Confidence in data accuracy: high (green), medium (yellow), low (red); for more details see colors
 FASEB slides, from NSF GSS
gives 37K, but 37K is an underestimate
because by design the survey includes only postdocs at academic degree-granting
institutions, and the survey may include only a subset of postdocs at those institutions, i.e. those with the title "postdoc," which is overrepresented by postdocs on NRSAs/T32s or those early in their careers. To adjust for these biases, we combine information from the NSF GSS and NIH data, as well as estimates of industry/nonprofits and government postdocs, to get 68,000.
 Estimated from Graduate Student Snapshot (of 9,000 graduates, 92% stay in the US; the estimate assumes that 70% take postdoctoral positions and that US citizens/PR and temporary residents are equally likely to take postdocs. If fewer than 70% take postdocs, then this number, as well as the number of foreign trained postdocs, will be smaller, and an estimate of the length of the postdoc derived from stocks and flows will be longer; conversely, if more than 70% take postdocs, then this number, as well as the number of foreign trained postdocs, will be larger, and the derived estimate of the length of postdocs will be shorter).
 Estimated from a 2:1 ratio of US Citizens/PR:temporary resident graduates and a 1:1 ratio for postdocs; calculation assumes that US citizens/PR and temporary residents stay in postdocs for an equal length of time. The foreign-trained estimate of 1900
assumes that the graduate students on temporary visas do not convert to permanent residents; deviations from this assumption would increase the estimated number of foreign postdocs. For example, if all the graduate students who were temporary residents became permanent residents at the end of graduate school, then in order to still have a 1:1 ratio of permanent:temporary residents in the postdoc pool, about 5800 foreign-trained postdocs need to flow in per year. As another example, if half of temporary resident graduate students become permanent residents at the end of graduate school, then on average 3900
foreign-trained postdocs need to flow in to maintain the 1:1 ratio.
 Median length of time with the title postdoc, estimated from the SDR. Because the SDR only covers US-trained doctorates and because foreign doctorates may do longer postdocs, this number may be an underestimate. For comparison purposes, with 8700 postdoctoral entrants per year and 68,000 total postdocs, we would expect a mean postdoc length of 7.8 years (due to the distribution being right-skewed, we do expect the mean the to be larger than the median).
- By Citizenship
- By Type of Support
- FASEB Education and Employment Data Compilation