The Race To Develop Technology

AMERICAN industry - one of the most competitive, entrepreneurial cultures on earth - is increasingly turning to cooperation as a way to blunt foreign competition. In the race to develop everything from the latest computer chip to plastic cars, United States companies are teaming up with each other, and sometimes universities and government labs, to meet the growing threat posed by Japan and Europe.

Yet even as the country dabbles with cooperative capitalism, questions loom. Do consortiums really work? Will they improve the competitive position of the US? Should the government encourage such ventures?

The questions are not trifling.

As competition becomes more global, joint research and development is likely to become an increasingly popular tool in the US. This is especially true of new technology: Developing high-tech widgets is a difficult and costly process for even the largest companies.

Congress is abuzz over finding ways to ensure US competitiveness in emerging new frontiers, such as supercomputers, high-definition television, and superconductivity. This may include legislation that would make it easier for joint ventures to form.

``There will be more consortia in a whole lot of areas,'' predicts Stephen Cohen, director of the Berkeley Roundtable on the International Economy.

Although long common in Japan, industry cooperation in its modern incarnation didn't take off in the US until after 1984. That was the year Congress revised the nation's antitrust laws to make it easier for competing companies to engage in joint research.

Some 125 cooperatives now exist, conducting research in areas ranging from new types of glass bottles to artificial intelligence. Some ventures involve only a couple of companies. Others, like the six-year-old Microelectronics and Computer Technology Corporation (MCC) in Austin, Texas, the grandfather of modern cooperatives, have well over a dozen corporate participants.

In May, International Business Machines Corporation teamed up with American Telephone & Telegraph Company and the Massachusetts Institute of Technology to plumb superconductivity. The most unusual consortium, however, is also the latest: The computer-chip venture announced here in Silicon Valley - normally a fount of radical American individualism - on June 21 by some of the country's top electronics firms.

Unlike the other consortiums, US Memories Inc. will not be a research cooperative but a manufacturing venture. The aim of the seven companies proposing it is to build a cutting-edge plant to produce computer-memory chips used in a wide array of electronic devices.

Consortiums are alluring for several reasons. Companies figure that by joining forces they can pursue research they normally could not do alone. It spreads around the risks and avoids duplicating work.

Pooling funds and talent is the only way some companies feel they can keep up with Japan, where joint research is more ingrained and often underwritten by the government. Europe, too, has been moving in this direction: More than 100 companies are looking at joining JESSI, a $4 billion European venture to develop memory chips launched in June.

``One company by itself can't do everything,'' says Thomas Mullen of the National Center for Manufacturing Sciences, a Michigan-based consortium.

Yet buddy-system capitalism has its problems. There is no overwhelming evidence that consortiums in the United States work, though advocates contend it is too early to gauge their worthiness.

``The jury is not entirely in on whether they work - and which ones under what circumstances,'' says a Senate staff member.

Overcoming the culture of competition and developing technology that will benefit the diverse interests of individual members is difficult.

``The tendency is to simply do what has been done in the past rather than ask what this consortium can do better than each individual company,'' says Robert Noyce, head of Sematech, one of the most visible cooperatives, formed to improve semiconductor-production technology. ``If you do that, you just become another research lab.''

Recruiting workers can be a problem. MCC had trouble getting member companies to give up top researchers for the effort. Only 16 percent of the cooperative's 430 workers are from participating companies, though MCC staff members say the consortium is starting to produce results.

``Companies are careful about allowing their best people to leave,'' an MCC official says.

As companies consider forming cooperatives to produce products rather than engage in research, other questions arise. Here US Memories Inc., the nation's first major manufacturing consortium, will be a test case.

Led by IBM but also backed by such electronics notables as Digital Equipment, Hewlett-Packard, Intel, and National Semiconductor, US Memories intends to produce the latest generation ``dynamic random-access memories.'' DRAMs are used in nearly all computerized devices.

In recent years, there has been widespread concern about growing dominance of the chip industry by Japanese firms, which control 90 percent of the world DRAM market.

Before the proposed consortium turns out silicon wafers, however, it will have to deal with antitrust concerns. The US Justice Department seems unlikely to challenge US Memories on these grounds, but that doesn't mean private suits won't arise.

Even here, some analysts don't think the consortium will run into problems, since the US occupies such a small corner of the world market.

Still, the idea of joint manufacturing ventures raises policy questions. Convinced that the US needs to move in this direction to stay up with the Japanese, some lawmakers are pushing bills in Congress that would give joint manufacturing ventures a degree of protection from antitrust action, similar to what it did for research cooperatives.

But these moves are opposed by free-market economists, among others, who argue that relaxing the rules will only lead to monopoly industries and stifle innovation.

Other concerns could arise over funding. Until now, Sematech has been one of the few cooperatives to receive federal backing. Sandford Kane, a former IBM executive who is heading up US Memories, says he doesn't plan to seek direct aid for the consortium but will likely ask for government-guaranteed loans. If he follows through, warns Claude Barfield of the American Enterprise Institute, there would be nothing to prevent the three current US manufacturers of DRAMs from seeking similar support.

Still, philosophical concerns aside, the success of US Memories will turn more on its ability to attract member companies and compete with the Japanese.

As Harvard University's Robert Reich, who supports the idea, puts it: ``It is going to be difficult to gain the process technology advances that the consortium is aiming for. No one should have any illusions about how difficult it is going to be.''