Postdoctoral fellowships make a lot of sense in theory: They offer recent Ph.D.s, especially those aspiring to careers in academic research, a place to develop professionally and – hopefully -- build a research profile before or while hitting the job market. But too often, these fellowships are underpaid, under-mentored positions where young academics languish during what are potentially their most creative, productive years.
That’s the upshot of a new report from the National Academy of Sciences, National Academy of Engineering and Institute of Medicine, which is highly critical of the structural factors driving the growth of postdoctoral ranks, and which recommends a series of reforms – including a big raise for postdocs working on federally-funding biomedical research. The report focuses on postdocs working in the science, technology, engineering and math (STEM) fields, but also cites the growth of postdocs in the social sciences and humanities.
Postdoc Report on "This Week"
On "This Week," Inside Higher Ed's free weekly news podcast, Gregory A. Petsko, lead author of the National Academies study, will be joined by Kenneth Gibbs, a postdoc, to discuss the findings. Sign up here to be notified of new editions of "This Week."
It’s unclear what, if any, difference the report will make in the long term, since an earlier, 2000 report on postdocs from the National Academics drew many of the same conclusions, but didn’t force systemic change – although more universities have introduced support services for postdocs. The postdoc problem is also highly complex, and its causes extend beyond any one group of actors or institutions. But advocates say the report comes at watershed moment and echoes recent warnings from other prominent scientists that the current postdoc system is unsustainable.
A Significant Shift
“When you talk to people my age, they usually say their postdoctoral years were some of the best of their lives,” said Gregory A. Petsko, lead author of the report and the Arthur J. Mahon Professor of Neurology and Neuroscience at Weill Cornell Medical College. “They had relatively strong academic freedom, and they were working on challenging problems under low pressure -- not having qualifying exams or a thesis to worry about – and they weren’t particularly concerned about what happened to them when they finished. You never expected you wouldn’t be able to find [a job].”
All that’s changed, of course, since the 1970s, when Petsko was a postdoc, he said, “and not in a good way.” The plight of postdocs worsened in the last decade in particular, he said, due in part to the failure of federal science agencies to have budgets large enough to keep up with demand. Recent Ph.D.s also face a much tougher tenure-track academic job market, even in the STEM fields.
“Today’s postdocs are not as happy as we were,” he said. “Their level of anxiety is much greater, and their workload is more burdensome and, and in some cases, it’s really very tough.”
Petsko said his fellow report committee members – mostly biological and physical sciences professors, with a few administrators and social science faculty members from institutions across the country – wanted to see just how much the landscape had changed, and make suggestions about improving it. But they faced what is perhaps the most significant and telling finding of the study: an overall lack of institutional-level data about how many postdocs are working in which fields.
In an “astonishing number of cases,” universities couldn’t even come close to an accurate estimate about how many postdocs they employed, Petsko said. That’s due in part to the fact that postdocs are called different things in different places, he said, but the lack of data added to the committee’s sense that postdocs are the “invisible people on their campuses.”
In the absence of comprehensive, institutional-level data, the committee looked at federally funded surveys of Ph.D.s, including the Survey of Earned Doctorates. Members also talked to postdoctoral researchers, senior officials from the National Science Foundation and National Institutes of Health, and leaders of various research programs and institutions.
The committee found that the number of postdoctoral researchers in science, engineering and health increased 150 percent between 2000 and 2012, “far surpassing” both the percentage increases in graduate students and in tenure and tenure-track faculty positions over the same period. In 2012, some 40 percent of all doctorate recipients said they planned on postdoctoral study; the rate was 50 percent for all life science, physical science, social science and engineering Ph.D.s.
It’s estimated that there are between 60,000 and 100,000 postdoctoral researchers working in various research fields in the U.S. According to 2012 data, most (65 percent) work in the life sciences. Some 13 percent work in the physical sciences and 11 percent work in engineering. Geosciences postdocs make up 3 percent of the population, as do those in math and computer sciences. Psychologists account for 2 percent, as do social scientists. “Other” accounts for the last 1 percent.
Although the number of postdocs in the social sciences is still relatively low compared to STEM fields, the number has “increased a great deal” in the last two decades, said Paula Stephan, a fellow committee member and a professor of economics at Georgia State University -- from 18 percent of new Ph.D.s in 1992 to 38 percent in 2012. Like STEM Ph.D.s, most economists and other social scientist Ph.D.s work exclusively in research, but some humanities postdocs involve teaching, she said.
The overwhelming majority of postdocs work in academe, but approximately 11 percent work in national labs and other federally funded research facilities. Those working outside academe tend to have better salaries, shorter appointments and better chances at long-term employment.
Over all, according to 2006 data, postdocs tend to have two-year appointments but spend a median of three to four years in the position. “It’s not unusual to find biomedical researchers who have completed several postdoctoral appointments that total more than five years,” the report says.
Median pay for recent science, engineering and health doctorate recipients working as postdocs is much lower than median pay for recent Ph.D.s not working as postdocs: $43,000 versus $76,000, according to 2010 data.
The report also says that while the NSF has required postdoctoral researcher mentoring plans since 2009, data suggest that mentoring is uneven and goes largely unevaluated.
Recommendations for Change
Petsko said he was “sympathetic” to principal investigators, especially junior faculty members, who have little time or experience to provide adequate mentoring as they chase grants and try to publish career-making articles in “boutique” journals that often require rigorous revisions. But he said that the dynamic often makes postdocs inexpensive lab workers, as opposed to trainees building their careers. The report recommends stronger mentoring and academic and alternative academic career training for postdocs. It says graduate students also need better guidance on whether seeking a postdoctoral position is a next, logical step, based on their career plans.
That’s reflected in part by how postdocs are funded: While the number of postdocs working in institutional fellowships and traineeships has stayed relatively steady, at about 5,000 since 1980, the number of postdocs funded by federal research grants has risen dramatically, from less than 15,000 in 1980 to nearly 35,000 in 2009.
The report makes other recommendations for change, including a term limit of five years, with “cumulative” research experience. Universities also should be more accurate in how they track and employ postdocs, reserving that title only for people receiving advanced training in research. College and universities should create staff scientist positions for longer-term employees who don’t fit that profile, and compensate them appropriately, the report says.
Regarding compensation for postdocs, the report says current salaries are too low. It recommends a minimum salary of $50,000, saying that the NIH’s National Research Service Award for postdocs of $42,000 has become the “de facto” pay at many institutions across disciplines.
Petsko said it would be great to see how many institutions played “copycat” if NIH raised the award amount to $50,000, and adjusted it each year for inflation.
The report says institutions should make salary data, along with all other data about postdoc employment, publicly available, and that the NSF should serve as their primary collector.
Praise, and Hope (If Not Doubt)
Jessica Polka, a second-year postdoctoral fellow in systems biology at the Harvard Medical School who is involved in postdoctoral outreach in the Boston area, said the report “succinctly and eloquently summarized the feelings of a lot of postdocs,” and that the recommendations addressed “some of the biggest problems facing postdocs today.”
Polka said she particularly supported the report’s recommendation to create staff scientist positions for people who wish to remain in research without pursuing a faculty position. She said the only missing recommendation, in her view, was a discussion of how many postdocs are appropriate for a given field, "since these positions should be tied to the number of relevant jobs rather than to demand for hands in the lab."
Petsko said the the report doesn't recommend quotas for Ph.D. program admissions or postdocs, since, in his view, there's no such thing as "too many" educated scientists. Instead, he said, the report focuses on the idea that labs should only hire as many Ph.D.s as they can properly train -- not simply as many as they can afford.
Overall, Polka said she was optimistic that the report would lead to some reform for postdocs. “Nothing can be done without starting a conversation,” she said, noting that she was happy in her position but knew of other postdocs elsewhere who were not. She said she hoped to be on the research-intensive, tenure-track job market within the next few years, and was working to bolster her research record before that time.
Kiernan Mathews, director of the Collaborative on Academic Careers in Higher Education at Harvard, was equally supportive of the report, but slightly more skeptical that it would lead to long-term change.
“Ultimately, my reaction to the NAS report is ‘Yes!’ But who is going to make all of this happen?” he said via email. “Many of the changes they are advocating are changes in departmental cultures and institutional/market incentives. That kind of change doesn’t just happen because we all agree it should. It takes changing one person at a time, all at the same time, and that’s damn near impossible.”
Still, Mathews said the report touched on an important topic with ramifications for the future of the profession as a whole, namely that the growing expectation for postdoctoral experience in faculty hires is leading to an “older and more financially constrained” academic workforce. That has a disproportionate effect on women, he said, who often put off starting families until they achieve stability in their careers.
Both Petsko and Stephan, the committee members, said real change would require long-term commitment on the part of federal agencies, faculty members, institutions and postdocs themselves. But they said an immediate, helpful first step would be increasing the salary of postdocs.
Addressing additional skepticism surrounding change the report might effect, Stephan said it comes at a time when many scientists feel that their “backs are to the wall,” and the research treadmill system has to slow down. She noted a much-cited article from this year in the Proceedings of the National Academy of Sciences, which raised many of the same concerns discussed in the postdoc report.
Bruce Alberts, former president of the National Academy of Sciences and current Chancellor's Leadership Chair in Biochemistry and Biophysics for Science and Education at the University of California at San Francisco, co-authored that article, called “Rescuing U.S. Biomedical Research From Its Systemic Flaws.” He said he hadn’t read the entire postdoc report, but said it was “in general agreement” with his concerns. He emphasized that if graduate student admissions won’t be reduced, academe needs to “much more transparently report” career outcomes and other data for Ph.D. programs. He also said professors need to do a better job of connecting students and postdocs to those in other careers, suggesting that the NIH might require such activity.
In an emailed statement, NIH's Office for Extramural Research wrote that "NIH believes that postdoc compensation should correspond to their advanced skills and many years of training" and "has taken steps to increase [stipends] to better reflect the level of experience that postdoctorates bring to their positions, and to provide for regular cost of living increases." The office noted that institutions are free to "supplement those levels for trainees and fellows," and set their own postdoc salary levels.
Regarding the report's call for more mentoring of postdocs, the office said that "NIH’s extramural and intramural programs have long recognized the importance of mentorship in research training, and especially given constrained resources NIH supports and expects that faculty will mentor postdocs so that they can set and achieve goals related to their career progress." NIH said it encourages mentorship by "encouraging grantee organizations to develop an institutional policy requiring an [individual development plan] for graduate students and postdocs supported by any NIH grant," for example, not just training grants and fellowships.