Report: Young scientists need more support

A blue-ribbon committee of scientists and science policy experts has urged federal agencies, universities, and industry to step up their support for the research projects of early-career scientists and engineers and for high-risk, “potentially transformative” research. The panel also broke new ground in calling for research universities to pay a greater portion of their faculties’ salaries.

The committee of the American Academy of Arts and Sciences says that early-career scientists are being squeezed out by their senior colleagues in the competition for limited federal research funding. Sponsoring agencies should establish new targeted programs and grant mechanisms or adapt existing grant programs to foster potentially transformative research, which could produce breakthroughs but also carry a high likelihood of failure. Proposals for such research are typically rejected by the traditional peer-review process in favor of less risky applications.

Other reports in recent years have made similar recommendations, and NSF and the National Institutes of Health have instituted new grant mechanisms that begin to address the issues. But the “boldest and most controversial” of the committee’s recommendations, said committee chair Thomas Cech, is its call for universities to pay a greater portion of their faculties’ salaries. Currently, up to 100% of a faculty member’s salary can be charged to a research grant, depending on the field and how much time that individual spends on the research. And universities typically require faculty members to obtain grants that cover at least their summer salaries. Salary support for established researchers reduces the number of new projects federal agencies can fund.

“A number of university leaders have called this naive, impractical, or were shaking their heads when we proposed this,” admitted Cech, a Nobel laureate and president of the Howard Hughes Medical Institute. “It came out of the realization that many of the problems encountered by the early-career investigators and the lack of support for potentially transformative research were directly tied to the fact that salaries had to be obtained, 100% in many cases, from federal research grants.”

A related problem has been the tendency for universities to build new research facilities without consideration for the operating costs. Institutions are allowed to recover their construction costs through the indirect costs they tack onto the direct costs of individual research grants. But much of the construction in recent years was based on unrealistic expectations about a continued expansion of the federal research enterprise, particularly at NIH.

The committee urged NSF, NIH, the Department of Energy, and other federal granting agencies to adopt career-stage-appropriate expectations when conducting merit reviews of their grantees. Universities should strengthen or develop mentoring programs to encourage early-career faculty, reconsider promotion and tenure policies for young investigators, and address the needs of scientists who are primary caregivers.

New investigators face increasingly stiff competition, especially at NIH, where the average age of a first-grant recipient climbed from 39 in 1998 to 42.4 in 2006. At NSF, success rates for first-time grant applicants plummeted from 22% in 2000 to 15% in 2006, and the average time from receiving a degree to a first NSF award has risen from 8.5 years in 1990 to 9.3 years in 2006.

Transformative research, the sort that the committee said led to the invention of the transistor and the discovery of angiogenesis, has been stifled by an increasingly conservative peer-review system, by the need for investigators to demonstrate publishable progress in order to keep new awards coming, and by other factors.

Panel member and former NSF director Neal Lane said the agency’s program officers need to “get out in the community”—attend scientific gatherings and make site visits as they once did regularly. “The decisions these program officers are making are critical to the future of science, engineering, and medicine in this country,” he said, and jobs need to be attractive to the best people. Lane, who also was science adviser to President Bill Clinton, said agencies need to collect more data about their funding activities so that comparisons can be made among programs. “At NSF half the people who get a new grant don’t show up again. So what happens to them? Perhaps they’re getting funding from other agencies. We know anecdotally that many of them do,” Lane said. “But if we believe that the future of the science and engineering enterprise depends on the career development of these young people, we need to know who they are, how they’re doing, and how the system that’s supporting them is doing.” The report is available at http://www.amacad.org/ARISE .