American Scientist

In the second half of the 19th century, the Ph.D. degree emerged as the culmination of academic training for scientists anticipating a research and teaching career at research universities. Postdoctoral training, rare before the 1950s, became increasingly frequent thereafter, particularly in the United States. One or two years of postdoctoral tenure were the norm in the 1970s and 1980s; currently, the average is more than three-and-a-half years, but longer in the biological than in the physical sciences. Thirty percent of the postdocs hold two or more positions before moving on to independent careers.

How is the postdoc experience valued? A Sigma Xi Survey of Postdoctoral Scholars, reported in "Doctors Without Orders," a supplement to this issue of American Scientist, shows that 70 percent are satisfied, 22 percent are dissatisfied, and 7 percent are neutral. The two main factors contributing to satisfaction are the quality of the training and how well structured the postdoc program is at the training institution, while salary and benefits play a diminishing role.

An objective measure of the postdoc experience is publication. Overall, postdocs average 1.2 peer-reviewed publications per year, but postdocs in well-structured programs average 1.4 publications, and those in superior training programs average 1.3 publications. Remarkably, 43 percent of first authors of research articles in Science were postdocs, according to a 1999 study. No wonder, the trainees as well as their sponsors highly value the postdoc experience.

Postdoctoral training in the United States is valued not only by U.S. citizens and permanent residents, but also and increasingly so by foreign scientists. According to the National Science Foundation 2001 Survey of Graduate Students and Postdoctorates in Science and Engineering (GSS), postdocs increased between 1977 and 2000 from 14,000 to 30,000. The increase was 33 percent (from 9,000 to 12,000) for U.S. citizens and permanent residents; for temporary residents, it was 260 percent (from 5,000 to 18,000), so that by 2000 foreign postdocs numbered 50 percent more than U.S. postdocs.

Why has postdoc training become extended in the last two decades? One reason is that tenured positions in academic institutions have remained approximately constant for 20 years, while Ph.D. degrees have increased. From 1981 to 2001, bioscience Ph.D.'s in tenured positions decreased from 47 to 30 percent; in the physical sciences, the decrease was smaller, from 70 to 62 percent. But biotechnology and other industries now offer well-paid research positions. Bioscience Ph.D.'s working in industry increased from 11 to 30 percent between 1973 and 2001, according to an NSF 2001 Survey of Doctorate Recipients.

Another trend in the two NSF surveys is a greater postdoc increase in the biosciences than in other disciplines. In 1977, bioscience postdocs were 7,000, about half of the total in science and engineering; in 2000, they were 17,000 versus 13,000 for the other sciences and engineering. During those 23 years, the number of temporary-resident postdocs in the biosciences increased fivefold, from 2,000 to 10,000. There are now about 56,000 postdocs in the United States, including those in the social and behavioral sciences.

Other factors contribute to lengthening postdoc training. Longer tenures allow postdocs to accumulate peer-reviewed publications and thus strengthen their qualifications. National Institutes of Health granting patterns also have impacted the biosciences. The age distribution of principal investigators for competing NIH grants has shifted upwards over the years. From 1991 to 2002, the grants awarded to those under 36 years decreased from 8.9 to 4.1 percent and from 23.6 to 12.5 percent for those between 36 and 40 years. The NIH grants awarded to principal investigators over 55 years old have increased from 12.3 to 20.8 percent.