Chapter 4 Taiwan
“The government does not intend to tell businesses what to do or what not to do,” President Tsai Ing-wen explained in response to criticism by TSMC Chairman Morris Chang in October 2017 (Taipei Times 2017). Chang had lashed out after the announcement about industrial policies that would focus on building new sectors, like biotechnology, green energy, and smart machinery, rather than champion the all-important semiconductor sector. Chang’s reasoning centered on how important semiconductors were to the Taiwanese economy and society. According to Chang, the country faced an existential crisis in which the “Silicon Shield” supplied by the global position of semiconductor manufacturing firms (Klingler-Vidra 2023; Cheng and Li 2021), especially TSMC, was being threatened. Tsai’s response that she did not intend to help Taiwan’s semiconductor giants was consistent with Taiwan’s policy approach since the onset of the twenty-first century to widen the pool of entrepreneurs in new industries rather than to shore up the competitive positioning of its dominant firms and technologies. It was also in line with Taiwan’s long-established orientation toward encouraging technology-centric startups rather than coordinating capacity-building among large leading firms.
As this semiconductor vignette suggests, Taiwan offers a different starting point, and thus a distinct continuity-change trajectory, in comparison to the two previous cases. The key difference is that the Taiwanese developmental state is characterized as being oriented toward small high-technology firms in select (critical) technologies. Volumes of research assert that Taiwanese industrial policies favored small firms (see Amsden 1985, 2001; Simon and Kau 1992; Yu 2012).1 For instance, Karl Fields (1997, 125) notes that the Taiwanese state’s ties with its “less concentrated private-business sector” were “decidedly more particularistic, diffuse and distant than those in Korea.” In placing Taiwan in East Asian business systems scholarship, Richard Whitley (1992, 53) similarly asserts that the “dominant feature” is the “small size of most firms.”
Thus, the Taiwanese case offers the opportunity to analyze continuity and change for the developmental state characterized as closest to the Mark I mode in its antecedent period. As this chapter shows, though its lead firms, notably TSMC, have adopted open innovation tactics, the Taiwanese state has largely exhibited continuity in its aim of widening entrepreneurial markets rather than fostering oligopolistic competition.
Antecedents: Taiwan (1949–1970)
Taiwan’s economy has long been depicted as small-firm dominant, relative to the other East Asian developmental states of Japan and Korea. The small-firm orientation was fostered by public policies as the state drew on a variety of “financial regulations, tax and labor laws” to provide “strong incentives to limit firm size” (Fields 1997, 130). The early focus on small firms is explained, by Robert Wade (1990) and others, as being due to the Kuomintang (KMT)’s desire to ensure that indigenous Taiwanese entrepreneurs would not pose a rival center of power to the martial law–era government.2 Analysts have observed that the Taiwanese state, from the martial law era, pursued technology- and small-firm-led “economic growth as congruent with, if not central to, enhanced national security” (Pempel 2021, 15).3
The regime’s interests in encouraging a flexible labor market focused on encouraging entrepreneurs leading export-competing firms. There’s a well-known saying in Taiwan that it is “better to be the head of a cock than the tail of a cow” (Kung and Yen 2018, 59). Said another way, it is better to be an entrepreneur and lead your own small company than it is to be an employee of a large company. Analysts note that in Taiwan, “workers generally did not picture themselves as lifetime industrial laborers; rather, a sizeable number were using their job to accumulate enough capital to start their own enterprise” (Gold 1986, 89). Despite the cultural values favoring entrepreneurship as a form of employment, in Taiwan’s early years, the talent pools, especially for science and technology-based entrepreneurship, were limited. J. Megan Greene (2008, 41) notes that “relatively few highly trained scientists and technicians had fled to Taiwan with the KMT”; they instead moved to the United States and Europe.
Rather than focusing on catch-up aims, policy sought to boost innovation capacity toward the world’s technological frontier. Analysts observed that they could see the gradual emergence of an overarching science policy from when the father of Taiwan’s economic miracle, Kwoh-ting (K. T.) Li, pursued strategic efforts to boost the talent pool for technologically oriented entrepreneurship (Kuo 1983). Li’s perspective was that the success of policies to incentivize small firms’ science and technology prowess were “vital to Taiwan’s future economic prosperity, and perhaps, political survival” (Winckler and Greenhalgh 1988, 219). Technology-centric economic competitiveness would help Taiwan to survive as a state and demonstrate the performance of the non-Communist approach (Amsden 2001). Yet Taiwan’s small firms were said to not naturally invest in significant R&D or upgrading. Instead, Whitley (1992, 55) characterizes the tendency for Taiwan’s small firms’ “technology investments” to be “short term so that they can be written off after a few years.”
In this spirit, the state encouraged small firms to export in particular industries through the Nineteen-Point Economic and Fiscal Reform Program (Nineteen-Point Program) and the Statute for the Encouragement of Investment (SEI) in 1960.4 Crucially, the lists of industries eligible for these tax relief initiatives reflected policy makers’ assessment of which technologies and industries were at the technological frontier at given points in time and central to national capacity upgrading (Wade 1990). The SEI was updated more “than a dozen times by 1980” and specified “categories and criteria of strategic productive enterprises singled out for special encouragement” (Gold 1986, 78). To further encourage Taiwan’s small firms to invest in R&D, the government offered a series of tax incentives, including an R&D tax rebate, a five-year tax holiday for new industries, the accelerated depreciation of investments, and special tax credits for new strategic industries (World Bank 1993, 229). In addition, part of efforts for improving small firms’ technological capacity was to secure technologies from abroad, attracting multinational companies to Taiwan and state firms and institutes transferring technology (see the China Data Processing Center, which was the country’s first data-processing and software company, for instance; Breznitz 2007, 102).5
Unlike the Japanese and Korean cases in which significant lines of credit were made available to (large) firms, the little money provided to Taiwan’s (small) firms came as a mix of equity and debt. The Chiang Kai-shek regime was careful to not allocate too much to any company, as the authoritarian leadership was worried that a concentration of capital could be used to mobilize support against the KMT. Finance for industry was “tightly controlled” (World Bank 1993, 20), allocated in small increments and through state-run banks. Small firms instead sought equity-based financing and loans from friends, family, revolving credit associations, and the unrelated “curb” loan clubs (Fields 1995). Another alternative to domestic banks as sources of financing in the 1960s was the US Agency for International Development Small Industry Loan Fund and Model Factory Program (Gold 1986, 71). With this funding came influence on policy in terms of American demands that “pushed for start-up companies,” given the capital and individuals available on account of land reforms (Rubinstein 2013, 30). This again augured for the early fostering of equity-based, Silicon Valley–styled financing.
Thematic Analysis
Firm Size
By the 1970s, the Taiwanese electronics industry—and local firms that were competing in this global market—was booming. Thomas Gold (1986, 88) remarked that a motivating credo for the era was that “everyone wants to be chairman of his own company.” Policy continued to encourage startups’ upgrading capacity by helping them “to locate, purchase, diffuse and adapt new foreign technologies” (Lall 1996, 208). Policy leaders like Sun Chen noted that there was a need for the state, through Industrial Technology Research Institute (ITRI), to provide R&D services, as the SMEs did not “enjoy sufficient economies of scale to justify extensive spending on R&D” (Greene 2008, 121). With the establishment and take-up of the Hsinchu Park by returning overseas Chinese, the government’s startup-centric innovation efforts were crystalizing. Gold (1986, 104) explains that the state hoped that “small, innovative companies headed by dynamic Chinese with experience abroad, buttressed by government technical and financial support,” offered an important niche for Taiwan in global computer and consumer electronics markets.
In line with its wider Mark I character, small firms remained essential to the technological upgrading effort. The tax rate for profits on high-technology investments was set at zero from the 1980s, as small high-tech firms were categorized as “strategic industry enterprises” that received a tax holiday (Wang 1995, 5–6). Policy makers, particularly those on the Council for Economic Planning and Development (CEPD), encouraged growing Taiwanese companies such as Formosa Plastics and Acer to invest in local startups as venture capitalists. One aim was to foster integration between growing firms and fledgling firms and entrepreneurs.
The term startup, rather than SMEs or small firms, began being used by science and technology policy makers in the early 2000s. In 2002, Taiwan’s SME Administration launched the Business Startup Award, stating that the “purpose of the award is to create more Startup companies” (MoEA SME Administration 2019a). Efforts were taken to boost available talent and financing for startups and to position startups as the organizational means for competing in innovative sectors in the knowledge economy. Already by the mid-2000s, the SME Administration had financially and administratively assisted seventy-nine startup incubation centers, sixty-five of which were within, or attached to, colleges and universities (Lee and Lai 2005, 2). The 2005–2008 Science and Technology Development Plan expanded on these efforts, with the Executive Yuan specifying the promotion of startups through the support of: (1) spinouts from R&D undertaken at universities; and (2) incubation services for technology-based startups (Executive Yuan 2005). The 2009–2012 S&T Development Plan took a similar tack, noting that encouraging “young people to establish their own businesses, and thereby transforming R&D capabilities into spin-off companies, will be a key method of fostering entrepreneurial innovation” (Executive Yuan 2009, 37). In a similar vein, efforts by the Ministry of Education to incentivize entrepreneurship included instruction, teaching, and competitions at universities (Executive Yuan 2009, 30).
Contemporary policy for startups has increasingly centered on fostering links with global ecosystems. For instance, in 2015, the National Development Council established Taiwan Startup Stadium (TSS) as an ecosystem builder; it aims to “cultivate globally-minded entrepreneurs and showcase the most innovative Taiwanese startups to the world.”6 Another example of a bridge-building initiative is the Taiwan Innovation and Tech Arena (TITAN), which was launched by the Ministry of Science and Technology (MoST) as a “tech startup ecosystem building program” in 2018 and partners with overseas innovation organizations in order to provide training and exposure for Taiwanese entrepreneurs (MoST 2021). Mentorship in the TITAN program is provided by foreign experts, especially in Silicon Valley, as a means of helping Taiwan’s startups to go global, and access is equally given to foreign talent looking to establish or grow their startup in Taiwan. TITAN, for its inbound efforts, is described by FutureWard as aiming “to boost the diversity of Taiwan’s innovation ecosystem by inviting talents from all over the world to create the next big thing with Taiwan’s startups” (FutureWard 2019). In 2022, TTA ran the first iteration of an accelerator program with 500 Startups in which twenty startups (or those planning to enter the market) from Taiwan participated (Kong 2022).
While startups persist as key beneficiaries, policy efforts increasingly include attempts to foster an open innovation approach alongside large firms. This has often been via partnerships with foreign companies rather than big domestically based businesses like TSMC. In 2019, the Executive Yuan partnered with AWS to create a Joint Innovation Center at the Startup Terrace innovation park (Executive Yuan 2019a). The rationale for the collaboration is that with AWS’s “discerning vision, technology services and capable guidance,” Taiwan’s startups “will be able to grow more quickly and connect to the global market to hone their competitive advantages” (Executive Yuan 2019b). In 2021, ITRI partnered with Arm, a British integrated circuit design firm bought by SoftBank in 2016, to run the “IC Design Platform for Startups” to “assist startups in accessing critical IP and accelerating the launch of competitive products for the global market” (ITRI 2021).
The contemporary open innovation setting is more software- and internet-based than the strategy of attracting multinational companies for hardware-focused technology transfer, initiated in the 1960s, but it does offer similarities. It also echoes some of the tech transfer aims of ITRI; rather than Taiwan’s government entities playing this role, collaboration with leading international firms is expected to enable technology upgrading opportunities for Taiwan’s startups. The beneficiaries remain local startups, not large firms. In this sense, there is not a shift toward Mark II, as Taiwan’s small firms, not its leading oligopolistic firms, remain an essential engine for innovation in new cutting-edge technologies.
Separate from government policy, large Taiwanese incumbents, including TSMC, have embraced open innovation approaches. In 2009, TSMC made headlines when it announced the Open Innovation Platform, in which the company moved to share “R&D efforts with clients” (Kwong 2009). While using the language of open innovation, however, the platform did not specifically target startups but rather their clients in a bid to share R&D costs. On its corporate blog, TSMC’s director of market development and emerging business management, Lucas Tsai, explained that “since its founding 35 years ago, TSMC has been working with startups across the industry, providing access to its technologies and manufacturing capacity to help them grow and thrive” (Tsai 2022). Tsai added that at the TSMC Open Innovation Summit, they brought together “startups, investors, and other chip innovators to discuss transformational silicon designs” and would work in partnership with Silicon Catalyst to run a semiconductor-focused incubator program (Tsai 2022). TSMC, as part of these activities, made headlines in 2023, when a former Kleiner Perkins partner, Wen Hseih, announced he was leaving to launch a Silicon Valley–based VC fund, Matter Venture Partners, backed by TSMC (Hu and Lee 2023). Thus, while Taiwan’s startup policy continues to operate largely in line with its antecedent period in terms of targeting and striving to benefit a widening pool of startups, some of its lead firms have embraced startups as a resource for their open innovation systems.
Employment
Taiwan’s labor market was more fluid and oriented toward entrepreneurship in its antecedent period. Over time, though, this propensity has been encouraged by policies that further promote startup activity. As such, Taiwan’s focus on upskilling and incentivizing technology-centric entrepreneurs has only grown.
Already in the 1970s, Taiwan’s emphasis was on skill development. To increase the availability of high-tech talent, Taiwan trained 50 per cent more engineers per 1,000 population than the United States (Wade 1990). In addition to domestically trained talent, overseas talent was also regarded as a bank in a brain circulation process, with policies trying to bring computer science PhDs back to Taiwan. This was possible, as Taiwanese students had begun moving to the United States to study (primarily PhDs in engineering) by way of American-Taiwanese funding programs, such as the US AID, in the 1950s (Greene 2008, 55). From the early 1980s, policies saw this overseas-trained talent as a resource that should be recruited back to Taiwan.
A concerted effort was made to foster an environment that would attract such talent with highly paid and intellectually interesting roles at high-growth firms. In December 1980, Hsinchu Park was established near ITRI and two elite engineering and science-focused universities in a bid to consolidate a hub of high-technology startup activity in Taiwan.7 To attract entrepreneurs to set up in Hsinchu Park, the state offered “preferential tax and other treatment, including low-interest loans and tax-free privileges” as well as subsidies for up to 50 percent of R&D costs (Wen and Chen 2014, 229). The state also offered financial assistance to entrepreneurs in the form of equity investment in ventures operating in the Park. Given these advantages, Hsinchu Park was a crucial means of attracting overseas talent (back) to Taiwan’s growing startups, as it quickly came to be “a beacon seen by Taiwanese engineers and scientists” working in the United States at firms such as IBM, RCA, and Texas Instruments (Rubinstein 2013, 41). Working at public research institutes in Taiwan, particularly ITRI, also proved a boon for technology-oriented entrepreneurial talent and inclination (Wen and Chen 2014, 228). Once engineers gained experience at ITRI, they were open to starting and scaling innovation-centric technology businesses. AnnaLee Saxenian (2006, 190) notes that between its founding in 1973 and 1998, “approximately 10,000 ITRI researchers left the agency to join the private sector.”
Encouragement for fluidity in labor markets came from the productive movements between Hsinchu and California, coined the Silicon Valley–Hsinchu Connection in reference to the interactions between the two startup clusters (Saxenian and Hsu 2001). The semiconductor connection was initially boosted by six Taiwanese engineers quitting their jobs at Silicon Valley’s leading semiconductor firm at the time, Fairchild Semiconductor, to return to Taiwan to set up their own firms. The connections played a role in UMC’s success, as it “was able to conclude agreements with three Chinese-owned startups in Silicon Valley” to develop chip design (Lall 1996, 73).8 Pioneer returnees “became role models for hundreds of subsequent returnees” (Saxenian 2006, 2). Through these efforts, by the late 1980s, 180,000 Taiwanese engineers had returned, many from Silicon Valley (Fuller 2002, 16). The scale of returnees was so immense that there were even associations of former employees of American tech companies, like Bell Systems. In 1992, for instance, the “Taiwanese Bell Systems Alumni Association had some 120 members” (Yeung 2016, 45). At the peak of the impact of this connection, in late 1998, 109 companies in Hsinchu Park were founded by returnees from the United States (Wen and Chen 2014, 234–235).
Despite the positive trend, and because of the phenomenal growth of Taiwanese firms in the critical technology of semiconductors, concerns about a talent shortage were a common refrain at the beginning of the twenty-first century. Growing opportunities in mainland China, among other overseas destinations, had meant that the engineering graduates coming from Taiwan’s universities would not satisfy hiring demands (Miller 2022). For instance, “a 2005 ITRI study found that the country’s semiconductor industry would need 37,500 new skilled workers over the next three years,” but the country’s university system could only supply 21,800 such graduates (National Research Council 2013). This shortfall was worrying, as analysts remarked that Taiwan’s universities, including those around which Hsinchu Park was based, had been “viewed mainly as skilled labor–creating mechanisms” (Breznitz 2007, 137). Also, while ITRI had acted as a crucial training ground for would-be tech-centric entrepreneurs, in the 2000s, it was increasingly seeing its staff leave to embark on opportunities in mainland China rather than stay to start a company in Taiwan.
To retain talent in Taiwan, the government launched education and training efforts around startups. In 2003, the SME Online University was created with a startup-focused segment as one of six core areas, with others including IT, marketing, finance, human resources, and comprehensive knowledge (MoEA SME Administration 2019a). Toward the end of the period, the Science and Technology Development Plan (2009–2012) specified universities for their role in boosting entrepreneurship. The plan presented the observation that European and North American countries had promoted entrepreneurship in universities, so “Taiwan should keep up with this global trend by strongly promoting entrepreneurship education and related entrepreneurial activities at universities” (Executive Yuan 2009, 38).
In addition to promoting Taiwan-based entrepreneurship, efforts to encourage global connections and international exposure were made.9 The Ministry of Economic Affairs (MoEA)’s SME Administration cited the Global Entrepreneurship Monitor (2019) report, which found that Taiwanese entrepreneurs lack awareness of opportunities. To remedy this, the agency aimed to deliver internationally linked programs to help ensure that “by taking part in international activities as well as making visits to multinational incubators and accelerators, the ecosystems will be filled with diverse possibilities and connecting with worldwide resources” (MoEA SME Administration 2019b). This aim was reflected in the From IP to IPO program launch in 2013, “which includes mentoring on venture capital negotiations by entrepreneurs and venture capitalists in Taiwan and Silicon Valley, to 40 selected start-up teams each year” (Klingler-Vidra 2018, 90).
Labor market fluidity has advanced through a concerted push to increase the availability of foreign talent for the startup ecosystem. One step toward this aim was the creation of the Contact Taiwan Program by the National Development Council in 2015. The program is an “all-out campaign to attract talent globally” to Taiwan (NDC 2015). With the program in place, the government passed the Act for the Recruitment and Employment of Foreign Professionals, which created an Entrepreneur Visa (Contact Taiwan 2018). Since its passage in 2015, the Entrepreneur Visa encourages foreigners to establish a startup in Taiwan that is focused on either serving the local market or having Taiwan as its headquarters (MoEA SME Administration 2023). Qualifications include the foreign national’s ability to secure VC funding or obtain a recommendation from an incubator or science park. Also, in 2020, the MoEA established Taiwan Accelerator Plus, which includes an international program that offers help to international startups interested in expanding to Asia using Taiwan as their point of entry.10
Startup policies encourage this widening pool of domestic and international entrepreneurs to focus on specific emerging industries and technologies, like biotechnology and financial technology (fintech). In 2018, for example, the Financial Supervisory Commission, in partnership with its private-sector partner, the Taiwan Financial Service Roundtable, launched FinTechSpace, the country’s first fintech-focused coworking space. Speaking at the launch event, the Financial Supervisory Commission Chairman Wellington Koo said that “the authorities will continue to seek more workspaces across the country for startups at affordable prices to enable them and Taiwan’s economy to flourish” (Yang 2018). FinTechSpace opened with startups founded by foreigners constituting more than half of the accepted tenants.
To be sure, Taiwan’s employment setting has remained largely aligned with the Mark I variety, with policies striving to widen the pool of highly skilled entrepreneurial talent by encouraging internationally minded entrepreneurship among Taiwanese citizens and attracting foreigners to the ecosystem. It remains fluid rather than striving for lifetime employment with large firms. Unlike Japan and Korea, incumbent Taiwanese firms are not included in the judging process or named as intended partners for the startup talent development, nor are government programs encouraging secondments. Instead, the flexible labor market initiatives continue to motivate high-technology startup employment and moves across geographies and firms.
Finance
In the antecedent period, Taiwan’s state-owned banks directed little credit to small firms. These state-owned banks, especially before financial liberalization in the 1980s, had significant control over the provision of credit. As a result, startups had to seek alternative means of financing, including unregulated and unofficial sources (the grey market) (Chu 1999). Policy also provided other forms of help for startups, including R&D infrastructure and tax incentives (Fields 2012). The state, continuing the SEI practice, offered tax credits of 20 percent for high-technology sector R&D in the 1970s as well as a five-year corporate income tax holiday for “newly established capital or technology-intensive projects” (GIO 1986, 237).
Unlike policy makers in Japan and Korea, those in Taiwan worked to ensure that equity-based venture capital in the early 1980s was available to technology startups. They aimed to foster a local VC market, like that in Silicon Valley, to encourage Taiwan’s growing technology firms (Yeh 2006). For K. T. Li, the expectation was that a VC market could expand Taiwan’s incomplete financial services sector, promote its technology startups, and advance the local use of modern management techniques (Saxenian and Li 2002, 140). In 1981, Li organized for a group of senior policy makers and select industry leaders to take a study trip to the United States and Japan to learn more about VC.11 By the end of the trip, Li and his colleagues concluded that venture capital would “integrate capital, technology, talent, and management for the purpose of upgrading Taiwan’s technological developments” (Klingler-Vidra 2018, 81). Two years after their study trip, Taiwan’s CEPD passed a Ministry of Finance bill—the Regulations for the Administration of Venture Capital Enterprises—that gave a 20 percent tax credit for first-time VC investors (Wang 1995, 2). The 20 percent tax credit was offered to local first-time VC investors if they maintained their high-technology VC investment for a minimum of two years.
In the years that followed, additional efforts sought to boost the availability of equity-based VC for Taiwanese startups. In 1985, the Chiao Tung Bank formed a VC fund by providing capital along with the Development Fund and the Sino-American Foundation, with the remaining half of capital coming from the American investment bank H&Q. To bring local incumbents into the growing VC market, tax credits were expanded to incentivize local corporations to invest as VCs in 1991 (Kenney et al. 2002). VC reinvestment was encouraged, as tax exemptions were then offered on the capital gains earned by venture capitalists exiting from their investments in high-technology startups (Koh and Wong 2005, 26).
With the Democratic Progressive Party government coming into office in 2000, after having run on a platform that promised to lessen the allocation of state resources to established industries, except the R&D tax credits for breakthrough technologies, the state discontinued the 20 percent venture capital tax credit. The rationale was that Taiwan’s VC industry no longer needed such help, as it had become the fourth largest VC market in Asia (behind Japan, Hong Kong, and Singapore) and the world’s third most active, in terms of deal volume, behind only the United States and Israel (AVCJ 2005). It was also said to be the Asian venture capital market that was “most Silicon Valley–like” (Gulinello 2005, 845).12 While tax credits were ceased, additional government funding for this early-stage equity market came in October 2001, as the Development Fund participated in the National Development Plan. The fund raised the remaining 70 percent of the money from private-sector investors (Klingler-Vidra et al. 2016).
Despite its vibrant venture capital market, the MoEA oversaw several regulatory changes as a means of trying to encourage even more private capital in the VC market. For instance, in 2004, the Amendment of Regulations on the Scope and Guidance of Venture Capital Enterprises effectively expanded funding channels and eased restrictions on investment scope and fund utilization. Then, in 2006, a Relaxed Scope was passed, which made it easier for investors to exit their positions by decreasing the required holding time of company securities and lifting the limit on share sales. Collectively, the VC and stock market policy changes helped to improve the exit environment for equity investors in Taiwanese startups.
Policy makers have worked to expand stock market access for Taiwan’s high-growth startups and even for foreign startups. The Emerging Stock Board was created within the Taipei Exchange in 2002 to help startups access the capital market as a “pre-IPO” platform. The ESB was created in response to specific concerns about the stringency of rules around listing on the Taiwan Stock Exchange and the requirements for several years of revenue, which high-growth startups struggled to meet. With this, stock market access was expanded such that the number of companies listed on exchange grew to five hundred by 2005, and in 2009, the first foreign company listed on the Emerging Stock Board (Taipei Exchange 2022).
The launch of the InvesTaiwan Service Center in 2010 would further attract equity financing—from foreign sources—for the country’s growing technology startups. The InvesTaiwan Center had ITRI representatives, as the MoEA was targeting foreign investment into growing companies in the information technology and biotechnology sectors. Given the contemporaneous signing of the Economic Cooperation Framework Agreement with China, the InvesTaiwan Center was expected to help address the “many foreign investors [who] have expressed interest in investing in Taiwan” (Lin and Huang 2010). As part of the HeadStart Taiwan initiative, the National Development Fund offered US$400 million in matching funds for venture capitalists investing in Taiwanese startups (NDC 2014). The HeadStart funding included a 40 percent match funding provision, as had previous National Development Fund initiatives, to entice international VCs to invest in Taiwanese startups. Also, in a bid to attract foreign VC, in 2015, the NDC announced further efforts to boost local startups, with plans to co-invest alongside four international funds, including 500 Startups and AppWorks (Fulco 2015). Policy makers also created a bilateral VC fund, run in partnership with New Zealand, in March 2012, and in 2017, the National Development Fund established the NT$2 billion Business Angel Investment Program (NDC 2019).
Continuing with the thrust of equity investment as a primary means of boosting innovation in new industries and firms, a major funding development came in 2017 with the launch of Taiwania Capital by the National Development Fund and private investors (NDC 2017). Taiwania describes itself as the government’s venture capital arm; it aims to “boost Taiwan’s economic growth by partnering cutting-edge startups in sectors including information and communication technology (ICT), biotech, material, energy and new agriculture.”13 At the other end of the equity investment cycle, the Taipei Exchange created the Pioneer Stock Board to help companies operating in priority industries access the public equities market.
Collectively, since the deliberate growth of venture capital markets since the 1980s as a means of providing smart money for technology-oriented startups, policies have remained equity focused and have sought to expand the set of equity investors geographically (by eliciting international investors) and across the early stage (e.g., angels and VCs) and to advance exit venues available for selling equity stakes (e.g., startup-friendly stock markets). Thus, Taiwan remains closest to the Mark I variety in its equity-based financial system.
Innovation
Already in its antecedent period, Taiwanese policy aimed for the technological frontier rather than catch-up technologies. It established technology-specific labs, notably the crucial Industrial Technology Research Institute (ITRI) in 1973, as a key lab for technological advances. ITRI “received government contracts to conduct research programs, develop key technologies and transfer the results to industry in a non-exclusive manner” (Tsai 1999, 73). ITRI’s aim was to advance cutting-edge innovation capacity, particularly for integrated circuits via the 1974 creation of the Electronic Research Service Organization (ERSO). Taiwanese engineers called the ERSO mode of technology transfer “R&D: reverse and duplicate,” given the emphasis on reverse engineering rather than exploratory research (Saxenian 2006, 189).
As high-profile initiatives like ITRI and the Hsinchu Park signify, Taiwanese policy has long striven to achieve high-tech industrial development. Complementing the concerted push in Hsinchu Park, strategies such as the Four-Year Plan for 1982–1986 and the Ten-Year Plan for 1980–1989 constituted a continuing shying away from capital-intensive activities and toward technology-intensive industries, especially computers, telecommunications, and robotics, in which small firms could compete. Competing at the technological frontier was already essential from the 1980s, given the rise of quotas or tariffs in foreign markets. Owing to Taiwan’s early success, its firms were already facing protectionism in key international markets, so new technologies that would not yet be subject to trade barriers were backed. Industrial policy demonstrated the push for R&D so that Taiwanese firms were competing at the international technological frontier. In particular, the Statute for Encouragement of Investment was revised (again) in January 1981 by K. T. Li, such that “industries receiving benefits under the statute” had to “spend a standard amount of money on R&D” (Gold 1986, 103).
The focus on radical innovation, though, was not entirely smooth, as it still relied on the acquisition of foreign technology or on the manufacturing end of technologies designed elsewhere (e.g., in semiconductor manufacturing). While the acquisition of technology was often done on the backs of key geopolitical connections, there were cases, especially around the personal computer (PC) industry, where it received growing scrutiny. For instance, in 1982, Apple “clamped down on the thriving Taiwanese Apple II cloning industry” (Breznitz 2007, 114). There were challenges around the use of intellectual property from US companies, as in “1984, US Customs agents seized Taiwan-made, ERSO-designed IBM-compatible personal computers as fake IBM” personal computers (Gold 1986, 104). Despite the issues with IP infringement in the development of Taiwan’s PC industry, these early imitation experiences helped to establish technical abilities. Acer, in particular, had gleaned competitive capabilities in the PC market by the mid-1980s (Wen and Chen 2014, 231).
Competition at the technological frontier was fostered by the state’s help in creating and scaling new firms—or, in Mark I terms, in creating new entrepreneurial pools. Arguably the most successful illustration of this approach is TSMC, which was founded in 1986 as a pureplay semiconductor chip fabrication foundry spun out of ITRI’s ERSO.14 TSMC’s approach—which separated semiconductor chip design from fabrication, enabling it to produce chips for a wide range of companies, as it did not pose a competitive threat—aimed to super charge the centrality of Taiwan in the global semiconductor industry, the critical frontier technology of the time. Though the state helped create TSMC initially, it also restricted its strategy, as TSMC was “forbidden to design products of its own” so that it would “not rival smaller, private firms” (Fields 1997, 146). Furthermore, the “foundry model suited small firms, such as TSMC, trying to advance technologically because these firms could learn through serving their customers” (Fuller 2013a, 52). TSMC’s focus on foundries—and not integrated chip design—suited its international customers and made space for would-be Taiwanese startups competing in the space.15
The state continued to focus on the technological frontier. Owing to the success of its semiconductor industry, the state availed largesse for startups competing at the frontier of emerging technological areas, including biotechnology, optoelectronics, and green technologies, and not semiconductors (Wong 2005). For instance, biotechnology and digital content were named as “twin stars” in the 2002 launch of the Challenge 2008 project (Lee and Lai 2005).16 In 2006, ITRI established a science park in Tainan to boost one of the new priority industries. Rather than serving as an extension of Hsinchu and its focus on semiconductors, startups in Tainan were encouraged to focus on displays and optoelectronics, particularly liquid crystal displays (LCDs) (Kung and Yen 2018, 61). The Tsai administration, upon taking office in 2016, offered tax breaks aimed at specific emerging technologies (Kuo and Han 2017). It prioritized five industries: green energy technology, smart machinery, internet of things, biotechnology, and defense industry. As mentioned at the beginning of this chapter, Morris Chang, founder of TSMC, expressed his disappointment that the semiconductor industry, a critical technology and significant contributor to the Taiwanese economy, was to receive little assistance from the government’s plan (Chang 2017; Klingler-Vidra and Kuo 2021). As the Tsai administration entered its second year (in 2017), this priority list developed into the so-called 5+2 Industrial Innovation Plan, adding Asia Silicon Valley and the circular economy. The policies focused on new technologies and emerging industries rather than semiconductors and Taiwan’s oligopolistic firms further leading the market.
Social Purpose
Compared to the cases of Japan and Korea, the Taiwanese case has a more explicit focus on external motivators, relative to domestic social purposes. This stems from the centrality of economic competitiveness to national security—even its very survival as a state—from the 1970s. Taiwan’s economic statecraft was piqued when the United Nations voted to recognize the People’s Republic of China (PRC) in October 1971 and simultaneously expel Taiwan. Taiwanese policy makers feared the implications of the “vulnerability of Taiwan’s de facto sovereignty” (Pempel and Tsunekawa 2015, 6) and prioritized strategies for helping ensure trade continued. Reflecting this reality, the “regime increasingly based its legitimacy on its ability to promote economic growth” (Gold 1986, 90). Economic competitiveness based on high-technology prowess was seen as essential to the survival of the (autonomous but little recognized) Taiwanese state.17 At the same time, innovation performance had domestic imperatives, as newly imposed tariffs and the increasing cost of Taiwanese labor was causing an increase in the costs of made in Taiwan products. This expedited the need to move up the value chain rather than compete on catch-up technologies on a cost basis.
Faced with what we would now call a middle-income trap, in the 1980s, Taiwanese policy makers, in consult with domestic and foreign advisers, focused “on high-technology industries: information, biotechnology, electro-optics, machinery and precision instruments, and environmental technology industries” (World Bank 1993, 133). The “authorities had made it clear that they were completely behind the policy of rapid industrialization and trade” and intervened to develop, and transfer, technologies (Gold 1986, 90).
By the dawn of the twenty-first century, the social purpose underpinning its startup capitalism remained a mix of domestic and external drivers. Rather than being too reliant on the Silicon Shield afforded by its position in the global semiconductor industry, the state was keen to diversify to emerging technologies. This included biotechnology, the green economy, and fintech (Executive Yuan 2001; Williams and Chang 2008). The aim was to build global links in these ascending technologies. This was evidenced in September 2016, when the NDC announced the Asia Silicon Valley Development Plan, which would promote Taiwan’s innovation and R&D activities and upgrade its startup ecosystem. The goal of the plan is to promote “innovation and R&D for devices and applications of Internet of Things (IoT)” and globally integrate “Taiwan’s startup and entrepreneurship ecosystem” (Asia Silicon Valley Development Agency 2024). The government has been promoting Startup Island Taiwan as a global brand. As part of these efforts, in October 2021, President Tsai announced the selection of nine startups as the country’s “Next Big” representatives, in a bid to both showcase startups and help these companies “expand to global markets” (Yen 2021).
The externally oriented social purpose of fostering global links revealed inherent tensions regarding how to reap opportunities in mainland China while not exposing the Taiwanese economy to too much of a national security risk. Especially considering the Economic Cooperation Framework Agreement’s 2010 furthering of ties, concerns were heightened. Perhaps the most visible tension came from the “the student-led Sunflower Movement” in 2014, which “successfully blocked the ratification of an agreement that would have allowed greater Chinese investment in Taiwan” (Lin 2021, 143). The aim was to navigate global opportunity, including that in mainland China, while working to retain talent and the technological edge in Taiwan.
Domestically, startup capitalism has been motivated as a mechanism to assuage rising living costs and slower growth. For instance, in 2008, the KMT candidate, Ma Ying-jeou, won and sought to enable closer cross-strait economic relations in order to address “slower growth, stagnating wages, demographic decline, a high youth unemployment rate, and the inequalities and risks produced by financialization” (Lin 2021, 142). Efforts were formalized in the Economic Cooperation Framework Agreement, which was signed in June 2010, to boost trade with mainland China. This opportunity to boost operations in China and receive investment from Chinese companies and VCs brought risks, too; Taiwan was ushering in unification, so the social purpose of subsequent policies was to achieve the delicate balance of benefiting from enhanced access to the mainland while not giving away Taiwan’s independence and unique competitive advantages.
As in Japan and Korea, greater diversity and inclusion also increasingly inform the rationales given when startup policies are launched. For instance, the MoEA SME Administration launched the Women Entrepreneurship Awards in 2018 and the Women Entrepreneurship Program in 2021. The program offers a range of services to bolster the presence and ability of women in the Taiwanese startup ecosystem, including “training courses, group counselling, accelerators for women, women entrepreneurship competition, financing, and networking, to emphasize enhancing the technological and international power of the female enterprises” (MoEA SME Administration 2021).
Taiwan’s motivating social purpose has remained more explicitly at the nexus of technological competitiveness and national security. The aim of fostering national security through its competitive position in critical and emerging technologies has adorned Taiwan’s policy making for decades. While startup policies are also construed as a means of job creation and social inclusion, the centrality of national security as a motivator tips the balance toward the external end of the continuum.
Continuity and Change in Taiwan
Taiwan’s policy orientation was firmly established toward small firms already in the antecedents era. In this sense, Taiwan offers a different trajectory from Japan and Korea, which were large firm–centric at the outset. Taiwan was the East Asian state most like a stylized startup nation, given the early encouragement of startups rather than conglomerates and the corresponding availability of early-stage equity investment. Figure 4.1 offers an illustration of Taiwan’s change and continuity examined in this chapter and shows it has largely remained closer to the Mark I mode.
FIGURE 4.1.Analyzing Taiwan’s institutional evolution: antecedents to startup capitalism
Figure 4.1 illustrates Taiwan’s different trajectory as it begins, and remains, at 3.5 in the size of firms central to innovation arena. This is owing to small firms having long been integral to the Taiwanese state’s growing export-led growth competitiveness.18 Policy in even the antecedent period emphasized the boosting the capabilities of small firms by incentivizing R&D and export activity. Taiwan’s innovation economy includes giants, like Acer and TSMC. Unlike in the previous cases of Japan and Korea, however, startup policy has not aimed to boost the competitiveness of its leading firms and established industries. Open innovation efforts have been undertaken by oligopolistic firms independently, and the state has invited established firms to serve as investors in VC vehicles. But the thrust of Taiwan’s startup capitalism is around widening the entrepreneurial pool by inviting foreign talent and directing attention to emerging technologies. Thus, small firms—not oligopolies—remain central to its approach.
Employment has seen a shift toward Mark I’s fluidity (to 3.5), even though it began with a score of 2.5 in the antecedent period. The relatively high, and outward, movement of the Taiwanese case’s employment context reflects its policy orientation that has encouraged high-technology entrepreneurship. In the antecedent era, manpower programs sent students to study overseas, especially in the United States, and programs like the creation of Hsinchu Park worked to entice them back to join and lead high-growth firms. A thrust of employment policies has been around upskilling talent rather than encouraging entrepreneurship or midcareer movements. Over time, this ambition has only grown, targeting entrepreneurial capacity building in specific technologies (e.g., biotechnology, fintech, and green tech) and striving for an increasingly international pool of entrepreneurs. Thus, Taiwan’s labor market is, so far, the most flexible, earning it a score of 3.5, as job movements, rather than lifetime employment, have only grown in its startup capitalism era.
Taiwan was an early mover, among the East Asian cases, in promoting early-stage equity markets for high-growth startups, with VC policies being implemented from the beginning of the 1980s. Given this orientation toward equity financing for innovation in the antecedent period, the starting point for financing is 3.5 on figure 4.1. That means it is relatively closer to a Mark I character than the Japanese and Korean cases, which both began much closer to Mark II (scores of 1). Over time, Taiwan’s equity financing has only grown with the advance of VC markets and the expansion of startup-friendly stock markets. We illustrate this shift with a slight move outward, to a score of 4. This means that Taiwan began with, and still has, the financing for innovation scores closest to the Silicon Valley Mark I ideal type.
Continuing with this theme of having more of a Mark I character from the start, figure 4.1 depicts Taiwan’s innovation in the antecedents period with a score of 2.5. This reflects our observation that, early on, Taiwanese policy focused on competing at the world’s technological frontier, not on catch-up technologies. Taiwanese policy makers helped fledgling firms compete in cutting-edge industries by acquiring foreign technology and transferring it out of public institutes. Technological developments by ITRI were spun out to private companies, notably UMC and TSMC, so Taiwanese firms could compete in global value chains in the technology industry. The very success of Taiwan’s semiconductor-centric miracle motivated policy makers to encourage startup-centric clusters around emerging technologies, such as biotechnology, LCDs, and green tech. The goal remains explicitly focused on the technological frontier, with startup capitalism aiming to bolster capabilities in new technologies. Owing to this approach, even incremental advances in the dominant semiconductor industry are not enabled by the state. Thus, the startup capitalism score for Taiwan’s innovation is 4, placing it close to the radical innovation end of the continuum.
Also at an elevated level relative to Japan and Korea, which each scored 3, Taiwan’s score for the mix of external and domestic social purpose is a 4 across the antecedent and startup capitalism eras. This is due to the more explicit national security link for Taiwan’s startup-centric innovation efforts from the 1970s onward. In this sense, economic statecraft that links techno-competitiveness with national security is most striking in the Taiwanese case. Competing at the frontier of radical innovation was pursued in the context of semiconductors from the 1980s. As Taiwan established a foothold in semiconductor manufacturing, policies have focused on emerging technologies. The value of diversifying innovation capabilities in critical technologies has been central to policy making more so than job creation and social inclusion aims. Startup prowess does serve these domestic aims, especially at times when growth stalls and cost-of-living concerns are heightened. But techno-competitiveness retains primacy. For this reason, Taiwan’s balance remains closer to the Mark I end of the spectrum, with a score of 4.
In sum, Taiwan remains closer to the Silicon Valley ideal than the Japanese and Korean cases. However, we note that creative destruction is not necessarily the aim of startup policy in Taiwan. Rather than disrupting large incumbents’ positioning, the objective of startup policies is to foster new entrepreneurial pools in emerging technologies. The goal is to catapult capabilities in the next generation of critical technologies, not to disrupt the dominant position of oligopolies in established industries.