For now, the United States still excels at teaching science and engineering at the graduate level, and also in university-based research. But as the Chinese get more feeder stock coming up through their improving high schools and universities, “they will get to the same level as us after a decade,” said Intel chairman Barrett. “We are not graduating the volume, we do not have a lock on the infrastructure, we do not have a lock on the new ideas, and we are either flatlining, or in real dollars cutting back, our investments in physical science.”

Every four years the United States takes part in the Trends in International Mathematics and Science Study, which assesses students after fourth grade and eighth grade. Altogether, the most recent study involved roughly a half million students from forty-one countries and the use of thirty languages, making it the largest and most comprehensive international study of education that has ever been undertaken.

The 2004 results (for tests taken in 2003) showed American students making only marginal improvements over the 2000 results, which showed the American labor force to be weaker in science than those of its peer countries. The Associated Press reported (December 4,2004) that American eighth-graders had improved their scores in science and math since 1995, when the test first was given, but their math improvement came mainly between 1995 and 1999, and not in recent years. The rising scores of American eighth-graders in science was an improvement over 1999, and it lifted the United States to a higher ranking relative to other countries. The worrying news, though, was that the scores of American fourth-graders were stagnant, neither improving nor declining in science or math since 1995. As a result, they slipped in the international rankings as other countries made gains. “Asian countries are setting the pace in advanced science and math,” Ina Mullis, codirector of the International Study Center at Boston College, which manages the study, told the AP. “As one example, 44 percent of eighth-graders in Singapore scored at the most advanced level in math, as did 38 percent in Taiwan. Only 7 percent in the United States did.” Results from another international education test also came out in December 2004, from the Program for International Student Assessment. It showed that American fifteen-year-olds are below the international average when it comes to applying math skills to real-life tasks.

No wonder Johns Hopkins University president Bill Brody remarked to me, “Over 60 percent of our graduate students in the sciences are foreign students, and mostly from Asia. At one point four years ago all of our graduate students in mathematics were from the PRC [Communist China]. I only found out about it because we use them as [teaching assistants] and some of them don't speak English all that well.” A Johns Hopkins parent wrote Brody to complain that his son could not understand his calculus professor because of his heavy Chinese accent and poor English.

No wonder there is not a major company that I interviewed for this book that is not investing significantly in research and development abroad. It is not “follow the money.” It is “follow the brains.”

“Science and math are the universal language of technology,” said Tracy Koon, Intel's director of corporate affairs, who oversees the company's efforts to improve science education. “They drive technology and our standards of living. Unless our kids grow up knowing that universal language, they will not be able to compete. We are not in the business of manufacturing somewhere else. This is a company that was founded here, but we have two raw materials-sand, which we have a ready supply of, and talent, which we don't.” (Silicon comes from sand.)

“We looked at two things,” she continued. “We looked at the fact that in disciplines that were relevant to our industry, the number of U.S. students graduating at the master's and Ph.D. levels was declining in absolute numbers and relative to other countries. In our K to twelve we were doing okay at the fourth-grade level, we were doing middle-of-the-road in the eighth grade, and by the twelfth grade we were hovering near the bottom in international tests related to math. So the longer kids were in school, the dumber they were getting... You have teachers turning off kids because they were not trained. You know the old saw about the football coach teaching science-people who do not have the ability to make this accessible and gripping for kids.”

One of the problems in remedying the situation, said Koon, is the fact that education in America is relatively decentralized and fragmented. If Intel goes to India or China or Jordan and introduces a teacher education program for making science more interesting, it can get into schools all over the country at once. In America, the public schools are overseen by fifty different state governments. While Intel does sponsor research at the university level that will benefit its own product development, it is growing increasingly concerned about the feeder system into those universities and the job market.

“Have we seen any change here? No, not really,” said Koon. So Intel has been lobbying the INS for an increase in the number of advanced foreign engineers allowed into the United States on temporary work visas. “When we look at the kinds of people that we are trying to hire here-the master's and Ph.D. levels in photonics and optics engineering and very large-scale computer architecture-what we are finding is that as you go up the food chain from bachelor's to master's to Ph.D.'s, the number of people graduating from top-tier universities in those fields are increasingly foreign-born. So what do you do? For years [America] could count on the fact that we still have the best higher-education system in the world. And we made up for our deficiencies in K through twelve by being able to get all these good students from abroad. But now fewer are coming and fewer are staying... We have no God-given right to be able to hire all these people, and little by little we won't have the first-round draft choices. People who graduate in these very technical fields that are critical to our industries should get a green card stapled to their diploma.”

It appears that young people wanting to be lawyers started to swamp those wanting to be engineers and scientists in the 1970s and early 1980s. Then, with the dot-com boom, those wanting to go to business school and earn MBAs swamped engineering students and lawyers in the 1990s.

One can also hope that the marketplace will address the shortage of engineers and scientists by changing the incentives.

“Intel has to go where the IQ is,” said Koon. Remember, she repeated, Intel's chips are made from just two things-sand and brains, “and right now the brains are the problem... We will need a stronger and more supportive immigration system if we want to hire the people who want to stay here. Otherwise, we will go where they are. What are the alternatives? I am not talking about data programmers or [people with] B.S. degrees in computer science. We are talking about high-end specialized engineering. We have just started a whole engineering function in Russia, where engineers have wonderful training-and talk about underemployed! We are beefing that up. Why wouldn't you?”

Wait a minute: Didn't we win the Cold War? If one of America's premier technology companies feels compelled to meet its engineering needs by going to the broken-down former Soviet Union, where the only thing that seems to work is old-school math and science education, then we've got a quiet little crisis on our hands. One cannot stress enough the fact that in the flat world the frontiers of knowledge get pushed out farther and farther, faster and faster. Therefore, companies need the brainpower that can not only reach the new frontiers but push them still farther. That is where the breakthrough drugs and software and hardware products are going to be found. And America either needs to be training that brainpower itself or importing it from somewhere else -or ideally both—if it wants to dominate the twenty-first century the way it dominated the twentieth-and that simply is not happening.