Last month, Wendy Baldwin, associate director for extramural research for the National Institutes of Health (NIH), released statistics that show a 20-year decline in the number of NIH research grants awarded to young scientists. In 2001, NIH gave out 6635 "competing" research grants; of those, only 251--3.8%--went to scientists age 35 or younger. This compares to 22% in 1980. During the same period, the number of NIH awards going to older scientists increased dramatically; scientists over 55 more than doubled their slice of the pie. The portion of awards to scientists over 46 nearly tripled.
The data, which were reported by Science magazine last week in an article that was reposted on Next Wave, may exaggerate the problem. In a subsequent telephone interview with Next Wave, NIH statistician Bob Moore, who produced the statistics, pointed out they count only PIs, not co-PIs. NIH grants can have only one PI, and in collaborative work this is often the collaboration's most senior participant. Moore speculates that the decline in the number of young PIs is likely offset by an increase in the number of younger co-PIs on multi-investigator grants. Moore acknowledges that NIH has no good data on co-PIs; still, the decline in the number of under-35 PIs may reflect, in part, the increasingly collaborative nature of biomedical science research. Still, the data remain, in the words of Princeton University President Shirley Tilghman (quoted in Science), "appalling."
The main reasons for the decline are complex and unclear. For years, some young scientists have claimed discrimination, but Tilghman, a long-time champion of the interests of young scientists, says that there's no evidence to support that claim. When a National Research Council panel headed by Tilghman looked into it in the early and mid 1990s, Tilghman says, "We could find no data at all [supporting the idea] that young people are being discriminated against."
If discrimination isn't the reason for the decline, what is? Everyone agrees that one major factor is the increasing age at which scientists get their first tenure-track job. Although some scientists apply for research grants while they are still postdocs and research associates, most institutions don't allow this; the great majority of proposals are written by faculty members.
As the GrantDoctor pointed out in last week's column, it's not that under-35 scientists aren't writing proposals; it's that the proposals they are writing are being submitted by their postdoctoral supervisors.
NIH doesn't have good data on the age of first tenure-track appointment, but Martin Ionescu-Pioggia, program officer for the Burroughs Wellcome Fund (BWF; a sponsor of this publication) offers a snapshot of the current situation. According to Ionescu-Pioggia, since 1995, the first year of BWF's Career Awards program, the average BWF award winner has commenced his or her tenure-track appointment at age 34.7. When you consider that BWF fellows (with $500,000 in hand) are likely to have an easier time than most getting a faculty job, it isn't surprising that rather few scientists under the age of 35 win the NIH sweepstakes. Fewer young scientists are playing.
So why are scientists taking longer to reach the tenure track? Baldwin told Science that it's because they have more to learn. Science, she points out, has become more complex. But Princeton's Tilghman and others find a more sinister explanation; they claim that young scientists are getting stuck in the pipeline because their elders--their more experienced colleagues--routinely take advantage of an abundance of skilled labor by hiring them on the cheap, with NIH's implicit blessing. NIH, Tilghman says, hasn't done enough to control the practice. "I think there's been a real failure of leadership at NIH," Tilghman said. Baldwin, though, defends NIH. "That is something that universities have to deal with," she said in the Science interview. "We don't have any control over it."
But employment decisions aren't the only factor. Young investigators lag their senior colleagues in part because they have a harder time getting funded. The funding rate for first-time applicants (22.1%) is considerably smaller than the rate for the whole pool (31% -- click here for a clarification of these statistics). There's no need to invoke discrimination to explain the discrepancy: More experienced proposal writers should be expected to write more effective proposals. But, while they may be inexperienced proposal writers, today's first-timers have been doing science far longer than their more experienced colleagues at the same point in their careers. There is no reason at all to think that they will perform less well than their senior colleagues. Why, then, should they be less successful? Simply by changing its reviewer's guidelines, NIH could rapidly reduce the age discrepancy.
Funding rates for new investigators are low in part because NIH discontinued the R29 program. Compared with R01 grants, FIRST awards were small, but new investigators did about 10% better in getting them. In 1998 the funding rate was about 32%, about the same as the R01 rate for previously funded investigators. In 1997 an NIH committee considered (among other things) whether the R29 program was doing what it was designed to do. They concluded that it wasn't, mainly because the benefit of the award for new investigators (higher funding rates) didn't offset the awards disadvantages (in particular, its small size and the fact that R29 winners didn't compete as well for R01s as NIH had hoped).
NIH might have concluded instead that R29s were too small to get young researchers off to a good start and tried to fix the problem by increasing the size of the awards. Or they might have concluded that reviewers did not give R29s sufficient weight in future funding decisions (in part, perhaps, as a result of the awards' small size). Rather than trying to fix the program, though, NIH decided to scrap it, and now first-time applicants must compete head to head, possibly on unfavorable terms, with their more experienced colleagues, not to mention their former supervisors.
Baldwin's comments notwithstanding, NIH does have a role to play in employment decisions. NIH can do more to help the most promising young scientists get tenure-track jobs when they're younger. These days, there's no credential that's quite as impressive as money in your pocket--preferably money from an NIH grant. Indeed, these days at some institutions an NIH research grant is a virtual requirement for a tenure-track job. If NIH made more awards to young scientists, those young scientists would get hired sooner. If NIH made enough of these awards, they could have a meaningful impact on the grim statistics.
NIH's K-series "Career Development" awards offer a great illustration of NIH's missed opportunities. One category of K-series awards, the K22 "Career Transition" awards, work much the same way as BWF's Career Awards; they are intended to assist young scientists in the transition from postdoc to the tenure track. In principle, such an award can make a young scientist--even a scientist under the age of 35--very attractive to potential employers.
Unfortunately, NIH doesn't make very many of these awards, and the ones they make are rather small. In 2001, only 52 K22s were awarded across all NIH institutes, with an average value of $136,000 per award. Although the terms of the award vary among institutes, these awards typically last 2 years and are not renewable. Compare this to the K02 awards, which are intended for already funded "independent investigators." These awards typically last for 5 years, and in 2001 NIH awarded 65 of them. Another 15 (K05) awards were made to "senior investigators"; these, too, typically last 5 years. These awards are also fairly small, but their longer term makes them more attractive. The K22 program should be expanded, perhaps by reducing or eliminating the "career development" awards for senior scientists--who, it would seem, are doing just fine without these programs.
The aging of young investigators is a problem of great importance, both for science and for scientists. It's a cliché, but one that is supported by ample evidence: many scientists are at their most creative when they are young. But instead of being brilliant, many young scientists are spending many of their most creative years working as glorified lab technicians, following their directors' instructions. But the key issue here is one of balance: Diversity is good, and age is a legitimate category to consider. Just as we should have more women doing more science, and more underrepresented minorities, we also should have more young people. Who knows how different biomedical science would be if some of it were being done by younger scientists?
But there is one concern that, to me, is at least as serious. This week's Career Development Center feature article discusses the effects of stress on young scientists, especially women. It seems likely that the same young scientists who are working hard to cure diseases and mitigate human suffering are themselves suffering stress-related health problems. As anyone who has spent much time around young scientists knows, one of the greatest sources of stress is the pursuit of tenure and, by extension, of research grants. Many young biomedical scientists don't begin to feel relaxed about tenure until they at least win their first NIH research grant; it is, at a minimum, a key milestone. The fact that young scientists are getting older means that more of their lives have passed before they can set aside the fear of being fired and focus entirely on their lives and the work they care passionately about.
[ This figure--31%--includes competitive renewals, which in 2001 were awarded at a 53% clip; the 2001 funding rate for all new applications was 27%--just a few percent higher than the rate for new investigators.]