"I want our postdocs to know that we as a nation appreciate what they do," Rita Colwell, director of the National Science Foundation (NSF), told Science's Next Wave in January 2001. "We need to reward those who are in the training phase," she added. To date, however, the early-career scientists who form what she termed "the lifeblood of the nation" have yet to see that appreciation take on the tangible form of higher pay.

But NSF-supported postdocs may get bigger paychecks starting in federal fiscal year 2005, which corresponds roughly to the 2004-2005 academic year. The 2005 NSF budget request to Congress, now in preparation, is likely to propose a sharp rise in postdoc stipends, Colwell told Next Wave in a recent interview. The base stipend for NSF postdoctoral fellowships now stands at $33,000 per year, but "we are arguing strenuously for [a] $45,000 minimal postdoc salary," Colwell said.

In addition to fellowships that pay the base rate, a variety of special early-career fellowship programs provide stipends that can be as much as $54,000 a year. Colwell hopes that these levels will also rise, but exactly when and by how much has not been determined. "We are in the process of working on the 05 budget and that will be one of the many topics that we will be discussing," she explained. "I think right now just trying to establish a [higher] minimum will be a huge step forward."

If Colwell's hopes for the 2005 NSF budget come true, NSF fellows will also get more generous research allowances. "Right now ... some of our postdoc fellowships provide no research allowances," she said. For those that do, allowances "can range from $4000 to $10,000." Colwell emphasized that "we want them large enough that the health benefits or fringe benefits can be paid from them," a point particularly crucial for fellows working at institutions that do not provide health insurance or other benefits to postdocs not considered university employees.

If NSF's base postdoc stipend remains unchanged until FY 2005, however, the pay of NSF graduate fellows will be bumping up against it. NSF graduate fellowships now pay $27,500 a year, and the 2004 budget request seeks a raise to $30,000. Asked about the equity of so small a pay differential, Colwell said, "we will try to do something [for postdocs] in the 04 budget if we can."

A higher base fellowship stipend, however, will affect only 5% of the 5600 postdocs supported by NSF. The remaining 95% are paid out of the research grants awarded to principal investigators. "We do not require" PIs to pay these postdocs at the fellowship stipend rate, Colwell says, "but they must in their grant instruments indicate what they will pay." The hope is that the NSF base stipend serves as "a floor" for postdoc pay, says Alan Lightbourne, an NSF Special Advisor, but the actual salaries vary among institutions and disciplines. A survey published in 2000 by the National Academy of Sciences found that 45% of institutions have no salary minimums or maximums for their postdocs, with individual postdocs' "experience/potential" and the "norms of the field" determining pay.

Still, Colwell anticipates a rise in postdoc salaries paid out of grants because she foresees those grants become larger and longer. "One of our major priorities will be [increasing] grant size and duration," which will create the "opportunity to increase the stipends," she said. Grants have been growing in both dimensions in recent years, confirms NSF Public Affairs Specialist William Noxon. The average size was $113,600 in FY 2001 and $115,660 in FY 2002, the second figure representing a 22% increase over 1998 values. The average length of award has traditionally been 3 years, but NSF is now "concentrating on trying to get to four," Noxon said. The goal for FY 2004 is an average grant of $150,000 that lasts 4 years. NSF sees larger and longer grants as more "efficient" because they permit better planning and less time writing proposals, Noxon said.

"We feel strongly that if we can increase grant size ... principal investigators will spend the money on graduate students, undergraduate students, and postdoctoral fellows," Colwell continued. "We did a very extensive study of about 6000 PIs and we got a 94% return, and [for a study] of 300 grants offices at universities, we got about a 90% return. ... The message was clear and unequivocal: If [PIs] have more money, longer time, they would spend it on students and postdocs."

Despite higher postdoc pay, Colwell does not foresee a cut in the numbers funded. "We feel strongly that we ought to maintain the number," she said. Some commentators argue that many postdocs' chronic difficulty finding permanent academic positions indicates that the nation's need to train new scientists and engineers has been exaggerated. Colwell, however, called this a "unidimensional perspective" because "there's a whole lot more to this than numbers."

"Scientific and technical training isn't just a pipeline to professorships in universities," she asserted. The security threats now facing the nation, for example, call for more technically trained persons with U.S. citizenship. "The Department of Defense and the intelligence agencies need highly technical and skilled individuals" eligible for high-level security clearance, says Colwell.

Even the best-laid plans of federal funding agencies must pass congressional muster, however. So, can the increase needed to boost postdoc pay survive in today's tight budgetary environment? "I'm fairly sanguine," Colwell said. "I do think that it's understood [in Congress] how important it is to retain scientists and engineers and to reward them in a way that they will be attracted to and stay in science and engineering. I think that we will come out of the 04 budget well. And I think that in the 05 budget we will be able to justify an increase."

And will Congress agree that additional money should go to postdocs? "I think that it is very, very likely because we can justify [the need for a raise] just as we did for the graduate stipend. Congress, both House and Senate, really resonated to it," she said. "They got it without a whole lot of hard work and explaining. And I think that it will be well understood that we do have a workforce need. ... So I do think that they will respond favorably."

Beryl Lieff Benderly writes from Washington, D.C.