On a weekend in June, with cottony white clouds hanging so low in the sky that they seem almost within reach and the piercing California sun shining down, the campus of Stanford University, the incubator for so many of Silicon Valley's ideas and people, is almost deserted. The students are in the library or their dorms preparing for their final exams, and the campus belongs to its visitors.

Chu Cheng-tao, a 27-year-old Taiwanese man who has just completed a Master's degree in computer science at Stanford and stepped down as head of the school's Taiwanese Student Association, appears at our appointed meeting place in front of the Memorial Chapel. Chu got his bachelor's degree at Taiwan's National Chiao Tung University, before learning three years ago that he had been admitted to Stanford. When he packed up his things and flew across the Pacific, he became only the third person in his class of more than 50 to pursue an advanced degree abroad.

With his affinity for the sciences, Chu took to Silicon Valley like a fish to water. In March, with his graduation looming, he took an internship with Google, the world's leading Internet search engine company. His formal entry into the working world went equally swimmingly: He submitted his resume to Google's jobs portal, and was hired to a paid position after just one interview.

Chu Cheng-tao, a graduate of Taiwan's National Chiao Tung University with a Master's degree in computer science from Stanford, now has the job he had hoped for with Google, the world's leading search engine company.

Fewer students going abroad

Chu looks set to earn a very good return on the two years he invested in a Master's degree from a Western university. Meanwhile, Wang Fu-min, who has just embarked on the sixth year of a grueling PhD program, is torn between continuing his academic career or leaving it to become an entrepreneur.

Wang sighs. He began his study abroad at the age of 31, late by most standards. After getting a Master's degree from UCLA, he caught the attention of a renowned professor in Stanford's Department of Electrical Engineering, who got him a scholarship and took him under his wing. When Wang was in his third year at Stanford, this professor unexpectedly left for MIT, leaving Wang to find a new advisor. This led to difficulties with his research on the use of semiconductors to study nanoscale device structures, and he now estimates that he'll need at least three more years to complete his PhD.

As the president of the North California chapter of National Taiwan University's alumni association, Wang is well aware of how few Taiwanese students are currently studying at prestigious American universities. "During the last five years," he sighs, "not one other Taiwanese student has entered my department. Stanford's standards for admission have always been high, so there have never been many. These days, on the other hand, about 70% of the ethnic Chinese studying in the less selective Cal State university system are from mainland China."

Due to new opportunities and changes in the market, these "tech-industry gypsies" are engaged in a third wave of technological emigrations, heading from Silicon Valley to Asia.

Losing ground

The pursuit of a PhD is a test of time, intellect and willpower.

"Most Taiwanese who decide not to pursue a PhD," says Wang, "make the decision because they haven't received funding." Wang explains that in addition to the sharp decline in the number of Taiwanese PhD students, there are few Taiwan-born academics in the prime of their careers teaching in the US, whereas there are growing numbers of faculty born in China and Korea who, to some extent, tend to look out for young scholars from their own homelands.

Students from mainland China have been flocking to other nations since the liberalizations of 1979. According to the US's Institute of International Education, China became the number-one source of foreign students in the US in 1993. At that time, Taiwan still had 37,000 students in the US, a figure not far short of China's 45,000.

The number of Chinese studying in the US has continued to grow rapidly over the last decade and reached 62,000 in 2005. China's huge contingent of students in other nations have become a force to be reckoned with, one of the drivers supporting the mainland's economic development. In contrast, the number of Taiwanese studying abroad has been shrinking. In 2005, there were only 25,000 Taiwanese attending school in the US, a decrease of more than 10,000 over ten years.

These numbers are allowing the mainland to establish beachheads at major US universities, corporations and research institutions, while Taiwan's reservoir of international resources is drying up.

According to Chen Shin-horng, head of the Chung Hua Institution for Economic Research's (CIER) International Division, mainland China ranked first in the number of foreign scholars in the US, accounting for more than 15,000 of the total of 84,000 who were there in 2003. Taiwan ranked 14th with only 1,241 scholars in the US. Korea, meanwhile, which has a population only double that of Taiwan, ranked second with 7,286 scholars in the US, or more than five times the number of Taiwanese. Given that India ranks third and Japan fourth, it's clear that Asian scholars are a force to be reckoned with in US academia.

Knowledge is the sine qua non of industry. Silicon Valley's Stanford University has been a model of how to manage academic-industrial partnerships. Its faculty members encourage their students to start their own businesses, and many even run their own businesses.

The semiconductor black hole

Over the last 20 years, students pursuing advanced degrees abroad have been a crucial source of human capital in Taiwan's efforts to keep pace with the rest of the world and develop its technology businesses.

Wave after wave of Taiwanese students returned from the US to start their own businesses in the 1980s and 1990s, helping to build Taiwan's own Silicon Valley--the Hsinchu Science Park. The number of returnees establishing themselves in Hsinchu hit a high of 6,000 in 1995. As of 2004, Taiwanese who had studied abroad had founded 113 of the park's 384 firms (29%) and accounted for 4% of its 100,000 workers. In recent years, the numbers returning to Hsinchu have ebbed, a fact which some attribute to the science park having reached capacity. They note that these days, it is the Central and Southern Taiwan Science Parks that are attracting major corporations and start-ups.

Others feel that the decline in the numbers returning is a result of the US eco-nomy's gradual recovery from the dot-com crash, coupled with the declining number of Taiwanese students going abroad.

Why is it that fewer and fewer Taiwanese have been going abroad to study over a period in which Taiwan's economy has taken off and local incomes have been climbing? If Taiwan's elite students don't go abroad, who will establish Taiwan's next-generation industries? And how should we go about training the kinds of engineers these industries will require?

Many believe that Taiwan Semiconductor Corporation and United Microelectronics Corporation have been almost twin black holes for skilled workers. Their voracious appetite has forced other firms--LCD panel makers, optoelectronics companies and digital content providers--to do what they can to poach personnel. All are competing to draw critical personnel from what is a very small pool, making it very difficult for small companies in particular to attract quality people.

Another reason for the declining number of returnees is that there are more disincentives to study abroad now. For example, the stock options that UMC began offering their employees as bonuses since 1983 have been very effective at retaining people. And, in 1999, the government began permitting graduate students in scientific and technical fields to perform their military service as military representatives at private firms. Those who are selected for this kind of service receive salaries and bonuses comparable to the company's actual employees, the only caveat being that they are required to serve for four years. The government has been allotting more spaces for this type of service every year, and there are about 3,900 positions this year. Finally, Taiwan's declining birth rate has reduced the pool of graduate-school applicants just as the number of graduate schools has increased, making entry into local graduate programs less competitive. This and other factors have made local programs more appealing to the top students in scientific and technological fields.

Successful companies require outstanding employees. Ethnic Chinese and Indians account for 25% of the Silicon Valley engineers who were born outside of the US. Valley residents like to joke that a strike by these Asian engineers would paralyze the valley.

Widening the "smiling curve"

If Taiwan's young scholars stop going abroad to study, the island's most talented personnel could soon find themselves without people to whom to hand over the reins. To date, Taiwan's information technology industries have advanced themselves through the efforts of personnel trained abroad. And manufacturing, management, and academia continue to recruit from Silicon Valley's talent pool.

For three consecutive years, the National Science Council has led a group of more than 40 industry representative to San Diego, Los Angeles, San Francisco, Vancouver, and Tokyo to recruit talent. The biggest names in Taiwanese technology, including TSMC, Compal, Realtek and Hon Hai have all participated.

It's expensive to hire from abroad. What is our strategic objective, and what kind of people do we need to attract?

The National Science Council projects that over the next three years, six major tech industries--semiconductors, displays, communications, information services, digital content and biotechnology--will create 106,000 jobs, and that the relevant departments at local universities will graduate only enough people to fill 96,000 of them. To meet the demand for personnel at the two extremes of the smiling curve--high-end research, and sales and marketing--Taiwan must recruit abroad.

These efforts have yielded results. The 2003 international recruiting efforts brought in 3,000 resumes and ultimately led to the hiring of 599 people. In 2004 the figures were 1,800 resumes and 656 hires, and in 2005, 2,590 resumes and 667 hires. Clearly, however, these numbers are nowhere near the projected demand.

Wang Pai-ling, assistant director of human resources at Compal, notes that Taiwan is the world's leading manufacturer of notebook computers. While Compal recruits both at home and abroad every year, it hires 90% of its employees in Taiwan. Those few it hires from abroad have experience with crucial technologies. For example, the company spent a year trying to hire an expert in the 802.11 wireless networking standards locally before finally finding someone in Silicon Valley. And it found the expert it needed to test notebook PC yields in Tokyo.

The head of research at a flash memory maker in the Hsinchu Science Park says that there are several thousand companies around the world doing IC logic and circuit design, at least several hundred of which are in Silicon Valley. But there are only 20-some companies developing flash memory, and the necessary people are exceedingly hard to come by.

His company is targeting the kind of experienced, capable people in this field who can propel the company's R&D to the next level. They made the rounds in Silicon Valley last year, arranging several interviews, but ultimately came away empty handed. They'll use the NSC's recruiting trip again this year, and continue their patient search. "Semiconductor R&D is beginning to shift from Europe and the US to Asia," he says. "And Taiwan has a lot of strengths in this arena."

Wang Fu-min, a PhD student in the Department of Mechanical Engineering at Stanford University, is deeply concerned about the steep decline in the number of Taiwanese students going abroad to study in recent years.

On the frontlines

CIER's Chen Shin-horng studies policies that attempt to foster the development of innovators and says that successful international brands have to be able to set the "rules of the game"--establishing prices, product standards, and the structure of the system. Their approach to innovation is different from that of a contract manufacturer; they don't simply respond to the technical issues their clients raise.

Building a brand is no simple task. You have to have a technological innovation and you have to have it at the right time. New brands tend to emerge during periods of product or industrial transition, which may involve disruptive innovations, new industry platforms (e.g. Korea's use of bandwidth compression technologies to launch its mobile communications industry), new product standards (e.g. the shift to clamshell cellphone handsets), or new applications (e.g. the Internet telephony company Skype). Companies wishing to establish themselves in a market can also purchase a brand or distribution channel, as in BenQ's purchase of Siemens' cell-phone division.

"Innovation in Taiwan remains largely a process of incremental technological improvement," says Chen. "But taking industries to the next level requires a focus on truly original work." Especially while China, India and Korea are all aggressively recruiting in Silicon Valley, Taiwanese firms must foster more talent and transform themselves if they are to remain competitive.

The personnel a company has on staff determine what direction it can take. These days, the world's leading corporations are targeting ethnic Chinese with international experience and competence in Chinese- and English-language settings.

Last July, two of the US's technological giants, Google and Microsoft, took their battle over the Taiwan-born Lee Kai-fu to the courts. Lee, whose research involves speech recognition, was Microsoft's first ethnically Chinese vice president. When he left the company to become Google's vice president of engineering and president of its China operations, it was big news in the US's Internet industry and sent shockwaves through Asia as well.

Lee had excelled as the founder of Microsoft's Asian research center in Beijing, but when he turned in his resignation in 2005, the company began a series of attacks. These attacks were visceral proof that the Microsoft-Google rivalry was beginning to extend beyond Seattle and Silicon Valley to the emerging markets of Asia--the two companies were taking up arms on the far side of the Pacific, and bringing the battle for key staff to the frontlines.

Silicon Valley's many natural advantages have attracted talent and capital from around, turning it into the heart of the global tech industry. The photo shows city hall in San Jose, one of the cities that comprise the valley.

A "blue ocean" personnel strategy

With the rise of China, what inducements can Taiwan offer to attract experienced ethnic Chinese from abroad to our shores?

Wang Ting-an, who, as executive secretary to the Executive Yuan's Science and Technology Advisory Group, runs the overseas recruiting trips. Many Chinese are unwilling to talk about the glass ceiling that confronts Asian emigres in the US. But the fact is that if former TSMC chairman Morris Chang had chosen to remain at Texas Instruments rather than come back to Taiwan, or Academia Sinica president Lee Yuan-tseh had stayed in the US, neither would have achieved as much.

"You can only attract people if you've established your own brand," says Chang Peng-heng, TSMC's vice president for human resources, offering industry's perspective. "Taiwan is already recognized as a leader in the IT industry." Chang believes that people like Lee Kai-fu don't want to limit their careers to Taiwan. If Taiwan's firms are to attract the talents of ethnic Chinese, ethnic Indians, and the globetrotting professionals now working in Japan, Korea, and elsewhere, they must give them more opportunity to "show their stuff" on the international stage. Chang speaks with authority on this point--TSMC currently has about 200 foreign white-collar staff working for it in Taiwan.

CIER's Chen believes that when thinking about personnel issues, Taiwanese firms need to abandon their traditional approach in favor of a "blue ocean" strategy. That is, they can't simply focus on attracting skilled workers, but must also themselves seek out new opportunities and directions. His reasoning is that no matter how much money the government offers to students to study abroad, Taiwan is still going to be left facing intractable structural problems, such as the expansion of higher educational opportunities in Taiwan, and the island's declining birth rate.

Chen recommends invigorating the economy and, more importantly, opening up the brain "fence" by providing skilled foreign workers with a welcoming environment in which to live and work. His other recommendations include strengthening national planning and government/private-sector cooperation on product development; building long-term relationships between Taiwanese research groups and foreign research parks; and encouraging local firms to establish overseas research and production facilities, the diverse foreign workforces of which will help their Taiwanese parent companies integrate their transnational operations.

Ethnic Chinese parents are committed to their children's education. Things are no different when they immigrate to the US--they squeeze into good school districts to ensure that their children can get into good high schools.

Tech-industry gypsies

Ethnic Chinese are engaged in a third wave of technological emigrations and returns--professionals from Silicon Valley and Taiwan are racing to China and circulating among all three locations.

Many of the upper management positions at foreign ventures in China are currently filled by individuals born in Taiwan. But as China produces more and more executives at home, and as others return from abroad, these Taiwanese are likely to be replaced.

Take Duh Jia-bin, for example. Once an executive with HP in Taiwan, Duh was sent to China in 1985 as vice president of the company's China operations. In 1994, he jumped ship, becoming the first president of Microsoft's China operations. He spent nearly five years with Microsoft, successfully fighting to build demand for the company's products in a Chinese market that was resistant to them. Then, in 1998, Cisco's John Chambers hired him as CEO of Cisco's China operations. Duh again excelled. In 2001, Cisco was reducing its global headcount, but its China operations remained unscathed. In fact, the company was experiencing phenomenal growth in the mainland Chinese market (sales had doubled again), leading it to increase its workforce there from 100 to 600, 95% of whom it hired locally.

Last August, Duh stepped down as president of Cisco, China. Mainland media speculated that less than ideal revenues were the reason, but the company's third-quarter China sales were up 10% from the previous year. Cisco's explanation was that the company's strategy required personnel adjustments.

After much thought, Duh declined Cisco's offer to transfer him to their corporate headquarters in the United States. He had spent 21 years in China, and his parents were still in Taiwan. His resignation took effect in May, ending his eight years with the company. Duh recently flew back to Kaohsiung from Beijing to visit his parents and is still thinking about what he will do next. Will he continue to work for major corporations? The 48-year-old Duh might well start his own business, striking while the iron is still hot. Regardless of what direction he takes from here, his dedication to his work has already produced tremendous successes.

The right people are crucial to national competitiveness. How much longer can Taiwan maintain its edge in personnel? It takes years to develop highly skilled workers. Time is pressing, and we can't afford to drag our feet.

Silicon Valley's many natural advantages have attracted talent and capital from around, turning it into the heart of the global tech industry. The photo shows city hall in San Jose, one of the cities that comprise the valley.

Silica, which is found in sand, stone and soil, is the tech industry's "secret" ingredient. High-purity silicon is a semiconductor of electricity. When transistors replaced vacuum tubes in the 1950s, it heralded a new era in electronics technology, one that companies such as HP, Intel and Apple were to take advantage of. The photo shows the Intel Museum, which has exhibits covering the entire history of Silicon Valley.