NSF’s Scientists and Engineers Statistical Data System (SESTAT) provides detailed employment, education, and demographic data for scientists and engineers under age 76 residing in the United States. The 2010 SESTAT defines scientists and engineers as individuals who have college degrees in S&E or S&E-related fields or who are working in S&E or S&E-related occupations.* (See table
The NSCG is the central component of SESTAT, providing data that detail the characteristics of the entire college-educated population in the United States (regardless of their S&E background). Its population of college graduates includes individuals trained as scientists and engineers who hold at least a bachelor’s degree. Because it covers the entire college graduate population residing in the United States, the NSCG provides information on individuals educated or employed in S&E fields as well as those employed or educated in non-S&E fields. The data presented in this chapter for all college graduates (regardless of S&E background) are based on the NSCG.
Whereas NSCG data cover the general college-educated population, the NSRCG supplements SESTAT by adding recent college graduates at the bachelor’s and master’s degree level. The 2010 NSRCG data represent almost 1.5 million recent bachelor’s and master’s graduates in science, engineering, and health (SEH) fields from academic years 2008 and 2009.
The SDR supplements SESTAT by adding doctoral scientists and engineers who earned their SEH doctorates from U.S. academic institutions. Data from the 2010 SDR were collected from doctoral graduates who received SEH research degrees from a U.S. academic institution before 1 July 2009.
*For details on the 2010 SESTAT see http://www.nsf.gov/statistics/sestat/ and http://www.nsf.gov/statistics/infbrief/nsf13311/.
The most recent Bureau of Labor Statistics (BLS) occupational projections, for the period 2010–20, suggest that total employment in occupations that NSF classifies as S&E will increase at a faster rate (18.7%) than employment in all occupations (14.3%) (figure
BLS also projects that, for the period 2010–20, job openings in NSF-identified S&E occupations will represent a slightly larger proportion of current employment than openings in all other occupations: 39.6% versus 38.3% (figure
Of the BLS-projected job openings in NSF-identified S&E occupations, 59% are in computer and mathematical scientist occupations, the largest sub-category of S&E occupations (table
In addition to S&E occupations, table
Employment projections are uncertain.* Many industry and government decisions that affect hiring are closely linked to national and global fluctuations in aggregate economic activity, which are difficult to forecast long in advance. In addition, technological and other innovations will influence demand for workers in specific occupations. The assumptions underlying projections are sensitive to fundamental empirical relationships, and, as a result, may become less accurate as overall economic conditions change.
* Although BLS does a reasonable job of projecting employment in many occupations, the mean absolute percentage error in the 1996 forecast of employment in detailed occupations in 2006 was 17.6% (Wyatt 2010). The inaccuracies in the 1996 projection of 2006 employment were primarily driven by not projecting the housing bubble and increases in oil prices (Wyatt 2010).
Although the Scientists and Engineers Statistical Data System (SESTAT) provides detailed information on college graduate scientists and engineers, it lacks similar data on individuals who do not have a bachelor’s degree. The Census Bureau’s American Community Survey (ACS) provides nationally representative occupational data for workers at all levels of education.* In 2011, about one-fourth of S&E workers age 25 and older did not have a bachelor’s degree. This sidebar looks at the demographic, educational, and employment characteristics of these S&E workers without a bachelor’s degree.†
Relative to college graduate workers employed in S&E occupations, a disproportionate number of those without a bachelor’s degree employed in S&E occupations were black or Hispanic and native U.S. born. In 2011, about 9% of S&E workers without a bachelor’s degree were black, and another 9% were Hispanic. In contrast, 6% of college-educated S&E workers were black and 5% were Hispanic. Asians represented only 3% of S&E workers without a bachelor’s degree, compared to 19% of S&E workers with a bachelor’s degree. In 2011, only 8% of S&E workers without a college degree were foreign born, compared to about one-fourth of college-educated S&E workers.
S&E workers without a bachelor’s degree were mostly concentrated in computer occupations, with 69% employed in the field. In comparison, 44% of the college-educated S&E workers held computer jobs. Among computer occupations, computer support specialists, network and computer systems administrators, and other computer occupations together represented about half of the S&E workers without a bachelor’s degree employed in computer occupations. Unlike the computer field, life sciences, physical sciences, and social sciences occupations had much smaller proportions of workers without a bachelor’s degree. About 3% of the S&E workforce without a bachelor’s degree were employed in these areas combined, compared to about one-fifth of the college-educated S&E workforce.
Relative to other occupations, S&E occupations provide stable employment with good earnings for workers without a college degree. In 2011, the median earnings among workers 25 years of age and older, without a bachelor’s degree, and employed in S&E occupations ($60,000) was twice as high as the median earnings among comparable workers employed in other occupations ($30,000). The unemployment rate among these workers in S&E occupations was 6%, about half the rate in other occupations (11%).
Workers employed in S&E occupations had more formal training (even if they did not have a bachelor’s degree) than those employed in other occupations, so it is not surprising that salaries were higher in S&E jobs. About one-third of the workers without a bachelor’s degree employed in S&E occupations had an associate’s degree, compared to 14% of those employed in other occupations.
* For methodological reasons, estimates from ACS and SESTAT differ slightly even for the college graduate population, which both surveys cover. For example, the two surveys vary in the level of detail collected on work activities, which affects how workers are coded into standard occupational categories. In addition, ACS collects data from households, whereas SESTAT collects data from individuals. Finally, the analysis using ACS data counts postsecondary teachers of S&E as working in non-S&E occupations because the Census Bureau data do not identify them by field.
† This sidebar defines the S&E workforce by workers in S&E occupations (except postsecondary teachers in S&E fields). The ACS data do not allow for separate identification of postsecondary teachers by fields. See appendix table
Employment patterns in the biomedical sciences have changed in the past two decades. The growth in the number of doctorates trained in the field has far surpassed the growth in academic positions, contributing to lengthy postdoc appointments, stiff competition for academic jobs, and an increasing proportion of doctorates going into positions that are not research-intensive (National Institutes of Health [NIH] 2012). According to the Survey of Doctorate Recipients (SDR), between 1993 and 2010, the number of U.S.-educated doctorate holders in the biomedical sciences substantially rose (from about 105,000 to nearly 180,000).* Over this same time, the proportion employed in academia declined (58% to 51%) as did the proportion employed in tenure or tenure-track positions (35% to 26%) despite the fact that both increased in absolute number. The proportion of U.S.-educated doctorate holders who reported research (basic or applied) as their primary or secondary work activity also declined in the education sector (from 75% to 70%). In contrast, the proportion of biomedical sciences doctorates employed in the business sector rose (from 31% to 39%). The majority of the increase in the business sector was driven by those whose jobs did not involve research as their primary or secondary work activity. The proportion of biomedical sciences doctorates reporting that they are employed in jobs closely related to their doctoral degree has declined over this same time (from 68% to 60%), whereas the proportion employed in jobs “somewhat” related to their doctorate has increased (from 24% to 32%). The available data cover the U.S.-educated doctorate holders; the data on foreign-trained doctorates in the field, a segment of the workforce that has grown significantly (NIH 2012), are not comprehensive. The information on postdoc researchers is also not comprehensive.
Despite the persistence of generally favorable employment indicators for biomedical sciences doctorates (the unemployment rate was around 2% in 1993 and 2010, and the rate of working involuntarily out of field was around 3% in both periods), the changes in the employment patterns have generated significant concerns in the profession. Concerns center on the rising number of research doctorates unable to find tenure-track academic research positions, the increasing number and length of postdoc appointments, the influx of foreign-trained doctorates seeking academic positions, and the rising number of early career doctorates taking positions that are not research-intensive and for which current graduate programs may not provide appropriate preparation. In addition, the overall training period, including PhD and postdoc research, is longer in the biomedical sciences than in other comparable disciplines, such as chemistry, physics, and mathematics (NIH 2012). Furthermore, average starting salaries are lower among doctorates in the biomedical sciences than in other fields, such as chemistry, clinical and health fields, and economics (NIH 2012).
In light of the changes in the profession and the resulting concern in the science community, NIH convened a working group consisting of biomedical educators and other experts on the biomedical workforce to develop a set of policy recommendations to support a robust and viable workforce.† The working group recently presented specific recommendations targeted at enhancing graduate training, postdoc research experience, and data collection and dissemination regarding the biomedical workforce. The following is a summary of the main recommendations of this working group:
* See NIH (2012) for a discussion on the fields of science considered as biomedical sciences. Based on the report, the following degree categories from the SDR are included in the data presented in this sidebar: biochemistry and biophysics, bioengineering and biomedical engineering, cell and molecular biology, microbiological sciences and immunology, zoology, biology (general), botany, ecology, genetics (animal and plant), nutritional science, pharmacology (human and animal), physiology and pathology (human and animal), and other biological sciences.
† For detailed information, see the NIH report available at http://acd.od.nih.gov/Biomedical_research_wgreport.pdf (accessed 16 November 2013).
Among college-educated individuals, a significantly higher proportion of men than women are employed in S&E occupations. For example, among S&E highest degree holders working full time, 26% of women, compared to 43% of men, hold positions with formal S&E jobs. This gender gap in S&E employment is found in all racial and ethnic groups. For example, among S&E highest degree holders working full time, S&E jobs are held by 43% of Asian women compared to 58% of Asian men, 22% of black women compared to 32% of black men, 19% of Hispanic women compared to 37% of Hispanic men, and 24% of white women compared to 41% of white men. The participation gap exists despite the trend that increasing proportions of women in all racial and ethnic groups are graduating from college. In most racial and ethnic groups, for example, a higher percentage of women than men have college degrees.
Field of degree, level of highest degree, employment sector, and other characteristics that are typically believed to be associated with occupational fields vary between men and women. As a result, it can be misleading to directly compare S&E employment rates by sex. Compared with men, women tend to have many characteristics—such as degrees in the life and social sciences, highest degrees at the bachelor’s level, and employment in 2-year academic institutions and in the non-profit sector—that are associated with working outside S&E occupations. Statistical models can estimate the size of the male-female participation gap in S&E occupations when various occupation-related factors are taken into account. However, estimates of these differences vary somewhat depending on the assumptions that underlie the statistical model used.
After accounting for differences between men and women in field of degree, level of highest degree, and employment sector, the participation gap in S&E occupations declines significantly (from 17 to 6 percentage points) but does not attenuate completely (figure
National Science Foundation |
National Center for Science and Engineering Statistics (NCSES) |
Previous Releases |
Science and Engineering Indicators 2014 Arlington, VA (NSB 14-01) | February 2014