HALF EMPTY, HALF FULL: HOW BAD ARE JOB PROSPECTS?
by Rebecca Kolberg

Astonishingly enough, this quote did not come from a postdoc who has been searching for employment with no luck for the past two years, on the verge of eviction and having her car repossessed. It was actually made by a former NIH postdoc who recently found a permanent, science-related job in another government agency and whose spouse, also a former NIH postdoc, has secured a research job in the private sector. They didn't even have to relocate.

Figure 1. Underemployment rates of biomedical and physical sciences PH.D.s 1973-1991
(Courtesy: National Academy of Sciences)

What the statement does reflect is the fundamental clash between reality and perception that occurs when many NIH postdocs discover that, despite years of diligent training and research, they may not get their dream jobs. Are there really no jobs out there for biomedical Ph.D.s? Or is the problem really one of perception, that is, "there are no good jobs out there," such as tenure-track positions at Harvard or UCSF?

To address such questions, it would help to consider some quantitative measures of the biomedical employment situation. Unfortunately, hard statistics on the fate of former NIH postdocs do not exist. Although it is hoped that such information will be included in an NIH-wide database now in the planning stages, the intramural program currently keeps no statistics on what sort of positions its approximately 3,000 postdocs and 225 clinical associates land when they leave NIH. "It is important to make available the resources to track fellows," says Michael Fordis, director of NIH's Office of Education. "Such information is vital to determining the outcome of the training you are providing."

However, the general employment outlook for America's next generation of Ph.D. and M.D. researchers has been touched upon in several recent studies, including the National Academy of Sciences' (NAS's) June 1994 report, "Meeting the Nation's Needs for Biomedical and Behavioral Scientists," the Committee on Science, Engineering and Public Policy's (COSEPUP's) April 1995 report, "Reshaping the Graduate Education of Scientists and Engineers"; and Stanford University's August 1995 report, "The Production and Utilization of Science and Engineering Doctorates in the United States."

Figure 2. Fraction of biomedical science Ph.D.s at career age
4-5 years on postdoctoral appointments, 1973-1991.
(Courtesy: National Academy of Sciences)

At first glance, the picture appears bleak. The Stanford report estimates that about 28% more bioscience Ph.D.s are currently being produced than can find employment in academia, industry, or government over the long term. According to the COSEPUP analysis, less than one-third of people awarded Ph.D.s in science and engineering from 1983 to 1986 were in a tenure-track position or had gained tenure by 1991. However, as the NAS findings illustrate, not all Ph.D.s are the same, and lumping biomedical Ph.D.s in with Ph.D.s from other fields such as mathematics, physics, and whole-organism biology may lead to an unduly grim career outlook.

According to the NAS report, the U.S. basic biomedical work force expanded dramatically in the 1980s, soaring from 64,000 Ph.D.s in 1981 to 92,000 Ph.D.s in 1991óa 44% increase that was double the rate of employment growth in other sciences and four times as great as the rate of growth of the entire U.S. work force. The NAS panel concludes that "expanding opportunities in health research" was the prime engine behind this "robust growth," which was accompanied by a 10% increase in the annual number of biomedical Ph.D.s awarded to U.S. citizens and permanent residents. The Stanford report, on the other hand, says that the production of science and engineering Ph.D.s may have far more to do with academia's need for teaching and research assistants than it does with the job market.

Despite the spurt in Ph.D. production, the NAS study shows the unemployment rate for biomedical Ph.D.s never surpassed 1.6% from 1973 to 1991ócompared with a rate of 4.9% to 6.7% for the entire U.S. workforce during the same time frame. In a perhaps more relevant analysis, the report finds that the past two decades have seen a slight, but steady, increase in the percentage of biomedical Ph.D.s who are underemployed, meaning they are working part-time when they would prefer full-time jobs or they are working at jobs that do not fully utilize their science skills. About 0.8% of biomedical Ph.D.s were underemployed in 1975 compared with about 1.4% in 1991 (see Fig. 1).

The number of new Ph.D.s with postdoctoral positions is sometimes considered to be a reflection of employment conditions, with postdoc appointments rising when the job market is tight. However, the NAS panel concludes that the dramatic rise in postdocs in the 1970s, followed by erratic swings in the 1980s, may instead reflect the availability of funding for postdocs, mirroring the increasing availability of funds in the 1970s followed by periods of constraint and relaxation in the 1980s (see Fig. 2).

Figure 3. Employment sector of the U.S. biomedical science work force, 1981-1991.
(Courtesy: National Academy of Sciences)

Although figures from the past decade are reassuring, the NAS committee warns that the biomedical employment boom is rapidly trailing off and the job outlook may be far less rosy for newly minted biomedical Ph.D.s than it was for even their recent predecessors. Confirming the suspicions of many NIH postdocs, the NAS analysis found that employment opportunities for basic biomedical Ph.D.s have grown more slowly in the academic sector than in "nontraditional settings"-with only about half of the biomedical science work force currently employed in academia compared with two-thirds in 1981. Although industry has picked up much of that slack, with almost 30% of biomedical Ph.D.s now working in industry compared with less than 17% in 1981 (see Fig. 3), the report says, "The best predictions for economic activity and R&D funding in the near future suggest that demand for basic biomedical scientists will grow slowly at best." Unless there is a sudden upswing in demand, the current rate of entry of Ph.D.s into the biomedical science work force-about 3,400 in 1990-should prove adequate, the panel concludes.

In its examination of clinical research, the NAS report reaches no strong conclusions on the job outlook for what it calls "physician scientists." The panel notes that total budgeted medical school faculty vacancies have grown at an average yearly rate of about 6% since 1989, with most of that growth coming in the clinical science departments as medical schools have grown increasingly dependent upon clinical income to support their activities. In fact, between 1981 and 1991, the number of full-time faculty employed in clinical departments mushroomed from about 38,000 to more than 59,000-an annual growth rate of nearly 5%. On the other hand, the health-care market's increasing emphasis on cost containment may place academic medical centers at a disadvantage in obtaining clinical income, possibly cutting into support for clinical investigators. Furthermore, the panel observes, the pressure to decrease the proportion of specialists in medicine may dampen enthusiasm for clinical research spending.

From his vantage point at the Office of Education, what does Fordis see in his crystal ball for the hundreds of postdocs and clinical associates now toiling at NIH? "I'm cautiously optimistic about employment prospects in industry, but I see a very guarded outlook in academia. The role of tenure is being reevaluated, and restructuring related to managed care is adversely affecting academic health-care centers across the nation. However, there will always be opportunities for people doing outstanding science. The job of all of us is to make certain that we are providing fellows with the best possible training and mentorship in laboratories doing outstanding science."