Fueled by a rapidly expanding supply of scientists and engineers, America's mighty economic engine powered the nation through more than a century of remarkable wealth creation. But Stanford Graduate School of Business economist Paul Romer warns that in the coming decades, the economic engine could begin to run out of gas. Our universities are not keeping up with the growing business demand for highly skilled workers, Romer says. Building on his research, he argues that increased incentives -- via government subsidies -- for universities to train scientists and engineers will be necessary to sustain, and certainly to increase, growth in the 21st century.

The proposed incentives, an offshoot of Romer's pioneering "new growth" theory of economics, are generating a fresh wave of interest reaching all the way to Congress. The pending Tech Talent Bill, based on Romer's recommendations, proposes a $25 million pilot program aimed at increasing the number of university math, science and engineering majors.

Romer, who is the STANCO 25 Professor in the Graduate School of Business and a senior fellow at the Hoover Institution, gained notoriety in the field of economics with his 1983 doctoral thesis that made two basic points: Economic growth is driven by new ideas and advances in technology; and by creating appropriate economic incentives, the government can increase the trend rate of growth in a way that can make all citizens better off. Prior to his theory, economists assumed economic growth in a physical world was constrained by diminishing returns and scarcity of resources, and the only offsetting force -- "serendipitous" discovery -- could not be understood using standard economic tools. Nor could it be influenced by government policy.

In Romer's view, nurturing scientific talent is the key to the nation's technological superiority and continued wealth creation. Support for higher education is the lever by which the government can move the entire economy.

The problem is this: Evidence shows that undergraduate institutions are a critical bottleneck in the training of scientists and engineers, and graduate schools produce people trained only for jobs in academic institutions. "Some in academic circles maintain that scientists selflessly pursue the greater good, that money -- or the lack of it -- has no bearing on the career a promising student chooses or the effort that a professor devotes to teaching such a student," says Romer. "But the fact is, incentives do matter for what happens on university campuses."

The government has done a good job of creating incentives that increased the amount of scientific research done on university campuses, says Romer. For the good of the nation, the government also must create incentives to increase the amount of scientific education offered to undergraduates and graduate students. Different programs will be required at these two levels.

To boost the number of undergraduates majoring in science, mathematics and engineering, Romer proposes a new program with new funding. Universities would compete for government grants on the basis of their success in increasing the fraction of their students who receive undergraduate degrees in these fields. Romer's ideas provide the foundation for legislation recently introduced in Congress. The $25 million a year pilot program proposed in the Tech Talent Bill (S. 1549 and H.R. 3130) potentially could grow to $200 million annually or more if proven successful.

"Performance along this dimension is easy to measure and can be closely linked to an easily stated national policy goal," Romer points out. "The United States should lead the world in the fraction of 24-year-olds who receive science and engineering degrees. Unfortunately, by this measure, we now lag far behind many other nations."

Effective incentives that will supply the business sector with more scientists and engineers at the postgraduate level will have to take a different form, says Romer.

"The challenge is not to increase the numbers of Ph.D. degree recipients," says Romer, pointing to recent growth in the total number of Ph.D.s trained each year, "but to increase the fraction of them that can put their skills to work in private-sector research and development." His proposal would make it easier for graduate students to shift away from the traditional career track in academia to one that prepares them for innovation in the business sector.

Without reducing current levels of government funding for research assistant positions, Romer advocates a dramatic increase in the number of "portable" fellowships that the government awards directly to scholars who show promise. Such fellowships might pay $20,000 per year for three years of graduate training in natural science and engineering, and would be portable in the sense that they could be applied toward any field at any institution the student chooses.

"Students could vote with their feet by selecting disciplines at universities that offer them the best training for the careers that most of them will follow, working outside of academia," he explains. Universities that develop innovative new training programs would be rewarded with an influx of students and tuition dollars.

At a magnitude of 50,000 fellowships annually, the $1 billion price tag would make this a major new initiative in science and economic policy. But as Romer points out, the government already spends more than this on direct grants to firms. These grants are intended to increase innovation in the business sector, but unless universities train more scientists and engineers, they will go to waste. "Ultimately, it does no good to subsidize business-sector demand for highly skilled workers if the university system cannot respond with an increased supply."