I have been encouraging my students to use transitional words in their topic sentences. Lest they think that only writers in the humanities need care about such things, I offer a passage from the President’s Council of Advisors on Science and Technology’s November 2012 report, Transformation and Opportunity: The Future of the U.S. Research Enterprise. Asking for better treatment for early-career researchers, this group of scientists makes its case with such transition words as indeed, likewise, and thus, straight off the University of Wisconsin’s helpful list of Transitional Words and Phrases. (Emphasis added.)
The United States is the world leader in basic research, performed at universities, Federal laboratories, and independent research institutes; led by principal investigators, university faculty, and laboratory staff; and populated by graduate students, postdoctoral researchers, and other research staff. The high standing of U.S. universities, research institutes, and laboratories derives both from the quality of the graduate programs (i.e., the training of students as researchers) and from the quality and volume of research results, measured by publications in leading journals, citations, patents, etc.
Indeed, the research enterprise depends on the relatively inexpensive labor provided by students and post-doctoral scholars. These early career scientists are a crucial part of the U.S. research enterprise. This dependence creates a potential conflict between the need to maintain the size of this early career workforce and the need to advance the careers of the talented researchers who will keep the American research ecosystem the most productive in the world. While compensation should reflect the nature of junior positions as researchers in training, the issues of graduate student and postdoctoral researcher numbers and the duration of doctoral programs and postdoctoral appointments are serious ones that need to reflect realistic expectations for future career opportunities. We need to be training people for science and engineering jobs that are likely to exist, for example in the private sector, and less for jobs that will be more scarce, assuming no significant increase in university tenured faculty.
Universities need to take a stronger role in managing the time that graduate students take to get their degrees. Overly long times to degree are costly, waste precious graduate education resources, and serve as a disincentive for attracting bright minds to STEM fields. Several reports over the last decade have documented the long time interval needed for degree completion. The best data comes from the NSF’s Survey of Earned Doctorates, although the most recent data are from 2003. The median total years spent in graduate school, less reported periods of non-enrollment, was 6.9 for the life sciences, 6.8 for the physical sciences, and 6.9 for engineering. This is too much time. While the last decade has seen some progress on the issue, in many fields, times to degree in absolute terms remain too long. There is no evidence that the quality of the graduate education is higher in those departments in which the time to degree is longer. University administrations need to take a critical look at the time to degree for Ph.Ds. when reviewing a department’s performance.
The growing number of post-doctoral scholars and the length of time spent in postdoctoral positions, most notably in the life sciences, must likewise be confronted. A recent NRC report documents that the number of biomedical Ph.Ds. increased from 5500 per year in the period 1997-2003 to 7100 in 2008, a 30 percent increase in 6 years. Over the same period the number of tenure and tenure-track faculty positions has been constant at about 30,000, and there is no reported prospect of significant expansion in the foreseeable future. While academic positions are the dominant employment for biomedical researchers immediately after completion of a Ph.D. (the post-doctoral fellowship years), beginning 6 years after the completion of a Ph.D. in biomedical research, nonacademic positions employ the majority rather than academic positions (including tenured, tenure-track, and staff positions).
It thus seems that a significant fraction of today’s postdoctoral fellows in biomedical research are essentially in training for jobs that do not exist in academia or for jobs in industry of other sectors into which they could move sooner. They are, de facto, low-paid university research staff. During the period in which the NIH budget doubled in real terms, postdoctoral salaries barely increased to keep pace with inflation. A decade-old recommendation of the National Research Council that the NIH increase post-doctoral salaries substantially has largely not happened. Recent increases in NIH’s NRSA postdoctoral fellowship stipends, of which about 3000 are awarded annually, have amounted to less than the concurrent inflation rate and have not reversed the previous erosion in compensation. We see no long-term satisfactory equilibrium unless universities, or their funding agencies, limit the number of postdoctoral researchers to be consistent with feasible career advancement to faculty or private-sector positions. A replacement system, better matched to the increased role of the universities in the innovation system, is an expansion of non-faculty career research and research management positions. These may be short-term or long-term as university resources and sources of funding may dictate, but they should be career, not training, positions. In short, there is a mismatch between the large number of Federally supported postdocs and graduate students who perform much of the biomedical research at universities and the opportunities and needs of the non-university workforce.