US Needs a Policy For Technological Future, CEOs Say

SMACK in the middle of California's Silicon Valley, computer legend Gordon Bell is on his living room floor, crouched over a bright red fire truck.

"Start the engines!" he says, pressing a button on the toy. It whirs, engine-like, for a few seconds. He presses the button again and says: "Let's move it now!" The truck runs forward about two feet.

"An American company probably made the [computer] chip," he says. But the truck was made in China. Every other electronic gadget in his living room is foreign-made too. "You can't even buy an American product," the former Digital Equipment Corporation executive sighs.

That fire truck says a lot about America's high-technology challenge. It's not just Japanese electronics companies racing against the United States. It's a new world where Chinese toymakers can use leading-edge technologies to turn out competitive products. Technology is not just spreading abroad, it is diffusing into more and more industries, blurring the line between what is high-tech and what's not.

Faced with this new world, the Clinton administration's challenge is to define America's place in it. It is a grand experiment: a nation of 250 million people trying to maintain the high-skill industries and keep the high-wage jobs that billions of other people are competing for. "Increasingly, what we can do, they can do," says Intel chief executive Andrew Grove. "We've got a big, big challenge ahead of us."

Even if the administration didn't act, powerful currents of change are pushing the nation toward a new and as-yet ill-defined place in this high-tech order. President Clinton promises a firm hand on the tiller. No one knows how well he will be able to steer.

In his office outside Washington, D.C., Dave Lehman pulls out a sheet of graph paper and draws a gently downward sloping line. "This is the analog world," he says. "Sony and Panasonic are driving down it. Every year your VCR gets a little cheaper."

He draws a second line that slopes down sharply, crossing the first. "This is the digital world," he says. "The Apples and AT&Ts are rapidly driving down this line." US companies are nearing the intersection at which their new digital technology will be as cheap as Japan's older analog technology. After that, Mr. Lehman warns, watch out! The companies that build digital products will have the advantage.

"I am unabashedly American in this," he adds. His own start-up company, TV Answer, hopes to manufacture digital products for interactive television.

Other high-tech executives are not so sanguine. After all, Japan and Europe are building more digital products too. But the technological shift is allowing the US to get back into the race. Economics may help. With the value of the dollar at record lows against the Japanese yen, US products are more competitively priced. "We may, with the yen-to-dollar situation, have a slight edge on cost," says Ed Sack, chief executive officer of Zilog Inc., a small chipmaker in Campbell, Calif. "But I think it would b ecome very easy to become complacent and lose it again."

US policy is undergoing a sea change. The cold war - the underlying rationale for 45 years of federal research and development - has given way to an era of defense cuts. The US can take a tougher stance on trade issues with allies. Mr. Clinton can also shift federal R&D resources from defense to civilian projects (national labs and other agencies cooperate with industry, Page 10).

There is plenty of skepticism about current R&D policies.

Last August, a National Science Board study concluded that the nation's civilian R&D spending has grown at a slower pace than that of major competitors and that important areas of research, such as process technology, were falling between the cracks. States, as well as the federal government, are trying to address these issues (state technology policies, Page 10).

In the middle of these shifting currents, the Clinton administration has launched a policy emphasizing: 1) environmental technology, especially a more benign automobile 2) a high-speed computer and communications infrastructure (so-called "information superhighways," 3) technology transfer from federal labs to factory floors, 4) modest increases in tax breaks for business investment, 5) modernizing federal agencies through technology and 6) more funds for basic research at universities.

"The good news is we know what needs to be done," says Vice President Al Gore Jr. "We know how to do it.... And we have established models to help us avoid mistakes. So for all these reasons I think there is cause for tremendous optimism for the beginning of a new era of US competitiveness in all of these fields."

In fact, the US has fairly few established models to rely on for government-industry joint ventures. So, while high-tech leaders

generally support the more tried-and-true aspects of the Clinton program, they are more skeptical about its experimental components.

Some high-tech industries, such as software, are so strong that they are looking for government support only in trade and tax issues.

Others decry the government's contradictory policies. The White House's 1994 budget targets biotechnology as one of several critical areas for federal support. But the president also wants to stem drug-price increases as part of his health-care reforms, a move the industry says would slow innovation.

Yet many executives are willing to have government experiment.

"We need greater alignment among industry, government, educational institutions, and labor than ever before," says Motorola chairman George Fisher.

When Clinton and Mr. Gore talk about improving the nation's information infrastructure just about everyone applauds. "We think that kind of leadership at the administration level can bring that vision to the fore," says AT&T chairman Robert Allen.

It's when the administration proposes to move government closer to the marketplace that debate sharpens.

Buried within the bowels of the Defense Department is the tiny Advanced Research Projects Agency. ARPA funds experimental technologies that it considers vital to national security.

So a few years ago when T. J. Rodgers, the outspoken boss of a Silicon Valley chipmaker, set out to expose federal high-tech boondoggles, the agency was one of his targets. The conservative Mr. Rodgers (his office sports a poster of former President Reagan) was surprised at what he found. Instead of waste and inefficiency, he found technical expertise. Instead of pork-barrel projects, he found a tiny agency with worthy projects and tremendous support from venture capitalists.

"You can sum it up in one sentence: small science," Rodgers says. He crossed DARPA - as it was then known - off his list of government boondoggles. (Clinton recently dropped the word "Defense" from the agency title, making it ARPA.)

Why is ARPA so successful? Observers say it is well-managed, technically expert, out of the limelight, and small. With a budget mostly free of political wrangling, the agency has effectively promoted critical new technologies that companies hadn't dared to fund.

Thus, ARPA belies the old argument that government can't pick winners and losers. "In technology, we are already making choices," says Dan Burton, executive vice president of the private-sector Council on Competitiveness. "So let's just review our priorities."

What's unknown is whether such small-time experiments will work in the big time. Complaints have cropped up around one of the agency's successes: massively parallel processing or MPP computers. The General Accounting Office is expected to release a report soon that, while not finding any wrongdoing, criticizes ARPA for promoting some MPP computers over others.

"The government has had a very significant role in stimulating the development of the technology and putting a halo of correctness around massive parallelism," says Jeff Kalb, president of MasPar Computer. But "if the government continues to support machines after you are past the research phase, then the best ideas don't necessarily win out - just the ones that get subsidized." His company did not get federal aid.

The criticism has caused ARPA to clean up its act, Mr. Kalb adds. But the battle is a portent of things to come.

As Washington pours in more dollars and encourages more commercial activities, the stakes surrounding federal high-tech will go up. Companies will lobby harder and complain louder when their machines aren't picked. Will congressmen resist the temptation to lobby for home-district companies? And perhaps a bigger question: When does the government pull out of a project to commercialize a technology? The line between the research phase and the market phase is awfully hazy.

"Yes, there's a great risk of pork," adds Bennett Harrison, political economy professor at Carnegie Mellon University. "And it's worth the risk."