China's new status as a global center for low-cost manufacturing has turned it into the world's third largest market for semiconductors. The figures vary depending on who is counting, but by one estimate China consumed around $35 billion worth of semiconductors in 2004.

China's domestic production of chips, however, doesn't come anywhere close to matching its appetite. Domestic production for 2004 has been valued at $5.5 billion of chips. More important, the rate of consumption is growing faster than the rate of production. The gap is getting bigger.

"I don't think China is a threat," says Nicholas Lardy, a China specialist who is a senior fellow at the Institute for International Economics in Washington. "China's semiconductor imports are many times domestic production. I'm not saying that's a permanent condition, but the gap has to start narrowing before we can talk about China being a threat."

For all its vaunted growth, China is generally at least one, and usually two, generations behind the state of the art in chip production. Currently, many of its chip plants, or "fabs," are stocked with used equipment sloughed off by other nations, such as Taiwan, that have moved on to higher ground. China does not yet make its own equipment for fabricating chips, and most observers argue that it will be much harder to make a dent in that industry than in simply manufacturing the chips themselves.

While the vast majority of computers, cellphones, DVDs and other electronic devices are now assembled in China, the top-of-the-line chips that run those devices are imported, either from East Asian neighbors like Taiwan, South Korea and Japan or from the good ol' United States. Intel, to pick just one very huge example, earns as much as 10 percent of its total $30 billion a year in revenue from selling computer microprocessor chips to China.

This is worth noting because the United States suffers from a well-publicized trade deficit with China that is vast and growing, engendering huge economic and political pressures. But don't blame the chip business! Semiconductors are America's second largest export to China; U.S. companies export far more chips to China than they import, mainly because the United States dominates at the very top of the line -- the manufacture of highly complex microprocessors that are the brains of new computers. The United States may not have a "diverse" base of producers of these chips, as the Defense Science Board report notes, but the companies that do exist, such as Intel, are formidable forces in the global economy.

A close look at China's chip industry also reveals that some of the warning signs stressed by nervous observers don't bear a whole lot of scrutiny. Venture capital may be flooding into chip design companies in Shanghai and Guangzhou, but the companies themselves, says Kuo, are mostly targeting low-end markets that do not require the latest technology. SMIC, China's chip-making leader, is struggling to become profitable. Western companies will continue to hold tremendous advantages by virtue of their ownership of key intellectual property involved in the design of both chips and chip-manufacturing equipment.

This is not to say that even those sectors won't eventually also move offshore. But as Borrus observes, "I'm not sure we are in any worse shape now than we were 10 years ago, frankly. Then it was IBM and Intel and Motorola and [Texas Instruments] and one or two others. And today it is IBM and Intel and Freescale [formerly Motorola] and T.I."

So why is the United States doing well in semiconductors while other industries, such as textiles, are getting shredded? One explanation is that chip making, with its fast-moving production cycles and constant emphasis on packaging an ever greater number of transistors into an ever smaller space, requires an ongoing investment in research and development that keeps a company, or an industry, at the top of the heap. Borrus points out that companies like Intel, by manufacturing the brains for new computers, will always be at the leading edge of semiconductor-manufacturing-process technology.

The pertinent question is whether the United States, as a whole, is making the kinds of investments in the future that are necessary for its long-term economic health. America has always thrived by inventing new markets, new businesses, new technologies -- and it has historically done so with fairly significant government help. The commercialization of the transistor was hugely influenced by federal funding, motivated by national defense. So was the development of basic computer technology and the Internet.

But there are signs that the federal government has become distracted. A report released in June by the National Bureau of Economic Research documented trends that appear to be very troubling for anyone concerned with the future of U.S. dominance in science and technology. Titled "Does Globalization of the Scientific/Engineering Workforce Threaten U.S. Economic Leadership?" the report answers its own question with a qualified yes.

"The U.S. share of the world's science and engineering graduates at all degree levels is declining rapidly," states the report's author, Richard Freeman. As one consequence, "the [U.S.] faces a long transition to a less dominant position in science and engineering associated industries."

Another recently published book, Ernest Peeg's "The Emerging Chinese Advanced Technology Superstate," reports that as a percentage of gross domestic product, federal spending on basic science has stayed basically unchanged for 30 years, while spending on physical science research has actually decreased. Meanwhile, Chinese government support for basic science, R&D, and science and technology education is surging. Chinese R&D expenditures grew by 22 percent per year between 1995 and 2002, compared with 6 percent in the United States. Chinese science and doctoral degrees increased by 14 percent per year from 1995 to 2001, compared with a 1 percent decrease in the United States. Certainly, the Chinese are starting from a smaller base and growth is naturally faster, but the long-term consequences of such numbers are eye-opening. By every measure deemed important -- patents filed, technical papers published, engineers who graduated -- the Chinese are surging.

It is difficult to see how any of the punitive measures being proposed, such as tariffs and export control measures, will address this fundamental shift. Protectionist moves, say most economists, would likely leave U.S. workers and citizens even worse off than they are now, as prices for all kinds of products would rise, while there would be no guarantee that U.S. firms would be better able to compete globally. Likewise, attempts to prevent China from gaining access to advanced technologies, through measures such as the Wassenar Arrangement, are widely seen as ineffectual. Even if the Bush administration succeeds in preventing U.S. semiconductor equipment manufacturers from selling their wares to China, it will be handing a golden opportunity to competitors such as Japan and Germany.

"I'm not sure that controlling the frontier can be technically done," says Lardy.

Instead, the United States must push at the frontier. But in the absence of consistent government support, who is going to make that push? Ten years ago, one might have looked to the venture capital firms of Silicon Valley, busily funding the Yahoos and Googles of the world. But guess where Silicon Valley is headed today?

Recent Stories