Now they have this place where all they do is research, with great computers and lots of resources. They have administrators-we hire people to do the dirty work. They just could not believe it. They voluntarily work fifteen to eighteen hours a day and come in on weekends. They work through holidays, because their dream is to get to Microsoft.“ Li, who had worked for other American high-tech firms before coming to Microsoft, said that until starting Microsoft Research Asia, he had never seen a research lab with the enthusiasm of a start-up company.

“If you go in at two a.m. it is full, and at eight a.m. it is full,” he said.

Microsoft is a stronger American company for being able to attract all this talent, said Li. “Now we have two hundred more brilliant people building [intellectual property] and patents. These two hundred people are not replacing people in Redmond. They are doing new research in areas applicable worldwide.”

Microsoft Research Asia has already developed a worldwide reputation for producing cutting-edge papers for the most important scientific journals and conferences. “This is the culture that built the Great Wall,” he added, “because it is a dedicated and direction-following culture.”

Chinese people, explained Li, have both a superiority and an inferiority complex at the same time, which helps explain why they are racing America to the top, not the bottom. There is a deep and widely shared view that China was once great, that it succeeded in the past but now is far behind and must catch up again. “So there is a patriotic desire,” he said. “If our lab can do as well as the Redmond lab, that could be really exciting.”

That sort of inspired leadership in science and engineering education is now totally missing in the United States.

Said Intel chairman Craig Barrett, “U.S. technological leadership, innovation, and jobs of tomorrow require a commitment to basic research funding today.” According to a 2004 study by the Task Force on the Future of American Innovation, an industry-academic coalition, basic research performed at leading U.S. universities-research in chemistry, physics, nanotechnology, genomics, and semiconductor manufacturing-has created four thousand spin-off companies that hired 1.1 million employees and have annual world sales of $232 billion. But to keep moving ahead, the study said, there must be a 10 to 12 percent increase each year for the next five to seven years in the budgets of key research-funding agencies: the National Institute for Science and Technology, the National Science Foundation, the Department of Energy's Office of Science, and the Department of Defense research accounts.

Unfortunately, federal funding for research in physical and mathematical sciences and engineering, as a share of GDP, actually declined by 37 percent between 1970 and 2004, the task force found. At a time when we need to be doubling our investments in basic research to overcome the ambition and education gaps, we are actually cutting that funding.

In the wake of the Bush administration and the Republican Congress's decision to cut the National Science Foundation funding for 2005, Republican congressman Vern Ehlers of Missouri, a voice in the wilderness, made the following statement: “While I understand the need to make hard choices in the face of fiscal constraint, I do not see the wisdom in putting science funding behind other priorities. We have cut NSF despite the fact that this omnibus bill increases spending for the 2005 fiscal year, so clearly we could find room to grow basic research while maintaining fiscal constraint. But not only are we not keeping pace with inflationary growth, we are actually cutting the portion basic research receives in the overall budget. This decision shows dangerous disregard for our nation's future, and I am both concerned and astonished that we would make this decision at a time when other nations continue to surpass our students in math and science and consistently increase their funding of basic research. We cannot hope to fight jobs lost to international competition without a well-trained and educated workforce.”

No, we cannot, and the effects are starting to show. According to the National Science Board, the percentage of scientific papers written by Americans has fallen 10 percent since 1992. The percentage of American papers published in the top physics journal, Physical Review, has fallen from 61 percent to 29 percent since 1983. And now we are starting to see a surge in patents awarded to Asian countries. From 1980 to 2003, Japan's share of world industrial patents rose from 12 percent to 21 percent, and Taiwan's from 0 percent to 3 percent. By contrast, the U.S. share of patents has fallen from 60 percent to 52 percent since 1980.

Any honest analysis of this problem should note that there are some skeptics who believe that the sky is not falling and that scientists and the technology industry might be hyping some of this data, just to get more funding. A May 10, 2004, article in the San Francisco Chronicle quoted Daniel S. Greenberg, former news editor of the journal Science and author of the book Science, Money and Politics, who argues that “inside-the-Beltway science (lobbying) has always been insatiable. If you double the NIH (National Institutes of Health) budget in five years (as recently happened), they're (still) screaming their heads off: 'We need more money.'” Greenberg also questioned the science lobbyists' interpretation of a number of statistics.

Quoting Greenberg, the Chronicle said, “To put scientific publishing trends in context... it's important to look not only at overall percentiles but also at the actual numbers of published papers. At first, it may sound startling to hear that China quadrupled its scientific publication rate between 1986 and 1999. But it sounds somewhat less startling if one realizes that the actual number of Chinese papers published rose from 2,911 to 11,675. By comparison, close to a third of all the world's scientific papers were published by Americans-163,526 out of 528,643. In other words, China, a nation with almost four times the population of the United States, published (as of 1999) only one-fourteenth as many scientific papers as the United States.”

While I think a dose of skepticism is always in order, I also think the skeptics would be wise to pay more heed to the flattening of the world and how quickly some of these trends could change. It is why I favor Shirley Ann Jackson's approach: The sky is not falling today, but it might be in fifteen or twenty years if we don't change our ways, and all signs are that we are not changing, especially in our public schools. Help is not on the way. The American education system from kindergarten through twelfth grade just is not stimulating enough young people to want to go into science, math, and engineering. My wife teaches first-grade reading in a local public school, so she gets Education Week, which is read by educators all over America. One day she pointed out an article (July 28, 2004) headlined, “Immigrants' Children Inhabit the Top Ranks of Math, Science Meets.”

It went on to say, “Research conducted by the National Foundation for American Policy shows that 60 percent of the nation's top science students and 65 percent of the top mathematics students are children of recent immigrants, according to an analysis of award winners in three scholastic competitions... the Intel Science Talent Search, the U.S. team for the International Mathematical Olympiad, and the U.S. Physics Team.” The study's author attributed the immigrant students' success “partly to their parents' insistence that they manage study time wisely,” Education Week said. “Many immigrant parents also encouraged their children to pursue mathematics and science interests, believing those skills would lead to strong career opportunities and insulate them from bias and lack of connections in the workplace... A strong percentage of the students surveyed had parents who arrived in the United States on H-1B visas, reserved for professional workers. U.S. policymakers who back overly restrictive immigration policies do so at the risk of cutting off a steady infusion of technological and scientific skill,” said the study's author, Stuart Anderson, the executive director of the foundation. The article quoted Andrei Munteanu, eighteen, a finalist for the 2004 Intel competition, whose parents had moved from Romania to the United States five years earlier. Munteanu started American school in the seventh grade, which he found a breeze compared to his Romanian school. “The math and science classes [covered the same subject matter] I was taking in Romania... when I was in fourth grade,” he said.

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.

“There are two things that worry me right now,” said Richard A. Rashid, the director of research for Microsoft. “One is the fact that we have really dramatically shut down the pipeline of very smart people coming to the United States. If you believe that we have the greatest re-seach universities and opportunities, it all has to be driven by IQ. In trying to create processes that protect the country from undesirables, [the government] has done a much better job of keeping out desirables. A really significant fraction of the top people graduated from our universities [in science and engineering] were not born here, but stayed here and created the businesses, and became the professors, that were engines for our economic growth. We want these people. In a world where IQ is one of the most important commodities, you want to get as many smart people as you can.”

Second, said Rashid, “We have done a very poor job of conveying to kids the value of science and technology as a career choice that will make the world a better place. Engineering and science is what led to so many improvements in our lives. But you talk to K through twelve kids about changing the world and they don't look at computer science as a career that is going to be a great thing. The amazing thing is that it is hard to get women into computer science now, and getting worse. Young women in junior high are told this is a really wretched lifestyle. As a result, we are not getting enough students through our systems who want to be computer scientists and engineers, and if we cut off the flow from abroad, the confluence of those two will potentially put us in a very difficult position ten or fifteen years from now. It is a pipeline process. It won't come to roost right away, but fifteen or twenty years from now, you'll find you don't have the people and the energy in these areas where you need them.”

From Richard Rashid at Microsoft in the Northwest to Tracy Koon at Intel in Silicon Valley to Shirley Ann Jackson at Rensselaer on the East Coast, the people who understand these issues the best and are closest to them have the same message: Because it takes fifteen years to create a scientist or advanced engineer, starting from when that young man or woman first gets hooked on science and math in elementary school, we should be embarking on an all-hands-on-deck, no-holds-barred, no-budget-too-large crash program for science and engineering education immediately. The fact that we are not doing so is our quiet crisis. Scientists and engineers don't grow on trees. They have to be educated through a long process, because, ladies and gentlemen, this really is rocket science.