Made in Taiwan: How Developing Countries Can Leverage High Tech
- April 27, 2023
- Lukasz Bednarski
Taiwan’s hi-tech revolution began with a vision – now the small island-state has captured around sixty per cent of the world market in semiconductor chips, used in everything from medical equipment and military hardware to smartphones and washing machines. Just how this happened can provide valuable lessons for others.
Taiwan’s economic success is linked to the development of hi-tech industry. Technological achievements helped the country grow rapidly, while avoiding the middle-income trap if a country fails to transition to a high-income economy due to rising costs and declining competitiveness. Taiwan, with its semiconductor cluster – especially TSMC, the manufacturer at the forefront of the industry – provides companies such as Apple and NVIDIA with the microchips that are critical to their success. Taiwan’s flagship position in the industry allows the country to punch above its weight, as its geopolitical security depends on its position in the global supply chain.
A crucial role in hi-tech supply chains turned Taiwan and South Korea into notable players in both their region and on the international stage, and so it is worth considering what benefits might accrue for other countries currently at the fringes of technological competition. In order to gain similar success, they would need to fulfil a list of prerequisites in terms of economic development, social capital, and institutional stability. True frontier markets, low on the Human Development Index and plagued by internal conflicts, have a long way to go before such opportunities might materialise.
Yet nobody expected the Taiwanese semiconductor industry to succeed. The island was far from the electronics industry trade routes and Silicon Valley, where most innovation takes place. In the 1970s and 1980s, when Taiwan started manufacturing chips, China, which later grew to become one of the biggest buyers of Taiwan’s semiconductors, was grappling with agricultural reforms and only just beginning to open up to foreign trade. Taiwan is affected by earthquakes and water scarcity: semiconductor manufacturing companies need significant amounts of pure water to operate, while earthquakes risk destroying expensive and sensitive manufacturing equipment. Even slight vibrations can result in costly defects during manufacturing.
But the semiconductor industry was a transplant that took, thanks to the firm actions of government officials. Free market forces played a limited role in the very beginning; instead, it was as if the Taiwanese government willed the industry into existence. How did they do it? What lessons were learned, and can this process be repeated for similar hi-tech manufacturing industries in other countries?
Taiwan’s hi-tech revolution started with a vision: moving the economy away from textiles, agriculture, and low-value manufacturing to another level, propelled by a sense of urgency. Taiwanese economic growth had slowed down and plateaued, just as China, its big neighbour to the north, was waking up from its commercial slumber. Technocrats in Taiwan’s government asked its National Science Council to propose industries that would allow the economy to move up the value chain and compete with China. Four areas were identified: electric vehicles; pharmaceuticals; geothermal energy; and semiconductors. All these proposals, in retrospect, had the potential to make Taiwan successful. But electric vehicles seemed too embryonic, and geothermal energy technologies were difficult to commercialise. The government picked semiconductors. The industry was considered a bit esoteric, but technocrats had noticed the success of the Japanese electronic watch industry and the opportunity to supply chips for this market. Taiwan could produce semiconductors for Japanese watches cheaper than anybody else, with its low labour costs. The groundwork for success was laid out by combining low-cost production inputs with cutting-edge research, providing the industry with an in-depth understanding of legacy technologies and creating a platform for future innovations.
The Industrial Technology Research Institute (ITRI) was established by the government in 1973 to pursue research and development. At first it lacked the relevant expertise, but a government-funded institutional framework meant knowledge could be accumulated and shared with the private sector. Simultaneously, the Technical Advisory Committee was convened in the US, gathering the foremost Chinese industry experts, eager for pro bono work on behalf of Taiwanese industrial development. This initiative to harness the power and talents of the Chinese diaspora helped shape ITRI’s objectives.
ITRI tried to partner with the world’s established semiconductor companies, but companies such as Texas Instruments or Intel were not interested. Other potential partners understood the Taiwanese plan to be a turnkey project, and proposed to build a semiconductor manufacturing plant in Taiwan and teach ITRI staff to operate it. That was not, however, what the Taiwanese envisioned. They wanted to be taught how to design and build the plant as well as to be helped to master the process. From the beginning, the project was commercially driven. In the 1970s, the Taiwanese economic landscape had been dominated by state-owned companies, who played a significant role in the early industrialisation of the island, but technocrats saw the limits to the role of the state.
Radio Corporation America (RCA) eventually won the partnership bid. At that time, RCA was still a significant player in the electronics industry, but its boom years were behind it. The semiconductor industry was also experiencing a downturn, which prompted the firm commit to such an unconventional project. RCA was competent in two semiconductor technologies of the time: CMOS (Complementary Metal-Oxide-Semiconductor) and NMOS (N-type Metal–Oxide–Semiconductor). NMOS was closer to the cutting edge, as CMOS was already a legacy technology. ITRI understood it would be difficult to compete in NMOS, so they focused on the legacy tech that could still generate revenue. However, to hedge risk and its position for the future, part of the workforce also learned how to make NMOS.
ITRI staff were sent to the RCA plant in Ohio to learn the entire process hands-on, from design to production. There was no syllabus; everyone had a mentor and was shown how to operate the machines. Since the technology was already far from cutting-edge, the Ohio staff felt free to share their know-how.
As the next step, ITRI decided to build its first semiconductor plant in Hsinchu Industrial Park. It was almost an exact copy of Ohio’s plant, but with a few innovations. It turned out to be operated so well that in a few months it surpassed RCA’s production yields. The plant was so successful that RCA was even interested in acquiring it.
To help the plant secure its market, RCA was offered chips at a discount. The second buyer was the Taiwanese military. According to the initial plan, the third buyer was to be a Japanese electronic manufacturer, one big enough to make a difference to profits of the plant, but small enough to look for solutions for its manufacturing needs outside of Japan – which was dominated by large companies, which blocked access to semiconductors with large orders. The ITRI’s commercial success led to a spin-off of the first private Taiwanese semiconductor company – United Microelectronics Corporation (UMC), still one of the most prominent semiconductor industry players. This development gave the signal that chips were an actual business opportunity. Soon large Taiwanese companies followed suit, trying to break in into the market. The industry’s talent was in such high demand that the employee turnover rate in ITRI reached 80 per cent.
Nevertheless, the Taiwanese semiconductor industry enjoyed little innovation at the time, so when Morris Chang, a big name in the semiconductors industry, came onto the job market after being passed over for a CEO job at Texas Instruments, he was immediately offered the position as boss of ITRI, with a mandate to innovate. The government was willing to put in money, at least once cornerstone investors among foreign companies had been tempted. Chang, with his network, seemed like the perfect candidate to attract them. He understood that Taiwan would face hard times competing in chip design, research and development, and decided to focus on production. This strategy, though hardly innovative, attracted smaller and middle-size semiconductor companies, who could not afford to build their own manufacturing facilities and had a hard time securing capacity at competitor plants – as this risked revealing their trade secrets.
Chang offered the world’s first pure-play foundry – one that only makes rather than designs chips for other firms. Industry leaders, such as Sony, Texas Instruments, Intel, and Hitachi, were initially uninterested, but, he eventually managed to secure Phillips to invest twenty-eight per cent. The company was not exactly his first choice. In Chang’s words, they were ‘first-rate second-raters.’ He had little choice, though; it was Phillips or nothing. Even wooing Phillips took eight months. Once Phillips was in, the Taiwanese government pledged forty-eight per cent and the rest of the equity came from local investors, who were almost coerced by government technocrats to put money in, and a development bank. The government also provided the new company – TSMC – with staff and equipment from ITRI, substantially decreasing the start-up costs. It was not easy at first to convince clients about this new business model. But soon, the idea caught on. It empowered new so-called ‘fabless’ companies — firms that outsource fabrication — such as NVIDIA (currently a leader in chips for AI applications), which have grown into industry giants. And it decreased barriers to entry for semiconductor start-ups that did not have to spend on manufacturing. Now TSMC produces over half of the world’s chips, counting Apple, Intel and NVIDIA as its customers. Due to TSMC’s ubiquity in electronics, consumers are more likely than not to find a TSMC product in their household, something that could not be said with any level of certainty about any other company in the world. Taiwan became a semiconductor powerhouse. Four of the world’s top ten chip foundries are in Taiwan, adding up to sixty per cent of the global market.
This Taiwanese semiconductor case study provides us with unique insights. It proves that a hi-tech industry can be established successfully in country without experience and know-how in the field, far away from the existing supply chains and trade routes. It is a precedent that allows for more creative thinking and brings opportunities for nations in different stages of industrialisation.
Input from government and academia is crucial, particularly in the early stages. Such efforts should begin with identifying a broad area for hi-tech industry development. The details can be worked out later. The help of experts and industry insiders in the diaspora can be harnessed, but support can be also sourced from agencies of intergovernmental organisations or independent consulting firms. The hi-tech industry identified for development should harness the existing comparative advantage of the country in the form of the location, resource base, talent base or production costs. Government and academia should create an institution with a high level of independence, aiming to pursue fundamental research in the identified field. While such an institution has to be developed jointly by government’s administration and institutions of higher learning, it must also enjoy a high level of independence from both when making decisions.
Know-how for industry development cannot be theoretical in scope and can rarely be built from the ground up. Cooperation with foreign industry players is needed, therefore, to transfer knowledge. The arrangement needs to be practical. It should allow access to production facilities and an on-site, hands-on learning experience. The shared knowledge must be commercially applicable. The acquired skills should include not only operational but also facility design skills.
As a next step, a foreign partner is recommended to assist in designing and constructing the country’s first hi-tech manufacturing facilities. Third party contributions to gradual innovation during the design and construction are to be encouraged. Countries cooperating with the foreign firm should target legacy technologies with commercial opportunities. Such technologies are typically easier to master and can provide a solid head start into the industry, while also supplying a return on investment. It is unreasonable to believe that any company will share their cutting-edge technologies.
While such a facility, in the beginning, needs to be state-owned, it should provide a platform for private industry spin-offs, or become privatised at some point through listing on the stock exchange, with an opportunity to acquire stocks by minority investors. The foreign partner should help with the marketing of the facility’s first products or undertake to acquire them at a discount. The country’s public sector and military are the natural next customers. As the industry develops, the government can assist in ensuing projects, but only if these find private cornerstone investors and are innovative in relation to the country’s previous industry projects.
The success of Taiwan in developing its chip industry offers valuable lessons for other countries looking to move up the value chain and achieve economic growth. It is a remarkable case study of how a vision, pragmatic solutions, government support, and foreign investment can transform a country’s economic landscape. While participation in high-tech supply chains is not a guarantee of success, it has provided Taiwan with benefits that extend beyond economic gains. As technological competition intensifies, the lessons learned from Taiwan’s experience may prove crucial for countries looking to avoid the middle-income trap and establish themselves as important players on the global stage.