*The headline of this story has been changed, see note at end.
It's an article of faith: the United States needs more native-born students in science and other technical fields. The National Academies' influential Rising Above the Gathering Storm  report in 2006 said the nation should "enlarge the pipeline of students who are prepared to enter college and graduate with a degree in science, engineering, or mathematics" to remain competitive. The U.S. Chamber of Commerce had a similar message  on the gap in so-called STEM (science, technology, engineering, and mathematics) students a year before. President Barack Obama has pushed  for more science teachers and training for the same reason.
But a new paper  contradicts the notion of a shrinking supply of native-born talent in United States. "Those who advocate increasing the supply of STEM talent should cool their ardor a little bit," says one of its authors, B. Lindsay Lowell, a demographer at Georgetown University in Washington, D.C.
The supply has actually remained steady over the past 30 years, the researchers conclude from an analysis of six longitudinal surveys conducted by the U.S. government from 1972 to 2005. However, the highest-performing students in the pipeline are opting out of science and engineering in greater numbers than in the past, suggesting that the threat to American economic competitiveness comes not from inadequate science training in school and college but from a lack incentives that would make science and technology careers attractive.
The researchers—led by Lowell and Harold Salzman, a sociologist at the Urban Institute and Rutgers University, New Brunswick—argue that boosting the STEM pipeline may end up hurting the United States in the long-term.
This happens, they say, by depressing wages in S&T fields and turning potential science and technology innovators into management professionals and hedge fund managers.
The way to promote US competitiveness in STEM fields is to "put more emphasis on the demand side," says Lowell, noting that U.S. colleges and universities produce three times more STEM graduates every year than the number of STEM jobs available. Cranking out even more STEM graduates, he says, does not give corporations any incentive to boost wages for STEM jobs, which would be one way to retain the highest-performing students in STEM.
The surveys analyzed by the researchers tracked students as they graduated from high school and entered college, following up with them 3 and 10 years after they had left college. That gave the researchers a full education and employment profile for three different groups of students: those who graduated college in 1977, 1986 and 1993 respectively.
The researchers found that the percentage of high school students who were enrolled in a STEM program or had earned a STEM degree 5 years after graduation dipped only slightly between 1972 and 2000, from 9.6% to 8.3%. The percentage of those STEM graduates who were working in STEM occupations 3 years after college increased over the period--from 31.5% for the 1977/80 cohort to 45% for the 1997/00 cohort. Similarly, the percentage of STEM graduates who continued to work in STEM occupations 10 years after college rose from 34.8% in the 1977/87 cohort to 43.7% in the 1993/2003 cohort.
That's not the case for the highest performing students, however, as measured by college entrance test scores and college grades. Although the percentage of those in the top quintile who pursued STEM in college climbed from 21% in the 1972/77 cohort to 28.7% in 1992/97, it plunged to 13.8% in 2000/2005. Likewise, the share of the top quintile of STEM graduates still holding STEM jobs 10 years out of college dipped from 44.8% in the 1977/87 cohort to 43.2% in the 1993/03 cohort.
The authors say those findings square with anecdotal evidence of STEM graduates being drawn to careers in management and finance careers starting in the early 1990s. "Maybe the competition rather than being with the East," Salzman says, referring to emerging economic powers like India and China, "is between different sectors of industry; with Wall Street."
The conversation about the STEM gap "hasn't been grounded in a sufficient body of evidence," Salzman says. Michael Teitelbaum of the Sloan Foundation, which funded the study, adds that claims of shortage are "often issued by parties of interest" such as employer associations. In the past, some U.S. businesses have been accused of using the shortage argument to justify outsourcing and hiring of foreign workers.
Susan Traiman of the Business Roundtable criticizes the new study, saying that it gives an illusion of a robust supply because it bundles all STEM fields together. There may be an oversupply in the life sciences and social sciences, she argues, but there is no question that there are shortages in engineering and the physical sciences. The findings "are not going to make us go back and re-examine everything we've been been calling for," she says.
*This story's headline has been changed. The previous headline said that the study argued that the U.S. needs "fewer" science students; the study itself does not argue for fewer science students, though it negates the idea that more science students are needed. Study author B. Lindsay Lowell, in his remarks to ScienceInsider, did suggest, however, that fewer students in the pipeline may help boost salaries and attract higher performing students into STEM jobs.