Technology Transfer from Taiwan to China: Is there a Risk?

 Technology Transfer from Taiwan to China: Is there a Risk?

by John J. Tkacik, Jr.

For two decades, American foreign policy operated on the premise that trade with China would have an inevitable liberalizing effect. This persisted after the Tiananmen crisis, and even after the 1996 Taiwan Strait missile crisis when Beijing attempted to intimidate Taiwanese voters from casting ballots for President Lee Teng-hui. It was at the foundation of Clinton's China policy and apparently undergirds the Bush policy as well.

But are there risks that liberalized trade with China will result in the leakage of defense-critical technologies? The "China Threat" is now a fact of life according to a 209-page annual Report to Congress of the U.S.-China Security Review Commission (USCC) . Similar concerns are evident in the Pentagon's smaller but still formidable 56-page Annual Report on the Military Power of the People's Republic of China. Both documents describe a vast and dynamic military establishment that is taking advantage of China's modernizing economy to modernize its warfighting capacity. In fact, the thesis of the USCC paper is that China's burgeoning defense budget, now viewed as the second-largest in the world is directly funded by the country's export-driven foreign direct investment (FDI).

Both Washington and Taipei are unsettled by the substantial leakage of advanced technology to China -- but for different reasons. Washington worries that advanced semiconductor fabrication technology will enhance China's military modernization. Taipei is even more worried that the wholesale migration of advanced industries to China will "hollow out" Taiwan's economy, leaving it a service industry shell with little manufacturing capacity. The loss of manufacturing has already taken a tremendous toll on Taiwan's labor market.

This paper will attempt to describe the concerns in the United States and in
Taiwan about technology leakage to Mainland and assess their validity. The preliminary conclusions are that fears in both the United States and Taiwan are well-founded, but that neither country has a coherent strategy to deal with the risks.

I: The impact of Foreign Direct Investment in China

In the year 2002, China surpassed the United States as the top destination for direct foreign investment attracting over $50 billion in FDI including more than $5 billion from U.S. firms. In the short span of a decade, China has turned itself into Asia's manufacturing center, with components and raw materials from all over the world converging on Chinese factories for assembly and other processing for re-export to the international market. Meanwhile, low growth and the war on terror cut global inflows of FDI to $735 billion in 2001, less than half the amount in 2000, with FDI to economies in the Asia and Pacific dropping 24% from 2000 to 2001. Yet in 2001 and again in 2002, China emerged as the top Asian destination for all foreign direct investment. In a zero-sum game of global capital flows, China is sucking the oxygen out of the markets in Asia, and elsewhere for that matter. One worrisome result, in the words of one Hong Kong analyst, is that competition from China is reducing Southeast Asia's economic growth rate.

Clearly, the gross numbers show that China's economic growth is driven almost entirely by international trade. Between 1992 and 2001, China's overall trade-to-GDP ratio went from 18.9% in 1980 to 34% in 1990 and to 49.3% in 2000. Increasingly, China's manufacturing base is moving from simple labor-intensive assembly up the ladder to capital-intensive fabrication of basic components. Yearly inflows of realized foreign direct investment rose from a microscopic $60 million in 1980 to $3.49 billion in 1990 and to $42.10 billion in 2000. In 1985, foreign-invested enterprises (FIEs) produced about one percent of China’s exports. In 1990, 12.5% of China's exports came from foreign-invested production lines, but by 2000 nearly half of China's exports were from FIE's. In 1995, China began to drop its average tariffs, inducing FIEs to import increased amounts of capital goods for the establishment of new production lines, import increased amounts of components for the new production lines. In a single five-year period between 1995 and 1999, export-processing trade went from 49.5% of China's total exports to 56.9%.

To supply this vast export sector, China imports vast amounts of production equipment, components, parts and raw materials. But little of these imports reach Chinese consumers. According to the USCC report, 20% of China’s total imports reach China’s domestic markets while the other 80% consists of capital goods and industrial inputs used for the country’s export sector. Much of the capital expended in FDI projects in China, therefore, has gone into financing imported capital equipment. So it should not be surprising that China's total import statistics grow as FDI increases. China reported a record $46.8 billion in FDI actually absorbed in 2001, over $50 billion in 2002, with much of the new FDI focused on high-tech sectors.

Taiwan's Advanced Technology Investment Stake in China

By the end of 2001, Taiwan firms had invested an estimated cumulative US$70 billion in China. It may be even more. Taiwan's Central Bank of China believes over US$100 billion in Taiwan-controlled assets are in China. Chinese figures are a bit more modest, showing a total of US$29.134 billion. Chinese government figures reflect that 2001 was the biggest year for Taiwan investments since 1995, with US$2.979 billion in Taiwan direct investment. With this increasing flow of Taiwan FDI to China, there has been a major change in Taiwan's investment patterns. No longer is the bulk of Taiwan money coming from small and mid-sized manufacturing operations. Instead, it is coming from Taiwan's largest firms. In the first seven months of 2002, for example, Taiwan firms invested US$1.94 billion in China, 51.9% of that was new investment in the electronics sector alone. By the end of 2002, the PRC Ministry of Foreign Economics and Trade reported that Taiwan was the second largest foreign investor in China. In 2002, Taiwan invested US$3.85 billion in Chinese ventures, up 38.6% from 2001, accounting for 53.3% of all Taiwan outbound investments. In January 2003, alone, government-approved Taiwan investments in China hit a new monthly high of US$455.29 million, 220.78% over the January 2002 figure.

The United States is a somewhat smaller investor in China than is Taiwan, but of higher quality. By 2000, American firms had invested a total of US$9.58 billion in China, with over a third in high-tech manufacturing. In the year 2000 alone, U.S. direct investment outflow to China and Hong Kong hit a record high of US$4.4 billion. U.S. investment in China alone rose to US$4.86 billion in 2001. With the US$5 billion in 2002, cumulative US investment in China by the end of 2002 must have been near US$18 billion.

The significance of these trade patterns is that China's advanced-technology (AT) infrastructure is being built largely with foreign direct investment. Although foreign-invested enterprise (FIE) production lines heretofore had been designed to assemble imported advanced technology components into electronic and information technology (IT) hardware and appliances for subsequent export, foreign suppliers are now setting up the actual component production lines in China, supposedly to be closer to their assembly lines and the assembly plants of other AT product customers.

Semiconductor technology is a prime example of this phenomenon. Already, Motorola manufactures mobile-phone specific integrated circuits (ICs) -- particularly global positioning ICs -- at its MOS-17 waferfab in Tianjin. The MOS-17 facility includes full-up R&D, design and manufacturing centers, and employed 2,400 workers at startup in 2002. By April 2003, Motorola had 13,000 employees in China, including 1,500 researchers at the company's 18 labs. With the MOS-17 fab, Motorola's total investment in China was about $3.4 billion in 2002, making the largest foreign investor in China. NEC Corp. of Japan has begun to produce 128-Mbit DRAMs at 0.25-micron linewidths (the industry standard) at its NEC Shanghai Huahong fab.

The Motorola and NEC investments have prompted Taiwan's top waferfabs, Taiwan Semiconductor (TSMC) and United Microelectronics (UMC), to petition the Taipei government to permit them to transfer existing waferfabs to China. While Taipei has agreed in principle to the TSMC and UMC requests, it has conditioned its approval on the assurances that the Taiwan waferfabs will be replaced in Taiwan by more advanced facilities. As yet, the Taipei government has not issued formal rules governing the transfer of semiconductor facilities to China.

Unauthorized Taiwan Semiconductor Investments in China

Foreign chip manufacturers, including those in Taiwan, have long argued they must set up production lines in China "to be near their customers." Taiwan's companies see the investment of several hundred million dollars in waferfabs as part of the cost of doing business in China's low-wage environment.

This is an unsettling phenomenon because the whole idea of a highly capital-intensive (i.e. in the billion-dollar range) "foundry" waferfab is that it does not need to be near the customer. Because their output is small, light and high-value, they are easily and cheaply transportable. In fact, most American semiconductor producers subcontract some or all of their integrated circuit (IC) designs to "foundry" fabs in Taiwan, and air-ship them back to the United States or onward to foreign end users. That is, there is little, if any, economic benefit to investing a billion dollars in a modern foundry waferfab in China when the semiconductors can be available to any end-user production line in China within a matter days for a fraction of the cost.

Indeed, one American writer argues that Taiwan ought not permit semiconductor investments in China because it is in Taiwan's interest to control the supply of integrated-circuit commodities to industrial consumers in China. China will not used economic or military force against Taiwan because it would interrupt the "supply chain" of this vital commodity to China's exporting sector.

While the People's Republic of China certainly welcomes and encourages broader foreign semiconductor investments in China like Motorola's or NEC's, Beijing's real goal is to bring advanced "foundry" waferfab, R&D, design, packaging and testing centers under its own control, and has devoted billions of dollars to acquiring foreign technology to do just this.

One way China hopes to gain advanced semiconductor technology without drawing the attention of American (or Taiwanese) export control officials is to set up "false foreign devil" (假洋鬼子) FIEs with nominal outside management, but controlled by Beijing. In 2000, Winston Wen-yang Wang (王文洋), son of Formosa Plastics Group (FPG) chairman Wang Yung-ching, joined Chinese president Jiang Zemin's son, Jiang Mianheng (江綿恆), to start up a joint-venture company called Grace SMC (GSMC宏力半導體) to build an 8-inch waferfab facility in Shanghai. According to "knowledgeable sources" in Shanghai, Winston Wang's GSMC received over two-thirds of Grace's operating capital from government-directed investments in China, including a US$1.1 billion dollar low-interest loan from Chinese banks, which comprise the bulk of the US$1.65 billion up-front investment for the first phase of the project. The Chinese government also granted the venture special software development tax status which dropped the tax rate to 3% from the original 17%.

Taiwan authorities became alarmed at the potential for unauthorized leakage of advanced technology encompassed in the Grace SMC deal. Taiwan's Ministry of Justice Investigation Bureau (MJIB) interviewed Wang concerning a possible diversion of capital from Taiwan to set up the Shanghai operation. Wang reportedly told the MJIB he agreed to become Grace SMC president due to his close personal relationship with Jiang Mianheng, and that his "major task was to raise capital overseas." Winston, however, does not appear to have been very successful. In the end, it appears that virtually all of Grace SMC's capital is coming from the Chinese government, and that Winston Wang is a figurehead who gives the firm a "foreign-invested" patina.

A Taiwan-invested waferfab firm that may indeed have some outside investment is Richard J. Chang's (張汝京) Semiconductor Manufacturing International Corp. (中芯, SMIC). Chang is the former president of one of Taiwan's smaller chipmakers Worldwide Semiconductor and a former engineer at the world's leading chip foundry, Taiwan Semiconductor (TSMC). He launched SMIC in Shanghai, reportedly with some legitimate private funding. The US$1.48 billion venture, however, seems to be a totally Chinese government-controlled operation -- Chang complained mightily to the media about the strictures Beijing placed on the company as it was raising venture capital. In October 2001, Chang told the Financial Times, "The authorities said how much money we could borrow, and from which Chinese banks - this is very new to us." Said the FT, "Chang has noticed another difference to doing business in China compared with Taiwan; he had had to employ 11 public relations officers to keep local officials informed, compared to just one in Taiwan." One wonders what these PR people do. SMIC's website carries very little useful information.

SMIC reportedly purchased five 257-nanometer (roughly 0.25 micron) lithography machines made by ASML of Netherlands, giving SMIC access to levels of technology for which the United States still refuses export licenses. U.S. guidelines still reportedly limit lithography equipment with capabilities under 0.35 microns, although Motorola has reportedly been granted a license to produce chips at the MOS-17 at 0.25 linewidths. SMIC is scheduled to advance to 0.13-micron production in 2004 and introduce 0.18-micron silicon germanium (SiGe) production technology soon.

Chang seems also to be the rain-maker for SMIC's second new eight-inch waferfab in Beijing, a project valued at US1.25 billion, for which Chinese government and government-owned businesses committed US$200 million seed money. The main investors are said to be Beijing's Capital Steel Corp. (首鋼) with US$50 million, the Beijing Municipal Government gave US$90 million and a group of other Beijing enterprises came up with US$60 million. SMIC and Richard Chang, however, are still trying to pull together another US$400 million to get the project off the ground. Reportedly a Chinese-American investor Chang Fu-hsing (張復興 ) is also involved. To put the project together, Richard Chang has hired Marco Mora, a noted Italian chip manager to oversee the fab's construction, hire personnel and organize the new operation.

But for all this money and effort, it is clear that the SMIC investment was not well thought-out, particularly in a world-wide economic slump when most chip customers purchase their chips only from reliable suppliers. Either that, or SMIC was not intended to compete in the international chip market in the first place. By September 2002, SMIC admitted it was headed for large losses. In order to keep their production lines running, both SMIC and Grace have resorted to turning out low-end DRAM chips. Yet by April 2003, one Shanghai-based semiconductor expert told reporters "I think they'll rack up incredible losses in DRAM . . . [SMIC is] building production lines, but they have no customers." SMIC's lack of customers persists despite a growing demand for chips by foreign firms in China. In March 2003, most international semiconductor companies remained puzzled by the nature of China's chip sector. Infineon's CEO Ulrich Schumacher, commented "China is this big phenomenon. Is it the biggest market of the future? Or is it the biggest threat? Nobody has a clue what China really is. What do you do now?"

A migration of Taiwan's "crown jewel" technologies to China via Taiwan entrepreneurial exiles has alarmed Taipei's government. In a top secret report entitled, "An Analysis on how the Chinese Communist Party Attracts Taiwanese High Tech Investment for the Suzhou Industrial Park," Taiwan's intelligence agency reported in July 2001, that the Chinese authorities have a blueprint to actively develop semiconductor and high-tech industry "clusters" which include the entire spectrum of each industry. The result, the report said, was that China has effectively attracted the key sectors of Taiwan's computer industry, from downstream component makers like computer motherboard and monitor producers to PC cases and mouse makers. The report suggested that the Taiwan-invested high-tech sector would be a virtual "puppet" of Beijing, and recommended that the Taiwan government adopt policies to curb high-tech investment in China. Indeed, the one high-tech area in China in which Taiwan's government still prohibits local investors from investing is semiconductor fabrication, but that ban, too, appears to be eroding.

China's theft of semiconductor intellectual property

Whether Taiwan's official bans on chip-fab technology transfers to China are maintained is only half the problem. Clearly, Chinese semiconductor firms will try to get the technology however they can. In March 2002, TSMC accused a former manager, Liu Yun-chien, of stealing information on its 12-inch wafer fabrication technology after discovering that Liu had fled to China and was employed by the Richard Chang's SMIC. Later, TSMC discovered that, without authorization, its entire documentation for its 8-inch waferfab production line had been put on-line with Taiwan's National Chiao-tung University library, and another employee had downloaded and copied it. China wants the technology and China is going to get it.

Nor is any country's intellectual property in advanced technology safe in Chinese hands.
In March 2003, Infineon CEO Ulrich Schumacher admitted that intellectual property was in peril in China. His answer was to stick with DRAM technology, and "do things on paper to protect your technology, but ultimately, you can't. You have to build in the possibility that things will happen."

For years, U.S.-based Cisco Systems had suspected that China's Huawei Technologies had illegally copied Cisco firmware and software for Huawei's routers. To throw the Chinese off the scent, Cisco even told an American scholar in January 2001 that "Cisco's covert disassembly of Huawei routers yielded no Cisco intellectual property or evidence of reverse engineering, suggesting that Huawei was now producing top-quality, indigenously-produced products."

But this was false. Cisco had indeed discovered that Huawei had pirated vast swathes of Cisco software, with "hundreds of smoking guns" among over 15 million lines of code which included exact matches between Cisco software and that of the Chinese machine. Even Cisco's software "bugs," glitches, and misspellings matched. In January, 2003, Cisco sued Huawei in U.S. federal court seeking monetary damages and a bar against Huawei selling in the U.S. Huawei quickly withdrew its routers from the U.S., and admitted in court filings that some of its software "appears to match" Cisco's. The copying, said Huawei, was unintentional and involved only about 2% of its software. Cisco, however, accuses Huawei of "pervasive and systematic" efforts to "steal and use Cisco's technology."

When Huawei couldn't decode Cisco's sourcecodes, it attempted to hire Cisco programmers. Between April 2001 to December 2002, Chinese directors at Huawei twice tasked the company's U.S. personnel chief to recruit Cisco employees to complete projects beyond the capacity of Huawei engineers. One Cisco employee said Huawei asked him for detailed information on Cisco's fastest router chip, and offered to pay his way to China, with a handsome stipend, to present his report. “I have no doubt that this was a direct solicitation of Cisco's confidential and proprietary information,” he told the court.

Obviously, the Cisco case is only the tip of the iceberg. Other U.S. companies acknowledge that their Chinese joint venture partners, and the Chinese government siphons off proprietary technology, trade secrets and intellectual property, but they are reluctant to pursue the issue for fear that it will impact their business standing in China. Apparently, the Chinese government's preferred way of gaining access to U.S. advanced technology is via human sources, particularly students and technicians from US-Chinese joint ventures in China who want to work in advanced technology laboratories and production facilities in the United States.

According to the U.S. Congress's General Accounting Office, one measure of this risk is reflected in the number of "deemed export licenses" for temporary workers in the United States. In 2001, the most common "country/technology" combination for "deemed export licenses" -- that is, export licenses authorizing foreign nationals (including nationals from Canada, the UK, and other US allies) to work in America's high-tech sector -- was "China/electronics technologies". About 44% of all "deemed" licenses approved in fiscal year 2001 authorized citizens of China to work with electronics technologies, including semiconductor technology. Despite the obvious risk of leakage of critical defense technology via Chinese technicians in domestic U.S. high-tech labs and production lines, the GAO said, the Commerce Department has never denied "deemed export licenses" to any Chinese applicants.

II. The Defense Risks of a modernizing China and Ways to Mitigate Them

Is there, in fact, a clear and present risk that advanced, critical defense technologies are leaking to China? A number of studies indicate there is.

Semiconductor technology is a key component of China's defense modernization, because it is an essential building-block for defense cybernetics, telecommunications, automation and advanced materials fabrication. Although the Chinese government has had a comprehensive science and technology R&D strategy in place since March 1986 (the so-called "863" and "Torch" programs), the Ninth Five-Year Plan launched in 1995 clearly identified areas where China's leadership believed its security and strategic interests would best be served. They included:

-aerospace
-lasers
-advanced materials
-microelectronics
-biotechnology
-information technologies
-industrial automation

At the time, the US Department of Commerce declared enthusiastically that "these technologies target areas of U.S. strength. China’s emphasis on technology transfer will result in many initiatives and incentives for acquiring these technologies" and encouraged U.S. companies to vie for contracts in these areas.

That enthusiasm was missing from a General Accounting Office conclusion in April 2002 that

Since 1986, China’s efforts to improve its semiconductor manufacturing capability have narrowed the gap between U.S. and Chinese semiconductor manufacturing technology from between 7 to 10 years to 2 years or less. According to our analysis of information obtained from semiconductor manufacturing facilities in China and industry experts, China’s most advanced commercial manufacturing facilities can produce chips that are only one generation behind current, commercial state-of-the-art technology. China has made improving its semiconductor manufacturing capability a priority for national and economic security reasons and plans to build as many as 20 multibillion-dollar manufacturing facilities over the next 5 to 10 years with substantial levels of foreign investment. The growing sophistication of China’s semiconductor manufacturing facilities, which has improved its ability to develop more capable weapons systems and advanced consumer electronics, has been fueled by China’s success in acquiring manufacturing technology from abroad.

By 2002, it was clear that China's industrial modernization was focusing on "pockets of excellence" where select technologies would benefit China's military industries. China's Commission for Science and Technology Industry for the National Defense (COSTIND) reportedly identified several such "pockets" which include preemptive long-range precision strike capabilities; information dominance; command and control; and integrated air defense. As a result, Beijing identified the development of an indigenous microelectronics industry as one of its highest priorities. An advanced domestic microelectronics sector, the Chinese leadership believes, will support both military modernization as well as commercial demand, and in recent months, China's defense emphasis has been on development of very large-scale integrated circuits will have direct application in future military systems, for example, advanced phased-array radars. For such systems to be able to track 200 targets 100 times per second at resolutions of one meter and accurately compute trajectories, predict g-force pulls and predict probable flight paths, hyper-high speed semiconductors are required.

China may not yet have the design or lithography capabilities, but they will as soon as they acquire the same industrial equipment as is available at the industry standard in the United States: 0.25 micron linewidths, with international standard design software. The advances in these technologies require micro-circuitry of increasingly finer linewidth lithographies. Up to 2000, China's domestic semiconductor fabs could etch integrated circuits down to about 0.6 microns, and their clock speeds and parallel processing capacities were limited accordingly. Moreover, China's domestic IC design talent and capabilities were crude.

The movement of 0.25 micron lithography into foreign-controlled China-located waferfabs, like the MOS-17 and the NEC-Huaneng facilities did not directly advance China's defense modernization because they reproduced IC-designs from abroad and were under the control and supervision of foreign technicians. But 0.25 micron linewidths remain -- as of August 2002 -- the industry standard, and the Grace SMC and SMIC fabs which will soon go into mass production at 0.25 microns are under the complete control of Chinese entities without foreign monitoring or supervision.

In August 2002, SMIC said it was able to acquire 0.18-micron rule equipment for dynamic random access memory (DRAM) chip production from European suppliers after the U.S. banned exports of the advanced chip-making equipment to the People's Republic. The U.S. government reportedly had denied an export license for advanced chip-making equipment to SMIC on grounds that it would serve China's military industries. By December, 2002, SMIC announced that it had developed 0.13 micron rules "with technological assistance from the Texas Instruments (TI), using "chip-making equipment provided by ASML of the Netherlands." The company also said it had launched a production line for 0.18-micron rule chips.

In March 2003, SMIC CEO Richard Chang claimed his company would retool its eight-inch waferfab into its first 12-inch silicon fab by the end of 2003. A Taipei press report said Infineon CEO Ulrich Schumacher "pledged" to license Infineon's 0.14-micron process to SMIC and may license its 0.11-micron process to SMIC during 2003.

In other words, the horses may already have bolted from the barn. Cutting-edge foundry semiconductor fabrication infrastructure may already be available to People's Liberation Army defense laboratories.

Do Chip Fabs directly aid China's Military?

Some have pointed out that these new fabs are of marginal utility to China's defense modernization without a robust IC design capability. But indeed Motorola and NEC have already set up design laboratories in China, and Taiwan's Via Technologies has the largest design house in the country. Several U.S. based chipmakers, including Motorola have also launched sophisticated software labs in China targeted on advanced telecommunications transmission codes, global positioning software as well as IC design software. However, whether foreign-owned IC design shops in China can actually maintain control of their own intellectual property is problematic. Either it will be pirated with impunity because it is not patented in China, or patents will be filed with the Chinese government where the military will have access to the property and utilize it for its own purposes -- or it will just be stolen outright.

What is not in question is that China's military modernization will be built largely on its rapidly expanding semiconductor industry, and China's semiconductor industry will get its technology almost exclusively from abroad -- particularly from the United States, Japan and Taiwan, as well as the European Union.

Can Taiwan export controls enhance its security?

There is an ongoing debate surrounding Taiwan's domestic legislation related to permitting Taiwan's semiconductor industry to transfer wholesale its production, testing, packaging, design and marketing operations to China. On the one hand, Taiwan politicians leery of China's motives and unwilling to allow Taiwan businesses to partake actively in China's military modernization are adamant that Taiwan chip-fabs be banned from giving China access to any technology that it doesn't already have.

On the other hand, Taiwan's semiconductor firms argue that in order to upgrade Taiwan's industrial base, they must find productive, profitable uses for their existing fabs. Upgrading to multi-billion dollar fabs utilizing 12-inch wafer production (300mm) discs which can hold two and a half times as many chips as existing standard 8-inch (200mm) fabs, requires that construction of the new facilities be financed by profits from the old lines.

IC entrepreneurs in Taiwan also argue that China's light-industrial sector is the world's fastest-growing market for chips, and that foreign fabs are moving into the China market already. The pressure on Taiwan's government to approve the migration of these key semiconductor technologies, fabrication equipment, design talent and capital is enormous. In March 2002, Taiwan's Premier Yu Shyi-kun agreed in principle for local semiconductor companies to build 8-inch wafer plants in China that use 0.25 micron and less-advanced manufacturing technology. But the government would allow only three eight-inch plants to migrate to China by 2005, and local chipmakers will be allowed to export only old equipment using 0.25 micron technology or wider (i.e. older technologies). Moreover, total investment by Taiwan's chip industry will not be permitted to exceed NT$70 billion (roughly US$2.1 billion) by 2005, with new capital investments accounting for NT$30 billion. Taiwan's military and intelligence analysts fear Taiwan will lose its competitive edge and have urged that no technology finer than 0.25 micron goes to China. The political debate in Taiwan has therefore shifted away from the previous policy of Taiwan President Lee Teng-hui of "Don't Hurry, Be Patient" which banned all semiconductor investments by Taiwan companies in China, to the more lenient policy of the Chen Shui-bian government of "Active Opening, Effective Management". The new policy was intended, no doubt, to stress the "effective management" of key technology exports to China, not the "opening" of China as an investment destination. But the vague foundations of the new policy prevent articulation of a clear strategy of why Taiwan wants to limit such exports. Indeed, Taiwan has two good reasons to limit these exports; 1) they are militarily sensitive and 2) they tend to "hollow-out" Taiwan's only industrial sector where it retains a significant international presence.

Pressures on the United States government are also significant. Despite an unofficial "policy" that "certain exports of semiconductor manufacturing equipment to China are limited to two generations behind state-of-the-art levels to address national security concerns" and a determination by the Department of Defense that technology more advanced than 0.7 micron (i.e. one-twentieth the thru-put of 0.13 micron lines) is "military critical", there is no consistent ban on US exports of semiconductor technology to China. U.S. firms argue that they face stiff foreign competition, and apparently the U.S. Department of Commerce agrees.

The facts of China's rapid military build-up, and the fact that the build-up is focused on the two-fold objective of investing Taiwan and deterring American aid to Taiwan are explicated in well-documented detail by the Pentagon's "Report on the Military Power of the People's Republic of China" of July 12, 2002. Further evidence of China's military intentions is exhaustively presented in the USCC report. Together, the two reports are more than sufficient grounds for the Bush Administration to order a review and tightening of the slap-dash and ad hoc export control policies that now pertain to dual-use exports to China. Indeed, this much was strongly recommended by the General Accounting Office report cited above.

If the United States takes the lead in reviewing the "China Threat", Taipei's government will follow suit -- and would likely breathe a sigh of relief if given the opportunity to base their investment ban on a request from Washington instead of on its own assessment of the national security risks -- which would be painted by Taiwan's opposition parties as politically motivated.

Notes

1. For an extended discussion of this topic, please see "Strategic Risks for East Asia in Economic Integration with China" by John J. Tkacik, Jr. Heritage Foundation WebMemo #171, November 12, 2002. Available at http://www.heritage.org/Research/AsiaandthePacific/WM171.cfm
2. For example, during his State of the Union Message on January 29, 2002, President Bush said, "America is working with Russia and China and India, in ways we have never before, to achieve peace and prosperity. In every region, free markets and free trade and free societies are proving their power to lift lives. Together with friends and allies from Europe to Asia, and Africa to Latin America, we will demonstrate that the forces of terror cannot stop the momentum of freedom." See http://www.whitehouse.gov/news/releases/2002/01/20020129-11.html.
3.  Report to Congress of the U.S.-China Security Review Commission-The National Security Implications of the Economic Relationship Between the United States and China, released July 15, 2002, and available at http://www.uscc.gov/anrp02.htm. Cited hereafter as USCC Report. It should be noted that the USCC went out of its way to avoid using the word "threat" when referring to China's military posture toward the United States; however, the initial "key finding" of Chapter 9 (p. 167) is that "China's defense spending...is funding a strategic buildup that is aimed at U.S. interests in the area."

4.Report to Congress Pursuant to the FY 2000 National Defense Authorization Act, Annual Report on the Military Power of the People's Republic of China, July 12, 2002, available at
http://www.defenselink.mil/news/Jul2002/d20020712china.pdf.
5. See especially USCC Report, Chapter 9.
6. See EXPORT CONTROLS Rapid Advances in China’s Semiconductor Industry Underscore Need for Fundamental U.S. Policy Review, a United States General Accounting Office Report to the Ranking Minority Member, Committee on Governmental Affairs, U.S. Senate, dated April 2002, and available at <http://www.gao.gov/new.items/d02620.pdf.
7. This figure was supplied by the U.S. Department of State. See the remarks at the World Affairs Council in Washington, DC of James A. Kelly, Assistant Secretary of State for East Asian and Pacific Affairs, on "U.S. Policy on China and North Korea", January 30, 2003, at http://www.state.gov/p/eap/rls/rm/2003/17164.htm
8. Hugo Restall, "Asia's Giant Sucking Sound", Asian Wall Street Journal, June 20, 2002, p. 14.
9. See an analysis by Arvind Panagariya "Why we lag behind China", Economic Times published by the University of Maryland, May 22, 2002 at <http://www.bsos.umd.edu/econ/panagariya/apecon/ET/et40-May22-02.htm>
10. U.S.-China Security Review Commission, Technical Briefing on Business, Trade and Economic Issues, Oral Testimony of Nicholas Lardy, 9 May 2001, 177. Cited at USCC p. 42
11. Yuh-jiun Lin 林昱君 "Reflections on the eve of the Great Deluge of Mainland imports" [大陸進口品大舉入侵 前夕的思考], in China Affairs (中國事務), Taipei, April 2002, p. 128. ．
12. U.S.-China Security Review Commission, U.S.-China Current Trade and Investment Policies and their Impact on the U.S. Economy, Oral Testimony of Kevin Kearns, 14 June 2001, 125. Cited in USCC report p. 42
Ing-wen Tsai, Ph.D., "A New Era in Cross-Strait Relations? Taiwan and China in the WTO," Heritage Foundation Lectures, No. 726, 14 January 2002 (lecture delivered 13 December 2001).
See Chen Junyi and Ye Lingling (陳駿逸、葉玲伶),"Central Bank says that Taiwan investment in mainland estimated to top US$100 billion" (央行：台商投資大陸 估逾千億美元), China Times Taipei (中國時報) November 10, 2000. 　On May 31, 2002, the Taipei Times cited a PRC report which claims Taiwan businesses have US$140 billion invested in China. See "Taiwan Businesses have US$140 billion invested in China" at http://www.taipeitimes.com/news/2002/05/31/story/0000138441.
See Wang Zhuozhong 王綽中, "PRC Statistics show Taiwan investments in China are at highest level since 1995" (中共統計1995以來 台商投資大陸增幅創新高) China Times (Taipei) April 5, 2002　. Also see "Taiwan's total mainland bound investments amount to USD29.56 Bn." from China Economic News Service (CENS) Taipei, April 16, 2002
"Taiwan's indirect investment in mainland China up 47% in April" from China Economic News Service (CENS) Taipei, May 21, 2002
(No author cited) "大陸外資排名 台灣躍居第二" (Taiwan ranks second among outside investors in the Mainland); World Journal internet edition, January 29, 2003, at http://www.worldjournal.com/publish/030129/11_0900.4w/m/4wms(030130)01_tb.htm
(No Author Cited), "Taiwan's investment in mainland China hit new high in 2002"; China Economic News Service (CENS), January 24, 2003
(No author cited), "Approved foreign investments in Taiwan down 41 pct in Jan."; China Economic News Service (CENS), February 21, 2003
USCC report Chapter 2, figure 2.5
See figures from the US-China Business Council at http://www.uschina.org/statistics/03-01.html.
See ASIAN TECHNOLOGY INFORMATION PROGRAM (ATIP) REPORT: ATIP97.04. It states that "Chinese GPS companies are manufacturing products for advanced navigation and positioning systems that are designed, developed, and integrated wholly in China using a wide range of GPS receivers from Rockwell, Motorola, GarMin, Trimble, etc." Report available at <http://www.atip.or.jp/services.html>
Anthony Cataldo "Motorola cleared to build 0.25-micron fab in China"; EE Times, August 21, 2000, web-posted at 12:19:57 PM EDT. See also the Motorola careers website for China at http://www.motorolacareers.com/moto.cfm?cntry=China. See also http://apspg.motorola.com/mos17.html.
Ibid
Craig Addison, Silicon Shield: Taiwan's Protection Against Attack; Fusion Press, Irvine Texas, 2001, pp 147-179.
Bai Dehua, 白德華 "宏力設廠 中方低利貸款至少11億美元" (Groundwork for Grace Factory Begun, Chinese side to provide at least US$1.1 billion in low interest loan); Taipei Commercial Times (工商時報) November 21, 2000
Taiwan Weekly Business Bulletin for November 16 to November 23, 2000. The Taiwan Weekly Business Bulletin is a publication of the USA-ROC (Taiwan) Business Council.
The General Accounting Office appears to have been told that both Grace SMC and SMIC have no Chinese investment (see GAO Semiconductor report page 12), but this is clearly false. Indeed, there is no evidence that they have anything but Chinese government financing at this stage. Shanghai Grace's website is singularly uninformative -- see http://www.gsmcthw.com/hrapp/.
"Semiconductors Mainland chipmakers aim to outpace Taiwan counterparts by 2004", CENS, Taipei, March 20, 2002. U.S. investors were listed as Goldman Sachs and Hambrecht & Quist. However, SMIC's website, only slightly better than Grace's, gives no indication of ownership -- except to note that the Chairman of the Board is Dr. Wang Yangyuan, "Department Director and professor at Beijing University's Micro-Electronics Research Center". See http://www.smics.com/website/enVersion/index_800.jsp. Professor Wang likely represents the majority shareholder -- the Chinese government.
Richard McGregor, "Apec-Asian Pacific Cooperation - New Plants Open On Fertile Ground - Technology In China", Financial Times, October 16, 2001
http://www.smics.com/website/enVersion/index_800.jsp
Tu Zhihao, Wang Shiqi 涂志豪、王仕琦工商時報 "Settling in the Yizhuang Economic Development Zone, SMIC's Beijing 8-inch Fab moves in and sets up shop" 落腳亦庄經濟技術開發區 中芯北京八吋廠 易地起灶 Commercial Times Taipei, June 17, 2002. However, the Wuhan Evening News describes Zhang Fuxing as a "Chinese visitor in the United States" (旅美華人張復興) and suggests that Zhang is not investing, he is merely "arranging funding". See an internet-archived version of that article at http://book.people.com.cn/big5/paper87/77/class008700001/hwz52857.htm
(no author cited), "SMIC expects losses to widen"; China Economic News Service (CENS), September 6, 2002.
(no author cited) "China's fabs face uphill struggle"; Reuters, reprinted in Taipei Times, April 3, 2003, at http://www.taipeitimes.com/News/biz/archives/2003/04/03/200684
Bill Heany, "China spurs chip market growth throughout Asia"; Taipei Times, April 7, 2003, page 10; at http://www.taipeitimes.com/News/biz/archives/2003/04/07/201195
Bruce Einhorn, "Infineon's 'Second Marriage'"; Businessweek, Online Asia, March 17, 2003,
http://www.businessweek.com/technology/content/mar2003/tc20030317_1400_t...
Dan Nystedt, "Top secret report sets off alarms in the tech sector", Taipei Times, July 4, 2001, <http://ww.taipeitimes.com/news/2001/07/04>
See Jimmy Chuang "Industrial espionage investigation turns to e-mail recipient"; Taipei Times, March 10, 2002; at http://www.taipeitimes.com/news/2002/03/10/story/0000127077. Also see Jimmy Chuang, "Ex TSMC employee suspected of selling secrets to Shanghai"; Taipei Times, March 7, 2002; available at http://www.taipeitimes.com/news/2002/03/07/story/00001266.
See (no author cited), "Company's secret theft claims leaves TSMC looking dumb", Taipei Times, March 20, 2002 at http://www.taipeitimes.com/news/2002/03/20/story/0000128440
Bruce Einhorn, op. cit.
James Mulvenon, "The Digital Triangle: A New Defense-Industrial Paradigm?"; Economics and National Security: The Case of China, Center for Strategic Leadership, US Army War College, Carlisle Barracks, PA, August 2002; page 278 (emphasis added).
Scott Thurm, "Cisco Ran Sting Operation To Nab a Copycat in China"; The Wall Street Journal, April 4, 2003. http://online.wsj.com/article/0,,SB104942212126672500,00.html
Scott Thurm, "Court Papers Show Huawei Tried to Hire Cisco Engineers"; The Wall Street Journal, March 17, 2003http://online.wsj.com/article/0,,SB104788003886352200,00.html
See EXPORT CONTROLS: Department of Commerce Controls over Transfers of Technology to Foreign Nationals Need Improvement; United States General Accounting Office (GAO) Report to the Chairman, Subcommittee, on National Security, Veterans Affairs, and International Relations, Committee
on Government Reform, House of Representatives, September 2002. (GAO REPORT GAO-02-972)
ibid; p. 11
Lian Junwei, "The West Indirectly Aids Chinese Military Science and Technology Improvements" Taipei, Commercial Times, April 16, 2002 2002.04.16　 工商時報
西方間接支援 共軍科技大幅提升, 連雋偉/外電報導
James Mulvenon's study for the Army War College offers a See U.S. Department of Commerce Technology Administration report of August 1995 entitled "China to Triple its R&D Investment" available at <http://www.ta.doc.gov/Reports/intlS&T/InterS&T.pdf>
GAO report page 2.
See Deputy Under Secretary of Defense for Technology Security Policy and Counterproliferation Lisa Bronson's Prepared Statement January 17, 2002 before the United States-China Commission hearing on
U.S. Export Control Policy Toward China. Transcript of USCC testimony is available at <http://www.uscc.gov/trn117.pdf>
(no Author), "SMIC skirts U.S. ban on 0.18-micron wafer equipment"; China Economic News Service (CENS), August 29, 2002.
(No author), "SMIC unveils 0.13 micron chip technology"; China Economic News Service (CENS), December 17, 2002.
(No Author),"SMIC plans to retool first 12 inch fab at year end"; China Economic News Service (CENS) March 27, 2003.
Shi Yuwen, 施鈺文"Japanese Waferfab R&D Centers move en masse Westward" (日本晶片研發中心 群起「西進」); Taipei China Times, April 22, 2002. Also see "Taiwan IC designers clamor for mainland investment rights", CENS Taipei, April 16, 2002. See also "VIA Tech to have more employees in mainland than Taiwan", China Economic News Service Taipei, April 29, 2002, which notes VIA's payroll in China "has ballooned to 450, making it the company with the largest team of IC designers on the mainland." Also note a statement by a China-based chip executive that "with many local IC design companies and multinational corporations moving in, China is an emerging and important market for Magma in the Asia-Pacific region," said Lung Chu, vice president of Asia-Pacific operations for Magma. "Because of its new architecture and advanced features, Magma's design solutions can not only handle large complex deep submicron SoC designs, but also greatly shorten time to market and improve designer productivity. We look forward to working with China's IC design community." See "Beijing National IC Design Industrialization Base Adopts Magma Design Solutions" a press release from Magma Design Automation Inc. of Cupertino, CA dated May 23, 2002, at <http://www.chipcenter.com/extframe/extpage.html>. But the Taipei Times notes that several Taiwan design firms will not use chip-manufacturing services in China nor will they set up R&D facilities there in the near future partially because of the lack of intellectual property protection in China. See Dan Nystedt "Chip design houses eye China market", Taipei Times, March 15, 2002, available at
http://www.taipeitimes.com/news/2002/03/15/story/0000127844
Xu Xiumei (徐秀美) "Motorola builds software center in Chengdu" (摩托羅拉在 成都建大型軟體中心), Taipei Commercial Times, May 20, 2002. See also Motorola Asia Pacific Semiconductor Products Sector press release "Motorola Opens China Predictive Technology Laboratory in Beijing" at <http://apspg.motorola.com/press/press/012999/beijing.html>. See also Intel China research Center press release at <http://www.intel.com/jobs/china/crcenter.htm>
Dan Nystedt, "TSMC moves to protect intellectual property in China", Taipei Times, May 24, 2002 at <http://www.taipeitimes.com/news/2002/05/24/story/0000137449.
"Cabinet gives go ahead for chip investments in mainland China"; Taipei China Economic News Service CENS, April 1, 2002
GAO report page 4
GAO report page 16
GAO report page 3 says, "the United States approves most licenses for exports to China of semiconductors manufacturing equipment and materials."