Still, the United States remains a central repository of talent and money, said Cradle Technologies' Chang. "There are certainly areas-the Internet, semiconductors-where we maintain such a dominance that it becomes a double pull, because you get the brightest minds who want to work here and maybe they will want to repatriate." The United States still leads the world in patent output, perhaps in part due to that general trend, he said.
More troubling, said Naveed Sherwani, CEO of Open-Silicon (Sunnyvale), is the U.S. failure to develop discoveries in basic research to the point where they can generate revenue that might find its way back to the research labs.
"We do the fundamental research to learn how to make a DVD, we cost-produce it," Sherwani said. "How much did we make? Maybe a billion dollars. How much is China going to make selling DVDs around the world? Maybe $60 [billion] to $70 billion. Now, suppose their margin is only 20 percent and they invest 10 percent back into research. They are going to be way ahead of us."
Universities are pivotal
U.S. educational institutions are in a unique position to foster talent and tackle the initial stages of research. Following those first steps, universities are also ideal launch pads for new companies, said Patterson of UC Berkeley. "Universities bring in these new people who don't know it can't be done and are not constrained by existing products," he pointed out.
For both Patterson and Sun's Edens, government support for university research is the pivot around which all innovation revolves. That is why Edens is worried about the falloff in Darpa's basic-research spending. "That really concerns us, because most of the basic research still occurs in universities," he said. "As graduate students move into the work force, basic research diffuses into the industry, via industrial labs, and then it eventually finds its way into products."
The role of the university lab is increasingly crucial as corporate labs vanish or scale back, Edens said. "We see that our peer research labs in other companies have really cut back in funding, going to a consulting funding model. At that point, in an industrial research setting, you don't really have a lab anymore."
The effectiveness of academic research was on view in the recent Darpa-sponsored autonomous-vehicle race, Patterson said (see Oct. 17, page 1). "Guess who won?" he asked. "It wasn't the defense contractors-academics [teams from Stanford and Carnegie Mellon] took the top three places. Here is the concrete evidence of the importance of having academic research participate, and the importance of government-funded programs like this. But this is the opposite of what people are doing [now]."
The corporate-sponsored university lab could potentially supplant the government-funded one. Intel Corp. has initiated such a program, with its network of labs in Berkeley, Seattle, Pittsburgh and Cambridge, England. Each lab tackles some critical area of information technology in close association with a nearby university.
Patterson thinks the lab network is an interesting concept that might someday help to correct current downward trends, but the problem is still money. "They might fund a staff member or two, but they are still relying on other funding for the graduate students who work on projects," he said.
A fundamental problem with relying on corporations to back basic research is the nature of economic pressures in the commercial world. Although company executives might be concerned about research, they are primarily accountable to the bottom line, and will never have the latitude to just pursue interesting ideas.
"Industry has been shortening its horizon in the past couple of decades," Patterson said. "In the short term, you don't get fired for not doing long-term research." It's hard to know ahead of time if your company is going to benefit from basic research, he said.
Of course, looking only at the United States produces a certain myopia. The basic engine of science and technology moves ahead in spite of national rivalries and political boundaries. With the advent of the Internet and sophisticated transportation, it is now possible to link with colleagues, share design data or outsource production without regard to geography.
"Whether you measure it in terms of patents coming out globally or you measure it from absolute dollars globally going into basic research, I think you will probably find that both of those numbers are going up, and will continue to go up," said Cradle Technologies' Chang. "I don't think there is a shortage of higher-end educational institutions around the world. Therefore, we are churning out people who are taking the kind of educational research that they have done at their schools and are starting to have money globally to do basic research."
No country can maintain a lock on technology for long, in terms of the larger cycles of history, Chang pointed out. And any country-or any individual company-may benefit from basic research conducted anywhere.
U.S. citizens, as taxpayers who supported much of the important research that made the country dominant in the 20th century, are actually in a poor position to judge whether enough is being spent on research today or whether it is being spent wisely. It's a tough problem even for the experts, and the worry is that politicians, like CEOs, are not going to be fired in the short term for not backing enough long-term research.