استراتژی‌های رقابتی برای اقتصاد نیمه‌هادی تایوان در اقتصاد دنیای نوین

Taiwan's semiconductor industry (TSI) has been a popular research subject. In particular, the small island country's characteristic fast follower approaches to starting a capital and knowledge-intensive high-tech industry – such as the semiconductor industry – and making it a success story provide valuable insight into the fast-changing dynamics of these industries and a role model for developing countries. For a long time, the primary competitive edges of TSI have been speed, cost, flexibility, and quality, enabled by policy formulation, bridging institutions, public infrastructure, vertical disintegration, entrepreneurship, and human capital. However, facing heightened competition, a change of status (i.e. TSI is no longer a follower but a forerunner in a relatively mature state), and a changing world economy, TSI needs to develop additional core competencies to remain competitive. This article discusses the approaches adopted by Taiwan's public and private sectors for such a purpose. By surveying a wide variety of data including laws and policies, national science and technology programs, industry news, market reports, and relevant literature, the study suggests that technology, value, sustainability, and brand are the additional competitive edges being developed for TSI. The paper also discusses potential obstacles for TSI in the foreseeable future.

As the epitome of high-tech manufacturing, semiconductor manufacturing produces a staggering volume and variety of chips each year. These chips are the critical components for making electronic devices in application categories including computing/information (e.g. computers and tablets), consumer electronics (e.g. videogame consoles, TVs, and digital cameras), telecommunications (e.g. smartphones), automobiles (e.g. GPS), and aerospace/defense. With these devices, human societies have been advancing at an accelerating speed. The process of semiconductor manufacturing starts from growing silicon ingots (the raw material for making wafers) followed by a range of activities, including integrated circuit (IC) design, wafer fabrication, IC test, and IC packaging. The placement of finished chips on a printed circuit board marks the end of this process [[1], Fig. 1]. This manufacturing industry is characterized by long manufacturing lead times, increasingly short product life cycles, complicated manufacturing processes, and substantial capital investments. For instance, the wafer fabrication stage for making high-end videogame chips takes several months to complete. During the fabrication of chips and modules, the binning process (categorizing parts according to certain performance criteria) and the substitution process (using alternative parts to continue a manufacturing step) make capacity allocation a serious challenge for production and supply chain planners [2]. And to construct a brand new 300-mm (12-inch) wafer fabrication facility (a.k.a. fab), the required capital investments are measured by billions of US dollars. These characteristics make semiconductor manufacturing a capital and technology-intensive industry. All semiconductor firms must achieve an optimal allocation of manufacturing resources so as to survive in this fiercely competitive industry. Called “Silicon Valley of the East” by Mathews [3], Taiwan is well known in the world for its highly successful high-tech industries, which have been the primary driving force of the national economy since the 1980s, and semiconductor manufacturing has unquestionably played the most significant role. How the small island country is able to develop such a successful semiconductor industry has been studied by numerous scholars, including Mathews [3], Liu [4], Chang et al. [5], Chang and Hsu [6], Tung [7], Chang and Tsai [8], Hung and Yang [9], Sher and Yang [10], Wu et al. [11], and Hu [12]. The significant status of Taiwan's semiconductor industry (TSI) is supported by the following statistics. According to the Industrial Technology Research Institute (ITRI) [13], Taiwan's largest public laboratory engaging in the R&D of advanced industrial technologies, TSI is ranked fourth in the world in terms of total revenues (US$48.8 billion in 2012 in a global market of US$297.6 billion). Currently, Taiwan owns the world's leading semiconductor contract manufacturing (a.k.a. foundry) service, leading outsourced IC test and assembly services, and second-ranked IC design services (Table 1). Currently, Taiwan houses the largest IC wafer fab capacity and the most complete semiconductor supply chain in the world (Fig. 1). As Hung [14] and Hung et al. [15] argued, the strong TSI has resulted in the country's unique position and competitiveness in thin film transistor-liquid crystal display (TFT-LCD) manufacturing because the two industries share similar IC fabrication processes [[15], Fig. 2]. In fact, by holding the second-place position in total revenues in 2012 (US$53.0 billion), Taiwan has continued to be one of the top TFT-LCD producing countries in the world. For the same reason of similar technologies being used, the strong TSI has also significantly contributed to the successful development of Taiwan's light emitting diode (LED) and solar photovoltaic (PV) industries [16], helping them to become the third and the second largest worldwide, respectively. Therefore, as Hou and Gee [17] and Mathews et al. [18] have argued, Taiwan's well-established IC infrastructure has provided a solid foundation to support the rapid development and advance of the other high-tech industries in the country. Table 1. Taiwan's world-leading semiconductor industry [13] and [19]. Value chain Category 2008 2009 2010 2011 2012 IC design Revenue (US$ billion) 11.90 11.68 14.37 13.10 13.90 Global share (%) 27.2 24.7 24.1 20.1 20.3 World ranking 2 2 2 2 2 Custom IC fabrication Revenue (US$ billion) 14.10 12.90 19.40 19.10 21.70 Global share (%) 68.9 66.7 68.2 68.8 67.8 World ranking 1 1 1 1 1 IC test & assembly Revenue (US$ billion) 11.11 9.57 13.11 13.25 13.29 Global share (%) 55.2 55.8 55.6 55.2 53.4 World ranking 1 1 1 1 1 Table options Full-size image (45 K) Fig. 1. Structure of Taiwan's semiconductor supply chain with the number of firms in 2013 [13], [19] and [20]. Figure options The structure of an industry is a key determinant of its competitive advantages [21]. As indicated by various government officials and academics including Chang and Tsai [8] and Wu et al. [11], the primary competitive advantages of TSI have been speed, quality, flexibility, and cost, enabled by vertical disintegration and cluster effects. Vertical disintegration appropriately describes the structure of TSI. As illustrated in Fig. 1, TSI is a complex network consisting of firms specializing in some specific stage in the semiconductor manufacturing process, such as chip design or foundry. These specialist firms constitute a vertically disintegrated semiconductor supply chain that not only has become the most distinct feature of TSI, but is also unique in the world. By contrast, firms in virtually every other major country or region that produce semiconductor products, including the United States, Europe, Japan, and Korea, have mostly remained integrated in designing and manufacturing semiconductor devices. These firms are generally referred to as integrated device manufacturers (IDM). When entering the semiconductor market in the 1970s, Taiwan was weak in related sciences (e.g. physics, electronics, and materials science), technologies, and management methodologies; furthermore, Taiwanese firms were significantly smaller than those in competing countries and these small or medium-size firms lacked adequate resources, such as sufficient capital, skilled technicians, and established international connections, to start a high-tech business producing, for example, semiconductors. However, Taiwan had a diligent workforce whose average wages were significantly lower than what their foreign counterparts were making. In addition, flexible organizational and operational response to change, such as changes in market demands and design specifications, was an asset widely possessed by Taiwanese firms. With appropriate strategies and policies formulated by the Taiwanese government – in particular, the adoption of a fast follower approach [4], [14], [15] and [18]; the establishment of ITRI to lead the execution of that approach; and the provision of public funding and taxation benefits/allowances to encourage entrepreneurship – TSI was able to grow and find a niche position in the fiercely competitive market. The success of Taiwan Semiconductor Manufacturing Company (TSMC) in pioneering its “Dedicated IC Foundry” business model [22] stimulated the mushrooming of local specialist firms and accelerated the shaping of TSI to become a vertically disintegrated semiconductor chain. The strategic placement of these firms in a designated area created agglomeration and cluster effects, which increased the overall competitiveness of TSI significantly [7], [9], [10] and [23]. Some of the renowned specialist firms in TSI are shown in Table 2, including TSMC (the world's leading semiconductor foundry), United Microelectronics Corporation (UMC, the world's fourth-ranked foundry), and The Advanced Semiconductor Engineering Group (ASE, the world's leading provider of independent semiconductor assembly and test services). Table 2. Major specialist firms in TSI. Value chain Company Founded Capitala Revenuea,b (′11/′12) Work-force World ranking Up-stream: fabless IC supply MediaTek 1997 0.42 2.97/3.37 6,600 5 Mid-stream: pure-play IC foundry TSMC 1987 8.51 14.60/17.17 37,149 1 UMC 1980 7.22 3.76/3.73 13,000 4 Down-stream: outsourced IC packaging & assembly ASE 1984 2.40 4.25/4.40 57,259 1 SPILc 1984 0.91 2.03/2.19 20,000 3 a US$ billion. b Revenues were reported on IC Insight and Gartner. c Siliconware Precision Industries Corporation, Ltd. Table options Since entering the twenty-first century, the global economy has been shaped dramatically by the rapid advancement of technology. The Internet has connected people from all over the world, affecting their daily lives in every conceivable manner. In the business world, the development of e-commerce has made planet Earth one giant marketplace and fundamentally altered the way people do business. The glory days of the PCs have been terminated by the arrival of innovative mobile devices such as smartphones and tablets. Coming in different shapes and with different functions, these devices are powered by advanced multi-core processors designed to function wirelessly with low-power consumption, capable of running multiple apps simultaneously to create a smooth and unforgettable user experience. In addition to mobile usage, semiconductor chips have also found growing applications in emerging markets such as power management, health care, and electric vehicles. These developments have resulted in fast-changing industrial dynamics and a dramatic reconfiguration of the market's product portfolio. For TSI firms, these changes present challenges in maintaining leadership in manufacturing technologies and continuing their world-class services in IC design and manufacturing. Most certainly, TSI firms need to maintain their traditional competitive advantages as they move forward. However, facing a heightened challenge from traditional rival countries such as the United States and Korea and newcomers such as China, those advantages will not suffice. Despite being a manufacturing expert for a long time, Taiwan has lacked expertise in software-hardware integration, branding, and global marketing; also, Taiwanese firms do not possess sufficient intellectual properties (IPs) and proprietary technologies in designing and manufacturing semiconductor products [24]. The timing of acquiring these assets and capabilities determines whether TSI will remain competitive in the new global economy or not. In addition, the semiconductor industry has entered a relatively mature state, which means serious work is required to reinvigorate the competitiveness of TSI. The overall semiconductor industry has displayed many of the hallmarks of maturity, including well-diffused technical know-how, vertical disintegration, increased firm concentration, knowledgeable customers, slower revenue growth, intensified price competition, and reduced profitability [25] and [26]. The mature state of TSI is further indicated by the success of those competitive advantages described earlier in making TSI a significant player in each stage of the industry value chain for many years (Table 1 and Table 2). However, there are also signs showing that the technology and business differentiation between TSI's leading firms and foreign competitors, which is reflected in their product or service attributes, has become more difficult, and these firms have competed for increasingly similar markets. As an example, the high technology dependence among TSMC, UMC, Vanguard, and GLOBALFOUNDRIES (all major foundries) since the 8-inch-wafer era has been quantitatively demonstrated by Wang et al. [[27], Fig. 2]. Although the presence of similar technologies and business models is a common phenomenon in a mature industry, it is also a worrisome sign. If TSI fails to de-mature in time, that is, altering its evolution course by deploying effective market-driven or technology-driven approaches to rejuvenate competitiveness [28], then according to the widely accepted industry life cycle model, TSI will inevitably see slow volume growth and diminishing returns and move into decline [29], [30], [31] and [32]. Such a development will certainly create tremendous damage to the entire national economy. Furthermore, global awareness of environmental protection and sustainable development are also major issues for TSI. In the 1990s, global warming (which causes a severe climate change), polluted environments, and depleting stocks of natural resources attracted worldwide attention. These problems have worsened in this century, pushing national governments to recognize the issues' serious impact on the well-beings of mankind. Increasingly, environmental impact assessment is required for economic development plans, and the greening of economies has been recognized by members of the international community as an effective solution to addressing the issues in economic development while meeting the challenges from environmental conservation. As Potts [33] described, green economies focus on climate mitigation (e.g. lowering greenhouse gas (GHG) emissions [34] and [35]), develop reliable and affordable supplies of clean energy [36] and [37], increase energy efficiency, reduce pollution, develop sustainable industries, promote industrial ecology (e.g. constructing eco-industrial parks [38] and [39]), and formulate policy to support sustainable development [33] and [40]. Countries all over the world have had various regulations or directives in place for the greening of their national economies ( Table 3). Table 3. Selected regulations and directives for environmental conservation and sustainable development. Category Region/country Regulation/directive Effective Energy conservation & security US Energy Independence and Security Act (EISA) 2007∼ EU Ecodesign Requirements for Energy-using Products (EuP) 2007∼2009 Ecodesign Requirements for Energy-related Products (ErP) (replacing EuP) 2009∼ Australia & New Zealand Minimum Energy Performance Standards (MEPS) N/Aa Control of hazardous substances and GHG emissions EU Restriction of Hazardous Substances (RoHS) 2006∼ Registration, Evaluation, Authorisation and Restriction of Chemicals (REACH) 2007∼ 2006/122/ECOF (Perfluorooctane Sulfonates, PFOS) 2008∼ US Consumer Product Safety Improvement Act 2008∼ Environmental Design of Electrical Equipment Act (EDEE) 2010∼ Recycling EU Waste Electrical and Electronic Equipment (WEEE) 2005b∼ China China WEEE 2011∼ a Different timeframes were set for different products. b The UK transposed WEEE into national laws in 2006. Table options In greening the global economy, environmental conservation directives such as WEEE, RoHS, and REACH (Table 3) have created the effect of non-tariff international trade barriers. That is, in addition to functionality, quality, and cost, products and services are also subject to a strict environmental impact evaluation [41]. To TSI firms, this means they must maintain constant compliance of those environmental standards. In the long run, the demand for green products and services will grow, but so will the pressure to reduce GHG emissions and disclose environmental impact data (such as carbon and water footprints). These developments present new challenges as well as emerging opportunities for TSI. The wake of global awareness of sustainable development has demanded that electronics be green, thus pushing TSI to develop core competencies in that direction. Therefore, this is the key research question addressed in this article: What are the strategies required for TSI to adapt to the world's changing economy, thereby reinvigorating its competitive advantage and enabling it to continue to proliferate well into the twenty-first century? This article adopted a broad historical and industrial case methodology to search for answers. Case studies are rich, empirical descriptions about specific instances of a phenomenon [42]. When used as a research methodology, the case study focuses on comprehending the dynamics present within single settings [43]. The cases selected for this type of research can be historical accounts or recent events, and they are used to form the basis from which theories are developed in an inductive manner. As Eisenhardt and Graebner [44] described, case study research builds theory “by recognizing patterns of relationships among constructs within and across cases” and the theory building process proceeds “via recursive cycling among the case data, emerging theory, and later, extant literature.” As summarized in [[43], Table 1], the process begins by defining the research question and ends when marginal improvement becomes small. The process is systematic with each step clearly defined. Therefore, this is a methodology which can be repeated. Case studies typically use rich data collected from various sources, thereby enabling researchers to develop theories objectively. The case study has been widely used to research industrial dynamics and competitive strategies of high-tech industries. For example, the works reported in many of the articles reviewed earlier, including [6], [8], [11], [14], [18] and [23], are all case study research. Most recently, Khor and Lalchand [45] presented their historical and current case analysis of Malaysia's sustainable power generation and recommended appropriate strategies. This research adopted the same approach because it is a pilot study on an industry with unique features. According to Eisenhardt [43], in the early stages of this type of research, case studies are particularly appropriate because the methodology does not rely on prior literature or empirical evidence. The data utilized in this work include information gathered from various governmental departments and other public institutions (e.g. industry statistics, national science and technology programs, laws, and policies), market studies conducted by professional organizations (e.g. IC Insights and Gartner), news releases issued by various TSI firms, and prior academic studies. This rich data set was used to reveal relationship patterns and reconcile evidence, which led to the framing of strategies for reinvigorating TSI's competitive advantage. Also derived from this data analysis is a framework which not only enhances the literature on industry life cycle but can also be used to assist decision makers in moving forward. This paper proceeds as follows. Section 2 reviews how Taiwan started its semiconductor industry, with a focus on the main strategies responsible for the current success of TSI. In Section 3, the concept of national innovation system and its application in the research of national innovative capacity and competitive strategy is discussed, followed by an introduction to selected national laws, national research programs, and industry alliances. They are selected because of the profound impact they have created in shaping the national innovation system for TSI; the most significant items are discussed in further detail. Section 4 follows the discussion in Section 3 and reveals the resulting competitive strategies for TSI. Finally, the paper concludes by indicating potential issues to be considered for future research (Section 5).

For decades, TSI has been arguably the single most important industry for the island country. With an appropriate fast follower approach to start the industry, followed by the successful development of a favorable industrial ecology and innovative business models, and supported by a diligent and well-educated workforce capable of responding to change with great organizational and operational flexibility, TSI has flourished to become a top player in the global semiconductor market. This success has also contributed to the success of other high-tech industries in the country, including TFT-LCD, solar PV, and LED. Entering the second decade of the twenty-first century, the success of TSI has continued. According to the SEMI World Fab Database, Taiwan reached US$9.0 billion in semiconductor equipment sales in 2012, the second highest in the world, and surpassed Japan to become the world's largest semiconductor material market, totaling US$10.0 billion. By the end of 2013, it is expected that there will be 26 12-inch fabs in the country with a total installed monthly capacity of more than one million wafers. The increasing competition, a relatively mature state of the industry, and the changing economies have pushed TSI to transform, from a manufacturing-oriented industry to a service-oriented industry. Similar to the start of the industry in the 1970s, the transformation requires policy formulation to lay down a roadmap and bridging institutions to lead the campaign. However, a blue ocean approach is appropriate because of TSI's current status. Through implementation of technological R&D programs at the national level and collaboration between public and private sectors, competitive edges in technology, value, sustainability, and brand are being developed. They, along with existing competencies (i.e. speed, quality, flexibility, and cost), would make TSI more competitive in the post-PC era. There are a number of issues that may create a negative impact on TSI. The most serious one is perhaps a shortage in human capital in the foreseeable future. Currently, Taiwan has the lowest birth rate in the world. It is estimated that Taiwan's population will reach zero growth by 2024 and will start to decline in 2025. The rapid aging of Taiwan's population has already caused problems for some private colleges and universities in the country, as they cannot attract sufficient students to enroll and may have to close in a few years. It also means that Taiwan's industries will soon face similar problems. Another problem is the decrease of profit margins for TSI caused by growing competitions and a world economy that has been slow to recover. This, along with some measures taken by the Taiwanese government (e.g. reducing tax benefits, which were long enjoyed by TSI employees when trading their company stocks), have significantly reduced the attractiveness of TSI in the minds of Taiwanese people. How to recruit and retain talented people in the industry is thus a serious issue for TSI firms as well as the government.