This website uses cookies and other technologies to help us provide you with better content and customized services. If you want to continue to enjoy this website’s content, please agree to our use of cookies. For more information on cookies and their use, please see our Privacy Policy.


切換側邊選單 切換搜尋選單

Reversing the Brain Drain

Time for Homecomings


Time for Homecomings

Source:Chien-Tong Wang

Thirty years ago, Taiwan sent a team overseas to study semiconductors and recruited Morris Chang back to the country. Can history repeat itself? Taiwan is hoping it can reverse its brain drain, but the government and private sector need to step up.



Time for Homecomings

By Yi-shan Chen
From CommonWealth Magazine (vol. 628 )

Ethan Tu, who founded the PTT bulletin board system when he was still in college, had a great job as Microsoft’s director of AI research for Asia-Pacific. But he resigned in April and left the United States, where he had been living for 13 years, to return to Taiwan and set up Taiwan AI Labs – news that stunned many in young ethnic Chinese communities.

“The PTT founder’s return to Taiwan is a main topic of conversation when Taiwanese engineers in the San Francisco Bay Area get together,” said Chris Liao, who graduated with an electrical engineering degree from National Taiwan University (NTU) in 2008 and now works for Google in its advertising division, in a speech to the Cabinet’s Board of Science and Technology.

The changes that have occurred in Silicon Valley in the past 10 years reflect the outflow of Taiwan’s software talent. Liao said that when alumni from NTU’s Department of Electrical Engineering would first get together for a meal in the Bay Area, one table would be more than enough, but now four tables are needed. Of the 200 people who graduated from the department in his year alone, 50 are in the United States, and of those who went into software, about 80 percent are in America.    

Whether Tu can help turn that around remains an open question, but speaking in an office behind Taipei’s main train station, he seems encouraged.

“I’ve been back in Taiwan for less than three months, but I’ve been surprised by three things,” Tu says. “First, it’s possible to quickly find talent. Second, the caliber of talent is very good. Third, government agencies and companies have been able to move pretty quickly.”

After his return, Tu rented a 230-square-meter office. He originally thought of setting up an operation with his own money and recruiting about 10 people. But to his surprise, he ended up receiving nearly 1,000 resumes a week from people with strong credentials, leading to his hiring of 20 people in one month. He continued his recruiting efforts and after a careful evaluation, decided to bring in investors and rent more office space.

Compared to the average pay of US$250,000 a year for an AI specialist in the United States, Tu is offering relatively reasonable compensation by Taiwan standards in recruiting employees, of whom one-third are people returning from abroad, one-third are new entrants to the job market, and one-third are senior hardware engineers.

“AI is like the internet in the 1990s; it will redefine the world. At this point in time, it’s about bravely creating dreams. As long as you are doing the right thing and there’s a challenge and you are creating demand and the right environment, talent will come,” Tu says.

Economic Growth Eroded by Depreciation

To young talent with high international mobility, “career development” is the biggest factor in determining where they decide to settle. So for Taiwan to reverse its talent deficit, addressing the problems with industrial transformation that have festered since the beginning of the new millennium is the top priority.   

“Many people think the problem of Taiwan’s low wages is the result of industries relocating overseas. I don’t buy that. The key is that Taiwan has not completed its transition from an industrial economy to a service economy,” says Academia Sinica academician and economist Ping Wang.

Wang and two other academicians, Chang-Tai Hsieh and Cyrus Chu, launched a study on Taiwan’s competitiveness and economic growth strategy in 2015. Among its conclusions, it identified two major obstacles to Taiwan’s industrial transformation: inadequate business startup momentum and neglect of soft power, leading to a lack of progress in the development of the software and service sectors.   

On Aug. 7, Wang and Hsieh led the Taiwan Economic Association in organizing a forum on innovation in tech services. It was one of the biggest gatherings of Taiwan’s economics community in recent times, signaling the urgency economists feel in addressing Taiwan’s stubbornly low wages and economic stagnation.

Wang says that during the transition from an industrial to a services economy, advanced countries don’t necessarily see a decline in the “added value” created, but that has not been the case in Taiwan.

Wang, Chu and Hsieh studied 16 existing sectors in Taiwan based on three indicators – total output, value-added, and number of people employed – and found that the only sector with high value-added and a leading position internationally was the semiconductor sector.

Of primary concern, they noted, is that while many sectors in Taiwan, including petrochemicals, financial services and the wholesale and retail sector, may have large outputs, many related industries and high numbers of people employed, they are plagued by low value-added and lack international competitiveness. The value-added ratio for Taiwan’s petrochemical sector is below 10 percent, well below Japan’s 22.8 percent, but it is 70.2 percent for Taiwan’s semiconductor foundry sector.

Wang contends that Taiwan’s low value-added problem was already becoming evident at the turn of the century. But the government’s response in the early 2000s was to invest in the “Two Trillion-dollar Industries” – DRAMs and flat panels – in an attempt to use public policy to fill in gaps in investment. At the time, Taiwan’s flat panel and DRAM sectors did not have their own technologies, yet the government gave them tax breaks and low-interest financing and continued to inject money into “zombie companies” even after they ran into trouble.

Framing the problem in statistical terms, capital depreciation has offset economic growth, and the productivity of capital has been even lower than the productivity of labor – meaning that neither companies nor workers have enjoyed the benefits of economic growth.

Solution No. 1

Creating a New Hardware-Software Ecosystem

“It’s really very ‘hard.’ Taiwan is the ‘hardest’ place I’ve seen,” Tu says after taking part in a discussion with government officials and people in the tech industry, referring to Taiwan’s mindset on “hardware.”

Tu believes the most practical way for Taiwan to make up for lost time and get a foothold in the software era would be for hardware and software circles to build mutual understanding and recognize that each side has considerable expertise. The two sides must respect each other and respect each other’s differences if they hope to attract top-flight talent, he contends.

“Software developers are people who stack building blocks. The most important thing to them is the kind of castle they will build. Having good building blocks at their side is an advantage, but whose blocks they use is not a priority,” Tu says bluntly.

Synology Inc., a Taiwanese company specializing in network-attached storage devices, is often cited as a prime example of hardware-software integration. But in an interview with CommonWealth, Synology CEO Vic Hsu openly defined his company as “selling software packaged in hardware.”

Engineer Ensky Lin (second from left) was made responsible for special projects in just his second year with Synology. He has appreciated a creative process that allows him to create something out of nothing.

It has 600 employees in Taiwan, of whom 70 percent are involved in R&D. Forty percent of the total workforce develops software, compared to only 10 percent dedicated to hardware.

Ensky Lin is a product developer in Synology’s Software Development Group who has been with the company for three years and was recognized as the company’s most valuable employee. In his second year there, he was put in charge of developing the software for Chat, an instant messaging service for Synology private cloud customers.

Lin compared his situation with that of a classmate working for a hardware manufacturer in Hsinchu who is also a software engineer. Lin said his classmate mainly writes software based on specifications provided by the customer, but Synology develops products based on users’ needs, and its software can also be updated at any time as necessary. The company is planning to recruit 150 people this year, with 80 of them software engineers.

“There’s no problem with Taiwan’s talent, but they need to be coached,” says Tu, who has interviewed countless engineers over the past two months.   

Replicating the ‘Morris Chang’ Experience

T.C. Wu, the chief technology officer of the Asian Silicon Valley Development Agency that is dedicated to developing Internet of Things applications, has observed the same thing.

Wu, the co-founder of Silicon Valley semiconductor manufacturer Atmel Corporation, cited the example of system architects. When developing a product in the United States, Wu says, a system architect first evaluates which problems can be solved by hardware and which ones can be solved by software. But Taiwanese tech companies don’t make big systems, so those processes and that kind of talent do not exist.

The result? When a problem arises, they force a hardware solution on it, even if it takes longer to develop hardware than software and the associated risks are higher.

Tu suggested replicating an old model.

“Thirty years ago, Taiwan’s government was willing to send a team to RCA in the United States to learn about semiconductor production, and asked Morris Chang to return to the country. Today, the government can spend resources to cultivate young people, recruit international talent or attract foreign companies to Taiwan to set up R&D centers. They don’t have to be in top cities; they can create a software talent cluster outside a major urban area,” he says.

One example Tu cited was a Microsoft AI lab being set up in Suzhou, a second-tier city in China.

Institutional Reforms

Beyond talent issues, Wu was also concerned that “Taiwan’s overall system is not that easy to maneuver” in terms of software and new industries.

Wu says Taiwan’s legal ecosystem needs an overhaul, but government agencies are generally conservative and move rather slowly, ultimately forcing business people to register their companies in tax havens.

The three Academia Sinica academicians believe that quick fixes to the regulatory environment are unlikely, especially with administrative foot-dragging having become the norm. But Wu suggested that “trial regulations on innovative startups” could be drafted to enable Ministry of Science and Technology or Ministry of Economic Affairs officials to bypass or break through the bureaucratic morass.

The Industrial Technology Research Institute (ITRI) and other government-sponsored research institutes are another part of the ecosystem that needs an overhaul.

Based on the three academicians’ research, the ITRI was once an incubator of new startups because researchers at the institute were willing to move around and were highly motivated to bring their innovations to the private sector. It was also very convenient to do, resulting in a constant flow of new enterprises. That transfer of knowledge and people has been less evident in recent years.

“Sources of capital and personnel mobility are keys to reform,” Cyrus Chu says.

At the end of July, Economic Affairs Minister Lee Chih-kung announced plans to reform the 16 research institutes under the ministry’s supervision. One important provision is the establishment of a “supervisor evaluation system” by the end of the year that will make it possible for institutes and companies to jointly hire or rotate personnel, creating a smooth flow of information and talent between academia and the private sector. This represents the biggest reform of research institutes in the last 40 years.  

Solution No. 2

Expanding the Service Sector

Beyond changing the technology ecosystem, Wang notes that there is an increasing amount of literature on economic development that stresses the importance of “soft power” and the service sector.

In the United States, for example, the most significant innovations and gains in productivity in the past 20 years have occurred in services.

According to Hsieh, “what the service sector needs to consider is not what people don’t know but what everybody does know but is not being done well,” pointing to a crucial aspect of the transformation of Taiwan’s service sector.

In Wang and Hsieh’s eyes, Taiwan’s globally recognized medical sector is the domestic service sector best-positioned to be exported to the world.

“In the future, Taiwan has to get involved in AI, smart machinery and information services and bring all of these trends together and move toward software and services. Only then can Taiwan harness its soft power advantage,” Wang suggested.

Changhua Christian Hospital has shown how a hospital can become a platform and export systems and knowledge and nurture new businesses, in the process helping employees find new stages on which to perform and offsetting a difficult domestic environment hurt by shrinking reimbursements from the national health insurance system. 

“Being called ‘teacher’ and helping students solve problems gives me a sense of accomplishment,” says Chen Hsin-chuan, who has taught in China for seven years as part of Changhua Christian Hospital’s team.

Twenty years ago, Changhua Christian Hospital began taking in community hospitals that were not doing well and turned them into branches, and it now has 10 of them in all. Its standard procedure after taking over another facility is to send a superintendent and chief administrator to run the branch while providing support behind the scenes. It has set up a system management center that nurtures a pool of management talent in everything from administration, medical supplies, and medical engineering to nursing, finance and nutrition.

It was also 20 years ago that the hospital started to institute international evaluations to force itself to standardize and systemize its management.

“The first 10 years paved the way for the following 10 years,” says Chan Tien-min, a vice superintendent who is also responsible for the hospital’s human resources division and International Medical Center.

This systematic development of a talent pool and knowledge management made it possible for hospital superintendent Kuo Shou-jen and another vice superintendent, Chen Hsiu-chu, to react quickly the first time a Chinese hospital came calling for help in 2010. They immediately sent more than 20 supervisors to the new Beijing Yanda Hospital and got it up and running in just 100 days, creating a name for Changhua Christian Hospital in China.

It later drew the interest of China’s biggest leasing company – Far Eastern Horizon Limited. Through a donation of shares, the Chinese giant set up a joint venture with Changhua Christian Hospital, called Horizon Healthcare Management (Shanghai) Co., to provide management consulting services to Chinese hospitals.

The Changhua hospital currently has 15 employees permanently stationed in China, and flies 100 qualified lecturers to China every month to teach courses on hospital management. They have served more than 60 Chinese hospitals.

Aside from instructors receiving double pay for their efforts, the hospital was able to distribute year-end bonuses equal to 2.6-months salary because of consulting income earned from Horizon Healthcare Management, keeping turnover in the nursing staff to a relatively low 7 percent.

More than Prayers Needed to Help Young People

Kuo admits that many people at the hospital initially opposed the idea of going to China and even cooperating with Chinese companies. “But I was committed to doing it, no matter what,” he recalls.

Kuo believed that continually shrinking reimbursements from the national health insurance system for services did not bode well for the hospital’s long-term prospects, and that real hope was needed to help young people rather than simply praying for them.

Two things were needed to provide that hope: first, a platform on which they could harness their talents, and second, money. “Money was still the most important thing. Why not make money from the Communists?” he says.

Kuo turned the hospital into a platform that not only exported medical expertise but also found value in back office services such as medical management, engineering projects and even cashier services.

Changhua Christian Hospital currently runs six companies, and in June, Kuo became the chairman of Swissray Global Healthcare Holding Ltd., a company listed on Taiwan’s over-the-counter market.

Solution No. 3

Recognizing Taiwan’s Strengths/Weaknesses

“Actually, Taiwan’s environment is pretty good,” says Arthur Huang, the CEO of sustainable solutions provider Miniwiz Co. who co-founded the company 12 years ago. “The key is taking a hard look at ourselves in the mirror every day, being critical of ourselves, knowing where we’re good and not so good, and finding the right road for us.”

Huang said Miniwiz knew from the beginning it needed to be active in international markets, but it was also aware that being Taiwanese or a Taiwanese brand offered no cultural advantage in the environmental realm. Yet to have the chance to attack the Asian market, Miniwiz had to establish a stable beachhead in Europe.

Academia Sinica academician Ping Wang believes Taiwan’s low wages are not because of businesses relocating abroad but because Taiwan has not made the transition from an industrial economy to a service economy.

Using Europeans is the best way to gain a foothold in Europe, so Huang recruited two Europeans among his five partners. Johann Boedecker, who now heads Miniwiz’s branches in Europe and the United States, was one of the earliest partners Huang cultivated.

Aside from paying them a good salary, Huang also rents housing for his European colleagues.  

“Even without financial resources, you have to do that because these people are necessary investments in the business. Only by doing this can I find the best people,” he says.

But when asked if he felt frustrated by that, Huang says no.

“If you want to promote your concepts to the world, why worry about who tells the story,” he says with confidence. In addition, even if being Taiwanese offers no cultural edge, it does come with a reputation of being “low cost, highly efficient and responsible,” he notes.

From reversing mindsets to overhauling systems, Taiwan desperately needs to get going on its long road to economic transformation because providing a suitable platform on which the new generation can fulfill its potential has become more urgent by the day in the fight to reverse the country’s talent deficit.

Translated from the Chinese by Luke Sabatier