Skip all navigation and go to page content

Chapter 5. Academic Research and Development

The Academic Doctoral S&E Workforce

S&E doctorate holders in academia influence the nation's academic R&D enterprise in two key ways. They work in institutions that conduct academic R&D and produce the bulk of academic articles and patents. Moreover, they teach individuals who then go on to earn S&E doctorates, many of whom will work in academia and contribute to academic R&D. The focus of this section is on the research aspects of the employment of doctoral scientists and engineers in academia.

This section examines trends in the doctoral S&E academic workforce in terms of its demographic composition and its deployment across institutions, positions, and fields. Particular attention is paid to the component of the academic workforce that is more focused on research, including graduate assistants, those employed in postdoctoral positions, and researchers receiving federal support.

The discussion in this section is limited to individuals, including foreign-born individuals, who received their S&E doctorate at a U.S. institution. (More than two-thirds of foreign-born doctorate holders employed in the United States earned their doctorate degree from a U.S. institution; see chapter 3 for more information on foreign-born doctorate holders working in the United States). Owing to the complex interrelationships among faculty and nonfaculty positions that jointly produce R&D outcomes, much of the discussion addresses the overall academic employment of S&E doctorate holders, including those in nonfaculty positions. At various points the characteristics of full-time faculty are discussed.

Trends in Academic Employment of Doctoral Scientists and Engineers

Academic employment of doctoral scientists and engineers grew over the past three decades and reached a record high of 272,800 in 2008, about the same as the employment numbers in 2006 (appendix table 5-16).[18] However, the change from 2006 was the smallest single-period increase in estimated total academic employment since at least 1973. The long-term growth rate in the number of doctoral scientists and engineers employed in the academic sector was slower than the rate of growth in the business and government sectors (table 5-6). As a result, the share of all S&E doctorate holders employed in academia dropped from 55% to 44% during the 1973–2008 period (table 5-7). In 2008, nearly half of those with recently awarded S&E doctorate degrees (that is, a degree awarded within 3 years of the survey year) were employed in academia, with 18% of recent doctorate holders employed in academic postdoc positions.[19]

Academic Employment of S&E Doctorate Holders

The academic doctoral S&E workforce includes those with a doctorate in an S&E field and employed in the following positions: full and associate professors (referred to as "senior faculty"); assistant professors and instructors (referred to as "junior faculty"); postdoctoral researchers (referred to as "postdocs"); other full-time positions such as lecturers, adjunct faculty, research associates, and administrators; and part-time positions of all kinds. Academic employment is limited to those employed in 2-year or 4-year colleges or universities.

Full-time faculty positions continue to be the norm in academic employment, but S&E doctorate holders are increasingly employed in other full-time positions, postdocs, and part-time positions (figure 5-13). The share of full-time faculty among all academic S&E doctorate holders fell from 88% in the early 1970s to 73% in 2008 (appendix table 5-16). Over the same period, the share of other full-time positions rose from 6% to 15%, the share of postdocs increased from 4% to 7%, and the share of part-time positions increased from 2% to 6% of all academic S&E doctorate holders.

The proportion of full-time faculty among S&E doctorate holders in higher education fell in all fields during 1973–2008, with the life sciences and psychology experiencing the largest relative declines. Growth in postdoc positions and other full-time positions accounted for the declining share of full-time faculty positions in the life sciences, whereas the growth in part-time and other full-time positions explained the drop in share of faculty positions in psychology (appendix table 5-16).

Over the past three decades, growth in the number of life scientists with academic employment was consistently stronger than for doctorate holders in other S&E fields (figure 5-14). Growth in academic employment slowed in the early 1990s for engineering, social sciences, physical sciences, and mathematics, but has increased since then in social sciences and mathematics (appendix table 5-16).

Women in Academic S&E Workforce

The number of women with S&E doctorates employed in academia grew from 10,700 in 1973 to 93,400 in 2008, more than an eightfold increase. In comparison, the number of male S&E doctorate holders increased 67% over the period, from 107,200 in 1973 to 179,400 in 2008 (appendix table 5-17).

These differential rates of increase are reflected in the steadily rising share of women in the academic S&E workforce. Women constituted 34% of all academic S&E doctoral employment and 31% of full-time faculty in 2008, up from 9% and 7%, respectively, in 1973 (table 5-8 and appendix table 5-17). Women's share of academic S&E employment increased markedly over time in all position categories, though to a lesser degree in part-time positions. Women have held a larger share of junior faculty positions (includes assistant professors and instructors) than positions at either the associate or full professor rank. However, as a result of the decades-long trend in the rising proportion of women earning doctoral degrees, coupled with their slightly greater propensity to enter academic employment, the share of women in all three faculty ranks rose significantly between 1973 and 2008. In 2008, women constituted 21% of full professors, 37% of associate professors, and 42% of junior faculty (figure 5-15).

Compared with their male counterparts in the academic doctoral S&E workforce, women were more heavily concentrated in the fields of life sciences, social sciences, and psychology, with correspondingly lower shares in engineering, the physical sciences, mathematics, and computer sciences. Women's share of doctorate holders in each of these fields grew during the 1973–2008 period (appendix table 5-17), with the most pronounced growth in share occurring in engineering, the field in which women were the least well represented.

Minorities in Academic S&E Workforce

Although the number of academic S&E doctorate holders who are members of underrepresented minority groups (i.e., blacks, Hispanics, and American Indians/Alaska Natives) has increased over time, they remain a small percentage of the total (appendix table 5-18). These groups constituted about 9% of both total academic employment and full-time faculty positions in 2008, up from 2% in 1973. Underrepresented minority groups have a relatively higher share of employment in other positions, which includes part-time positions, than in the full-time faculty and postdoc employment categories (table 5-9).

Underrepresented minorities were concentrated in different degree fields and different types of institutions than whites. Compared with white S&E doctorate holders employed in academia, underrepresented minorities were relatively concentrated in the social sciences and relatively less represented in the physical sciences and the life sciences (appendix table 5-18). Relatively fewer underrepresented minorities were employed at research universities than whites in 2008, and relatively more were employed at master's colleges and universities (table 5-10). (See chapter 2 sidebar, "Carnegie Classification of Academic Institutions," for a brief description of the Carnegie categories.)

The share of Asians/Pacific Islanders employed in the S&E academic doctoral workforce grew dramatically over the past three decades, rising from 4% in 1973 to 14% in 2008. Asians/Pacific Islanders were heavily represented in engineering and computer sciences, where they constituted 27% and 35%, respectively, of the S&E academic doctoral workforce in 2008. Far smaller proportions of Asians/Pacific Islanders were present in social sciences (8%) and psychology (5%) (appendix table 5-18). A larger share of Asians/Pacific Islanders than whites was employed at research universities and medical schools in 2008 (table 5-10).

Foreign-Born U.S. S&E Doctorate Holders

Academia has long relied on foreign-born doctorate holders, many of them with doctorate degrees from U.S. universities, to staff faculty and other academic positions. No current information is available about the number of foreign-born individuals with foreign doctorates who are employed at U.S. universities and colleges. The following discussion is limited to foreign-born individuals with U.S. doctorates.

Academic employment of foreign-born U.S. S&E doctorate holders has increased continuously since the 1970s at a rate that has exceeded the growth in academic employment of U.S.-born S&E doctorate holders (figure 5-16). As a result, the foreign-born share of the total academic employment of U.S. S&E doctorate holders increased from 12% in 1973 to nearly 25% in 2008 (figure 5-16), and reached particularly high proportions in engineering (46%) and computer sciences (51%) (appendix table 5-19). In all fields, foreign-born doctorate holders were a larger share of postdoc employment than of full-time faculty employment. Overall, 46% of postdoc positions were held by foreign-born U.S. S&E doctorate holders, compared with 23% of full-time faculty positions and 23% of other full-time positions.

Of the 39,000 Asian/Pacific Islander doctorate holders employed in academia in 2008, 9% were native-born U.S. citizens, 44% were naturalized U.S. citizens, and 47% were noncitizens. In 2008, Asians/Pacific Islanders represented 50% of the foreign-born faculty employed full-time in the United States and 62% of the foreign-born doctorate holders with postdoc appointments. In contrast, only 1% of native-born full-time faculty and 5% of native-born postdocs were Asians/Pacific Islanders.

Academic Researchers

The interconnectedness of research, teaching, and public service activities in academia makes it difficult to assess the precise size and characteristics of the academic research workforce by examining the employment trends in academic positions, because individuals employed in the same position may be involved in research activities to differing degrees or not involved in research. Therefore, self-reported research involvement is a better measure than position title for gauging research activity.[20] This section limits the analysis to "academic researchers"—academic S&E doctorate holders who reported that research is either their primary work activity (that is, the activity that occupies the most hours of their work time during a typical work week) or their secondary work activity (the activity that occupies the second most work hours per week).

Doctoral S&E Researchers

From 1973 to 2008, the number of academic researchers grew from 82,300 in 1973 to 184,700 in 2008 (appendix table 5-20). The 2008 total included 137,800 individuals employed in full-time faculty positions. The proportion of academically employed S&E doctorate holders that are researchers declined slightly from 1993 (70%) to 2008 (68%) (figure 5-17). A nearly identical pattern of decline was observed for the share of full-time faculty that reported research as a primary or secondary work activity. The proportions of researchers among all academic S&E doctorate holders and all full-time faculty were higher in the life sciences, engineering, and computer sciences than in the social sciences and psychology (appendix table 5-20). In most fields, the share of academic researchers declined between 1993 and 2008.

A different picture emerges when only considering researchers who report research as their primary work activity. In contrast to the declining share of academic employees who reported research as their primary or secondary work activity, the share who reported research as their primary work activity steadily increased from 1973 to 2008 (figure 5-17). Taken together, these trends suggest that while research as an important work activity is not becoming more widespread among S&E doctorate holders employed in academia, a growing share of academic S&E positions are becoming research intensive.

Among full-time doctoral S&E faculty, the increased share of doctorate holders reporting research as their primary work activity reflects a shift in priority from teaching to research for many faculty. From 1973 to 2008, the proportion of full-time faculty identifying research as their primary work activity climbed from 19% to 36%, while the share with teaching as their primary activity fell from 68% to 47% (figure 5-18). The balance of emphasis between teaching and research varied across the disciplines, with a higher share of faculty in the life sciences identifying research as their primary work activity, and a higher share of faculty in mathematics and social sciences reporting teaching as their primary activity. Since 1991, the proportion of doctorate holders who list research as a primary work activity declined in physical sciences, computer sciences, and life sciences fields, but grew in mathematics, psychology, engineering, and the social sciences (appendix table 5-20).

S&E Full-Time Faculty Researchers

Table 5-11 examines the relationship between research and the career stage of S&E full-time faculty. The smallest share of primary researchers occurred among the most recently degreed faculty (33%). The share of faculty who indicated research as their primary work activity increased with time since doctorate in the succeeding two cohorts, and then fell in the last reported cohort (12 years or more since doctorate). The higher share (48%) of primary researchers within the second cohort, 4 to 7 years since doctorate, coincides with the period during which many early career faculty would be preparing to apply for tenure at their university, and would have heightened motivation to complete research projects and publish results. In the last cohort reported in the table, 12 years or more beyond the doctorate, the share of full-time faculty reporting research as a primary activity fell to 35%. Other responsibilities—such as mentoring younger faculty, advising doctoral students, and accepting major committee assignments or faculty leadership roles—may become primary work activities for many faculty at this career stage.

A similar pattern prevailed in most degree fields—the share of faculty who indicated that research was their primary work activity increased through the early career cohorts and then fell as faculty approached mid-career. Research was more frequently a primary work activity for early career faculty in engineering and computer sciences than for faculty in other fields (table 5-11).

Collaborative Research

Research in many fields has increasingly involved collaboration. This section describes S&E doctorate holders' self-reports of their collaboration with others using data from the 2006 Survey of Doctorate Recipients.[21] Information on trends in coauthorship can be found later in this chapter under "Coauthorship and Collaboration."

In 2006, roughly 70% of S&E full-time research faculty employed in academic institutions reported working with an immediate work team or with others working elsewhere in the same organization, nearly 60% worked with individuals in other organizations in the United States, and nearly one-third worked with individuals located in other countries. Team work is most common among life scientists, physical scientists, and engineers, and least common among mathematicians and social scientists.

International collaboration was more common among foreign-born S&E full-time research faculty. Communication by e-mail or telephone was, by far, the most commonly used mode of international collaboration, followed by travel to the United States by the foreign collaborator(s), foreign travel by the U.S.-based collaborator, and communication through web-based or virtual technology.

For a more extensive discussion of these topics, see the "Collaborative Research" section in chapter 5 of the 2010 edition of Indicators (NSB 2010). For data on international collaborative activity in the S&E workforce more generally, see the "International Engagement by the Domestic S&E Workforce" in chapter 3.

Graduate Research Assistants

The close coupling of advanced training with hands-on research experience is a key strength of U.S. graduate education. Many of the 434,100 full-time S&E graduate students in 2008 (table 5-12) contributed significantly to the conduct of academic research.

The number of research assistants (RAs)—full-time graduate students whose primary mechanism of financial support is a research assistantship—has grown faster than graduate enrollment, both overall and in most fields. Graduate research assistantships were the primary means of support for 27% of graduate students in 2008, up from 22% in 1973. In the field distribution of RAs, there was a shift away from the physical sciences and social sciences and into the life sciences, computer sciences, and engineering. In engineering and the physical sciences in 2008, the proportion of RAs was high relative to graduate enrollment; 42% of graduate students in the physical sciences and 40% of engineering graduate students were supported in their graduate study primarily by research assistantships. In the life sciences, the proportion of RAs relative to graduate enrollment was similar to the overall proportion across all fields (27%), possibly reflecting the heavier reliance on postdoctoral researchers rather than RAs in the life sciences fields (table 5-12).

The majority of the academic research workforce remains employed in the intensive and very intensive research universities, although the research universities' shares of both academic researchers and of RAs have declined since 1973. (See chapter 2 sidebar, "Carnegie Classification of Academic Institutions," for a brief description of the Carnegie categories.) During the 2003–2008 period, the research universities employed 48% of all S&E doctorate holders in academic positions, 57% of those reporting research as their primary or secondary activity, and 79% of S&E graduate students for whom an RA was their primary means of support (table 5-13). Trends indicate a growing research presence by full-time academic researchers at institutions not classified as research universities, although RAs remain highly concentrated in the research universities.

Academic Employment in Postdoc Positions

The number of S&E doctorate holders employed in academic postdoc positions climbed from 4,000 in 1973 to 18,000 in 2008 (appendix table 5-16).[22] (See sidebar, "Postdoctoral Researchers.") During that time period, the share of postdocs increased from 4% to 7% of all academically employed S&E doctorate holders. Postdocs were much more prevalent in the life sciences, engineering, and the physical sciences than in social sciences, although the proportion of postdoc positions in physical sciences has declined since the mid-1990s (figure 5-19 and appendix table 5-16).

The demographic profile of individuals employed in academic postdoc positions has changed dramatically over time. The proportions of postdocs held by women, racial/ethnic minorities, and foreign-born individuals has climbed since 1973 (table 5-14).

Early Career Postdocs

A temporary postdoc appointment is a common stop along the career path of S&E doctorate holders, particularly during their early career stages. In 2008, 36% of recently degreed S&E doctorate holders in academia were employed in postdoc positions, a figure that approached the share (42%) employed in full-time faculty positions (appendix table 5-21). With the exception of 2003, the share of recently degreed S&E doctorate holders in academic postdoc positions has exceeded the share holding full-time tenured or tenure-track faculty positions since 1995 (figure 5-20). S&E doctorate holders 4 to 7 years beyond the doctorate degree were far less likely than their recently degreed counterparts to be employed in academic postdoc positions; in 2008, only 11% of these doctorate holders held postdoc positions.

The vast majority of academic postdocs are employed at very high research activity universities. In 2008, the share of all academic postdocs employed at these institutions reached 75% (table 5-15). At the research universities, 70% of S&E postdoc appointments in 2008 were held by recently degreed individuals, and 5% by doctorate holders who were 8 or more years past their degree. The postdoc populations employed at medical schools and other universities and colleges included a larger pool of doctorate holders who had not recently earned the doctorate degree.

In comparison to 1995, a larger share of S&E doctorate holders employed in academia in 2006, 45% versus 41%, had held a postdoc appointment at some point in their career, and a slightly larger share than in 1995 had been employed in postdoc positions two or more times (table 5-16). Postdocs and multiple postdocs are relatively more prevalent among early career S&E doctorate holders than among the total pool of S&E doctorate holders. Early career postdoc employment and multiple instances of postdoc employment are typical for academic careers in the life sciences and the physical sciences (table 5-16), the two fields of study that have had the highest incidence of postdocs over the years (figure 5-19).

Government Support of Academic Doctoral Researchers

The federal government provides academic researchers with a substantial portion of overall research support. This section presents data from S&E doctorate holders in academia who reported on the presence or absence (but not magnitude) of federal support for their work.[23]

Academic Scientists and Engineers Who Receive Federal Support

In 2008, 45% of all S&E doctorate holders in academia and 56% of those for whom research was a primary or secondary activity reported federal government support for their work (appendix table 5-22). For S&E as a whole and for many fields, the share of S&E doctorate holders and researchers receiving federal support has declined since the early 1990s.

Faculty and other full-time S&E doctoral employees were less likely than postdocs to receive federal support. Throughout the 1973–2008 period, fewer than half of full-time S&E faculty received federal support, whereas the share of postdocs receiving federal support was above 70%. Since 1991, the share of academic S&E doctorate holders receiving federal support has declined in all position categories (appendix table 5-22).

Federal support is more prevalent in very high research activity universities and medical schools. More than 60% of S&E doctorate holders and full-time faculty employed in research universities and medical schools received federal support in 2008 (appendix table 5-23). The percentage with federal support was less than 30% among those employed in doctoral/research universities, master's-granting universities, and baccalaureate colleges.

Federal Support of Early Career S&E Doctorate Holders

Federal support has been less available to early career S&E doctoral faculty than to more established faculty, and the percentage of early career S&E faculty with federal support has declined. From 1973–2008, S&E doctorate holders with recently earned doctorates (i.e., doctorates earned within 3 years of the survey) employed in full-time faculty positions were far less likely to receive federal support than those in postdoc positions (figure 5-21). In 2008, 27% of recent doctorate recipients in full-time faculty positions received federal support, down from 38% in 1991. Of recent S&E doctorate recipients employed in postdoc positions in 2008, 71% received federal support, which was a substantial decline from 1991 (84%).

S&E doctorate holders employed as full-time faculty who had received their doctorate 4–7 years earlier were more likely to receive federal support than those with more recently earned doctorates, and the same was true of those employed in postdoc positions (figure 5-21). As with recent doctorate recipients, the share of full-time faculty and postdocs 4–7 years beyond the doctorate who received federal support also declined from 1991. The shares of early career full-time faculty and postdocs with federal support were higher in some fields (life sciences, physical sciences, and engineering) than in others (mathematics and social sciences) (appendix table 5-24).


[18] Unless specifically noted, data on S&E doctorate holders in this section come from the Survey of Doctorate Recipients, a biennial NSF survey. All numbers are rounded to the nearest 100. Small estimates may be unreliable.
[19] The United States is unlike many other countries in the fraction of doctorate holders who are employed in academia. A comparison of 1990–2006 doctorate recipients in 14 countries for which data are available found that in most of these countries, more than half of doctorate holders were employed in academia, compared with 47% for the United States. Only the United States, Austria, and Belgium had substantial fractions of doctorate holders employed in the business sector, and the United States had one of the smallest fractions employed in government (Organisation for Economic Co-operation and Development 2009).
[20] Respondents were presented with a list of work activities and asked to identify the activities which occupied the most and second most hours during the typical work week. This measure was constructed slightly differently prior to 1993, and the data are not strictly comparable across the two periods. Prior to 1993, the survey question asked the respondent to select the primary and secondary work activity from a list of activities. Beginning in 1993, respondents were asked on which activity they spent the most hours and on which they spent the second most hours. Therefore, the crossing over of the two trends between 1991 and 1993 could partly reflect a difference in methodology. However, the faster growth rate for researchers in both the 1973–91 and 1993–2008 periods means that changes in question wording cannot fully explain the observed trend. Because individuals may select both a primary and a secondary work activity, they can be counted in both groups.
[21] On the 2006 Survey of Doctorate Recipients, respondents were asked to indicate whether they "Work with an immediate work group or team?"; "Work with others in the same organization (company, university, agency, etc.), but not the same group or team?"; "Work with individuals in other organizations in the U.S.?"; and "Work with individuals located in other countries?" For respondents who indicated that they had collaborated with individuals located in other countries, subsequent questionnaire items inquired about the nature of the collaboration (for example, sharing information, sharing facilities, preparing a joint publication) and the mode of collaboration (for example, collaboration via telephone or e-mail, travel to foreign country).
[22] These data include only U.S.-trained postdocs employed in U.S. academic institutions. A 2003 survey conducted by the Sigma Xi honor society, which was nonrepresentative and likely to undercount foreign-degreed postdocs, found that 46% of responding postdocs had received their doctorate from a non-U.S. institution.
[23] Interpretation of the data on federal support of academic researchers is complicated by a technical difficulty. Between 1993 and 1997, respondents to the Survey of Doctorate Recipients were asked whether work performed during the week of April 15 was supported by the federal government. In most other survey years, the reference was to the entire preceding year, and in 1985, it was to the month of April. However, the volume of academic research activity is not uniform over the entire academic year. A 1-week (or 1-month) reference period seriously understates the number of researchers supported at some time during an entire year. Thus, the numbers for 1985 and 1993–97 cannot be compared with results for the earlier years or with those from the 1999 through 2008 surveys, which also used an entire reference year.

The discussion in this edition of Indicators generally compares data for 2008 with data for 1991. All calculations express the proportion of researchers with federal support relative to the number responding to this question. The reader is cautioned that, given the nature of these data, the trends discussed are broadly suggestive rather than definitive. The reader also is reminded that trends in the proportion of all academic researchers supported by federal funds occurred against a background of rising overall numbers of academic researchers.