SOURCE: NSF, 1991:Table 1, for 1973-1989; NSF, 1994d:Table 1, for 1991.
NOTES: In this figure, postdoctoral appointees are included in the labor force.
The data are national estimates of the numbers of scientists and engineers with doctorates from US institutions. The estimates are derived from the biennial sample Survey of Doctorate Recipients conducted for the National Science Foundation by the Office of Scientific and Engineering Personnel, National Research Council.
In 1991, survey procedures and timing were changed in ways that improved the estimates but introduced major comparability problems. The response rate, which had fallen steadily during the 1980s (from 66% in 1979 to 58% in 1989), increased to 80% in 1991. Nonresponse bias in the earlier surveys had led to overestimates of 5% or more in the total number of scientists and engineers in the United States. The new procedures, which involved much more intensive followup of those who did not respond initially, no doubt reduced the overestimate, but the extent is not known. The drop in number of employed scientists and engineers from 1989 to 1991 is due at least in part to the change in survey procedures. For example, if the estimates in 1989 were reduced by 5%, the number of doctorates working in the United States would have increased by 3% instead of decreased by 3% from 1989 to 1991.