"Personal harmony, good location, favorable times"--these are the factors that have contributed to Taiwan's amazing success in the manufacture of computer hardware and peripherals. But customer requirements are changing. As the worldwide emphasis shifts to software, can the island keep up?
Let's do a little free association. Somebody says "Taiwan," and the knee-jerk response is...what? Most people would certainly come up with one or more of the following: "computer," "modem," and "high-tech." For Taiwan has long enjoyed the reputation of being at the cutting edge of the ultra-modern technology that powers the brave new electronic age. Most computer users have contact with Taiwan-manufactured hardware every working day, even though they may not be aware of its place of origin. So-called "information products" have become the island's most important exports. According to the 1997 Yearbook of the Taiwan Semiconductor Industry , published by the Industrial Technology Research Institute (ITRI), in 1996 the production value of PCs and related hardware totaled US$16.4 billion--$24.2 billion if the output of Taiwan-owned overseas plants is included. Monitors, notebooks, personal computers, and motherboards accounted for 80 percent of the total.
Many of Taiwan's PC products now rank first worldwide in terms of global sales. In 1996, the island's scanners enjoyed a mammoth 95 percent of the total world market, with motherboards taking 74 percent, computer mice 65 percent, keyboards 61 percent, and monitors 53 percent. "No wonder Taiwan is often called 'the PC kingdom,'" says Kuo Yun(果芸), vice chairman and president of the Institute for Information Industry (III). "Taiwan is the world's PC production center, producing about one-third to one-half of all PCs." Overall, the island's information industry has grown to the point where it now trails only those of the United States and Japan.
What is the secret of this astonishing success? There are several factors at work, but chief among them is the structure of the grassroots manufacturing industry that actually produces these high-value exports. Computer products are notorious for having a short life span. But Taiwan's information industry is largely made up of small and medium-size enterprises (SMEs), which are very flexible and can react swiftly to change. Unlike the large and sometimes lumbering manufacturing giants of Japan and Korea, Taiwan's companies are quick on their feet and famous for offering buyers a wide range of options on product specifications.
According to Chang Li (張笠), deputy secretary-general of the Taipei Computer Association, other big plus points are that the island's SMEs are willing to accept pretty much whatever orders come in, irrespective of size; and their CEOs are accessible to customers in a way that their Japanese or Korean counterparts are not. And Taiwan can offer streamlined, all-in-one service because, with the exception of Central Processing Units (CPUs), all peripheral parts are manufactured on the island. "This kind of flexibility and structure is just right for the information industry," says Hwang Jung-chiou (黃重球), director-general of the Department of Industrial Technology (DOIT) at the Ministry of Economic Affairs (MOEA). "Another advantage is that the industry has accumulated thirty to forty years of experience in quality manufacture. Taiwan's competitive advantage was and is in manufacturing and production technology. In manufacturing, there's still a lot of know-how involved."
Hwang's words highlight the importance of the human factor. Staff generally have a reputation for being well-trained, hardworking, and keen to upgrade their skills. Higher up the scale, there is the ongoing influx of talent from California's Silicon Valley--overseas Chinese who return to their roots to start up new companies in the Hsinchu Science-based Indus trial Park with the help of tax breaks and similar incentives.
Other countries can copy certain elements of Taiwan's high-tech success blueprint, but the human factor cannot easily be replicated. Malaysia, for example, produces only about 1,000 university graduates a year, whereas Taiwan turns out 7,000 graduates in the fields of electronics and electrical engineering alone. Hong Kong and Singapore boast great infrastructures and liberalized business environments, particularly where telecommunications and financial systems are concerned, but they lack Taiwan's manufacturing base and R&D facilities.
Chang Li sums up the success of the local information industry in a popular Chinese expression: "Personal harmony, good location, favorable times." He points out that when the Nationalist government moved to Taiwan in 1949, it brought over many people with management and engineering skills and thereafter paid serious attention to the information industry, developing industrial parks and promoting technical education. He also notes the importance of the Japanese connection. Japan ruled Taiwan between 1895 and 1945, establishing strong ties between the two countries at many levels, so it was natural that when Japan's resurgent industrialists were looking for overseas production bases or joint-venture partners, they should turn to nearby Taiwan. The net result was that in the 1970's, when the PC market really began to take off, Taiwan was in a strong position to ride the new wave, since personnel, manufacturing facilities, and technology were all in place.
So far, so wonderful. Unfortunately, however, there is an important missing link, and its name is software. Unlike the squat metal boxes and flickering monitors that sit on every executive's desk these days, software programs are culture-specific. They are also a dime a dozen in an already overcrowded international market. Consequently, such software as is developed on the island must find a market locally, and the business generated from such operations is nowhere near big enough to justify the effort.
Kuo Yun of III mentions a specific example to illustrate the difficulties. The accounting software produced in Taiwan cannot be used in the United States. If a Taiwan company wanted to develop software that targeted the US accountancy market, it would first have to overcome the language difficulties and then come to terms with very different accounting standards and procedures. The cost would be astronomic.
It might be argued that the cost of developing software in America is astronomic also--but the potential market is correspondingly much bigger. There are estimated to be about forty-two computer users in the United States for every one in Taiwan. A successful US software inventor could earn a lifetime's profit from one outstanding, durable program, while in Taiwan the inventor's skills, translated into exclusively local sales, would only sustain the business for a couple of years at most. It is very hard for Taiwan to compete in the global software market.
Local analysts are also aware of an important psychological perception that hinders future development in this field. According to Lin Bao-shuh (林寶樹), deputy general director of ITRI's Computer & Communication Research Laboratories, the public lacks confidence in the software business. People think software can be pirated too easily, thus depriving inventors of their profits. In other words, there is not nearly enough emphasis on the protection of intellectual property rights and innova tive attitudes.
"In Taiwan, the investment environment means people have no faith in intellectual property, and our stock market regulations discourage creativity," Lin says. "People are willing to invest only in factories and hardware facilities." He argues that US-based Yahoo! does not make money from its website services directly, but the company is nevertheless worth billions of dollars. Why? Because people see the founders' potential and value the stock accordingly. Such a situation would be impossible in Taiwan.
It is a vicious circle: the climate militates against software companies, so there are no big software success stories, so investors have no incentive to change tack and invest in software production. Few local software companies make money, with Trend Micro Device, Dynalab Taiwan Inc., and Unisys being notable exceptions. But only Trend, which makes world-class anti-virus software, and Dynalab, with its ideograph-based software, have succeeded in penetrating the Japanese market, and neither of these companies is listed on the Taiwan stock exchange. Taiwan's annual hardware production value already exceeds NT$100 billion (US$3.3 billion). In software, it would be a real achievement to hike it over the NT$1 billion (US$33.3 million) mark.
"Our software skills are relatively weak," Lin says. "We have good hardware engineers, but for software, which requires creativity and innovation, we lack the people." Why? Chang San-cheng (張善政), director of the National Center for High-perform ance Computing (NCHC) at the Cabinet-level National Science Council (NSC) believes he knows the answer. "The United States dominates about 80 percent of the global software market," he notes. "Japanese hardware manufacturing is so strong that it can easily knock out the US competition, but Japan isn't strong in software. Taiwan, like Japan, is strong in hardware. Both nations' educational systems are similar. Therefore, it's the rote-learning educational systems that stifle creativity. We can't blame the information industry; it's a long-term educational problem."
Chang also points out that information industry professionals are weak in knowledge of other fields, lacking system integration skills. Taiwan's most successful software product is probably Trend's anti-virus program, but that is primarily of use only to computer professionals. And Chang identifies another impediment to software development: The government procurement process, which in other countries often serves to stimulate development of new products, in Taiwan is bedeviled by bureaucracy and red tape. "Their obsession with preventing corruption has stifled development," he says. He cites a case where a software package initially offered to the government for NT$4 million was actually sold for NT$2.8 million (US$133,333/$93,333), because the vendor wanted the contract so badly. According to Chang, quality suffered and problems were legion. "The government doesn't respect professionalism," he says. "All it cares about is the bottom line."
Private sector criticism of government procurement systems is even more vocal. Chang Li of the Taipei Computer Association says that the government invariably purchases software from the lowest bidder. "But you can't compare software products just in terms of price," he says. "What's the use of buying the cheapest software if it doesn't do what you want it to?" He thinks the government should first choose the software that most closely meets its requirements and then negotiate the price.
NCR Systems Taiwan, a long-established foreign firm that was formerly part of AT&T, supplies software to a number of government departments and agencies, including the Department of Health, the Ministry of Foreign Affairs, and the National Library. The company's managing director, Cliff Wu (吳輔世), has huge problems with government purchasing regulations, which according to him have not been revised since the forties or fifties. "It's a total waste of public money if the government buys goods at a knockdown price and then can't use them," he says. "Although the rules are designed to prevent corruption in the civil service, these constraints in fact just cause civil servants to be overcautious and stifle their creativity." Wu believes the government should delegate wider purchasing powers to high-ranking officials. Since information products change so rapidly, he also suggests that the government would do well to outsource in the high-tech field, thus saving staff costs and other resources.
Hwang Jung-chiou of the DOIT argues that the government's purchasing system was designed to regulate the acquisition of mature products that were "measurable" or otherwise had easy-to-state specifications. Since software is hard to evaluate, and high-tech products are frequently marketed before they have matured, "the present bidding system is bound to be ineffective," he says. His preferred solution involves customers using software and other high-tech products for a trial period before proceeding to final acquisition.
Other problems faced by the information industry are, in Chang Li's words, "far beyond any solutions the industry itself could come up with." He is referring to across-the-board issues such as deregulation, the living environment, and internation alization. In his view, one reason why the island's PC industry has been so successful is that the government did not overprotect or overregulate it. The same is true of Taiwan's success as a transshipment center for computer accessories--an area where it is actually stronger than Singapore and Hong Kong, because the local import tax on such components is just 1 percent.
Internationalization and the living environment are both hot topics at present, largely because of the government's determination to push through the APROC plan, under which Taiwan is slated to become a regional operations center in six key industrial and commercial sectors. But the important thing is to persuade large companies to establish their regional headquarters in Taiwan, not just support services or subsidiaries; and that to a large extent depends on how easy it is to import key personnel from overseas. Silicon Valley is successful because it acts like a magnet, luring persons of outstanding talent from all over the world. If Taiwan can become a similar magnet, the outlook is rosy. So--what are the prospects?
Cliff Wu of NCR Systems Taiwan considers many of the island's laws outdated and inappropriate. One example: Six years ago, the regulations were relaxed to make it easier to set up a bank in Taiwan. Sixteen overseas banks came to the island but, because they could not set up an entire branch network overnight, at first they concentrated on installing automatic teller machines (ATMs) in as many locations as possible. Until last year, however, each bank was limited to installing thirty new ATMs annually. Wu finds that absurd. The regulation was obviously intended to prevent a proliferation of unwanted ATMs, but he points out that if a bank finds that people are not using a particular ATM it will remove the machine anyway.
Even if the legislative environment can be speedily updated to make Taiwan more user-friendly to would-be investors, there still remains the worrisome problem of R&D. While Taiwan's SMEs provide the information industry with flexibility and a viable structure, the other side of the coin is that they lack the resources to participate in effective R&D and thus develop new products. The DOIT's Hwang Jung-chiou says the information industry's average R&D budget is less than 2 percent of operating capital, which is high by reference to other manufacturing industries, but low compared to international standards.
This has particular reference to software development, where R&D costs are often stratospheric but can later be matched by outstanding profits. Chang San-cheng of the National Center for High-performance Computing points to the fact that Microsoft and other major software companies consistently rank above IBM and Apple in terms of assets. The future of the industry lies with software, in other words, and Taiwan is currently still tied to hardware manufacture.
Many observers now agree that the solution lies in concentrating on the mainland and overseas Chinese markets for both hard-and software. Kuo Yun of the III notes that in 1996 the Taiwan domestic market absorbed 600,000 PCs, while mainland China took 2.8 million. There are no significant cultural or language problems to be overcome, and the mainland's economy shows all the signs of heading for a major boom. Chang Li estimates that the mainland has some 6 million commercial enterprises, many of which will want to be computerized at some point. That adds up to a lot of potential hardware and software sales. Even if Taiwan companies cannot invent the software, there is no reason why they should not act as agents and distributors for it.
Lin Bao-shuh of ITRI says that Taiwan still has a chance as long as it is prepared to upgrade its investment environment and target the greater Chinese market. He points to what happened at Microsoft. When he was living in Seattle in 1987, Microsoft was just a 3,000-employee company catering to the US domestic market. Then Bill Gates hired a Japanese executive who pushed the company's international market share from zero to its present 51 percent. Microsoft now has 30,000 employees. Lin says Microsoft is strong today because of its personnel and software; and competition in software is keener than in the hardware arena. "In the software business, you can only survive if you're first or second, not third," he says. Microsoft netted US$3.5 billion in 1996, while the profits of the software companies that ranked from second to tenth in the league table totaled less than US$0.5 billion.
What is the government doing to help the industry? Some measures are already in place and more are planned. (See box opposite.) Certainly, no one underestimates the importance of keeping up the pressure for excellence. "The only hope for the survival of Taiwan's future development is maintenance of high-tech standards, and that means moving toward becoming a fully 'sci-tech' island," says Kuo Yun.
The Cabinet is in full agreement with that, as witness its approval earlier this year of the NSC's first "sci-tech" white paper, which provides a blueprint for future development. The most ambitious stated goal calls for nothing less than increas ing the production value of technology-intensive goods to US$300 billion by the year 2000, which would have the effect of making Taiwan a developed industrial nation. There is also an islandwide research fund: initially it was 1.81 percent of 1995 GDP, and that will increase to 2.5 percent of the then GDP by 2000 and 3 percent by 2010. The plan calls for dramatic increases in the numbers of research personnel; the creation of more science-based industrial parks, each housing 200 high -tech companies; and boosting the number of Internet users on the island to 3 million by 2000.
For his part, DOIT Director-General Hwang Jung-chiou foresees a future where both the government and the private sector pay much more attention to practical technological applications. He recognizes, however, that such a transition will necessarily involve dramatic changes in mindsets, systems, and regulations. His summary of the situation in which the island finds itself is very simple, but it underlines the enormity of the task facing Taiwan's technological movers and shakers: "I believe that Taiwan needs to invest, but it must get away from the obsession with hardware," he says. "In the future, Taiwan needs to invest instead in people, creativity, innovation, and intellectual property."