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 5
Japanese Investment and Technology Transfer: An Exploration of its Impact: Report of a Workshop 2 Assessing the Evidence — Japanese Investment and Technology Transfer DEFINITIONS AND APPROACH Indices of “foreign direct investment” seek to measure the “extent to which foreign firms and individuals control U.S. production. ”10 The very nature of multinational corporations raises difficulties in defining “foreign” firms, and there are legitimate grounds for disagreement over the extent of ownership necessary to exercise “control.” The approach used by the U.S. Department of Commerce is to count a firm with at least 10% of its capital held by a single foreign owner as a U.S. subsidiary of that company.11 The value of this definition is to clearly distinguish direct investment from portfolio investment, which is not undertaken with the aim of exercising management control.12 10 Edward M. Graham and Paul R. Krugman, Foreign Direct Investment in the United States (Washington, D.C.: Institute for International Economics, 1989), p. 7. 11 Ibid., p. 9. 12 Direct investment itself can be separated into acquisition of an existing company, the purchase of a minority equity stake, “green field” investment to build a new subsidiary from scratch, and new funding for the expansion or modernization of a subsidiary that comes from the home country rather than from the retained earnings of the subsidiary itself.
OCR for page 6
Japanese Investment and Technology Transfer: An Exploration of its Impact: Report of a Workshop There are a number of possible approaches to the analysis of “technology transfer.” Technology is “applied science” or the application of knowledge to the production of goods and services. The most common image of “industrial technology” is proprietary knowledge subject to intellectual property protection through patents or trade secrets. Technology trade statistics measure the flow of licensing fees to the owners of proprietary knowledge from those wishing to use it. But technology is embodied first and foremost in people — scientists, engineers and technicians employed by companies, universities and research institutes. The timely application of new (or improved) technology to a more efficient production of goods and services requires superior management and marketing practices. Weighing the significance of the organizational context for technology transfer is especially relevant to this discussion. It is widely held that Japan's competitive advantage in manufacturing is primarily due to innovations in organizing and managing the manufacturing process. Attitudes, ideologies and even “culture” affect the receptivity of individuals and organizations to such practices as quality control circles, just-in-time manufacturing and concurrent engineering. Organizational assets outside the R& D lab have a major impact on how effectively companies are able to improve their competitiveness through the utilization of technology. “Technology transfer” is the process by which organizations learn how to develop and manufacture new products or improve manufacturing processes. In the case of proprietary knowledge, technology transfer may be accomplished through an exchange of documents, equipment, personnel, or some combination. Transfers of “hard technology” between organizations are often accompanied by a financial transaction, but within organizations such as MNCs this is not always the case. “Soft technology” transfer within a single organization is mainly accomplished by training or personnel rotation. Between organizations, transfer of management practices can occur as a result of business relationships, labor mobility or competitive pressure. In the case of Japanese FDI in the U.S., there are several potential “technology transfer routes” that need to be considered. One route is the transfer of technology from a Japanese parent to an American subsidiary. The second is the flow of technology in the opposite direction, from the U.S. subsidiary to the Japanese parent, including transfers in which the subsidiary is a conduit or intermediary for technology developed by other organizations. The third is technology transfer from the American subsidiary of a Japanese firm to U.S.-owned firms. A fourth route is the transfer of knowledge between U.S. subsidiaries of Japanese companies. This report deals primarily with a review of the evidence concerning the first route and, to a lesser extent, the third route. The second route — technol-
OCR for page 7
Japanese Investment and Technology Transfer: An Exploration of its Impact: Report of a Workshop ogy transfer from U.S. subsidiaries to Japanese parents — will be addressed to a limited extent because it is a major focus of the concerns expressed by Americans about Japanese investment in the United States. A discussion of the policy implications raised by the workshop would be incomplete without consideration of these concerns.13 AGGREGATE STATISTICS Table 1 presents statistics that measure aspects of foreign and Japanese control of the American economy. There are at least 1,500 Japanese- TABLE 1 Measures of Extent of Japanese Control Over the U.S. Economy U.S. Affiliates of All Foreign Firms U.S. Affiliates of Japanese Firms FDI Stock as % of Total Net Worth of Non-financial Corporations (1990) 10.5 2.17 Employment of Foreign Controlled Affiliates as % of All U.S. Employment (1989) 5.0 0.57 Assets of Foreign Controlled Mfg Affiliates as % of All U.S. Mfg Employment (1989) 16.8 1.99 Employment of Foreign Controlled Mfg Affiliates as % of All U.S. Mfg Employment (1989) 9.7 1.26 Value-Added by Foreign Affiliates as % of U.S. GDP (1987)* 3.4 0.37 Value-Added in Manufacturing as % of all U.S. Manufacturing (1987) * 8.5 0.47 Note: * preliminary data Source: Edward Graham, compiled from Bureau of Economic Analysisdata 13 The workshop did not uncover new evidence of large-scale transfer of technology from the United States to Japan resulting from Japanese FDI, though conferees were invited to provide such evidence. Academic research on this aspect of Japanese investment has not been as extensive as research on other dimensions, such as supplier-manufacturer relationships in the automobile industry. In the absence of concrete data, perceptions will be based on anecdotal evidence. It is likely that concerns about the implications of Japanese FDI for outward technology flow will continue to figure prominently in policy debate.
OCR for page 8
Japanese Investment and Technology Transfer: An Exploration of its Impact: Report of a Workshop affiliated manufacturing firms operating in the United States,14 and the book value of Japanese direct investments in U.S. manufacturing now exceeds $17 billion.15 By these measures Japanese investment, despite rapid growth through most of the eighties, is still not an overwhelming influence on the overall U.S. economy or on U.S. manufacturing. What about technology transfer? The statistics that most directly measure technology transfer are those that cover licensing between foreign parents and American subsidiaries — the financial transactions that accompany the flow of proprietary technology.16 According to Commerce Department data, in 1989 U.S. subsidiaries paid their Japanese parents over ten times what they received in licensing fees and royalties — which indicates that Japanese firms transfer much more technology to their U.S. affiliates than they take back to Japan. The statistics do not point to a particularly large or rapidly increasing amount of technology transfer from U.S. subsidiaries to Japanese parents. Australian and Canadian companies received more in licensing fees and royalties from their parents than the affiliates of Japanese firms, even though Japan's FDI in the United States is much greater than that of Australia or Canada. 17 Statistics also indicate that U.S. parent companies export much more technology to their Japanese subsidiaries than Japanese subsidiaries in the United States export to their parents.18 14 See JETRO Monitor, April 1991 for a special issue on the JETRO Survey of Japanese Manufacturers in the United States. 15 U.S. Department of Commerce, Bureau of Economic Analysis, Survey of Current Business, June 1990, p. 64. 16 Japanese aggregate statistics are summarized in footnote 1. U.S. Department of Commerce statistics are published annually in the Survey of Current Business. Unlike the Japanese figures, the U.S. statistics distinguish between transactions of affiliated and unaffiliated entities. However, because the Commerce Department includes royalties and other fees for intangible assets in its country breakout figures for transactions between affiliated entities, they do not represent a pure measure of technology transfer. That distinction is made for unaffiliated transactions. In practice, usually around three-quarters of total licensing fees and royalties between unaffiliated companies are payments for “industrial processes ” —technology transfer. But special cases can easily distort the picture that emerges from unaffiliated transactions, and serve as a warning that analysts should be cautious about drawing conclusions from statistics on affiliated payments. For example, a large percentage of total 1988 U.S. payments of royalties and licensing fees consisted of payments to South Korea for rights to broadcast the Seoul Olympics. 17 Anthony J. DiLullo and Obie G. Whichard, “U.S. International Sales and Purchases of Services,” Survey of Current Business, September 1990, p. 47. 18 For a discussion, see Kan H. Young and Charles E. Steigerwald, “Technology Transfer and Foreign Direct Investment in the United States, ” unpublished paper, 1989.
OCR for page 9
Japanese Investment and Technology Transfer: An Exploration of its Impact: Report of a Workshop TABLE 2 Royalties and Licensing Fees Between the United States and Japan: Affiliated Entities (millions of dollars) Year U.S. Receipts U.S. Payments Total U.S. parents from Japanese affiliates U.S. affiliates from Japanese parents Total U.S. parents to Japanese affiliates U.S. affiliates to Japanese parents 1987 1,111 1,090 22 217 10 207 1988 1,432 1,404 28 275 5 269 1989 1,518 1,486 32 337 6 331 1990 1,702 1,665 37 397 12 386 Source: Anthony J. DiLullo and Obie G. Whichard, “U.S. International Sales and Purchases of Services,” Survey of Current Business, September 1991. TABLE 3 Licensing Fees Between the United States and Japan: Unaffiliated Companies (millions of dollars) Year U.S. Receipts U.S. Payments Total Royalties and Licensing Fees Industrial Processes Total Royalties and Licensing Fees Industrial Processes 1987 839 723 104 88 1988 1,022 883 123 108 1989 1,044 898 129 120 1990 1,205 1,048 184 165 Source: Anthony J. DiLullo and Obie G. Whichard, “U.S. International Sales and Purchases of Services,” Survey of Current Business, September 1991. Besides the fact that U.S. statistics lump royalties in with licensing fees in reporting transactions between affiliated firms, there are other reasons why data on licensing fees may not be the best measure of technology transfer, especially transfer within an MNC. Technology transfer within a single firm is usually simpler than transfer between firms, and there may be no obvious incentive for formal licensing agreements in such cases. However, MNC's do have a motive for recording licensing transactions between units in different countries because of national disparities in corporate tax rates, foreign exchange regulations, or other aspects of the policy environment. It may be in the interests of shareholders or management to record income in the country with the lowest corporate tax rate or
OCR for page 10
Japanese Investment and Technology Transfer: An Exploration of its Impact: Report of a Workshop to repatriate earnings in the form of licensing fees from subsidiaries in countries with foreign exchange controls. These motivations bear little relation to the value of the technology itself, so drawing conclusions about the actual technology transfer situation from intra-firm licensing transactions is problematic. It is at least conceivable that technology transfers from U.S. subsidiaries to Japanese parents are underreported and that transfers in the other direction are overpriced. Even though the rate of corporate taxation is not particularly high in the United States, Japanese companies only pay Japanese taxes on subsidiary income that is remitted as dividends, and do not pay taxes on licensing income.19 Therefore, manipulating the transfer prices used in transactions between the parent and U.S. subsidiary, including licensing fees, may be an effective strategy to minimize the corporate tax bill. 20 Besides technology transfer that can be measured in licensing statistics, there are other facets of the operations of Japanese MNCs that have an impact on the technological level of their U.S. subsidiaries and the U.S. economy generally. In some instances technology transfer is involved. For example, if a Japanese MNC's U.S. subsidiary does a significant amount of R&D or sophisticated manufacturing in the United States, this may contribute to technology development and productivity here. R&D expenditures and employment by Japanese subsidiaries in the United States provide some (admittedly limited) indications of potential technology transfers from Japan to the United States. In the absence of other aggregate indicators that directly measure technology transfer, it is necessary to turn to other statistics that might serve as proxies, or which measure the contribution of Japanese FDI to the technological level of the U.S. economy. One possible indirect measure is the R&D spending by U.S. subsidiaries of Japanese MNCs. Overall, the figures indicate that subsidiaries of Japanese firms do less R&D per employee than other affiliates but that Japanese manufacturing affiliates are comparable to U.S. manufacturing 19 Congressional Research Service, Japan-U.S. Economic Issues: Investment, Saving, Technology and Attitudes (Washington, D.C.: The Library of Congress, 1990), p. 44. 20 A certain amount of manipulation of this type is perfectly legal, but there are limits. A number of Japanese firms have come under increased scrutiny by the Internal Revenue Service recently. For an example, see “Bei, Kogaisha ni 87 Oku En Tsuicho, Fujitsu, Teiso de Taiko” (8.7 Billion Yen to U.S., Subsidiary--Fujitsu Responds to Suit), Nihon Keizai Shimbun, March 17, 1991, p. 1. It is important to point out in this connection that multinationals headquartered in the United States and elsewhere have long been the target of “transfer pricing” accusations by host country governments.
OCR for page 11
Japanese Investment and Technology Transfer: An Exploration of its Impact: Report of a Workshop enterprises as a whole. However, the rapid increase in Japanese FDI in recent years signifies that the situation is in flux and that the statistics should be interpreted with caution.21 There may be a time lag between establishing a foreign manufacturing base and performing R&D at the subsidiary. Table 4 shows that the amount of R&D undertaken by Japanese operations in the United States increased between 1987 and 1988. Indications are that Japanese companies are continuing to increase their investment in U.S. R&D activity.22 The question of whether R&D spending by the American subsidiaries of Japanese companies is a good proxy yardstick for inward technology transfer must also be raised. Some R&D expenditures made by the U.S. subsidiaries of foreign companies may represent technology monitoring or other activities that ultimately promote the flow of technology from the United States to the Japanese or other foreign parent.23 A positive relationship between the performance of R&D by subsidiaries of Japanese companies and technology transfer to the United TABLE 4 Expenditures on Research and Development (in $ thousands per worker) 1985 1986 1987 1988 1989* All U.S. Affiliates 1.83 1.98 2.02 2.04 2.08 All U.S. Affiliates (mfg) 3.08 3.55 3.61 3.78 3.88 U.S. Affiliates of Japanese Firms 1.26 1.32 1.01 1.34 1.53 U.S. Affiliates of Japanese firms (mfg) NA NA 2.88 NA NA All U.S. Manufacturing (Business expenditures) 2.60 2.74 2.87 3.11 3.33 Note: * preliminary data Source: Edward Graham, compiled from Bureau of Economic Analysisand National Science Foundation data 21 Edward Graham, from his presentation on “Overviews” at the Workshop on Japanese Investment and Technology Transfer: An Exploration of Its Impact. 22 Ibid. 23 See Evan Herbert, “Japanese R&D in the United States,” Research-Technology Management, November-December 1989, p. 11-20.
OCR for page 12
Japanese Investment and Technology Transfer: An Exploration of its Impact: Report of a Workshop States no doubt exists, but more study is needed before this complex process is fully understood. One final statistical measure of the possible contribution of Japanese firms to the level of technology in the United States is the output per worker, or productivity, in manufacturing operations. Since the transfer of both proprietary knowledge and managerial skills should eventually result in increased productivity, this may ultimately be the most useful measure. If Japanese management's vaunted skills in raising manufacturing productivity are transferrable across international borders, then the output per worker in manufacturing subsidiaries will be higher than U.S.-owned firms in the same industry. The statistics show that the productivity per worker of Japanese manufacturing subsidiaries is indeed higher than the average for all foreign affiliates and for U.S. manufacturing overall. Whether higher productivity is more a result of the relative “newness” of Japanese manufacturing facilities or of better management, it appears that Japanese affiliates are making a more than proportional contribution to U.S. manufacturing productivity. Until more recent data become available, we will not be able to assess the impact of the large acquisitions Japanese firms made after 1987. Some acquisition targets were “laggard” American firms.24 The TABLE 5 Value-Added (in $ thousands per worker) 1985 1986 1987 1988 1989 All U.S. Affiliates 47.1 48.4 48.1* NA NA All U.S. Affiliates (mfg) 43.4 47.1 48.6* NA NA U.S. Affiliates of Japanese Firms 66.3 64.8 59.1* NA NA U.S. Affiliates of Japanese firms (mfg) 42.7 47.3 49.8* NA NA All U.S. Manufacturing 39.3 42.4 44.8 47.2 48.3 Note: * preliminary data Source: Edward Graham, compiled from Bureau of Economic Analysisand Bureau of Labor Statistics data 24 Bridgestone's acquisition of Firestone may be a good example. See Thomas F. O 'Boyle, “Spinning Wheels-Bridgestone Discovers Purchase of U.S. Firm Creates Big Problems,” Wall Street Journal, April 1, 1991, p. 1.
OCR for page 13
Japanese Investment and Technology Transfer: An Exploration of its Impact: Report of a Workshop overall productivity growth of Japanese affiliates may stagnate in response to increased acquisition activity of this type.25 The real test will come a few years hence, when it can be determined whether or not Japanese management has succeeded in turning these acquired enterprises around. Part of the debate over Japanese FDI and the impact that it has on the U.S. industrial technology base relates to the adequacy of existing data. The Bureau of Economic Analysis (BEA) in the Commerce Department compiles the statistics, but some data series are not made public or disclosed to other parts of the government in order to ensure confidentiality of sources.26 Many Americans, including some members of Congress, believe that policymakers do not have adequate information on foreign investment and the activities of foreign affiliates for informed decision making. One key issue is that BEA data are reported on an enterprise rather than establishment basis. This hampers analysis at a detailed industry level. Legislation passed in November 1990 represents a compromise between the Administration and Congress, and is designed to allow the various data-collecting bureaus to make better use of the information that the government already possesses for the analysis of FDI.27 Recent administrations have been reluctant to adopt measures that contain disclosure requirements that place a differential impact on foreign firms. The fear is that such provisions would have a “chilling” effect on foreign investment and violate principles of national treatment. For the present, proposals affecting Japanese and other FDI will be debated and enacted on the basis of available data. Although there is nothing in those statistics that paints a particularly disturbing or encouraging picture of the impact of Japanese FDI on technology transfer, some experts warn of the danger of “flying blind,” and argue that a small commitment of resources to data collection and dissemination would significantly reduce the risk of policy mistakes based on incomplete information.28 25 Edward Graham, op. cit. 26 See Graham and Krugman, op. cit., pp. 135-143 (Appendix A). 27 The first report to Congress under the provisions of this legislation was released in August 1991. The “data link” project to allow Bureau of Economic Analysis data to be linked with Bureau of the Census and Bureau of Labor Statistics data is expected to produce concrete results in time for the next report, which is expected in late 1992. See U.S. Department of Commerce, Foreign Direct Investment in the United States — Review and Analysis of Current Developments (Washington, D.C.: U.S. Department of Commerce, 1991), pp. 88-89. 28 Edward Graham, op. cit.
OCR for page 14
Japanese Investment and Technology Transfer: An Exploration of its Impact: Report of a Workshop JAPANESE CORPORATE STRATEGY Concerns have been raised about large Japanese firms purchasing majority or minority interests in small, high-tech American companies in fields like semiconductors, software and biotechnology.29 Some see Japanese firms as “vacuum cleaners” that invest abroad in order to acquire foreign technology. What is the evidence for this assertion? There has certainly been an increase in the amount that Japanese companies have invested in U.S. high technology start-ups. According to Venture Economics, Japanese minority investments in U.S. venture businesses increased from $176 million in 1988 to $320 million in 1989.30 This trend has been most pronounced in California's Silicon Valley, where one publication recorded 42 cases of equity investment by Japanese companies in the area from 1986-1990.31 A recent study by the Economic Strategy Institute shows that Japanese investments made up more than half the total high technology investments (across all industries and regions) in recent years.32 These acquisitions may indicate that there are a number of Japanese companies, particularly in electronics, that are using direct investment to acquire technology as well as to diversify and further penetrate the U.S. market.33 Those who argue that “technology gathering” is not an inherent part of Japanese corporate strategy point out that the $320 million in minority investments in high technology start-ups constituted only about 2% of the $16 billion total for Japanese FDI during 1989. Experts on Japanese corporate strategy cite the figures on licensing transactions between Japanese firms and their U.S. subsidiaries that were 29 Frank Press, Scientific and Technological Relations Between the United States and Japan: Issues and Recommendations (Washington, D.C.: Commission on U.S.-Japan Relations for the Twenty First Century, November 1990), p. 1. 30 Venture Economics Inc., “Corporate Venturing News,” May 4, 1990, p. 1. The number of Japanese investments in small U.S. high technology start-ups declined from 60 in 1989 to 56 in 1990, with the amount falling from $320 million in 1989 to $306 million in 1990. It is expected that the number and amount of investments will post another decline in 1991, but the magnitude is uncertain. Communication from Venture Economics to Office of Japan Affairs, January 1992. 31 See “Digging Our Own Graves?” San Jose Mercury News, December 10, 1990, p. 7D. 32 Linda Spencer, Foreign Investment in the United States— Unencumbered Access, Economic Strategy Institute paper, May 13, 1991, Table 1. The Spencer report also contains a list of foreign investments in U.S. high technology companies. See “BT 100,” Business Tokyo, May 1991 for another example of a private attempt to compile information on this subject. This article contains a list of the top 100 Japanese investors in the United States. 33 See Phyllis A. Genther and Donald H. Dalton, Japanese-Affiliated Electronics Companies and U.S. Technology Development: 1990 Assessment, U.S. Department of Commerce, August 1991.
OCR for page 15
Japanese Investment and Technology Transfer: An Exploration of its Impact: Report of a Workshop TABLE 6 Partial List of Japanese Investments in Silicon Valley 1986-1990 Company Japanese Investor Equity amount or percentage Ardent Computer Kubota $51.5 million Akashic Memories Kubota 100% Aegis Olin Asahi 100% Advantage Prod. Tech. Sumitomo Metals 5% Auspex Systems Nissho Electronics 4.5% Auspex Systems Fuji Xerox 4.5% Atari Games Namco Ltd. NA AG Associates Canon Sales 30% Anacomp Magnetic Disc Division Hitachi Metals 100% C-Cube Microsystems Kubota 38.1% Centerpoint Computers Kyocera NA Cooper Lasersonics Ion Lasers Division Lexel Laserline 100% Domain Technology Kubota $66 million Exabyte Kubota $7.5 million Excel Microelectronics Rohm/Exar $5.7 million EICO Iwatani International NA Gould Nippon Mining 100% Komag Kobe Steel 20% Lam Research Sumitomo Metals 4.5% Literal Kawasaki Steel 21% Micro Mask Hoya $26 million Mips Kubota $36.6 million Maxoptics Kubota 25% Maxtor Kubota $17 million Mountain Computer Nakamichi 100% Materials Research Sony MRC 100% Mosaic Sumitomo Metals 10% Next Canon $100 million National Advanced Systems Hitachi 100% Novellus Systems Seki & Co. $1.9 million Poqet Computer Fujitsu 55% Precision Image Graphtec 100% Rasna Kubota 15% Siltec Mitsubishi Metals 100% Silicon Graphics NKK $35 million Sprague Semiconductor Sanken 100% Telemar Resources Chikyu Kagaku NA Telmos Rohm $1.5 million Varian Special Metals Division Tosoh 100% Verbatim Mitsubishi Chemical 100% Via Technologies Fujitsu Microelectronics 24% Vitelic Oki Electric NA Source: San Jose Mercury News, compiled from American Electronics Association Japan Office and UllmerBros. Inc. data
OCR for page 24
Japanese Investment and Technology Transfer: An Exploration of its Impact: Report of a Workshop The unmistakable conclusion is that almost twenty years of accumulated investment by Japanese companies in the U.S. consumer electronics industry has resulted in only limited technology transfer from Japan to the United States. The overall contribution to the American economy has been modest, particularly when compared to what the industry has contributed to the Japanese economy over the same period and the value of Japanese electronic products that American consumers have purchased over those two decades. Such skills as precision mechanical design know-how are almost absent in the U.S. plants operated by Japanese firms, but can be found in abundance in Japan.44 Increased domestic procurement, training and the manufacture of more sophisticated products by Japanese electronics companies in the United States has taken place only as a result of U.S. trade pressure. Semiconductors The Japanese semiconductor industry is the world manufacturing leader. 45 The status of the Japanese industry was achieved within a very short time period in the early and mid-1980s, mainly at the expense of U.S. firms. When discussing technology and technology flow in this industry, it is important to distinguish between the capital and technology intensive “front-end” steps in which circuits are processed onto the silicon wafer (“wafer fab”), and the less complex “back-end” assembly and testing steps. The less demanding back-end steps often take place in other locations. It is also important to distinguish between semiconductor design and manufacturing. The comments that follow focus primarily on semiconductor manufacturing. There are a number of factors that make the semiconductor business inherently global. The high cost of R&D and capital equipment must be 44 Ibid. 45 According to Dataquest, Inc., in 1990 Japanese chipmakers held 49.5% of the world market, compared to 36.5% for U.S. firms. The 2.6% Japanese loss of market share and the 1.6% gain for the U.S. were the first since 1982. See Evelyn Richards, “U.S. Firms Boost Share of Chip Market,” Washington Post, January 3, 1991, p. E1. These figures do not include “captive” production of semiconductors by firms for use in their own products. The exclusion of IBM, one of the largest semiconductor producers in the world, and other U.S. firms that do not sell their chips on the open market depresses the actual U.S. share.
OCR for page 25
Japanese Investment and Technology Transfer: An Exploration of its Impact: Report of a Workshop amortized against a global market.46 At the same time, firms must service regionally distinct markets for all but the most undifferentiated commodity products. This is especially true for application-specific integrated circuits (ASICs) and custom-designed chips. The raw materials and semiconductor manufacturing equipment markets are also regional because of the need for frequent, reliable service. Investment decisions are strongly influenced by the fact that technological costs are set at a global level and entry costs are set at a regional level.47 In view of these factors, it is not surprising that the semiconductor industry has witnessed a rise in the number of international alliances over the past several years. While many of these arrangements involve little direct technology transfer, the percentage that do is growing. 48 The rise of the Japanese semiconductor industry in the late 1970s and early 1980s was based on chips fabricated in Japan. Japanese companies did not begin to move the front-end process overseas until trade problems arose in the mid-1980s, and are still not as internationalized as their U.S. and European competitors. Front-end FDI was initially aimed at the United States, but there is increasing investment in Europe as Japanese firms seek to boost local content and establish a European image in anticipation of the unified European market. There are several exceptions to the general rule that Japanese FDI in semiconductor manufacturing has been largely motivated by political pressure. For example, the production of silicon wafers in the Pacific Northwest is advantageous because of low electricity costs. In the ASIC and custom chip markets the importance of contact with customers and rapid turn-around time necessitate a significant local marketing infrastructure and encourage the localization of some manufacturing capability as well. 46 These costs are likely to increase. A vice-president at Matsushita predicts that for the 256M DRAM, which will probably be introduced early in the next century, the cost for a wafer fab that could produce 10 million chips per month (twice the corporate-wide capacity of the largest 1M DRAM manufacturer today) could be as high as one trillion yen in current prices, or $7.1 billion at an exchange rate of 140 yen per dollar. See “Kane Kui Mushi Handotai” (Semiconductors: The Worm That Eats Money), Nihon Keizai Shimbun, September 19, 1990, p. 11. But U.S. experts call these estimates grossly inflated. A recent report by the National Advisory Committee on Semiconductors estimated the costs in 2001 of the equivalent of a 4 Giga-bit DRAM line at $2 to 3 billion. 47 Kenneth Flamm, from his presentation on “Industry Specific Experiences,” at the Workshop on Japanese Investment and Technology Transfer: An Exploration of Its Impact. 48 Ibid.
OCR for page 26
Japanese Investment and Technology Transfer: An Exploration of its Impact: Report of a Workshop FIGURE 3 Offshore Semiconductor Fabrication — Japan-Based Companies Source: Kenneth Flamm compiled from Dataquest data
OCR for page 27
Japanese Investment and Technology Transfer: An Exploration of its Impact: Report of a Workshop Technology acquisition can also motivate Japanese investment. There are examples of Japanese companies buying small, U.S. semiconductor firms in order to access their “embodied technology,” but it is fairly uncommon for technology acquisition to be the primary motive in semiconductor manufacturing. 49 Japanese investments that focus on software and design, however, appear to be motivated by technology acquisition. Japanese semiconductor firms also use FDI to acquire technology by setting up an R&D lab to plug into the local technology base. Does FDI in the United States by Japanese semiconductor manufacturing companies lead to an increased inward technology flow? There is undoubtedly some inward technology transfer resulting from the transfer of wafer fabrication — the front-end steps. Manufacturing low-cost semiconductors is as much an art as a science; important technology is embodied in the organization and management of the facility. If Japanese wafer fabs in the United States are near the state-of-the-art, that know-how is learned by U.S. production engineers.50 In the United States, where industry process technology has traditionally been transferred through labor mobility, it seems likely that some technology is flowing to U.S. companies through the movement of technical personnel between Japanese-owned and U.S. operations. In addition to manufacturing, many Japanese chipmakers are beefing up their R &D operations in the U.S. It is too early to tell whether these moves will lead to significant inward technology flow. One concern that U.S.-owned chipmakers have about Japanese FDI in the semiconductor manufacturing equipment industry, as opposed to FDI in semiconductor manufacturing itself, relates to “technology security.” Japanese equipment suppliers tend to move offshore with their customers, to which they are often linked by financial and other ties. A number of Japanese companies have acquired U.S. equipment makers. Even in non-acquisition cases, it appears that investment and increasing activity in the United States by Japanese semiconductor equipment and materials firms are leading to further competitive pressure on U.S.-owned suppliers. 49 Ibid. 50 Hitachi and NEC were scheduled to begin fabrication of 4M DRAMs at their U.S. facilities in the spring of 1991, a significant step in light of the fact that mass production of the 4M at their Japanese facilities was only launched in the autumn of 1990. See Mead Ventures, Inc., “Hitachi To Begin 4M DRAM Production in U.S.” East Asia High Tech Review, January 1991, p. 2. However, with a world-wide slump in semiconductor sales extending through the autumn of 1991, Japanese companies had announced 4M DRAM production cutbacks in their Japanese facilities, and plans to ramp up production in the United States appear to have been placed on hold.
OCR for page 28
Japanese Investment and Technology Transfer: An Exploration of its Impact: Report of a Workshop U.S. semiconductor manufacturers typically trust equipment suppliers to service installed equipment without telling competitors details of their manufacturing process. Because of the close ties between the Japanese equipment suppliers and semiconductor firms, U.S. manufacturers are concerned that they may not be able to place the same trust in Japanese equipment suppliers.51 In summary, there are both positive and negative implications for technology transfer of Japanese investment in the U.S. semiconductor industry. Japanese-owned wafer fabs may transfer some manufacturing technology to the United States, and R&D facilities may transfer more technology in the future. But the ancillary movement of Japanese semiconductor suppliers to the United States puts strong pressure on the U.S. equipment industry, which may result in long-term problems for domestic chipmakers. Japanese investments in design and software, not the subject of this discussion, appear to be motivated by the desire to strengthen Japanese industry in these areas. Japanese investment in this industry has been too recent, and evidence is still too limited, to draw a “bottom-line” conclusion. If it were possible to gather additional data and if there were no constraints on resources devoted to analysis, how would one answer the question of how much technology transfer is associated with Japanese investment in U.S. semiconductor manufacturing? One approach would be to examine in detail specific technologies (such as a particular process technology or quality control system) and trace transfer of technology among companies via movement of engineers, licensing, etc. This would provide a finer grained reading. Another approach might be based on econometric analysis. One might focus on measuring productivity of domestic firms and then assess the impacts of growth in nearby FDI on their productivity growth. Neither of these approaches would provide a complete answer to the question, but both promise to contribute to deeper understanding. 52 Automobiles In the semiconductor and consumer electronics industries, sophisticated processing and component steps can be distinguished from the relatively simple assembly portion of the manufacturing process. In contrast, automobile assembly is itself quite sophisticated. This guarantees a fairly high 51 Kenneth Flamm, op. cit. 52 Comments by Kenneth Flamm after the workshop.
OCR for page 29
Japanese Investment and Technology Transfer: An Exploration of its Impact: Report of a Workshop proportion of the finished product's value-added for the country in which a car is assembled. At the same time, auto manufacturers buy a significant proportion of their components from outside suppliers. The price and quality of components have a significant impact on the competitiveness of the end product. Management of the procurement process and supplier relationships is a crucial issue for auto companies.53 Management of relations with automobile components suppliers can embody “soft technology.” The question is whether standard Japanese practices in this area confer a competitive advantage on Japanese firms, and whether or not these practices are being transferred to the United States by the growing operations of Japanese auto companies here, the so-called “transplants.” A recent survey of U.S., Japanese and transplant operations conducted by Michael Cusumano and Akira Takeishi at MIT's Sloan School of Management addresses these questions. The Japanese transplants currently source critical components such as engines from Japan.54 The MIT study concentrated on four “medium-tech” parts (shock absorbers, gauge assemblies, front seat assemblies and instrument panels). These are not the highest value-added parts, which are usually imported from Japan, nor are they the components normally produced internally by U.S. automobile manufacturers. Automakers have considerable discretion in how much of the design work they assign to the suppliers of these parts. Two of the Big Three U.S. automakers, five of the six transplant operations and three of the top five Japanese automakers participated in the MIT study. Half of the U.S. sample of suppliers were divisions of the Big Three, one company was German-owned and the remainder were independent U.S.-owned. All of the parts were made in the United States. In the case of suppliers to the transplants, 37% were independent U.S. suppliers, and 85% of the parts were sourced in the United States. For the suppliers to Japanese plants, all the parts were sourced in Japan, mostly from independent suppliers. 53 Michael Cusumano and Akira Takeishi, The Global Challenge of Supplier Management: A Comparison of Japanese, Japanese Transplant, and U.S. Auto Plants, MIT Sloan School of Management Working Paper #3158-90/BPS, 1990, p. 1. Also available as “Supplier Relations and Management: A Survey of Japanese, Japanese-Transplant, and U.S. Auto Plants, ” Strategic Management Journal, Volume 12, No. 7 (November-December) 1991. 54 Nissan has announced plans to follow Honda and open an engine plant in the United States. See “Nissan, Bei ni Enjin Kojo” (Nissan, Engine Factory in the U.S.), Nihon Keizai Shimbun, January 17, 1991, p. 13. Toyota has announced plans to export engines from its Kentucky factory to Japan. See “Toyota, Bei ni Shinkojo” (Toyota, A New Factory In the U.S.), Nihon Keizai Shimbun, November 28, 1990, p. 11.
OCR for page 30
Japanese Investment and Technology Transfer: An Exploration of its Impact: Report of a Workshop The study indicates that Japanese supplier management practices do differ from American practices and that the differences appear to give Japanese manufacturers an advantage.55 The Japanese procure more parts from outside suppliers, have fewer suppliers per part, maintain longer relationships with suppliers, and assign more of the design and engineering of parts to suppliers than the American manufacturers. An important aspect of Japanese supplier management is the effort that is made by the manufacturer to work closely with the supplier to reduce prices and defects over the lifetime of a model. The Japanese tendency toward longer contracts and fewer suppliers per part leads to other positive results, such as earlier supplier involvement in design. This helps shorten the time needed for product development. The “problem,solving” approach common in Japan can be contrasted with the “bargaining” orientation of American firms.56 In general, the supplier management practices of the transplants fall between the U.S. and Japanese averages, showing more similarities to the practices of Japanese companies operating in Japan. The transplants had even fewer suppliers per part than the Japanese plants, and their contract lengths were longer than the U.S. average. They also appeared to follow the Japanese practice of giving suppliers more frequent feedback and suggestions for improvement than was customary for the U.S. manufacturers. The survey measured the performance of suppliers on price and quality dimensions. In contrast to the Japanese plants, suppliers to both U.S. and transplant manufacturers did not meet initial price targets on average. However, the transplants achieved a certain amount of success in reducing prices over time, while prices rose at the Big Three plants. The survey yielded interesting results on the quality dimension as well. Suppliers to the Big Three plants had an average of 1.81% defects. Suppliers to the transplants had a lower defect rate of .05%; the parts supplied to Japanese plants—which were sourced exclusively in Japan— had an even lower rate of .01%. Over the model lifetime, defects dropped at a 30% annualrate at the transplants, 9.5% at the Japanese plants and only 1.7% for the Big Three suppliers. 55 Cusumano and Takeishi, op. cit., p. 46. 56 It may also be the case that these business relationships lead to anticompetitive practices and higher prices for consumers. A recent survey by the U.S. and Japanese governments found that installed and uninstalled auto parts were significantly more expensive in Japan than in the United States. See U.S. Department of Commerce, “Summary Analysis, DOC/MITI Automotive Parts Price Survey,” June 27, 1991.
OCR for page 31
Japanese Investment and Technology Transfer: An Exploration of its Impact: Report of a Workshop TABLE 9 Summary of Main Survey Results: Supplier-Manufacturer Relations Dimension U.S. Japan/U.S. Japan Number of Suppliers Per Part Most (1.8) Fewest (1.2) Fewer (1.3) Type of Suppliers In-house and U.S. Independent Affiliated U.S./Japan U.S. Independent Independent & Affiliated Japanese Length of Relations Long (> 10 yrs.) Short (1-4 yrs.) Long (> 10 yrs.) Length of Contracts Short (1.7 yrs.) Longer (2.5 yrs.) Longest (3.2 yrs.) Length of Parts Transactions Long (3.2 yrs.) Short (1.6 yrs.) Longest (3.5 yrs.) Selection Criteria (Emphasis) Past contact, Affiliation Price & Quality Price Role in Development Inquiries at 30 months & selection at 26 months Inquiries at 33 months & selection at 24 months Inquiries at 27 months & selection at 23 months 70% Blackbox 30% Detail-controlled 64% Blackbox 23% Detail-controlled 96% Blackbox Source: Michael Cusumano The authors concluded that the low defect rate achieved by the transplants was partly due to their utilization of parts sourced in Japan and from Japanese-owned suppliers that followed the manufacturers to the United States. However, 85% of the parts covered in the survey were sourced in the U.S., and 37% came from independent U.S. suppliers. Further, U.S.-based suppliers performed better on price and quality for the transplants than they did supplying the Big Three. Anecdotal evidence indicates that this gap in performance may be partially due to more stringent inspection of parts destined for the transplants rather than to real process and productivity improvement. The price and quality improvements over time that were achieved by the independent U.S. suppliers to the transplants indicates, at least to a certain extent, that the “soft technologies” which comprise Japanese supplier management have a positive impact on performance even in the American context. The results also suggest that technology is being transferred to the U.S. operations of the Japanese automakers. There are positive implications for the quality and price of U.S.-made automobiles.
OCR for page 32
Japanese Investment and Technology Transfer: An Exploration of its Impact: Report of a Workshop TABLE 10 Summary of Main Survey Results: Supplier Performance Dimension U.S. Japan/U.S. Japan Pricing Practices Above Target (109%) Above Target (110%) Below Target (98%) Price Changes Increasing (+0.9%) Decreasing (−0.4%) Decreasing (−2.1%) Defect Rates High (1.81%) Low (0.05%) Lowest (0.01%) Defect Rate Change Decreasing (−1.7%) Decreasing (−30.1%) Decreasing (−9.5%) Information Exchanges Low, mainly statistical process control Higher, more on process steps Higher, more on process steps and costs Supplier Suggestions Few Many Many Source: Michael Cusumano In response to survey questions, the managers of the transplant factories pointed out a number of factors that prevented them from performing even better. Difficulties encountered in the American environment included insufficient product design capability of American suppliers, deficiencies in the abilities of second-tier U.S. suppliers, and a limited manufacturing infrastructure for dies and fasteners. American and Japanese suppliers have different expectations regarding profits and pricing, which raised obstacles for the transplants. Traditional American practices dictate higher profit margins for suppliers, annual price increases for the life of the model and higher prices for higher quality output. Japanese suppliers make lower margins and see no conflict between lower prices and higher quality. Another interesting finding of the survey is that the supplier management techniques used by Japanese companies and their U.S. subsidiaries are being transferred to the Big Three through competitive pressure. According to the evidence compiled by Cusumano and Takeishi, there is a marked trend among U.S. manufacturers toward fewer suppliers per part and longer contract lengths. The survey also found evidence that U.S. automakers are making progress in obtaining higher quality and lower priced parts.
OCR for page 33
Japanese Investment and Technology Transfer: An Exploration of its Impact: Report of a Workshop TABLE 11 Early Versus Later 1980s Comparisons U.S. Sample Japanese Sample Pre-1987 Post-1986 Pre-1987 Post-1986 Average Sample Size 8 10 7 20 Performance Variables Target Price Index 105.0* 116.4* 100.7 97.6 Price Change % 2.38** .148** −2.08 −2.07 Defect Rates 2.60 1.37 .014 .008 Defect Rate Change % −.590 −1.73 −14.3 −7.58 Descriptive Variables Suppliers Per Part 4.75 1.42 1.00 1.40 Contract Length (years) 1.00 1.58 3.08 3.25 Suggestion Index 2.75 2.46 4.00 4.80 Supplier Selection Criteria Initial Price Offered 3.63 4.36 5.00 4.90 Target Price Ability 3.88 4.27 4.43 4.75 Quality 4.75 4.82 4.58 4.75 Past Relations 4.63 4.27 3.29 3.05 Financial Affiliation 4.13 3.91 1.86 2.25 (scaled 1-5, 5 = very important; 1 = not important) Notes: All variables showed significant differences at 0.10 between pre-1987 and post-1986 responses for the combined sample. Pre-1987 = responses on models introduced before 1987; Post-1986 = responses on models introduced after 1986. * = Difference significant at 0.10 ** = Difference significant at 0.05 Source: Michael Cusumano Overall, the evidence from the automobile industry is encouraging because it indicates that U.S.-based suppliers and the Big Three are capable of change in the direction of superior Japanese practices. 57 Some Japanese automakers are making well-publicized efforts to build long-term relationships with U.S. suppliers.58 In addition, indications are stronger in the auto 57 There is evidence that the adoption of just-in-time inventory management by American companies — both auto makers and firms in other manufacturing industries — was spurred by the transplants as well. See Gary R. Saxonhouse, “Sony Side Up: Japan's Contributions to the U.S. Economy,” Policy Review, Spring 1991, p. 63. 58 Paul Blustein, “Doing Business the Toyota Way-Firm Offers Advice to Possible U.S. Suppliers,” Washington Post, November 1, 1990, p. C1.
OCR for page 34
Japanese Investment and Technology Transfer: An Exploration of its Impact: Report of a Workshop industry than they are in semiconductors or consumer electronics that the transplants are moving in the direction of greater vertical integration and increased technical activity in their U.S. operations. 59 But doubts remain about how far and how fast technology transfer from transplants to the Big Three will proceed. Technology transfer from Japan to the United States may be indispensable for the transplants to enjoy the same manufacturing edge that Japanese companies do at home.60 Whether or not “soft technology” is transferred from Japanese to U.S.-owned organizations and the impact of this technology transfer on the competitiveness of the domestic industry may depend to a great extent on whether U.S. companies — managers, unions, and suppliers — are flexible enough to change traditional relationships. 59 In addition to the reports on Nissan and Toyota's plans to build engines in the United States cited above, Honda plans to increase the amount of R&D it does in the United States See “Honda Giken: Bei no Kenkyu Kaihatsu Taisei o Kyoka” (Honda: Strengthening Its U.S. R&D Organization), Nihon Keizai Shimbun, September 10, 1990, p. 9. 60 Robert Lawrence, from his presentation on “Overviews” at the Workshop on Japanese Investment and Technology Transfer: An Exploration of Its Impact.
Representative terms from entire chapter: