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 15
U.S.-Japan Strategic Alliances in the Semiconductor Industry: Technology Transfer, Competition, and Public Policy 4 Forces Driving the Formation of Strategic Alliances GENERIC FORCES Why has the number of strategic alliances risen so dramatically over the past five or six years? To understand the reasons requires that the semiconductor experience be placed in a global context. From a global perspective, the semiconductor industry reflects an emerging pattern of closer international interdependence. The numbers of U.S.-Japan linkages have multiplied across nearly all trade-related sectors, including old-line manufacturing (e.g., automobiles), the services (e.g., banking and securities), and high technology (e.g., biotechnology and aircraft). As discussed in more detail below, an alliance boom in semiconductors occurred in the mid-1980s and again in the late 1980s. Within the past year or so, however, it appears that the number of newly formed U.S.-Japan alliances may be declining somewhat while U.S.-U.S. alliances are expanding. Data on venture capital investments from Japan provide some evidence but do not give a full picture of trends in other areas such as acquisitions.10 Still, today's level of activity remains far above the early 1980s. 10 Venture Economics reports that in 1990 the Japanese invested $23 million in the U.S. semiconductor industry, while in 1991 the level of investment dropped to $11.25 million (communication with Venture Economics, March 1992). It is also important to note that a downturn in U.S.-Japan alliances is occurring in the context of the growth of global consortia and standards-based groupings, as discussed in more detail later.
OCR for page 16
U.S.-Japan Strategic Alliances in the Semiconductor Industry: Technology Transfer, Competition, and Public Policy The increase in semiconductor alliances since 1985 requires explanation, because the semiconductor industry saw perhaps the highest level of activity across all industrial sectors (with the possible exception of biotechnology). What generic forces are at work propelling companies to forge alliances? What industry-specific factors have led to the proliferation of U.S.-Japan linkages in the semiconductor industry? Perhaps the most obvious generic force is a trend toward the ''globalization of markets.''11 Companies cannot afford to confine themselves to domestic markets, no matter how large they may be. Instead, they must compete in all major markets around the world or risk falling out of long-run contention. For most products, mass volume sales, involving global markets, constitute the sine qua non of low-cost production. Those companies failing to compete worldwide will lose the advantages of rapid movement down steep learning curves. If the world market cannot be partitioned into national units and if semiconductor producers aspire to survive in the crucible of world competition, it is essential that they find ways of getting close to foreign customers. Because the up-front costs and risks of breaking into foreign markets can be prohibitively high, there are strong incentives for companies to find foreign partners to distribute and sell their products through marketing agreements, one form of strategic alliance. To compete effectively in foreign markets over the long haul, however, companies must establish their own physical presence abroad so that they can understand and respond to local needs. A physical presence makes it possible for companies to do what is necessary to succeed: communicate continually with local customers; learn how to operate in foreign environments; cultivate long-term relationships; develop, adapt, or custom-make products for foreign end users; manufacture locally; and offer timely delivery and reliable after-service. Doing this from a distance is impossible. Very large corporations can project a physical presence in foreign markets by going multinational,12 this is what IBM, AT&T, NEC, Toshiba, Hitachi, Texas Instruments, Motorola, and others have done. They have built wholly owned subsidiaries that market, design, manufacture, and provide service for customers in their own local markets. However, because multinationalization of this type is beyond the reach of smaller companies, their only alternative is to develop a series of strategic alliances—such as marketing agreements and manufacturing licenses—that serve the same functional purpose. Such ambitious and fast-developing companies as Sun 11 Kenichi Ohmae, The Borderless World: Power and Strategy in the Interlinked Economy (New York: Harper Business, 1990). 12 Christopher A. Bartlett and Sumantra Ghoshal, Managing Across Borders: The Transnational Solution (Boston: Harvard Business School Press, 1989).
OCR for page 17
U.S.-Japan Strategic Alliances in the Semiconductor Industry: Technology Transfer, Competition, and Public Policy Microsystems and MIPS have concluded a series of strategic alliances that have positioned them to carve out significant shares of Japanese and European markets. Even for very large firms the costs and risks of having to compete around the world have become so high that they actively seek alliances not only for themselves but also for their multinational enterprises. IBM Japan, for example, has put in place an extended network of alliances with nearly a thousand local Japanese companies, which has helped it achieve a strong "insider" position in the Japanese market.13 HIGH-TECHNOLOGY FACTORS The heavy concentration of U.S.-Japan alliances in the high-technology sectors, especially computers, biotechnology, and semiconductors, can be linked to the special characteristics of technology-intensive industries. The rapid development of technology, wide scope for ongoing innovation, and continual coming on stream of new generations of products lead to short product life cycles and very high risks, which give rise, in turn, to cost-and risk-reducing alliances. The drive for innovation means that commercial competition is bound to be fierce and that the incentives to enter into strategic alliances are apt to be powerful. In semiconductors as in computers and biotechnology, no single company can dominate in all product markets; few, indeed, can even realistically hope to be active across a spectrum of markets. Given the need for continuous innovation, most high-technology companies build strategic alliances in order to: (1) compensate for in-house weaknesses or technological gaps; (2) fill out product lines and portfolios; (3) position the company to enter lucrative new markets; (4) better serve an established or targeted customer base; and (5) reduce the costs, risks, and time required to develop new products and process technologies. To achieve these goals, U.S. and Japanese companies are willing to swap technology (e.g., cross-licensing), second source, undertake joint development projects, and organize joint ventures. In more traditional fields, such as textiles and steel, where the technology is more mature and the time required for one generation of new products to supplant another is far longer, the incentives for companies to enter into strategic alliances are much less powerful (unless such alliances permit them to branch out in new directions). Semiconductor and computer companies find themselves trapped in the squeeze between the spiraling costs and risks of R&D and continuous capital investments, on the one hand, and the collapsing time intervals during which profits on current-generation products can be made, on the other. 13 Ohmae, op. cit., p. 131.
OCR for page 18
U.S.-Japan Strategic Alliances in the Semiconductor Industry: Technology Transfer, Competition, and Public Policy FIGURE 4 Building plus equipment costs for a high-volume fabrication line. Source: Dataquest, April 1991. The fixed costs of doing research and of building new plant facilities for the 16-megabit DRAM, for example, have soared to almost prohibitive heights—even for giants in the electronics industry (see Figure 4). Such costs would be tolerable if the time intervals within which companies could earn satisfactory profits were long enough; however, the contraction in product life cycles means that the window of opportunity for rent retrieval is exceedingly short—usually only the first year or so of a new product's introduction. Although there were widespread reports in early 1992 of plans by Japanese semiconductor companies to cut capital spending and perhaps R&D spending, the top five Japanese merchant semiconductor companies have consistently outspent the top five U.S. merchants in R&D in recent years.14 For dynamic random-access memory (DRAM) and other commodity very large-scale integration (VLSI) chips, perhaps the key competitive requirement is achieving economies of scale. Semiconductor companies must be able to cross a high production threshold to justify the costs of R&D and new plant investments. If companies want to stay competitive in the VLSI business, they must be able to pay for continuous increases in R&D and manufacturing investments at a rate of around 20 to 30 percent per year. For many companies, especially merchant houses, the escalating financial burdens are too heavy to bear. Accordingly, the incentives to enter into manufacturing alliances with Japanese firms are strong. The only other options for VLSI companies are unattractive ones: raising large sums for investment financing or ultimately having to withdraw from DRAM and commodity chip production. 14 National Advisory Committee on Semiconductors, Attaining Preeminence in Semiconductors, February, 1992, p. 23. This chart is based on Dataquest figures.
OCR for page 19
U.S.-Japan Strategic Alliances in the Semiconductor Industry: Technology Transfer, Competition, and Public Policy The continuous waves of innovation—the constant coming on stream of new products and of diverse, new families of products—mean that the financial, technological, and manpower resources of even the largest companies are bound to be stretched to their limits. Even highly diversified companies, possessing deep financial pockets, cannot escape the discomfort of being stretched in too many directions. Hence, the imperatives of commercial competition have forced semiconductor producers to be receptive to the idea of linking up with foreign partners, because alliances offer companies the opportunity to pool technological and manpower resources. To cite a recent example: the world's largest computer manufacturer, IBM, has decided to enter into a joint venture with Siemens to manufacture 16-megabit DRAMs in France. IBM has been actively forming alliances with small companies in Japan and Europe, as well as the United States, investing more than $1 billion in 200 companies over the past 10 years.15 So compelling have corporate incentives become that crisscrossing linkages of strategic alliances now tie together virtually all major (and most minor) semiconductor companies in ever denser and more complicated networks. One would be hard-pressed to find a single significant semiconductor company anywhere in the world today that has managed to remain isolated from the powerful pull of alliance linkage. Compare the current situation with the 1960s and 1970s, when strategic alliances were the exception rather than the rule, and the dimensions of the recent alliance explosion can be appreciated fully. U.S.-JAPAN ALLIANCES If generic and semiconductor-specific forces explain the ebb and flow of alliance formation, what explains the marriage of U.S. and Japanese firms? Why have companies from the two sides of the Pacific joined together in so many alliances—conspicuously more than American-European or Japanese-European cases? According to one study, European investments in U.S. high technology are rather limited (about 16 percent for all industries and about 5 percent of the foreign investments in semiconductors)16 (see Figure 5). According to the same study, Japanese investments in semiconductors make up about 90 percent of all foreign investments in 15 "Learning from Japan," Business Week, January 27, 1992, p. 55. 16 Linda Spencer, Foreign Investment in the United States: Unencumbered Access (Washington, D.C.: Economic Strategy Institute, 1991). The ESI data base includes a variety of print and electronic sources updated periodically. Other sources of data on foreign investments in U.S. high-technology industries include the American Electronics Association Tokyo Office, Ulmer Brothers, Venture Economics, and Dataquest. Dataquest compiles information on many aspects of the electronics industry, including investments, licensing, and other tie-ups that do not involve equity purchases.
OCR for page 20
U.S.-Japan Strategic Alliances in the Semiconductor Industry: Technology Transfer, Competition, and Public Policy FIGURE 5 Foreign investments in U.S. high technology. Note: Numbers in parentheses refer to industry totals. Source: Adapted from Linda Spencer, Foreign Investments in the United States: Unencumbered Access (Washington, D.C.: Economic Strategy Institute) May 1991, p. 10. this industry and about 60 percent of all foreign investments in U.S. high-technology industries. The most obvious explanation for the preponderance of U.S.-Japan alliances in semiconductors is that the U.S. and Japanese semiconductor industries are the biggest and best developed in the world. In 1991, Japan accounted for 38 percent of the world market, and North America 29 percent (see Figure 6). Japanese companies control 46 percent and North American companies 39 percent of total semiconductor sales (see Figure 7). The extraordinary level of economic integration and high degree of trade interdependence are background factors leading to alliances between U.S. and Japanese companies—not only in semiconductors but also across a broad spectrum of industrial sectors. FIGURE 6 World semiconductor market in billions of dollars. Note: Total greater than 100% due to rounding. Source: Dataquest, June 1992.
OCR for page 21
U.S.-Japan Strategic Alliances in the Semiconductor Industry: Technology Transfer, Competition, and Public Policy FIGURE 7 Regional shares of the worldwide semiconductor market. Source: Dataquest, June, 1992. Differences in national systems of industrial organization (levels of vertical integration and diversification within firms, corporate finance, and capital markets) have brought opposites together17 (see Table 2). Small, venture start-up firms in the United States, in need of funds, manufacturing foundries, and marketing outlets, look to large, deep-pocketed, vertically integrated, and diversified Japanese corporations to meet these needs, since large diversified U.S. corporations have shown far less interest until quite recently in forming alliances with small U.S. venture start-up companies. Many of the large Japanese giants, in turn, look to small U.S. start-ups to provide new product designs to fill niche markets or to compensate for certain deficiencies in their own innovative capabilities. Since the largest returns on alliance investments—such as the development of new technologies or company positioning for long-term diversification into whole new fields—are often reaped only after a long period of gestation, it is not surprising that Japanese companies (with their long time horizons) are among the most active in alliance partnerships.18 In the best of circumstances, opposite firms attract and combine in ways that overcome the respective limitations of different industrial systems. Although there are signs of change, it is true that during the past decade, large U.S. corporations, faced with the same opportunities as Japanese 17 Daniel I. Okimoto, Takuo Sugano, and Franklin Weinstein, eds., Competitive Edge: The Semiconductor Industry in the U.S. and Japan (Stanford, Calif.: Stanford University Press, 1984). 18 Many large U.S. corporations diversify into new sectors, such as financial services and information systems. However, it has been some time since a large U.S. company attempted to diversify into semiconductor manufacturing, perhaps because of less promising prospects for returns on investment than expected in other fields.
OCR for page 22
U.S.-Japan Strategic Alliances in the Semiconductor Industry: Technology Transfer, Competition, and Public Policy TABLE 2 A Taxonomy of the U.S. and Japanese Semiconductor Industries United States Large Systems Houses IBM, AT&T/NCR, Rockwell, Hewlett-Packard, Honeywell Companies that sell a small percentage of output on the open market but function as captive producers and purchase large quantities of semiconductors from the outside Systems/Merchant Houses Motorola, Texas Instruments Companies that sell systems, but whose merchant semiconductor sales account for a high percentage of total sales Full-Line Houses Intel, National Semiconductor, Advanced Micro Devices Newer Companies New Merchants: Cypress, LSI Logic Fabless Microprocessor Designers: MIPS, Cyrix Fabless Niche Customizers: Altera, Chips and Technologies New Systems Emphasis: Sun Microsystems Japan Large, Blue-Chip Corporations NEC, Toshiba, Hitachi, Mitsubishi Electric, Fujitsu, Matsushita Other Major Producers Canon, Sharp, Sanyo, Oki, Sony Lateral Entry Companies Minebea (NMBS), Kubota, Nippon Steel, Kobe Steel Smaller Component Makers Hosiden, Alps, Sanken, Toko NOTE: "Fabless" refers to companies which do not possess their own fabrication facilities. SOURCE: Compiled by NRC Office of Japan Affairs staff. companies, have been less active in developing strategic alliances with smaller U.S. firms. Whether this is the result of shorter time horizons or a shortage of capital (or both), is open to further study. Certain features of Japanese industrial organization, especially intercorporate shareholding, make it easier for them to operate on the basis of a longer time horizon than U.S. companies. The nature of America's stock market forces U.S. firms to march more closely in step to the drumbeat of quarterly profits.19 The dense network of intercorporate shareholdings protects Japanese management from the tyranny of short-term profit maximization. Corporate share- 19 Michael L. Dertouzos, Richard K. Lester, Robert Solow and MIT Commission on Industrial Productivity, Made in America: Regaining the Productive Edge (Cambridge: MIT Press, 1989), pp. 53–66.
OCR for page 23
U.S.-Japan Strategic Alliances in the Semiconductor Industry: Technology Transfer, Competition, and Public Policy holders, such as banks and insurance companies, do not buy and sell their stocks in response to short-term fluctuations in share prices or to standard indicators such as price/earning ratios.20 What they seek is the steady appreciation of stock value through sound, long-term company growth.21 The high degree of diversification and vertical integration for which Japanese electronics giants are known has also had the effect of heightening interest in strategic alliances because there are potential multiplier effects associated with the introduction of new technologies (through strategic alliances). New technologies can be used to upgrade the quality of both products and process technology for internal, captive markets, including components, manufacturing, software applications, and new product designs. For large Japanese firms, there is a captive market learning curve down which companies can move; the advantages of moving down this learning curve have been widely observed in the case of application-specific integrated circuits (ASICs). Thus, the characteristics of Japanese industrial organization have facilitated Japanese participation in strategic alliances. BUSINESS CYCLES, TECHNOLOGICAL CHANGE, AND POLITICAL FACTORS There appears to be a rough correlation between downturns in business cycles and peak periods of alliance formation. When business conditions turn bearish, U.S. companies appear more disposed to enter strategic alliances to meet their financing needs and to survive the sharp downturns in demand without having to revamp existing structures. The alternative is to cut back on R&D projects, new capital investments, and core technical personnel. Looking at the available data (refer to Figure 3), we see that the number of American-Japanese alliances hit a peak in 1986 and 1987, when business demand in the semiconductor industry slackened. When there was a revival of demand in 1988, the number of alliances plummeted. The empirical evidence is imperfect, of course, making it hard to establish a clear-cut case 20 Masahiko Aoki, Information, Incentives and Bargaining in the Japanese Economy (Cambridge, England: Cambridge University Press, 1988), pp. 99–149. 21 In the early 1990s, a series of Japanese stock market scandals, declining share prices, and deterioration in the positions of commercial banks put an end to the "bubble economy" and rapidly rising asset prices of the late 1980s. Japanese press reports in early 1992 indicated that Japanese electronics firms planned to cut capital investment during the coming year. See "4.9% hen, 6 nen buri mainasu" (4.9% Decline, First Decrease in 6 Years), Nikkei Sangyo Shimbun, March 10, 1992, p. 2. The long-term implications are unclear. Some analysts believe that Japanese companies may be forced to concede large portions of the DRAM business to South Korea and Taiwan. Japanese companies have weathered extended industry downturns in the past.
OCR for page 24
U.S.-Japan Strategic Alliances in the Semiconductor Industry: Technology Transfer, Competition, and Public Policy of cause and effect but the rough, macrolevel data suggest that business cycles do affect the microlevel disposition of U.S. and Japanese companies to take advantage of the benefits of strategic alliances. A closer look at the data suggests that the situation is more complex than the picture painted by simple time-series analysis. Factors other than fluctuations in business demand can also have significant effects. In 1988 the number of U.S.-Japan alliances fell abruptly from 124 the previous year to 46 (refer to Figure 3). This was the year following the Toshiba Machine Company incident (1987), which involved that company's sale of militarily sensitive technology to the Soviet Union. In the aftermath of the Toshiba incident, MITI officials quietly discouraged Japanese companies from entering strategic alliances or getting involved in overseas activities that would expose them to possible foreign criticism. Hence, political and strategic factors can have a direct impact on alliance formation. Why in 1989 was there a surge in strategic alliances, when business demand remained brisk? Technological developments leading to the emergence of new product clusters represent another intervening variable affecting alliance formation. The reason strategic alliances increased in 1986–1987 and again in 1989–1990 may have been related to the coming on stream of ASICs, a major new family of products (see Figure 8). FIGURE 8 American-Japanese semiconductor alliance by product type: 1984–1990. Source: Compiled by NRC Semiconductor Working Group using Dataquest data.
OCR for page 25
U.S.-Japan Strategic Alliances in the Semiconductor Industry: Technology Transfer, Competition, and Public Policy The emergence of ASICs and the passage of an interval of time following the Toshiba Machine Company incident may have provided a stimulus to alliance formation. At any given point in time, a complex mix of underlying and immediate factors comes together to affect trends in strategic alliances. Mention must be made of yet another major driving force, originating from the fierce competition in the computer industry but pulling semiconductor producers into its vortex: the struggle to define and establish broadly based standards in operating and applications software and the choice of chips (discussed in Chapter 8). This effort to establish operating software standards has driven the recent IBM-Apple alliance aimed against Microsoft. The adoption of MIPS' RISC (reduced instruction set computing) chip and Intel's X86 also provided the framework for the emergence of an ambitious, multifirm consortium known as Advanced Computing Environment (ACE). The struggle over market shares in computers and key semiconductor chips is being shaped by the rush to establish dominance in software and chip standards. This high-stakes race lies behind the alliance alignments and realignments that are helping to redefine the structural boundaries of the semiconductor and computer industries. RECESSIONS: LESSONS LEARNED The world semiconductor industry seems to go through a recession roughly every four years. Over the past 14 years there have been four recessions: in 1977, 1981, 1985, and 1990–1991. With each cyclical downturn, semiconductor companies have been forced to make a series of painful adjustments. Accordingly, semiconductor companies have learned some important lessons. Often it takes a serious crisis of some kind to break old, established patterns of doing business, force companies to devise new ideas and approaches, and generate adaptive modes of behavior based on collective learning. For collective learning to occur, there needs to be some stability in management and in the labor force; the more continuity, the greater is the learning capacity (other things being equal). Japanese corporations may be at an advantage because of their practice of lifetime employment. Owing to the stability of their work force and the continuity of personnel in middle and top management, Japanese corporations are in a position to learn and adapt in ways that elude many U.S. merchant semiconductor houses. Prior to the 1977 recession, U.S. companies dealt with downturns by cutting variable costs through belt-tightening measures: workers were laid off, operating budgets were cut, and the overall level of activity was curtailed. However, the 1977 recession taught these companies that by itself, variable-cost cutting could no longer be sufficient. The capital intensity of the semi-
OCR for page 26
U.S.-Japan Strategic Alliances in the Semiconductor Industry: Technology Transfer, Competition, and Public Policy conductor industry had reached high enough levels that mere manipulation of variable costs would not solve the problem. Attention had to be focused on reducing fixed costs as well. This realization prompted companies to attack fixed costs by canceling and slashing plans for expensive new capital investments. The combination of variable-cost cutting plus cutbacks in fixed costs became standard practice from 1977 forward, as evident in the 1981 and 1985 recessions. Out of the 1977 recession came the further realization that (1) U.S. merchant semiconductor producers share important interests in common; (2) these interests could be effectively served through the organization of an industrial association representing their views in Washington; and (3) although U.S. producers enjoyed technological and commercial dominance (in 1977), the large, diversified Japanese corporations posed a serious, long-term threat. In response to the first two perceptions, the Semiconductor Industry Association (SIA) was organized in 1977; since then, the SIA has shown itself to be among the most active and effective of America's industrial associations in representing member company interests and objectives. The third perception has led to a sustained focus on Japan as a competitive threat to the U.S. semiconductor industry, a ''strategic'' industry portrayed as pivotal to America's overall position in high technology. 22 The delineation of this threat has not changed over time; concern over dumping and restricted Japanese market access persists, with heightened concerns about the closed nature of Japanese industrial groupings. The general tenor of the industry's position toward Japan has not softened since the conclusion of the. Semiconductor Trade Agreement in 1986. However, improved relationships have developed between U.S. semiconductor executives and Japanese user executives as direct commercial contacts have been facilitated within the framework of the 1986 agreement and a successor agreement consummated in 1991. The evolution of U.S.-Japan trade negotiations has undoubtedly affected the growth of alliances, although in ways difficult to measure precisely. The 1986 U.S.-Japan semiconductor agreement did not open the door for U.S. semiconductor firms in the Japanese market until the U.S. government applied sanctions on selected Japanese equipment imports in 1987. Although the rate of increase in U.S.-based firms' share of the Japanese market has not been rapid enough to satisfy many observers, the relationships slowly and at times painstakingly nurtured between U.S. and Japanese companies under the agreement have built confidence over time that companies from the two countries can work 22 Semiconductor Industry Association, The Effect of Government Targeting on World Semiconductor Competition (Cupertino, Calif.: Semiconductor Industry Association, 1983).
OCR for page 27
U.S.-Japan Strategic Alliances in the Semiconductor Industry: Technology Transfer, Competition, and Public Policy together.23 In addition, the agreement probably also helped to encourage Japanese firms to form alliances with U.S. firms as a way to ensure their long-term viability in the U.S. market. Although there are many contentious effects of the agreement, such as alleged Japanese dumping in third-country markets in violation of the Agreement, the Agreement stabilized the policy context and thereby probably provided an incentive for alliance formation. In considering responses to the 1977 recession, it is important to note a different set of adjustments that took place in Japan. Although Japanese firms were subjected to the same underlying cost pressures as their U.S. counterparts, the established practice of lifetime employment precluded layoffs of large segments of their permanent work force (although they did pare down part-time and temporary workers). Japanese corporations were able to continue to invest heavily in new plant facilities, thanks to deep financial pockets (especially the availability of debt financing through the system of main banks). 24 Thus, differences in labor and capital markets led to different corporate responses to recession in the United States and Japan. The response of Japanese corporations was also affected by another structural difference—the close working relationship between government and industry. In the mid-to late 1970s, MITI's role in organizing national research projects, such as the VLSI project in semiconductors, had the effect of pumping R&D resources into companies at a time of urgent need. This helped them sustain ambitious R&D activities through bad business conditions. Beyond the welcomed infusion of capital, government funding also had the unintended but salutary effect of creating a new, centralized channel of funding for semiconductor R&D, one designed to funnel money centrally from corporate headquarters rather than individual semiconductor divisions. Participation in national research projects also meant that Japanese companies were firmly committed to the completion of these ambitious projects (which could not be terminated because of a public commitment made to the government). It also kept semiconductor R&D elevated at the highest level of priority within Japanese firms. Hence, the impact of government-organized research activities went well beyond the output of the projects themselves. As unintended but important consequences, such national projects left a lasting impact in terms of research commitments and priorities, changes made in the system of corporate R&D funding, and adaptive lessons learned in recession management. 23 Undoubtedly, there is a strong current of opinion in the Japanese semiconductor industry that resents the agreement and sees U.S.-Japan alliances aimed at increasing foreign market share as "gunshot weddings" necessary to avoid U.S. trade sanctions. 24 M. Therese Flaherty and Hiroyuki Itami, "Finance" in Daniel I. Okimoto, Takuo Sugano, and Franklin Weinstein, eds., Competitive Edge: The Semiconductor Industry in the U.S. and Japan (Stanford, Calif.: Stanford University Press, 1984), pp. 134–176.
OCR for page 28
U.S.-Japan Strategic Alliances in the Semiconductor Industry: Technology Transfer, Competition, and Public Policy In the recession of 1981–1982, when the combination of semiconductor revenues and government funding was insufficient to meet their needs, semiconductor divisions in Japanese companies were able to draw on allocations made available by corporate headquarters to underwrite their expenditures. The pool of general funds came from profits earned by the sale of downstream products (consumer electronics products such as video cassette recorders). Semiconductor divisions were thus able to support essential research through cross-subsidizing from profits earned by other divisions. Vertically integrated, diversified Japanese companies, in short, utilized the built-in advantages of cross-subsidization, a natural stabilizing factor during demand downturns in specific product markets. The biggest and most severe recession to hit the semiconductor industry struck in 1985. The reason this recession hit companies so hard was because business demand had soared to new heights prior to 1985, buoyed by brisk demand from the emerging personal computer market and by abnormally high book-to-bill ratios. The high book-to-bill ratios meant that companies were manufacturing at full capacity and piling up inventory at a fast pace in an effort to meet brisk bookings. When demand dropped abruptly, the feeling of being caught in a free-fall was magnified by the large inventories that companies had amassed (thanks to the false sense of euphoria associated with high book-to-bill ratios). In this sense, one could call the 1985 downturn an inventory-based recession. Despite harsh business conditions in 1985, Japanese companies maintained very high levels of R&D expenditures. Managers in some Japanese companies took 5 percent cuts in salary and bonuses and curbs were placed on aggressive capital investments. Unlike earlier recessions, there were also cutbacks in plans for the expansion of capital investments. Although Japanese companies went ahead with the construction of such facilities as clean rooms, they chose not to install expensive equipment until the recession had passed. Like their U.S. counterparts, Japanese companies were forced to focus on reducing fixed as well as variable costs. Japanese companies also discovered that the products embodying older technology tend to be the most vulnerable. Wherever possible, they tried to roll back manufacturing capacity in products based on slightly older technology and to push aggressively ahead in products incorporating newer technology. By moving from the rear to the forefront, Japanese companies anticipated gaining some measure of insulation against cyclical free-falls. American firms already understood that. As a result, companies in both countries moved to close down excess capacity in older products during the most recent recession (1990–1991). To make this move, attention turned again to the advantages of joining forces with well-chosen corporate partners. Mired in a deep recession, U.S. companies had no choice but to make deep cuts in both variable and fixed costs, laying off workers in record
OCR for page 29
U.S.-Japan Strategic Alliances in the Semiconductor Industry: Technology Transfer, Competition, and Public Policy numbers and canceling plans for large-scale capital investments. American and Japanese firms began to see strategic alliances as a practical, counter-cyclical option—as either an alternative or a supplement to other cost-and risk-reducing measures. Many U.S. firms came to see strategic alliances as a means of obviating the need to make certain manufacturing, marketing, or R&D investments that they might otherwise have felt compelled to make. Around the mid-1980s when the number of strategic alliances soared, the impetus grew to pool resources, to fuse one company's strengths with those of a partner so as to consolidate and expand competitive advantages while offsetting company weaknesses. The Toshiba-Motorola alliance is a typical example explored in more detail in the case study that appears in Appendix A. By bringing together complementary strengths, the partners filled out their product portfolios, enhanced their overall technological capabilities, and compensated for in-house gaps and weaknesses. The scope of strategic alliances broadened and extended into virtually all facets of the semiconductor business:
OCR for page 30
U.S.-Japan Strategic Alliances in the Semiconductor Industry: Technology Transfer, Competition, and Public Policy Once the advantages became apparent, no phase of the semiconductor business could stand outside the powerful pull of alliance linkages. Forming alliances became a fixture in the range of corporate options: Increasingly, the advantages of alliance formation came to outweigh the perceived risks and uncertainties. Indeed, the costs of going it alone (i.e., standing aloof from alliances) came to be seen as unacceptable. To survive in the competitive marketplace, especially during recessions, semiconductor companies had to scramble to find strategic allies. The rush to find allies, in turn, further intensified the level of commercial competition.
Representative terms from entire chapter: