SOMETHING is wrong with this picture:
In 1987, American semiconductor companies form a consortium to improve chipmaking in the United States. Only by banding together with government subsidies and relaxed antitrust laws, they argue, can they beat back foreign competition.
Fast forward to July 1992. IBM, still a consortium member, announces a memory-chip deal with Japan's Toshiba and Germany's Siemens. On the same day, consortium member Advanced Micro Devices Inc. announces its own chip partnership with Japan's Fujitsu.
Confused? Some analysts are. "It is inconsistent to, on one hand, work with foreign companies and, on the other hand, pursue protectionist strategies," says Robert Wilson, a professor at the L.B.J. School of Public Affairs at the University of Texas at Austin.
"I don't think you can construe it as anything but a contradiction," adds Paul Nesdore, manager of international planning services at International Data Corporation.
This contradiction has important consequences for legislators and others trying to formulate US technology policy. By fuzzing national borders, multinational firms are forcing them to confront intriguing new questions. What does a US company look like? Does it exist anymore? Is national technology policy even relevant in a global economy? Needed: different focus
The easiest - and probably the wrong - focus of debate is the multinational companies, analysts agree. For one thing, the chip consortium, called Sematech, and last month's semiconductor deals have different aims. Sematech wants to improve the US manufacturing process; the international deals involve developing next-generation chips. With international competition rising and development costs moving beyond the reach of any one firm, companies are anxious to spread their risk.
"The world has changed," explained W. J. (Jerry) Sanders,chairman of Advanced Micro Devices, when he announced the deal with Fujitsu last month. Mr. Sanders was a longtime and vocal critic of Japanese trade practices. Now he plans to build with Fujitsu a $700 million factory to make a new kind of computer memory chip.
Analysts - and even legislators - agree that globalization is changing the rules of the game.
"If you are a company, you need to play the game of international competition at two levels," says Daniel Burton, executive vice president of the private-sector Council on Competitiveness in Washington, D.C.
Companies "have got to survive in the international marketplace.... At another level, they have to have a closer relationship with their governments." Firm not blamed
Sen. Jeff Bingaman (D) of New Mexico, a key supporter of a national technology policy, doesn't blame Sematech members for pursuing international partners.
"In some ways the fact that US firms are competitive in the world market is what allows them to enter into international agreements like this," he says.
But "I think it's appropriate to look at each of these joint ventures and make a judgment as to whether they're in our country's best interest. The fact that it's an international marketplace makes it that much more necessary for us to have a national technology policy."
But for whom? That's the big question.
"If the policy is based on a presumption that there is such a thing as a US company ... that's a policy destined for failure," says Richard Lester, director of the Industrial Performance Center at the Massachusetts Institute of Technology.
Harvard Professor Robert Reich inaugurated this debate two years ago by asking a deceptively simple question: "Who is us?"
"Us," he argued, is not necessarily a US multinational firm. A Hitachi plant in the US delivers more economic benefits to American workers and the country than, say, a Motorola plant overseas. Thus, he concluded, government should not base national technology policy on companies but, rather, on the American work force.
Critics were quick to point out that American multinational firms are still primarily based in the US. In the manufacturing sector in 1988, US multinationals had 70 percent or more of their assets, sales, and manufacturing in US operations. But the trend toward partnerships is clearly increasing. Last year, US high-tech companies consummated an estimated 90 partnerships - up from 10 to 15 a year in the mid-1980s. And increasingly, those partners are foreign. Besides the IBM-Siemens-Toshiba and the AMD-Fu jitsu deals, US high-tech companies that have made foreign deals include Apple Computer (with Japan's Sony and Sharp), Intel Corporation (also with Japan's Sharp), Zenith (now a subsidiary of France's Groupe Bull, which is also linked to IBM), and Rockwell International Corporation (with Japan's Okuma Corporation of Nagoya). With the exception of Zenith and Apple, all these US companies are members of Sematech. Pace to increase
Analysts agree that the pace of foreign high-tech partnerships will increase. Part of it is a fad, Mr. Nesdore says. Perhaps a third of these deals won't work. But the rest of them promise to shake up old notions about technology policy. It is not hard to imagine how manufacturing techniques pioneered at government-subsidized Sematech could be quickly adopted in a foreign partnership.
US technology already slips over US borders at alarming speed. Cypress Semiconductor, a longtime critic of Sematech, recently announced it was moving its manufacturing plant to Asia. Why? Because Asian chip fabrication plants are now as high-tech as US facilities and still pay much lower wages. The challenge for policymakers will be to ensure, as much as possible, that the US still benefits from technology's rapid spread.
"Rather than bemoaning that, we should accept it," says Mr. Burton of the Council on Competitiveness. "What's the option? If we're going to just let it go, isn't that a recipe for decline? "We're going to have to live with greater ambiguity than we ever had before," he says.