industry: Technology Transfer, Competition, and Public Policy electronics (including high definition television), telecommunications, aerospace, and other end-user industries. Successful penetration of the Japanese and other rapidly growing Asian markets is imperative. If structured properly, strategic alliances with Japanese companies can contribute to the maintenance of manufacturing and fabrication facilities in the United States and to the expansion of market opportunities abroad. If the vigor of the U.S. semiconductor industry is important to the national interest, what specific responsibilities fall on the U.S. government and the private sector to preserve and enhance this national interest? COMPETITIVE ADVANTAGE: ISSUES FOR U.S. INDUSTRY One of the main benchmarks—and long-term objectives—of the U.S. semiconductor industry should be not simply the crossing of a market share threshold in the Japanese market (for example 20 percent in any given year) but, more fundamentally, the establishment of a permanent foothold in Japan's industrial structure, a breaking into the labyrinth of long-term, inter-firm relationships. In particular, U.S. producers must position themselves to get involved at the design-in phases of end user applications. Strategic alliances, effectively structured, constitute perhaps the best means of establishing a permanent foothold in the Japanese semiconductor market. Strategic alliances should be used to adapt to the rapidly changing commercial environment and to advance adaptive efficiency in the utilization of resources. Concretely, the prime objective should be to upgrade America's manufacturing and equipment-making capabilities. The continuing erosion of this infrastructure could undermine America's capacity to maintain a world-class semiconductor industry, and special efforts may be required by industry in partnership with government to meet this goal.40 American firms should make it a high priority to obtain a reverse flow of technology in forming strategic alliances, particularly in the area of manufacturing know-how. The construction of state-of-the-art fabrication facilities in the United States, which employ and train U.S. workers, could help to meet these goals. By strengthening America's own manufacturing infrastructure, U.S. firms can also deal from positions of greater strength in strategic alliances; this will have the effect of making the flow of technology better balanced. 40 One promising trend that might be encouraged is the emergence of contract manufacturing service companies such as Solectron, which recently won the Malcolm Baldridge quality award.
OCR for page 87
U.S.-Japan Strategic Alliances in the Semiconductor Industry: Technology Transfer, Competition, and Public Policy The semiconductor and related industries should analyze and communicate lessons learned from past experiences in strategic alliances with Japanese firms; the diffusion of knowledge should help companies with no prior experience; industry associations could establish a forum to accomplish these goals of diffusing organizational know-how; firms should work with government to foster these goals. Any U.S. firms contemplating alliances with Japanese companies should enter with the idea of learning as much as possible (from government and other U.S. firms) and of effectively applying what they learn to the development of new sources of competitive strength. In the past, a glaring asymmetry existed: too many U.S. firms entered with only short-term, quick-fix objectives in mind, whereas their Japanese counterparts approached the alliance in a spirit of moving down an organizational learning curve and creating long-term competitive strengths. To get the most mileage from strategic alliances with the Japanese, many U.S. firms will have to reexamine and revamp their organizational structures and strategies; this may require the establishment of new incentives to nurture individuals who can work effectively with Japanese counterparts and improve capabilities to develop technology-based strategies. NATIONAL INTERESTS: ISSUES FOR THE U.S. GOVERNMENT Sound macroeconomic policies are a necessary but not sufficient condition for world-class competitiveness in semiconductors. Serious consideration should be given to measures to rectify America's low savings and investment ratios, deficits, including revamping the tax framework in the United States. Because U.S. institutions and procedures are not well suited to formulate and implement comprehensive industrial policy, the U.S. government must be careful not to do long-term damage to the semiconductor industry or related end user industries by distorting the incentives to be efficient. In specific areas judged to be critical for national security or other reasons (such as semiconductor equipment manufacturing), government policy can have a positive effect in maintaining technological capabilities within the United States and can provide incentives for ''advantageous liaisons" between U.S. and Japanese semiconductor companies. Department of Defense support for SEMATECH should be continued. Policymakers could also consider measures that encourage the U.S. semiconductor industry to invest in capital equipment at a level that will maintain a vibrant manufacturing and process design infrastructure in the United States. Protection of U.S. semiconductor markets should be avoided in prin-
OCR for page 88
U.S.-Japan Strategic Alliances in the Semiconductor Industry: Technology Transfer, Competition, and Public Policy ciple, except when foreign practices clearly violate international rules and domestic laws (e.g., dumping and predatory pricing) or when national security is clearly jeopardized. Brandishing the threat to close the U.S. market as an instrument of diplomatic leverage or retaliation against bilateral trade imbalances ought to be resisted. The United States should develop a variety of approaches to strengthening semiconductors as a strategic industry through incentives for R&D, for companies that export, and for alliances that feature the transfer of advanced manufacturing technology to the United States. Such incentives might include preferential access for foreign-based firms involved in "advantageous alliances" to participate in government-supported R&D in the United States. The U.S. government's pressure on Japan to open its semiconductor market has been instrumental in helping American companies break through the structural barriers of Japanese industrial networks. As long as U.S. firms encounter impediments, appropriate pressures should be continued. In principle, flows of Japanese investment into the United States should not be obstructed unless such investments can be shown to pose serious dangers to competition, to maintenance of a U.S. production base in key industry segments such as semiconductor equipment manufacturing, or to U.S. security interests. Foreign capital has contributed to the development of many small U.S. firms and start-up companies. The U.S. government (in cooperation with the private sector) should improve mechanisms to collect and analyze information on investment and technology flows, and efforts should be made to assess their immediate and likely impacts. This information and analysis should be made available in a timely fashion to the public and to policymakers involved in the Committee on Foreign Investments in the United States (CFIUS). The U.S. government could work with industry to establish a West Coast facility to provide advice and information to U.S. firms contemplating strategic alliances. Federal funding in support of basic research and graduate training in semiconductor-related fields at universities should be increased, and efforts should be made to encourage cooperation between universities and industry. The specter of the worst-case scenario (the United States supplying new ideas to foreign manufacturers who derive most of the value added) is a real concern. This study of U.S.-Japan technology linkages in semiconductors suggests that considerable effort will be required to avert this outcome. Ensuring the long-term competitiveness of the U.S. semiconductor industry and of the United States as a place of production requires not only economic revitalization at the macroeconomic level, but also strategic alliances structured to produce positive, demonstrable benefits to the United States. If U.S.-Japan alliances can bring in vital technology, manufacturing know-
OCR for page 89
U.S.-Japan Strategic Alliances in the Semiconductor Industry: Technology Transfer, Competition, and Public Policy how, and resources—instead of serving primarily as conduits for technology transfer to Japan—the long-term interests of both countries will be served. The role of strategic alliances has grown enormously over the past decade and will continue to increase. Whether or not such alliances expand the overall pie and lead to an equitable distribution of benefits depends largely on the ability of the participants to make the most out of them. Their success will have important implications not only for the participants but also for the technology base, the economic well being, and the national security of the United States as a country.
OCR for page 90
U.S.-Japan Strategic Alliances in the Semiconductor Industry: Technology Transfer, Competition, and Public Policy This page in the original is blank.
Representative terms from entire chapter: