INTERNATIONAL comparisons of high school math and science scores have been dispiriting for some time, and the latest batch represents only a small improvement. This month the Program for International Student Assessment published data suggesting that the math skills of American 15-year-olds lag behind those of their counterparts in most other industrialized nations. Last week the second authoritative source of comparative research, the Trends in International Mathematical and Science Study, showed American eighth- graders doing better than in the past, though they still came in 15th out of 45 nations.
It's too simple to say that these trends threaten America's global leadership. Economic success depends on many factors -- efficient capital markets, flexible labor markets, a culture that celebrates entrepreneurial risk and regards a failed start-up not as a shame but as a learning experience. But it's nonetheless true that math and sciences are a key to economic growth. Mathematical competence boosts the productivity of mid-level workers, and mathematical stardom fuels industrial and financial innovation.
Moreover, the principal American strategy for addressing the shortage of star brainpower has crashed into a wall, and revival may be only half-possible. That strategy is to import foreigners and enroll them in math and science graduate programs. At many campuses, non-American students represent perhaps half of the enrollment in such programs -- or did until restrictions imposed after Sept. 11, 2001, took away their visas. University presidents are rightly pushing for the visa rules to be relaxed again: Far from posing a threat to national security, the practice of welcoming bright foreigners to the United States is a prime tool of public diplomacy. But the truth is that even if visa restrictions are lifted, the old expectation for foreign graduate students may need some rethinking.
That expectation was that, once admitted to the United States, the foreigners would want to stay: They would start companies in the high-tech clusters around Palo Alto, Calif., or Boston, creating desirable jobs and fueling America's industrial pre-eminence. But this expectation held good partly because the countries that sent graduate students were not attractive places to return to. Increasingly, economic takeoff in countries such as China and India is changing the equation. Moreover, it used to be that technologists wanting to serve the American market had to live here; but the growth in "offshoring" of technology services shows how times are changing.
In sum, it can no longer be assumed that foreign math and science stars will tend to remain in this country. Training them in the United States is still a good idea, both for foreign policy reasons and because their contributions to economic and intellectual advance abroad ought to be celebrated. But if the United States wants to stock its own industry with top-flight technologists, the surest route is to train Americans. That's an extra reason to hope that math at middle and high schools can be strengthened.