Shirley M. Tilghman
Recalling the meeting, Tilghman notes that the cohort of doctoral wanna-bes has never topped 10 percent of graduates. But she describes this year's yield as the worst ever. "It worries me because the future of science needs these kids opting to do science," she says. "And they're not opting to."
A new 120-page report from the National Research Council (NRC) helps explain why. "Addressing the Nation's Changing Needs for Biomedical and Behavioral Scientists" (www.nap.edu/books/0309069815/html) is the 11th in a series of reports since 1975 mandated by the National Research Service Award (NRSA) Act. Rife with graphs, tables, and demographics, this "National Needs" report differs from earlier ones by examining the entire workforce in the targeted sciences, not just NRSA trainees.
Drafted by an 11-member committee drawn largely from academia, the report discusses biomedical, clinical, and behavioral research, as well as inadequate minority representation in those fields. Its starkest conclusion about biomedical science is that the number of new Ph.D.s awarded annually "is well above that needed to keep pace with growth in the U.S. economy and to replace those leaving the workforce as a result of retirement and death."
The report notes that biomedical Ph.D. production swelled over the past quarter-century as the bulk of funding shifted from training grants to research grants. It recommends that "research training and overall Ph.D. production in these fields should not be increased." No particular strategy is advocated to achieve that goal.
Why didn't the committee call for a decrease in the number of Ph.D.s awarded? Chairman Howard Hiatt, a professor of medicine at Harvard Medical School, explains that "to change suddenly the numbers of people could be very disruptive to the research that's going on at the present time"--research that he stresses is "extraordinarily effective." Insisting that "the notion of keeping things constant is, in itself, a major step," he observes that a future committee could assess the outcome of a no-growth strategy and, if warranted, propose a decline in Ph.D. production as a further step.
As another observer points out, NRC reports are political documents whose messages are carefully calibrated to be taken seriously. If the committee had advocated a cutback, this observer continues, "All hell would have broken loose," even though the proposal may have lacked any practical means of enforcement.
Reports Have Impact
The findings and advice presented in the National Needs report should induce a sense of déjà vu. Two years ago, an NRC-sponsored committee chaired by Tilghman also depicted a Ph.D. glut in the life sciences and called for restrained growth of the graduate student population.1 And in 1995, a paper by William F. Massy, now professor emeritus of business administration and education at Stanford University, and Charles A. Goldman, a senior economist at RAND in Santa Monica, Calif., shoehorned science training into a theoretical framework consistent with the new report.2
Massy and Goldman argued that doctoral enrollment was driven more by the need for research and teaching assistants than by the labor market. The resulting Ph.D. glut, they explained, had led to chronic underemployment and deteriorating career attractiveness, particularly to American students. Though the paper encountered some hostility at first, Goldman points to "an accumulation of corroborating evidence and perspectives in the last five years."
Although these reports and papers may seem ineffectual--as well as obvious--to many biomedical trainees, their authors do perceive the work as having an impact. "I think there's a lot of interest in and concern about these issues" at the National Institutes of Health, says Michael S. Teitelbaum, a demographer at the Alfred P. Sloan Foundation in New York and a member of the panel that drafted the National Needs report. Tilghman notes that unpublicized proposals "are wending their way slowly through the NIH machinery."
Both the 1998 and the new NRC reports mention gradual increases in the duration of graduate and postdoctoral training. Tilghman points to a recent initiative to treat this "symptom of a system that is broken": In 1998, Cold Spring Harbor Laboratory established a small graduate program that expects its students to earn Ph.D.s in four and a half years. If the program succeeds, she hopes it will force other elite programs that compete for the same pool of applicants to speed up training as well.
According to Goldman, the NRC studies and his and Massy's paper have had two other notable effects: Graduate departments are increasingly exposing students to teaching skills, which are in greater demand than research skills. Departments are also much more conscious that "large numbers of their students will not be going into academic careers."
In the last few years, that realization has apparently taken hold among students and postdocs. "Now I go to Career Day meetings at universities, and people are admitting that they got into their Ph.D. program in order to get a job in industry," says David G. Jensen, managing director of Search Masters International, a Sedona, Ariz., recruitment firm for the biotech and pharmaceutical industries. Stephen Rosen, chairman of Celia Paul Associates in New York, notes that most scientists who consult him about career transitions "are trying to move out of the academic environment into an industrial environment," even if that means changing fields.
The National Needs report asserts that growth in industrial employment slowed in the mid-1990s, with the percentage of biomedical scientists working in industry slightly lower in 1997 (23.9) than in 1993 (25.1). But for now, Jensen sees a swelling demand from companies, which keeps recruiters "very, very busy." He acknowledges, however, that the job market still isn't hot enough to absorb all Ph.D. recipients, except for those in a handful of specialties.
Postdocs and Post-Postdocs
The National Needs report documents a drop in the fraction of new Ph.D. recipients planning postdoctoral study, from 73 percent in 1996 to 65.1 percent in 1997, the lowest such figure since 1977. But it's too early to tell if this downturn signals a trend or is merely a historical blip. The preliminary figure for 1998, according to NRC project officer James A. Voytuk, is 67 percent.
"Our initial reaction was that the figure for 1997 was a reflection of improvements in the economy," says Jennifer Sutton, NRC study director for the National Needs report who has since moved to the National Cancer Institute. Instead of planning to do a postdoc, "[Graduates] were more likely to find other types of more permanent or more attractive jobs." Graduates were never interviewed, however; they merely indicated (on a survey form) their plans rather than their subsequent actions, and Teitelbaum describes the numbers as "really hard to interpret."
For Ph.D. recipients who carry through on their intentions to become postdocs, the report suggests their plight better than it does their options. It briefly mentions a 1998 paper by Elizabeth Marincola, executive director of the American Society for Cell Biology, and Frank Solomon, a biology professor at Massachusetts Institute of Technology, that advised "the creation of respectable, reasonably paid professional scientist positions, to be held by fully trained researchers who neither write grants nor train others."3
"People genuinely want to do labwork at the bench as their career," Marincola remarks now. "But we've not yet made a place for them." While acknowledging that a few such positions already exist, she adds, "The vast majority lack the stability that we would like to see them have. And I think it's a critical mass issue, because unless jobs are available on a fairly wide scale, then they stand a good chance of being marginalized."
A case to consider are the hundreds of nonpostdoc, non-tenure-track staff scientists employed by NIH's intramural program. Generally appointed five years or more after they have received their doctorates, they lack independent resources and are supervised (usually one per lab) by senior investigators. Their appointments can last as long as five years, and annual salaries start at about $55,000. Noting that many staff scientists' appointments have been renewed, Michael M. Gottesman, deputy director for intramural research, asserts that he "can state unequivocally that this new professional designation has been a big success at the NIH" since it was introduced six years ago.
More Studies--and More Money?
According to Hiatt, NRC plans to send the National Needs report to the NIH director, who will forward it to the Secretary of Health and Human Services. Hiatt is set to discuss the report with the director of the National Institute of General Medical Sciences. An NIH spokesman says the report will be sent to congressmen who request it.
NRC issued the report six years after the last National Needs report and almost two years late. More frequent publication, however, may be in the offing. NIH research training officer Walter T. Schaffer says that the National Academy of Sciences (of which NRC is an operating agency) has proposed providing interim reports to NIH. These would appear every two years following release of the demographic data on which the National Needs reports are based.
Marincola and Solomon, meanwhile, are wrapping up a second phase of their work, in which they're collaborating with Richard B. Freeman and Eric R. Weinstein of Harvard and the National Bureau of Economic Research, a think tank in Cambridge, Mass. (Freeman was also on the committee that drafted the new National Needs report.) Focusing on 25 prominent cell and molecular biology labs in the United States, the group is asking all the principal investigators and many of the postdocs and graduate students detailed questions about their productivity and career choices. The study, which the group hopes to publish by year's end, examines subjective judgments of productivity rather than objective measures such as numbers of papers published.
Goldman says he and RAND colleague Valerie Williams may do a study pursuing topics in the National Needs report in greater depth. One idea meriting further analysis, he adds, is "increasing the [NIH] stipend in good times so that when funding is increasing, the number of positions that are created does not expand as fast as the funding. Then when funding is leaner, keeping the rate of stipend growth low so that the number of positions can be preserved until the next funding increase."
In a similar vein, the National Needs report recommends regular cost-of-living increases in stipends and other forms of trainee compensation. It advocates that such increases "be incorporated into budget planning, so that stipends are not again allowed to decline in real value." Recalls committee chairman Hiatt: "We urged these changes because they seemed so obvious."
Douglas Steinberg is a freelance writer in New York.
1. P. Smaglik, E. Russo, "NRC report: cap life sciences graduate school enrollment," The Scientist, 12:6, Sept. 28, 1998. The report is at the Web address books.nap.edu/books/0309061806/html/index.html.
2. W.F. Massy, C.A. Goldman, "The production and utilization of science and engineering doctorates in the United States," Stanford Institute for Higher Education Research Discussion Paper, 1995. This paper is soon due to come out as a book, The Ph.D. Factory (Bolton, Mass., Anker Publishing Co., 2000).
3. E. Marincola, F. Solomon, "The career structure in biomedical research: implications for training and trainees," Molecular Biology of the Cell, 9:3003-6, 1998.
Workforce Projections and Directions
The "National Needs" report just issued by the National Research Council does far more than acknowledge the Ph.D. glut and recommend holding growth constant. At the heart of the report is "Appendix D," which presents numerous demographic forecasts:
* In a high-growth scenario, the report predicts that the supply of biomedical workers will jump by 4.4 percent annually from 1995 to 2005; a low-growth scenario forecasts a 2.2 percent increase. A projection midway between these extremes has the biomedical Ph.D. workforce expanding from 91,440 in 1995 to 128,511 in 2005.
* In 1997, the U.S. Bureau of Labor Statistics (BLS) forecast a 2.1 percent growth rate in demand for researchers in the life sciences as a whole. If supply outstrips demand, as the NRC and BLS figures suggest, the National Needs report contends that the number of students entering these fields would presumably decline--a normal economic response. But the report then notes that the depth and timeliness of that decline are "impossible to predict."
* Forecasts of the future workforce are clouded by poorly charted factors, such as the "stay rates" of foreign graduate students and the behavior of Ph.D. holders who move into and out of science. The number of Ph.D. recipients in a given year is also notoriously hard to predict, even if one knows the number of graduate school enrollees. "When first-year enrollment rose," the report says, "the number of graduates six to nine years later decreased at least as often as it increased." It dryly proposes that the length of graduate education "allows students many opportunities to drop out."
* The report estimates how many recipients of biomedical Ph.D.s would have been needed in 1997 if the biomedical workforce had grown at the same rate as the entire United States labor force: 1,571. The actual number of Ph.D. recipients in 1997 was 5,420. For the workforce to have remained constant, the number of biomedical graduates would have had to plunge from 5,100 in 1995 to 650 in 1996.
* Since 1960, the number of female biomedical graduates has climbed slowly but steadily. Male graduates, while always more numerous, stayed constant from the early 1970s to about 1990, when they began increasing too. From 1995 to 2005, the report projects that the female workforce will grow at 2.5 times the rate of the male workforce. Starting from a smaller base, the population of female biomedical scientists will increase by 19,900, compared to a 17,200 increase in males. Women consequently will go from 28 to 36 percent of the total workforce.
* When the workforce is surveyed at two-year intervals, almost 5 percent of those employed in science at the earlier time point have shifted to jobs outside science, whereas a third of those originally outside science--a much smaller number--have returned to science. The peak ages for leaving science are from 39 to 49 years old. The number of biomedical Ph.D. holders outside science is projected to rise slightly from 9,850 (11 percent of the workforce) in 1995 to 10,199 (8 percent) in 2005.
Postdocs Get a Report of Their Own
Two weeks after releasing the National Needs report, the National Research Council issued a guide titled, "Enhancing the Postdoctoral Experience for Scientists and Engineers" (national-academies.org/postdocs). Produced by the National Academies' Committee on Science, Engineering, and Public Policy (COSEPUP) and supported by the National Institutes of Health and several foundations, the guide has two facets: It depicts the grim situation facing many of the nation's postdocs, whose population is estimated at 52,000. And it suggests possible ways of alleviating their plight.
Some of the guide's portrait is familiar: Current postdocs are older and likely to be foreigners. Their chances of moving into faculty positions have dropped significantly since 1987. Median earnings in 1997 were $30,000. Appointments in the life sciences, where most postdocs work, last longer than those in other fields. Less familiar details emerge from a recent COSEPUP survey of 30 academic and 10 nonacademic institutions: Only 52 percent of academic institutions guaranteed fully paid vacations to postdocs, 45 percent guaranteed them fully paid sick leave, and 17 percent required regular performance evaluations. Postdoc associations didn't exist at 58 percent of all institutions surveyed.
The guide offers plentiful advice on such practical topics as "Is a 'Hot Lab' the Best Lab?" and concludes with 10 pages of sensible recommendations. For example, it advocates appointment letters containing contractual terms and signed by postdocs, their advisers, and institutional representatives. Many passages establish that COSEPUP favors more generous postdoc salaries, but the guide does not go so far as to supply tables detailing higher pay scales. And though it recommends that institutions "not encourage unlimited growth in the postdoctoral ... population," the guide doesn't explicitly call for a freeze on growth, much less a contraction of that population.
Would the recommendations, if followed, improve the lives of postdocs, given other structural problems plaguing the life-sciences profession? Yes, says COSEPUP member Irving L. Weissman, a professor of cancer biology and pathology at Stanford University School of Medicine. He acknowledges that "it's a hard system we have," particularly for junior faculty, whose insecurity may tempt them to drive postdocs mercilessly even as the postdocs' own insecurity prompts merciless self-treatment. But that issue, he maintains, is separate from the report's goal of getting institutions "to look clearly at what postdocs are to them" so that the institutions will establish appropriate rules and pay postdocs as much as $10,000 more per year.
"We were addressing some of the immediate issues" of postdocs," agrees fellow COSEPUP member Brigid L.M. Hogan, a professor of cell biology at Vanderbilt University School of Medicine. Her early hope is that at institutions without postdoc associations, postdocs will "start shaking the guide in the faces of administrators and saying, 'We want one of those [associations].'"
According to associate director Deborah D. Stine, COSEPUP will present the guide at meetings of scientific and academic groups. It also plans to hold a convocation in Washington, D.C., on March 2, 2001, to discuss the guide's recommendations. All interested parties, from postdocs to presidents of institutions, are invited.