September 15, 2003
San Jose, California
Good afternoon. I am delighted to be here. I would like to thank both the U.S.-Taiwan Business Council and the Fabless Semiconductor Association for organizing this important conference.
My Bureau at the Commerce Department – the Bureau of Industry and Security – actually deals with a range of issues at the intersection of business and national security, including export controls, anti-boycott compliance, defense trade and offsets, imports and foreign investment that affect our security, and monitoring the health of our defense industrial base. The People’s Republic of China and Taiwan figure prominently in several of these areas. This afternoon I would like to focus my remarks on export control issues related to these important trading partners.
Let me begin by briefly sketching the historical context and geopolitical landscape that provides the underpinning for today’s export opportunities with these emerging growth markets. I will then turn to some specific issues related to the semiconductor industry.
This audience, of course, knows well about the ups and downs of U.S.-China
relations over the past several decades. The treatment of exports of sensitive technology has reflected the evolution of that relationship.
Strong political, economic, and security ties between the United States and Taiwan have endured for over 50 years. In 1979, when the United States established relations with Beijing, our close bond with Taiwan continued unabated and became even stronger following Taiwan’s political transformation to a full democracy. Our policy has always stressed that the differences between the PRC and Taiwan need to be resolved peacefully, and we have consistently opposed the use of force to decide Taiwan’s status. It therefore stands to reason that our system of export controls has always treated – and continues to treat – Taiwan as a friendly partner.
Our relationship with the PRC has been far more complex and has evolved significantly over time. In 1950, one year after the United States and its European allies established the Coordinating Committee for Multilateral Export Controls – known as COCOM – to restrict trade with the Soviet Union, the PRC also became a proscribed country for exports of sensitive Western goods. In effect, COCOM’s principal objective was to serve as the de-facto economic arm of NATO, and impede the Communist Bloc’s ability to develop its defense industrial base. Over the years, the scope of COCOM controls was quite expansive, reaching a wide range of commodities that were readily available over the counter at retail outlets. The fundamental supposition was that the export of any controlled item to proscribed destinations should be considered as an export for a hostile military use – in other words, that any export could and would be diverted to support hostile military operations.
These stringent measures were, in fact, effective for many years because the COCOM members possessed a virtual monopoly on the technologies that they controlled. But over time, we began to see our relationship with the PRC in a more nuanced manner – no longer as part of a monolithic, Communist movement, but as a potential strategic counterweight to the Soviet Union and the Warsaw Pact. Accordingly, in the mid-1980s, COCOM developed what became known as the "China Green Line" – denoting a higher level of technology that could be exported to the PRC at national discretion, without the need for unanimous approval of all COCOM members. Thus, we were beginning to take steps to treat the PRC in a more favorable manner than other COCOM controlled countries.
When the Cold War ended with the dissolution of the Soviet Union, the political consensus to maintain a rigid system of COCOM controls ended with it. The West replaced COCOM with the far more informal Wassenaar Arrangement – a regime that seeks to foster common export control policies among "producer states" with respect to conventional arms and what are known as dual-use items and technologies. Dual-use items are those that have both military and commercial applications, such as high performance computers, semiconductors, encryption software, and night vision equipment. Unlike COCOM, the Wassenaar Arrangement focuses on a narrow range of items. It does not have an East-West orientation, nor does it specify the PRC as a target country. And it has even admitted Russia as a member.
Although the Wassenaar Arrangement identifies items and technologies for export control, it leaves licensing decisions to the national discretion of its 33 members. Unlike COCOM, where unanimous consent of the membership had to be obtained before exports of specified items were permitted, each Wassenaar member today is free to make its own judgments regarding the security implications of proposed export transactions. And, as I have noted, there is no obligation to subject proposed exports to the PRC to any special scrutiny. The same principle, in fact, holds true for the three other nonproliferation export control regimes – the Nuclear Suppliers Group, the Australia Group (which deals with chemical and biological items), and the Missile Technology Control Regime. Each of these regimes lists items for export control, but each leaves the decision regarding the disposition of specific proposed transactions to the national discretion of its members. None specifies the PRC as a specific country of concern.
These realities, of course, affect our administration of controls on the export of dual-use items and technologies, including semiconductor manufacturing equipment and microprocessors. We fully appreciate the fact that the U.S. technology monopoly that made Cold War controls effective no longer exists. Not only have Japan and the countries of Europe developed high technology industries in numerous sectors, but so too are Taiwan, South Korea, and other Asian countries developing prodigious indigenous capabilities. There is not always consensus among this growing group about what technologies and which countries pose risks. It is thus far more difficult today to develop an international regime of technology denial than it was in the past. And export controls that limit market opportunities for U.S. companies, while permitting foreign competitors to sell the identical item, often hurt ourselves, and not the intended targets of our controls.
We also fully appreciate the fact that many of our regime partners have divergent views on whether the PRC might pose a strategic threat to regional and international security. Indeed, we have differences of opinion within our own country on that issue. And even though most of our partners share our assessment of China’s poor human rights record, this does not translate into a consensus on issues of high technology trade.
So where does this leave us with respect to China, and what are the implications for Taiwan? Scholars have long noted the pendulum swings in attitudes toward China, and in U.S.-China relations. All too often, one issue or another has dominated the relationship. In my view, one of the quiet accomplishments of this Administration has been to keep relations with China on an even keel and in a positive, cooperative framework.
At the outset of the Administration, it did not always seem that this would be possible. But since the effective handling of the EP-3 incident in April 2001, the U.S.-China relationship has greatly improved. The United States and China today are significant trading partners, peaceful competitors for regional influence, and nations with shared global interests. We seek to have constructive and cooperative relations with the People’s Republic of China, and to integrate China into a rules-based international community. In this manner, we and the Chinese can successfully address a wide-range of issues important to our mutual well-being.
For example, since September 11, the United States has worked with China and other coalition countries in the war on terrorism. We have also collaborated actively with China on efforts to restrict North Korea’s nuclear program. And we seek a productive dialogue with China on arms control and nonproliferation; human rights; efforts against drug trafficking, alien smuggling, and international crime; protection of the environment; and, of course, trade and investment. No single issue dominates our complex relationship. We seek to expand cooperation where possible and to address in a straightforward manner those areas where we have differences.
Since China first embraced market reform and significantly opened itself to foreign trade and investment, its trade with the United States has grown exponentially. China is now our fourth largest trading partner. Although we run our biggest trade deficit with China, and although the growth in U.S. exports to China has been much less pronounced than the tremendous growth in the U.S. imports from China, China is our seventh largest export market, with sophisticated manufactured products and machinery leading the list of U.S. exports.
With China’s accession to the WTO in December 2001, the President granted Permanent Normal Trade Relations (PNTR) status to China, thus terminating the application of the Jackson-Vanik provisions. The systemic reforms that China will undertake as a consequence of its WTO accession should facilitate our business dealings with the Chinese. Although to date China’s implementation of WTO provisions has been mixed, its commitment at the highest levels of leadership appears to remain strong. Given the breadth and complexity of the commitments undertaken by China, the U.S. Government will devote considerable attention and resources to monitoring and enforcing Chinese compliance with its trade obligations.
The increases in the overall level of trade with China have been paralleled both by a steady growth of U.S. investment in China and by increases in licensed trade of dual-use goods and technologies. The dollar value of licensed exports to China increased from approximately $515 million in calendar year 2001 to over $2.8 billion in 2002. And licensed trade would be substantially higher than that in 2003 had we not decontrolled the export of general purpose microprocessors, thereby leading us to return without action license applications valued at over $5 billion, because a license is no longer needed for the export of such microprocessors to China.
I should also note, however, that licensed exports to China constitute less than 15 percent of overall exports to China and less than 3 percent of the value of our trade deficit with China. Thus, the notion often put forward by the Chinese Government that, if we simply relaxed or eliminated export controls, our trade deficit would disappear or be greatly reduced, is simply not the case.
Although China is not a member of any of the four multilateral export control regimes, it has agreed to accept a number of critical international nonproliferation norms. The PRC has joined the International Atomic Energy Agency, and ratified the Nuclear Nonproliferation Treaty, the Chemical Weapons Convention, and the Biological Weapons Convention. China has also promulgated a series of regulations and restrictions on the export of missiles and missile-related technology.
Nonetheless, from a practical perspective, China’s performance in the area of nonproliferation and export controls has been less than satisfactory. A number of PRC entities have exported sensitive items to countries of concern. The U.S. Government has imposed sanctions on these entities, and we have urged the PRC to strengthen the implementation and enforcement of its export control program. In addition, we are seeking greater transparency regarding end-user assurances through our ability to conduct on-site checks in the PRC. Indeed, if we want to be able to make licensing decisions based on the commercial bona fides of end-users, we need to be able to conduct routine end-use checks in China to verify that the items we export are being used for the appropriate purpose by the appropriate entity.
Such end-use verification visits by the Commerce Department are an ordinary part of strategic trade. They ensure that strategically sensitive exports are not diverted to military end users, and thereby strengthen confidence in our trade relationship. We conduct such end-use verification visits, without problem, in over 85 countries. However, we have difficulty on this issue in China, where the government often restricts our ability to conduct this routine activity. Although we have made some progress with the Chinese in this area, much more needs to be done in order to have an effective system in place. Without further progress, our ability to license exports to certain Chinese companies will be increasingly circumscribed.
Let me now turn directly to the semiconductor industry. Your industry is not only a fundamental building block of the information technology sector, it is one of the greatest creators of wealth in the United States and in the world. Semiconductors have accelerated the development and productivity of industries as diverse as telecommunications, automobiles, and military systems. Although ten years ago Taiwan was a small player in the semiconductor market, today Taiwan’s semiconductor industry is an indisputable success story. Taiwan now possesses the most sophisticated technology and world renowned customer service. As you know, it is home to the two largest semiconductor pure play foundries in the world the Taiwan Semiconductor Manufacturing Corporation (TSMC) and the United Microelectronics Corporation (UMC).
The success of Taiwan in semiconductor production – especially the contract semiconductor business – is inextricably linked with that of U.S. companies. Our fabless semiconductor companies have benefitted tremendously from this business model. As this audience well knows, mutually advantageous partnerships and alliances between U.S. and Taiwanese companies in the chip making area abound.
In addition, Taiwan has long represented a huge market opportunity for U.S. chip equipment producers. In 2002, Taiwan imported from the United States semiconductor manufacturing equipment valued at over $1 billion. Today, Taiwan is the third largest market – after the United States and Japan – for such equipment. Because of our close political relationship, exports of fab tools to Taiwan have posed few security concerns, and the approval rate for license applications is extremely high.
As for China, it is poised today to become a major player in semiconductor production, much like Taiwan ten years ago. Although the Chinese share of global integrated circuit production is still tiny, it is forecast to grow rapidly. In particular, the Semiconductor Manufacturing International Corporation (SMIC) and the Grace Semiconductor Manufacturing Corporation (Grace) are hoping to achieve the levels of success that TSMC and UMC have enjoyed. These companies and others have attracted significant foreign investment and established high profile partnerships with major semiconductor firms. Key integrated device manufacturers are also establishing their own production facilities in China. All of these firms aim to take advantage of China’s ample supply of low cost engineers, cheap land, strong government backing and incentives, and, of course, huge internal market.
China also represents a large and growing market for semiconductor manufacturing equipment. Chinese demand for such equipment was $1.2 billion in 2002, about 6 percent of the world market. This is expected to increase over the next several years. With regard to U.S. exports of semiconductor manufacturing equipment to China, they totaled $350 million in 2002, making China the fourth largest export market for the U.S. industry.
The emergence of China as a major semiconductor producer – and the global integration of its semiconductor industry – pose somewhat of a dilemma for U.S. export control policy makers. While the U.S. Government very much wants to support economic opportunities for the semiconductor industry, some of these opportunities in China also raise potential security concerns, because advanced semiconductors are at the heart of today’s advanced weapons systems.
As I have noted previously, our trade relationship with China is far more complex than that with Taiwan, where trade has always proceeded with few restrictions. Our export control policy for China with regard to the semiconductor industry reflects the complexity of our overall relationship.
U.S. policy places stringent controls on the export of goods and technologies that could make a significant contribution to a potentially hostile military. While general purpose chips or microprocessors are no longer controlled, controls remain for those chips intended for military applications. In addition, the equipment used to make sophisticated semiconductors is tightly controlled for national security reasons by both the United States and Wassenaar Arrangement member countries. So what, you may ask, are the practical implications of this framework?
On the one hand, some commentators have asserted that the U.S. Government should try to keep China two generations behind state-of-the-art developments in the semiconductor industry. On the other hand, some members of industry argue that we should de-control semiconductor manufacturing equipment entirely. Our policy is at neither extreme. We do not take the Cold War stance that all exports to the PRC should be evaluated as exports that will be diverted to a hostile military. But, because of the security issues I have outlined, we also are not prepared to treat China as we would a partner in our export control regimes. Nor are we prepared to treat China as we do Taiwan.
In the case of China and its developing semiconductor industry, the United States maintains a "case by case" review policy – each proposed transaction is reviewed on its merits, paying particular attention to the risk that a PRC end-user would divert the equipment or technology to a military end-use. Accordingly, the analysis and review of the user and end-uses of the equipment, the types of semiconductor devices produced (for example, uncontrolled memory chips versus leading edge microprocessors), and the customers for these products are paramount to the license decision making process. We also pay close attention to the degree to which the PRC end-user develops and maintains a meaningful export control compliance program designed to prevent onward proliferation. There are no predetermined technology limits or "red lines" set forth in the Export Administration Regulations, but we carefully evaluate the quantity and quality of the equipment and technical know-how proposed for export to assure that it is necessary to support the end uses stated in the license application.
Given our growing overall trade relationship with the PRC, we want to reap the substantial economic benefits that U.S. exports to China provide for U.S. companies, and for the American consumers that purchase the myriad electronic devices in which the semiconductors are used. Moreover, with the revenue generated by exports, the United States gains security benefits through the continued health and competitiveness of its high-tech firms. Thus, the vast majority of license applications to export semiconductor-related production equipment to commercial end users in China have been approved, though often with numerous and stringent conditions. At the same time, exports of controlled items to entities that are linked to China’s military establishment will continue to be subject to a policy of denial.
From a business perspective, I understand that the process for seeking approval for exports to China’s semiconductor industry is sometimes prolonged and frustrating. However, as we in the export control policy making community become more familiar and comfortable with the main players in the growing Chinese commercial industry, I expect that the process itself may become smoother. Toward this end, we are currently studying possible measures that will ensure our security but at the same time result in reduced license processing times. Some of the measures we are evaluating include pre-screening the commercial bona fides of the small number of mainland foundries, and developing standard license conditions that keep pace with changes in technology. We also are expanding our contacts with industry trade associations and U.S. and foreign semiconductor firms in order to facilitate better understanding of the emerging PRC semiconductor sector.
However, this will only take us so far. We must see greater openness and greater cooperation by the Chinese Government on export control issues – especially on issues such as end-use verification visits, which help build confidence and trust that licensed exports are being used consistent with license conditions. One way in which you can be helpful in this process is by explaining to your commercial counterparts in China that end-use verification visits are an ordinary part of strategic trade, by encouraging the Chinese to cooperate with requests for such visits, and by impressing upon the Chinese that it is in their own interest to devote the necessary attention and resources to this issue. Another important contribution you can make is to urge the companies with whom you trade to develop effective export control compliance programs, and to urge the Chinese Government to continue to develop a comprehensive and effective system of export controls for sensitive goods and technologies and to improve its enforcement capabilities. All of this will serve to enhance U.S. confidence in our ability to engage in strategic trade with China.
We are also paying close attention to the growing investment and trade that Taiwan’s semiconductor industry is making on the mainland. In this regard, it is important to note that the re-export from Taiwan to China of U.S.-origin controlled semiconductor equipment, software, and technology is subject to U.S. re-export requirements – even if some of those items were exported to Taiwan under license exceptions. This means that a Taiwanese exporter would be required to apply to the Commerce Department for a re-export license to ship to China not only U.S.-origin equipment, but also foreign-made equipment in which 25 percent of its value consists of controlled U.S. content. Taiwanese companies need to keep these requirements in mind as they invest in semiconductor foundries on the mainland and increase the levels of technology they are prepared to transfer. And Taiwan itself also needs to continue to enhance its own export control system in order to ensure that sensitive goods and technologies do not fall into the wrong hands.
China’s ultimate success as a major world player in the semiconductor industry remains to be seen. Some analysts predict that China will become the second largest market for semiconductors and one of the largest markets for semiconductor equipment and materials in the next decade. Clearly, as the Chinese market grows, it will increasingly influence the global semiconductor industry, and U.S. policy makers must adapt to deal with this development. Our mission is to do so in a way that promotes U.S. economic interests, while at the same time protecting U.S. national security. I expect that our counterparts in Taiwan are pondering similar issues with regard to their economic relationship with mainland China in the semiconductor industry.
In sum, I think there is a broad highway of trade opportunities available for enterprising, yet prudent companies. Few highways offer so much promise, but also have so many speed traps along the way. I therefore urge you to work closely with us at the Commerce Department as you seek to expand your business relations in ways that promote our twin objectives of trade development and national security.