Postdoctoral fellowships for many life scientists have become marathons that delay their entry to their first “real” jobs by some years. To counter the problem, potential employers in government, industry, and universities have started to set term limits to the postdoctoral experience.

Affymetrix, Inc. ( http://www.affymetrix.com)

Memorial Sloan-Kettering Cancer Center ( http://www.mskcc.org)

National Postdoctoral Association ( http://www.nationalpostdoc.org)

University of California, San Francisco – Physiology Department ( http://www.ucsf.edu/physio)

Employers in industry, government, and academe increasingly require the postdoctoral fellows they recruit to have a broad base of understanding of their fields. In many cases, postdocs can achieve that base only by devoting more time to their initial fellowship or by taking more than one postdoctoral stint. As a result, increasing numbers of postdocs now reach their late 30s before they take their first permanent job. “When I travel around, I’m shocked by the number of 40-year-old and even older assistant professors,” says Michael Stryker, former chair of the physiology department at the University of California, San Francisco (UCSF). “It’s clear in a lot of places that the period of postdoctoral training has become very prolonged.”

Few statistics exist on the ages of postdocs when they gain their first employment. However, says Alyson Reed, executive director of the National Postdoctoral Association (NPA), “The average length of a fellowship seems to be in the neighborhood of four or five years.” Typically, Stryker adds, “the people we hire for faculty positions have had about three years postdoctoral experience. But it ranges from zero in two cases to up to five years in the longest one I can think of. The five-year ones are often more than one postdoc.”

Whatever the exact average, it represents an increase over the roughly three years that today’s full professors took for their fellowships. Of course, science has become more difficult and complicated since then. “But it’s not clear that the complexity of science should make a postdoc longer, because the tools have improved,” says David Scheinberg, head of the molecular pharmacology and chemistry program at Memorial Sloan-Kettering Cancer Center.

Setting Term Limits


Supervisors of postdocs are recognizing the problem of excessively long postdoctoral training, and taking measures to minimize it. “There is definitely a trend toward setting time limits for postdoctoral appointments,” Reed says. “Institutions that have been in the forefront of developing procedures to enhance the postdoctoral training experience are indeed aware of a holding pattern associated with some postdoctoral training programs. The National Institutes of Health has essentially set a term limit of five years for its training fellowships. And at least 22 of 74 academic institutions that have published data have their own term limits.”

Such limits offer benefits to postdocs and potential employers. After spending three or four years on their research, postdocs may fear that they won’t be employable. “If you wait too long, you begin to wonder whether your fellowship is an impediment to hiring,” Scheinberg points out. Stryker agrees. “You can only stay in training positions for so long until you get a formal consideration of career prospects,” he says.

That factor applies particularly to companies that seek to hire postdocs. “There are certainly many more opportunities in biotechnology and pharmaceuticals that are becoming interesting to the best and brightest postdocs, particularly in translation areas such as experimental therapeutics,” says Scheinberg. “Most postdocs say: ‘I need to do a second postdoc to get more published papers, or to stay another year in this postdoc so that I can get a better paper,’” Reed adds. “But if industrial recruiters looking at resumes see someone who’s been a postdoc for more than five years, they begin to wonder what’s wrong.”

Significantly, postdoctoral fellowships in industry almost always take shorter times to complete than those in academe. “Postdoctoral training periods are determined by the time required to complete both the training agreed to by the mentor and postdoctoral student and the research project that is taking place,” says Thomas Gingeras, vice president of biological sciences at Affymetrix. “Experience points to at least a two-year training period.” Scheinberg emphasizes the importance in academia of ongoing evaluation and communication between fellow and mentor about career goals and pathway opportunities.

Positions for Creative People


Employers, meanwhile, risk losing some of the ingenuity of young scientists if they wait too long to hire them. “I regard it as a necessity for us to be able to offer positions to people at the most creative times of their lives, which is shortly after their Ph.D.s,” Stryker says. “The consensus for our search committees is that if someone has done Ph.D. research at the highest level in one lab and has done the same in postdoctoral work in another, they know that a lot of the credit must go to the young person no matter how excellent the reputation of his or her principal investigator. That’s the time of life you want to have that person starting as an assistant professor.” Scheinberg and his colleagues at Sloan-Kettering take a similar approach. “The critical factor in hiring postdocs for faculty positions,” he says, “is the discoveries they have made.”

Observers see several reasons for the extended time that today’s postdocs take to complete their fellowships in life science. In a recent paper, Jennifer Ma of the TIAA-CREF Institute and Paula Stephan of Georgia State University attribute it to the increasing proportion of Ph.D.s being awarded in the life sciences and to adverse job market conditions that life scientists experience during their fellowships. Reed amplifies the latter point. “Where the job market is stronger, you see shorter postdocs or none at all,” she says. “In disciplines where there’s a glut, you see the holding pattern more for postdocs. That’s particularly the case for postdocs looking for tenure-track, jewel-in-the-crown jobs.”

Universities are also setting out to save money – and coincidentally reducing their recruitment of postdocs – by refusing to replace retiring professors in the traditional way. “They fill retired tenure-track positions with adjuncts and other nontenured individuals,” Reed explains. “Postdocs shouldn’t assume that every time a professor retires, it will be an opening for a tenure-track position.”

Grants and Other Concerns
Concern about gaining research grants also plays a role in dissuading some fellows from applying for jobs at early stages in their postdoctoral work. “There is a worry among some postdocs that, given the current contraction of federal funding, they will be launched into careers without the ability to rapidly gain external funding,” Sloan-Kettering’s Scheinberg says. “There is some trepidation about getting out too fast without a substantial body or work.” That fact applies particularly to postdocs in university departments outside the top tier. “At the best institutions we never consider even for a millisecond whether anyone has independent grant support in making an offer for them to become an assistant professor; we are confident that anyone we hire will eventually get support,” UCSF’s Stryker explains. “But in less competitive institutions, it’s a tremendous advantage to a young person to have obtained independent research support. For the institution hiring them, they are a lot cheaper.”


In certain cases, postdocs need extra time to complete their fellowships. “Some of the laboratories most in demand get very talented scientists who apply too late in their graduate careers to get in immediately. So they do something else in the interim,” UCSF’s Stryker says. “Some of the most brilliant ones have done this because they didn’t get their lives organized enough as graduate students to find a place in the lab they most wanted. They come to us after another brief postdoc.”

Another exceptional group needs more basic training. “In some of the most integrated areas of biology, notably systems biology as it evolves, there are people who come to life science laboratories with backgrounds in physics or mathematics or other hard sciences,” Stryker continues. “These people generally take longer for their postdoctoral training; they get the training that a biology Ph.D. would need in addition to the postdoc. For them, the average length of fellowship is probably five years. Some have had training periods as long as seven years.”

Approaches of the Elite


Elite academic institutions have largely led the way in methods of reducing the time taken for postdoctoral research. “UCSF, like many institutions, has rigidly distinguished between research staff and training positions – scientists who have all the perks of employees versus students,” Stryker explains. “That has brought home to the faculty that faculty members have an educational responsibility to their postdoctoral fellows.”

But the change of title from postdoc to research staffer can’t disguise the fact that faculty jobs for postdocs are rare and highly competitive. “If you’re an academic research associate, it may sound like you have a fulfilling position in academia,” Reed says. “But perhaps only 20 percent of those scientists are potential tenure-track faculty in biomedicine.”

Stryker points out the basic fact that supervisors and their postdocs must bear in mind. “There’s educational value in some period of years, such as three years – or five years for physicists and mathematicians – of being in a training position,” he says. “But at some point they’re not training any more; they’re in career paths as independent scientists.”

A former science editor of Newsweek, Peter Gwynne (pgwynne767@aol.com) covers science and technology from his base on Cape Cod, Massachusetts, U.S.A.

DOI: 10.1126/science.opms.r0600001

10.1126/science.opms.r0600001