|Make your work easier and more efficient installing the rrojasdatabank toolbar ( you can customize it ) in your browser.|
Left for Dead: Asian Production Networks and
|Network Interface Card||34%|
|Switching Power Supply||31%|
Source:Market Intelligence Center/III
Not surprisingly, given its deep tie in this period to US producers, as OC electronics
activity began to emerge it was concentrated in the personal computer (PC) and PC-related
product-markets. In turn, the nerve-centers of OC activity in PC electronics in this era
were Taiwan and Singapore, the home bases for emerging Asia-Pacific MNCs like the former's
ACER and the latter's Creative Technologies.(29) Taiwanese
producers were at the heart of the nascent alternative supply base. Ultimately, their
position would crystallize in the 3rd period, culminating by the mid-1990s, as Table 1
shows, with significant to dominant world market shares in 14 PC-related supply
categories. Singapore-based OC producers similarly began to emerge in this
second period as significant suppliers of hard disk drive-related components and services,
and of multimedia sound cards, PC subassemblies and PC assembly services.(30)
By contrast to both the US and OC developments in this second phase, the pattern of Japanese investment led to a dual production structure under the control of Japanese firms and premised on traditional product cycles--sophisticated products were produced at home with sophisticated processes to serve advanced country markets, while lower-end products were produced with simple processes in regional affiliates to serve local Asian markets. Both sets of operations sourced from a common supply base, located largely in Japan and controlled, directly or indirectly, by Japan's major electronics companies. Where Japanese companies responded to government or commercial pressures to localize, they did so, as Ernst suggests, from within their established supply base--that is, by transplanting the operation of an affiliated domestic Japanese supplier--not by sourcing locally from the emerging Asian supply base. In short, the Japanese production networks boasted redundant investment and remained relatively closed, even as the US networks became more open and entwined with indigenous OC producers, and more specialized.
These trends were fully elaborated during the third stage, from 1985-early 1990s. At home, US firms focused scarce corporate resources more intensely on new product definition and the associated skills (e.g., design, architectures, software) necessary to create, maintain and evolve de-facto market standards. In turn, they upgraded their Asian affiliates, giving them greater responsibility for hardware value-added and manufacturing, and significantly increased local sourcing of components, parts, and subassemblies. They even contracted-out design and manufacture of some boards and components. Thus, during this period, the Asian affiliates of US firms continued to migrate from PCB to final assembly with increased automation; to increase both component production and final system value-added; and to assume global responsibility for higher value-added systems (e.g., from monochrome desktops to color notebook PCs). Their production networks extended to more and more capable local Asian producers who became increasingly skilled suppliers of components, subassemblies and, in some cases, entire systems. Even in areas like memory chips and displays where Japanese firms remained important suppliers to US firms, there was sufficient competition from other Asian sources (e.g., Korea in memory chips) or sufficient political pressure to keep the supply architecture open.
Major US producers of PCs like Apple illustrate well these developments.(31) Apple Computer Singapore (ACS) opened a PCB assembly plant for the Apple II PC in 1981. By 1983 nine local companies were contract manufacturing PCBs for the Apple IIe and Lisa PCs. By 1985 ACS was upgraded to include final assembly of Apple IIes for the world market. From 1986-89, ACS was expanded and upgraded to begin some component design work. In 1990 ACS assumed final assembly responsibility for two of three new Macintosh PCs (and PCBs for the third) and designed (locally) and manufactured associated monitors. By then, essentially all components were sourced in Asia (except the US-fabbed microprocessor)--ACS's 130 major suppliers included local firms like Gul Technologies and Tri-M (PCBs). ACS had also demonstrated that its growing technical prowess could pay competitive dividends in speeding time to market: It was able to move from designs to production roll-out in up to half the time of Apple's other facilities. By 1992, ACS assumed responsibility for final assembly for all Asia-Pacific markets, including Japan, was designing and supplying boards globally, manufacturing monitors and some peripherals, and designing chips. Over $1 Billion was being procured annually through ACS. In 1993, ACS set up a design center for Macs for high-volume desk-top products--Apple's only hardware design center outside the US. By 1994, ACS had become the center for distribution, logistics, sales and marketing for the Asia-Pacific region, and was assembling the MacClassic II, LC III and IV, mid-range Centris, and Quadra 800 for global distribution. Regional sourcing reached $2 billion, half from Japan (LCD displays, peripherals, memory, hard disk drives), another quarter from Singapore, $250-500 million in Taiwan for OEM desktops, monitors, PCBs, Powerbooks, Digital Assistants, and chips. Korea's Goldstar also supplied monitors. By late 1994, ACS had begun to design the motherboard and tooling for, and assemble the multimedia system Mac LC 630 PC for worldwide export. Two new Mac products completely designed and manufactured at ACS were launched in 1995.
The value-added/local sourcing progression of other major US electronics
players in Asia is broadly similar.(32) For example,
Compaq Asia (hereafter: CAS for Compaq Asia-Singapore) established its Singapore factory
in 1986 for PCB assembly of components sourced from Asia (including Japan), for desktop
PCs to be final assembled in the US. By 1994, after terminating an OEM relationship with
Japan's Citizen Watch, CAS was designing and manufacturing all notebook and portable PCs
for worldwide consumption, and all desktop PCs for the Asia-Pacific. Similarly,
Hewlett-Packard's Singapore operations evolved from assembly of calculators in 1977 to
global responsibility for portable printers and Pentium desk-top PCs and servers, with
local manufacturing, process design, tooling development, and chip design. Motorola's
Singapore operations evolved from simple PCB assembly of pagers and private radio systems
destined for the US in 1983, to world-wide mandates for design, development and automated
manufacture of double-sided six-layer PCBs, for design and development of integrated
circuits for disk drives and other peripherals, for some R&D, and for sourcing of at
least $500 million of parts and components within the region. Similar kinds of stories
could be told for AT&T in telecommunications products, IBM and DEC in PCs and
peripherals, Maxtor, Connor, Seagate, and Western Digital in hard disk drives, and for TI,
Intel and National Semiconductor. Table 2: Taiwan Firms' 1994 OEM Relations in
PC-Related Products (representative sample)(33)
|Acer||Apple, Fujitsu, NEC, NCR, Data General, Siemens||Notebooks and/or monitors|
|Delta||Apple, Compaq, IBM,||Power Supplies|
|Elite||DEC, IBM, NEC, Siemens||Motherboards|
|FIC||ATT, Dell, Unisys||Motherboards|
|Inventa||Apple, Compaq, Dell,||PDA, Notebooks|
|Lite-on||Compaq, DEC, Dell||Power Supplies and/or Monitors|
|Tatung||Apple, Packard Bell, NEC,||PCs and/or Motherboards, Monitors|
Source: MIC/III and press accounts
As US Asia-based affiliates up-graded and specialized in this way during
the third period, their indigenous OC suppliers followed suit. Table 2 gives some
indication of this by examining the emergence, by the period's end, of OEM relationships
with major China Circle producers. In turn, by leveraging their link into the US
production networks and the global distribution capabilities thereby provided, the
strongest China Circle producers began to control their own production networks. Chung
Chin's chapter suggests this in showing how, in the early 1990s, intense competition and
growing needs for scale-intensive investment forced a shake-out and consolidation among
Taiwanese and Hong Kong-based electronics firms.(34) Firms
like ACER, the Formosa Plastics Group, and Tatung began to ride herd on an extensive
indigenous supply base of thousands of small and medium-sized design, component, parts,
subassembly and assembly houses throughout the China Circle and extending into Southeast
Asia. These firms form an intricate sub-contracting structure of affiliated and family
enterprises which comprise the local production network and supply base. The numerous
small firms are aligned vertically with the few larger scale enterprises that act as
intermediaries for foreign MNC customers.(35) Designs and
key components flow down from the large-scale enterprises; more labor-intensive production
activities flow up along the subcontract network leading to final assembly. Table
3: Domestic vs. Off-shore Production Value of Taiwan's Electronics Industry, 1992-1995
Toward the end of the third period, in response to steep rises in factor input costs in
the NICs, and exacerbated by currency appreciation, these emerging OC production networks
become more and more regionalized. For example, Table 3 suggests the extent to which
considerable PC-related production is now being carried on by Taiwanese MNCs within the
region but outside of Taiwan. As the table suggests, production outside of Taiwan but in
the OC Asian networks accounted for increasing shares of total production under Taiwanese
control, approaching one-quarter of the total in 1995. As Chung Chin argues in her
Chapter, the off-shore activity is concentrated in certain product segments, with about
two-thirds of 'Taiwanese' production of keyboards, half of power supplies, and about a
quarter of monitors and motherboards now taking place outside of Taiwan. Investment
targeted both Mainland China and Southeast Asia--partly a result, as other chapters
suggest, of the timing of both Taiwanese and Mainland policy reforms, and partly of
prudent geographical risk-spreading by OC investors.(37)
In sum, by the early 1990s, the division of labor between the US and Asia, and within Asia between affiliates and local producers, deepened significantly, and US firms effectively exploited increased technical specialization in Asia. In stark contrast, up through the end of 1993, Japanese firms still controlled their Asian affiliates' major decision-making and sourcing activities from Japan. More low-end process/product technology had been off-shored, including production of audio systems (cassette recorders, headphones, low-end tuners, etc.), under-20-inch televisions and some VCR models, cameras, calculators and appliances like microwave ovens. Local Asian content had risen toward 60%, but core technological inputs like magnetrons, chips and recording heads were exclusively sourced from Japan, and the 60% 'local' content was mostly supplied by the off-shore branch plants of traditional domestic Japanese suppliers. Local design activities were invariably to tailor Japanese product concepts for local Asian markets, and global mandates for advanced products, let alone their design, development, and manufacture, were nowhere to be found outside of Japan. In contrast to US producers, for example, Japanese PC producers sourced displays, memory, some microprocessors, drives, power and mechanical components, plastics, and PCBs from Japan (or in the case of some low-end components, from off-shore affiliates), and did PCB and final assembly, and essentially all advanced design and development in Japan. In short, Japanese firms intensified rather than rationalized their dual production structure, and, by exclusion from their production networks, failed to benefit from increasing, cheaper, and faster technical capabilities in the rest of Asia.
In Asia today, beneath the superficial similarity engendered by aggregate trade and investment data and macro-analyses, lie distinctly different electronics production networks under the control of US, Japanese and OC multinationals. The US networks tend to be open to outsiders, fast and flexible in decision-making and implementation, structured through formal, legal relationships, and capable of changing contour (and partners) as needs change--in an image: open, fast, flexible, formal and disposable. Their activities are centered in the NICs, especially Singapore, but increasingly reach into the rest of Asia and China. By contrast, the Japanese networks tend to be relatively closed to outsiders, more cautious to make and implement decisions which are generated from Japan, and structured on stable, long-term business and keiretsu relationships--that is, closed, cautious, centralized, long-term and stable. Despite the recent surge of Japanese investment into Asia, their networks are still most definitely centered in Japan.
The respective networks also rely on distinctively different supply bases, boast different product mixes, and, most significantly, constitute very different divisions of labor. The US networks rely on an open, competitive supply architecture in which Japanese, US, Taiwanese, Singapore, Korean and other Asian firms compete on cost, quality and time-to-market and, in some cases, provide significant value-added. By contrast, the Japanese networks rely on a largely domestic and affiliated supply base with little value-added by other Asian producers. The US networks produce (and in some cases design and develop) increasingly sophisticated industrial electronics like hard disk drives, PCs, InkJet Printers, and telecommunications products. The Japanese networks still mostly produce consumer audio-visual electronics and appliances. The US networks exploit a complementary division of labor in which US firms specialize in especially 'soft' competencies (definition, architecture, design--standards areas) and Asian firms specialize in hard competencies (components, manufacturing stages and design/development thereof). By contrast, the Japanese networks exploit a division of labor with significant redundancies in which domestic Japanese operations produce high-value, high end products using sophisticated processes, and off-shore affiliations produce low-value, low-end products. The US networks exploit increasing technical specialization throughout the production process in which the Asian contribution is maximized; the Japanese networks exploit a value-added specialization between products in which the Asian contribution is minimized.
By comparison, the emerging OC networks appear to combine features of both the Japanese and US MNC approaches, with distinctive characteristics of their own. Much like the Japanese, OC networks are difficult for outsiders to penetrate. Much like the US, OC networks are fast and flexible. Indeed, industry estimates of OC network business speed peg the time from conception to execution at a fraction of that of larger MNCs burdened with formal organization and layered decision-making.(38) In some cases, OC networks can design and execute in less time than it takes the Japanese giants just to make a go-ahead decision.(39) For the Taiwanese design houses in particular, this capability is apparently built on a high-value-added foundation, macro-cell based design methodologies and libraries of already-characterized component functions that can be combined and altered to implement new concepts.(40) The rapid design capability then joins with the hyper-competition among subcontractors in the network to implement the new designs as fast as possible.
Unlike either the US or Japanese networks, the OC networks seem especially focused on
intricate division of production tasks (e.g., components and subassembly steps) that can
be farmed-out all the way down to family job shops and home-workers. Individual units
within the network operate at small scale with minimal capital investment requirements,
and link on the informal bases of guanxi, that is, kinship or friendship ties. The
flexibility that results, mirroring the industrial district capabilities in Italy and
parts of Germany, makes it possible to increase or decrease production scale on short
notice, or to enter and exit niche product-market segments, all at minimal cost and with
minimal fixed investments.(41)
The best OC networks also run extremely lean in general, sales and administrative overheads where they match the best practices of MNC leaders like Hewlett-Packard (at about 10% of sales for microcomputers and printers), and are far superior to most advanced MNC performers (15%-upwards of 20% of sales).(42) Of course, such cost-minimization is inherent in the sub-contract structure of the OC production networks where affiliates and family enterprises can be squeezed (if necessary, in time-honored sweat-shop manner
In short, the OC networks appear to be insular, fast, flexible, guanxi-mediated, and fluid. They tend to be centered in the China Circle and increasingly focused on Mainland China. Like the Americans, the OC networks seek to exploit a highly competitive supply base and concentrate on industrial electronics. Much like the Japanese, OC networks retain in the home base high value-added products manufactured with more advanced processes, and off-shore to cheaper production locations lower value-added products assembled with simpler processes. Unlike the Japanese, however, the OC networks also self-consciously leverage increasing technical specialization through local relationships wherever possible. And unlike both, the OC network relationships are increasingly China-centered--rather than using a NIC base as the regional center, OC networks may end up with a China base as their global center, using demand and technical know-how in the domestic China market to achieve world-class scale, costs and innovation.
As argued at the outset, the competitive consequences of the differences between US, Japanese and OC networks have been significant. The US networks relieved the constraining threat of competitive dependence on Japanese rivals by re-constituting the architecture of supply in electronics. Simultaneously, the turn to skilled but cheaper Asian suppliers helped to lower overall production costs, fierce competition within the supply base helped to reduce turnaround times, and specialization and diversity within the network permitted US producers to keep better pace than Japanese rivals with rapid technological and market shifts. Growing Asian technical capabilities freed US firms to focus their efforts (and scarce resources) on new product definition and standards competitions, systems integration, software value-added and distribution. In the bargain, the US networks helped to spawn and sustain direct Asian competition to Japanese firms in several of their stronghold markets like memory chips, consumer electronics and displays. And while OC network capabilities grew prodigiously, they did not directly challenge revived US leadership in the last round of competition. Overall, US firms not only stayed the competitive course, they prospered.
Yet, the current US position is no more a guarantee of future success than was Japan's in the early 1980s. Much depends on how Japanese firms respond to their current competitive dilemmas and on how OC firms leverage opportunities in the China Circle.
The Ernst and Huchet chapters provide evidence of nascent Japanese adjustment which, at
least at first blush, appears to draw a different image from the closed network structure
emphasized here. Ernst and Huchet see some evidence of increased openness and increased
reliance on OC, Chinese and Korean suppliers as Japanese firms adjust to the competitive
success of US and indigenous Asian producers and target the China market. Whether those
changing characteristics are permanent or temporary is very much an open question,
however. In 1996 there has been anecdotal evidence that Japanese networks are snapping
back toward the more traditional, closed model as the yen again depreciates and as
Japanese firms absorb know-how from the partners they took on in Asia. In any case, there
is no evidence that the basic Japanese strategy of control of value-added through
ownership has changed; nor do Japanese firms appear intent upon exploiting increased
specialization in the rest of Asia wherever they can do the specialization themselves. And
whatever the precise characterization, neither Ernst, Huchet nor I would expect much
convergence of the Japanese and US models.(43)
They would also agree that the precise characteristics of Japan's Asia-based networks
that created vulnerability over the last decade--closed, cautious, Japan-centered,
long-term and stable--could be turned into competitive strengths with a dose of
rationalization and a pinch of vision. Japanese firms could decide to accept slower
domestic growth and the need to exploit technical capabilities in the rest of Asia as
givens. They could decide to selectively incorporate Asian producers into the family and
build stable, long-term, mutually advantageous ties focused on exploiting specific
technological capabilities in other parts of Asia. They could decide to invest for the
long-term. They could decide to drive their growth from Asia's: If Asia becomes a launch
market for new product concepts--and it's rapid growth and burgeoning wealth suggest that
it must in some market segments--Japanese firms might just then be better positioned to
exploit the development.(44)
Just as big a competitive wild card is the growing electronics capability in the China Circle linked to OC investments in the US, Southeast Asia, and eventually Japan.. A competitive China Circle scenario is easy enough to describe: The combination of Hong Kong-based financial and producer services, with Taiwan-based digital product and process design, Southeast Asian component specialization, highly skilled but cheap Mainland labor, and, of course, the Mainland market, provides a tantalizing scenario for regional dominance. The OC network characteristics identified above--insulated from outside control, fast, flexible and fluid--appear to be a compelling mix for exploiting the region's possibilities. And the sheer scale of production for the mainland and, from the mainland, for overseas markets would dwarf the leverage provided by any other home market base. To this potent brew should be added the self-conscious developmental intent of governments throughout the region to nurture indigenous capabilities, and of China's to move to the technological frontier as fast as possible.
The quite significant constraints on the emergence of such a scenario should not be underestimated, however. Unlike the Americans, who have retained capability in most core component technologies and a significant though diminished position in capital goods, the OC networks remain dependent on Japanese competitors for advanced manufacturing equipment and high value-added core components (e.g., for Taiwanese producers, $500 million of LCD displays and $3 Billion of memory chips in 1994). Even more of a constraint, however, is continuing dependence on the American networks for microprocessor architectures, advanced product concepts, and global distribution. It is likely that the Chinese market can eventually help to break those constraints--by providing the returns to invest to relieve core component dependence, the new product concepts that can become global standards, and leverage to develop indigenous brands and global distribution channels. But that is likely to take time, probably several decades. In the interim, the China Circle will witness one of the great market battles in memory as US, Japanese and indigenous production networks vie for 21st Century advantage.
1. This chapter is drawn from a larger work in progress on global competition in electronics. See, Michael Borrus, Punctuated Equilibria in Electronics: Microsystems, Standards' Competitions, and Asian Production Networks, forthcoming, 1997.
2. The government's support took two forms--direct financial support of $100 million per year to the industry's manufacturing technology consortium, Sematech--or half of Sematech's annual budget--and negotiation of the US-Japan Semiconductor Trade Agreement (STA). For details, see Michael Borrus, Competing for Control: America's Stake in Microelectronics, (NY: Ballinger, 1988). For an elaborate analysis of the STA see Kenneth Flamm, Mismanaged Trade? Strategic Policy and the Semiconductor Industry, (Washington DC: Brookings Institution Press, 1996)
3. This position is argued explicitly by industry consultant William F. Finan and his academic collaborator Jeffrey Frey in their Nihon no Gijyutsu ga Abunai: Kenshő, Haiteku Sangyő no Suitai [Japan's Crisis in Electronics: Failure of the Vision] (Tokyo: Nikkei Press, 1994).
4. Steven Jay Gould makes this point with respect to natural selection in Wonderful Life (NY: W.W.Norton, 1989) at p.236.
5. For a fuller analysis, see Borrus, Competing for Control, supra. The domestic market served as a launch market during the late 1970s-1980s for, among other products, the VCR, Camcorder, Walkman, hand-held TV, fax machine, portable copier, and notebook PC.
6. On the latter point, see Ichiro Uchida, "Restructuring of the Japanese Economy" in Eileen Doherty, ed., Japanese Investment in Asia: International Production Strategies in a Rapidly Changing World, (U.C. Berkeley: Berkeley Roundtable on the International Economy, 1995).
7. On the former point, see Uchida, supra. As argued below--and somewhat at odds with Ernst's perspective--rationalization does not necessarily imply radical change in the way Japan's industrial firms operate. For example, some Japan scholars like Greg Noble argue that Japan's lifetime employment system appears to have survived the recent economic shocks essentially intact, with only marginal modifications. Verbal remarks of Noble at the BRIE/Asia Foundation Conference, "Competing Production Networks in Asia," April 27-28, 1995.
8. By 'open', I mean that key product specifications, especially the interface specifications which permit interoperability with the operating system or system hardware, are published or licensed and thus available to independent designers of systems or software who can produce complementary or competing products.
9. By production network I mean the organization, across national borders, of the relationships (intra and increasingly inter-firm) through which the firm accomplishes the entire value chain of production including research and development, product definition and design, supply of inputs, manufacturing, distribution, and support services. Especially significant in Asia, are supplier relationships that include subcontracting, OEM (original equipment manufacturing), and ODM (original design manufacturing) arrangements between foreign MNCs and domestic suppliers of intermediate production inputs, such as materials, tools and molds, parts and components, subassemblies, and software--some of whom may also compete in final product markets. See the elaboration in Dieter Ernst "Networks, Market Structure and Technology Diffusion: A Conceptual Framework and Some Empirical Evidence," report prepared for the OECD, Paris, 1992
10. The next section defines the concept of supply base.
11. By 'continuous innovation,' I mean the capacity to add incremental advances in performance, functionality, or features within or between given product generations--e.g., from 75mhz to 250 mhz pentium microprocessors, or from 25mhz 386-based PCs with 4Megabytes of RAM and 100 Megabyte hard drives to 200mhz pentium-based PCs with 32 Megabytes of RAM, 2 Gigabyte hard drives and CD-ROM drives.
12. Applying and exploring the limits of this method are principle goals of BRIE research supported by the Alfred P. Sloan Foundation.
13. See the discussion in Laura Tyson, "They Are Not US," The American Prospect, (Winter, 1991), p.37ff. See also, Yao-Su Hu, "Global Corporations Are National Firms with International Operations", California Management Review, Winter, 1992, p.107ff. Of course, the debate addressed by these articles was popularly launched by Robert Reich in "Who is Us?" Harvard Business Review, (Jan.-Feb., 1990) p.53ff. More generally, the recent collection by Dennis Encarnation and Mark Mason, ed., Does Ownership Matter? (NY and London: Oxford University Press, 1995) supports the persistence of important differences among multinational firms based on national origin.
14. See "A Survey of Multinationals" The Economist, March 27,1993, p.6-7, citing United Nations data. The major exceptions are oil companies (because oil fields tend to be located abroad) and small country multinationals like Nestle, Unilever and ABB (because their markets are located abroad)--and the latter would fall into the 60-90% range if Europe was treated as their home base. By that measure, the most non-oil MNC is IBM, with about 50% of assets outside of the U.S. But because half of its assets are still concentrated in the U.S., even IBM can be said to have the U.S. as its home base.
15. This conclusion is easily reached from industry conversations and even a quick perusal of the annual reports of the 1000 largest US and 1000 largest non-US firms. More generally, the evidence in John Dunning's comprehensive work on MNCs supports this conclusion, as does Michael Porter's work. See, John Dunning, Multinationals, Technology, and Competitiveness,(London: Unwin Hyman, 1988); Michael Porter, The Competitive Advantage of Nations, (London: MacMillan, 1990).
16. My colleague Stephen Cohen coined the concept of national market logics in Borrus, Cohen, et.al, "Globalization and Production," BRIE Working Papers, #45 (Berkeley: BRIE, 1991).
17. For a discussion of this concept of technology trajectories see Michael Borrus, "The Regional Architecture of Global Electronics: Trajectories, Linkages and Access to Technology" and the sources cited there, in Peter Gourevitch and Paolo Guierrieri, eds., New Challenges to International Cooperation: Adjustment of Firms, Policies, and Organizations to Global Competition, (San Diego: UCSD, 1993).
18. For one effort to elucidate some of these variables--the state, labor relations and financial systems--as part of a formal analytic explaining national economic development, see John Zysman, "How Institutions Create Historically Rooted Trajectories of Growth," Industrial and Corporate Change, V3#1, 1994, 243-283. Some of the variables described above are similar to those used by Michael Porter, Competitive Advantage, supra, however, to different ends in a decidedly different, albeit complementary, analytic.
19. On the 'guanxi' network concept see Gary G. Hamilton, "Competition and Organization: A Reexamination of Chinese Business Practices," paper prepared for the IGCC Conference The China Circle: Regional Consequences of Evolving Relations Among the PRC, Taiwan, and Hong Kong-Macao, Hong Kong, December 8-10, 1994
20. For a fuller discussion of this US-Japan comparison, see Borrus, Competing for Control, supra, at chapters 4 and 5.
21. See the discussion of sequential increasing supply dependence in consumer electronics in Consumer Electronics Sector Working Group, MIT Commission on Industrial Productivity, "The Decline of U.S. Consumer Electronics Manufacturing: History, Hypotheses, Remedies," December, 1988.
22. For an extended discussion of the supply base and architecture of supply concepts, see Borrus, "The Regional Architecture of Global Electronics," supra.
23. For the broad range of major component technologies involved, see the discussion in Borrus, Ibid.
24. The characterization of US FDI is based on the BRIE US Electronics FDI database, compiled from public sources and maintained by Greg Linden, supplemented by industry conversations, and reviewed by senior managers with Asia responsibility from most of the firms mentioned in the text. The characterization of Japanese electronics FDI in Asia which follows is consistent with, and in part draws on data and detail in the chapters by Ernst and Huchet, as well as from Ken-ichi Takayasu and Yukiko Ishizaki, "The Changing International Division of Labor of Japanese Electronics Industry in Asia and Its Impact on the Japanese Economy," RIM, Pacific Business and Industries, V1, #27, 1995, p.2-21. See also the relevant chapters in Doherty, Japanese Production Networks, supra.
25. On the progression from consumer to industrial electronics in Taiwan, see also Scott Callon, "Different Paths: The Rise of Taiwan and Singapore in the Global Personal Computer Industry, " Japan Development Bank Discussion Paper Series, #9494, August, 1994. More generally, on the development of Taiwan's information technology industry, see Kenneth L. Kraemer and Jason Dedrick, "Entrepreneurship, Flexibility and Policy Coordination: Taiwan's Information Technology Industry," (Irvine: Center for Research on Information Technology and Organizations, 1995).
26. There were, of course, tremendous variations in the role played by state policy, and in the policies themselves, in the different countries of the region. In highlighting a few commonalties, I do not mean to slight those differences. The active role played in general by governments in the region has been explored in detail in a variety of scholarly works. See, e.g., Robert Wade, Governing the Market: Economic Theory and the Role of Government in East Asian Industrialization, (Princeton: Princeton University Press, 1990); Stephan Haggard, Pathways from the Periphery: The Politics of Growth in the Newly Industrializing Countries, (Ithaca: Cornell University Press, 1988). More recently, see the excellent contributions in Andrew MacIntyre, ed., Business and Government in Industrializing Asia, (Ithaca: Cornell University Press, 1994).
27. In focusing on OC electronics producers, I am ignoring the significant regional investment by the Korean Chaebol who emerged during this period as major, region-wide producers of consumer electronics and components. See, e.g., Martin D. Bloom, "Globalization and the Korean Electronics Industry," Pacific Review, v.6 #2, 1993.
28. The Table is drawn from a presentation prepared by Tze-Chen (T.C.) Tu, Director of Taiwan's Market Intelligence Center (of the Institute for Information Industries), "Upgrading Taiwan's IT Industry--New Challenges and the Role of International Cooperation," at the BRIE-Asia Foundation Conference, Competing Production Networks in Asia: Host-Country Perspectives, San Francisco, April 27-28, 1995.
29. This picture is perhaps a bit at odds with the emphasis in the rest of this volume on the importance of Hong Kong as a force for China Circle integration: Hong Kong was at best a distant third most important NIC site for MNC or indigenous electronics production. Moreover, its relative importance declined as investment spread into southeast Asia and Mainland China--though Hong Kong undoubtedly played an important role in helping to channel Taiwanese investment onto the Mainland during the third period, described below.
30. Callon, ibid.
31. Based on press accounts, company annual reports and SEC 10K filings as compiled by Greg Linden in the BRIE Asia FDI Database. See Linden, "Apple Computer East Asian Manufacturing Affiliates," November 7, 1994, unpublished summary.
32. For Compaq, see Linden, "Compaq East Asian Manufacturing Affiliates," November 7, 1994, unpublished summary; for Hewlett-Packard, see Linden, "Hewlett-Packard East Asian Manufacturing Affiliates," November 9, 1994, unpublished summary; for Motorola, see Linden, "Motorola East Asian Manufacturing Affiliates," November 7, 1994, unpublished summary.
33. Presentation by Tze-Chen (T.C.) Tu., supra, supplemented by press reports.
34. As she shows, the resulting industry concentration was most visible in Taiwan's largest domestic product sectors, notably monitors, PCs and PCBs, where the top ten indigenous producers now account for over 70% of the market.
35. For elaboration on the following, see, e.g., G.S.Shieh, "Network Labor Process: The Subcontracting Networks in Manufacturing Industries of Taiwan," Academia Sinica, Bulletin of the Institute of Ethnology, #71, Spring, 1991; Brian Levy and Wen-Jeng Kuo, "The Strategic Orientation of Firms and the Performance of Korea and Taiwan in Frontier Industries: Lessons from Comparative Case Studies of Keyboard and Personal Computer Assembly," World Development, v.19,#4, 1990. I have also drawn on an excellent paper by one of my graduate students, Fu-mei Chen, "From Comparative Advantage to Competitive Advantage: A Case Study of Taiwan's Electronics Industry," unpublished manuscript, May 21, 1994.
36. Presentation by T.C. Tu, supra.
37. The Taiwan straits crisis of 1996 does not appear to have substantially influenced the willingness of Taiwanese electronics firms to invest on the Mainland, though it undoubtedly reinforced their desire to spread the risk by simultaneously investing into Southeast Asia.
38. Representative estimates range from Apple's judgment that its Singapore operation can move a new product into production in half the time of its other operations, to Ming Chien, Chairman of FIC, who estimates that motherboards can be completely changed out (with all attendant alterations to the rest of the system) in Taiwan in 2-3 weeks vs. up to a year in the US. On the former, see Singapore Business Times, 11/27/90 p.14; on the latter, see Callon, "Different Paths," supra.
39. Louis Kraar, "Your next PC could be made in Taiwan," Fortune, August 8, 1994, p.90ff, inferring from Dataquest estimates.
40. See Callon, "Different Paths," supra, citing interviews in Singapore. Structured IC design approaches were pioneered in the US at Universities like Berkeley and CalTech, where many OC engineers were formally trained.
41. On European industrial districts, see the work on flexible specialization, notably, Michael Piore and Charles Sable, The Second Industrial Divide: Possibilities for Prosperity, (NY: Basic Books, 1984).
42. Based on industry discussions.
43. So long as Japanese, US, and OC firms continue to be driven by very different domestic linkages, strategies, industrial structures, and policy, capital market and labor market influences, their network differences are likely to persist even if they converge in competitive purpose.
44. In fact, the opportunity to drive development out of Asia is already appearing in a set of significant potential product markets. These include broadcast media where firms like HongKong's TVB and Murdoch's Star TV are pioneering direct broadcast TV transmission, software where indigenous concepts could lead in new directions, and segments of the wireless communication markets where, for example, Motorola projects that China will pass the US to become its largest market for pagers in the next few years.