Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
THE EDUCATION OF GRADUATE SCIENTISTS AND ENGINEERS 47 3 THE EDUCATION OF GRADUATE SCIENTISTS AND ENGINEERS 3.1 OVERVIEW The recent increase in annual production of scientists and engineers with graduate degrees extends a trend of steady growth. In 1993, more than 25,000 scientists and engineers received PhDs from US universities, up from about 18,400 in 1983 (NSF, 1994f). In the same year, some 80,000 scientists and engineers received master's degrees from US universities (including those who intended to continue toward the PhD degree), a number that has increased steadily from about 65,000 a year in the early 1980s (NSF, 1994b). Most of the recent increase in the number of science and engineering PhDs awarded annually can be accounted for by an influx of foreign students (discussed later in this chapter). Including those students, average growth in the total science and engineering graduate-student population has averaged about 2.5 % per year since 1982. The total number of graduate students in science and engineering in the United States rose from 339,600 in 1982 to 431,600 in 1992, an increase of 27% (Table 5 in NSF, 1994a). Figure 3â1 shows this growth by major field. In 1992, most graduate science and engineering students (87%) were enrolled in universities that grant doctorates, a percentage that has varied only slightly since the Survey of Doctorate Recipients began in 1975 (NSF, 1994a). Most (67%) were full-time students.
THE EDUCATION OF GRADUATE SCIENTISTS AND ENGINEERS 48 FIGURE 3â1 Science and engineering graduate student enrollment, by broad field, 1982â1992. SOURCE: Calculated from NSF, 1994a: Table 1. NOTES: The broad fields are defined as in the notes to Figure 1â1.
THE EDUCATION OF GRADUATE SCIENTISTS AND ENGINEERS 49 3.2 THE MASTER'S EXPERIENCE In some fields, a master's degree is the professional norm. A master's degree generally entails 2 years of coursework. Some master's-degree programs require a research thesis, others do not. In the latter case, the master's degree is not so much a terminal degree as a recognition of the coursework and qualifying examinations completed after about 2 years in a doctoral program. In recent decades, the 2-year master's degree has served in some fields as the terminal degree. For example, the American Society for Engineering Education in 1987 reaffirmed the appropriateness of the master's degree for engineering students not expecting to enter careers in research or university teaching (ASEE, 1987). About 4.6 times as many master's degrees in engineering are awarded each year as engineering PhDs (for comparison, the ratio in the physical sciences is close to unity) (NSF, 1994b). The master's degree is also a customary end point in public health, computer science, and bioengineering and for those who want to teach in high schools and community colleges. Data on the number of master's degrees by field, sex, race, and citizenship are included in Tables B-16 through B-19 in Appendix B and on the employment of new master's-degree recipients in Appendix C. 3.3 THE DOCTORAL EXPERIENCE Acquisition of research skills is central to the doctoral experience. The typical PhD program constitutes a two-part experience of great depth and intensity that lasts 4 or more years. The first part consists of about 2 years of course work. The second part focuses on a doctoral dissertation based on original research that might take 2 or 3 years or more to complete. The dissertation, as a demonstration of ability to carry out independent research, is the central exercise of the PhD program. When completed, it is expected to describe in detail the student's research and results, the relevance of that research to previous work, and the importance of the results in extending understanding of the topic (CGS, 1990). It is customary in most fields of science and engineering for a doctoral candidate to be invited to work as a research assistant (RA) on the project of a faculty member; an aspect of this research project often becomes the subject of the student's dissertation. A traditional expectation of many students (and their professors) is that they will extend this work by becoming university faculty members. If they do, promotion and tenure depend to a great extent on continuing research publication. A properly structured requirement for demonstrated ability to perform independent research continues to be the most effective means to prepare bright and motivated people for
THE EDUCATION OF GRADUATE SCIENTISTS AND ENGINEERS 50 research careers. Original research demands high standards, perseverance, and a first-hand understanding of evidence, controls, and problem-solving, all of which have value in a wide array of professional careers. Carnegie-Mellon University in Pittsburgh is one institution that is experimenting with a number of reforms. Paul Christiano, provost of the university, offered the committee a summary of trends affecting graduate education in science and engineering: â¢ More cross-fertilization between disciplines to exploit new opportunities. â¢ Somewhat greater emphasis on master's-degree programs. â¢ Greater interest in advanced nondegree programs. â¢ More teaching practice for faculty and graduate students, complemented by a new teaching center. Dr. Christiano said that new commercial and societal needs invite innovative approaches to graduate education. He cited a need for more interdisciplinary programs and for an appreciation of the value of graduate education by potential students and employers. He also identified some obstacles to change, including reduced interest of US students in science and engineering, institutional inertia, and the short-term view of industry sponsors. For example, in the case of industry, he has found that graduates of Carnegie-Mellon's research center for engineering design have not always been well accepted by some employers, because its graduates are not linked to a traditional field. He felt that such obstacles could be reduced by better communication and more interaction between universities and industry, which would demonstrate the benefits of interdisciplinary centers. Box 3â1: Experimentation and obstacles to reform In the course of their dissertation research, doctoral students perform much of the work of faculty research projects and some of the university's teaching. Therefore, institutions and individual professors have incentives to accept and help to educate as many graduate (and postdoctoral) researchers as they can support on research grants, teaching assistantships, and other sources of funding. By the time they receive PhDs, 63% of science and engineering graduate students have been RAs and half have been teaching assistants. This system is advantageous for institutions, to which it brings motivated students, outside funding, and the prestige of original research programs. And it is advantageous for the graduate students, for whom it supports an original research experience as part of their education. Although the research component of the doctoral experience is dominant, other components are also important. They include a comprehensive knowledge of the current state of knowledge and techniques in a field and an informed approach to career preparation. Because of the recent trend toward large group projects in some disciplinesâin which a research topic is divided among a number of students, postdoctoral fellows, and facultyâa PhD candidate can become so focused on a particular technique that there might be little opportunity for independent exploration of related fields or career options. When a graduate student becomes essential to a larger research project, completion of the degree can be unduly delayed. Furthermore, students working on tightly focused research might conclude that this is the only valued achievement for scientists and engineers.
THE EDUCATION OF GRADUATE SCIENTISTS AND ENGINEERS 51 In many fields, nonresearch jobs are accorded lower status by faculty; students who end up in such jobs, especially outside academe, often regard themselves as having failed (that is less true in engineering and chemistry, in which nonacademic employment is often the norm). If the number of academic-style research positions continues to level off or contract, as seems likely, a growing number of PhDs might find themselves in nonacademic careers to which they have been encouraged to give little respect. 3.4 TIME TO DEGREE The average time to complete a doctoral degree has increased for graduate students in all science and engineering fields. Over the last 30 years, the average time it takes graduate students to complete their doctoral programs, called the âtime to degreeâ (TTD), has increased steadily. One measure, the median time that each year's new PhDs have been registered in graduate school, has increased in some fields by more than 30%. (The time to master's degree does not seem to have increased, although no one collects national statistics on it.) The lengthening of the period of graduate work is accompanied by a second trend. It has become more common for new PhDs in many fields to enter a period of postdoctoral study (discussed at the end of this chapter), to work in temporary research positions, and to take 1-year faculty jobs before finding a tenure-track or other potentially permanent career-track position. We are concerned about the increasing time spent by young scientists and engineers in launching their careers. Spending time in doctoral or postdoctoral activities might not be the most effective way to use the talents of young scientists and engineers for most employment positions. Furthermore, because of the potential financial and opportunity costs, it might discourage highly talented people from going into or staying in science and engineering. The median number of years between receipt of the bachelor's degree and the doctorate in science and engineering has increased from 7.0 years during the 1960s to 8.7 years for those who received doctorates in 1991 (Table 5 in NSF, 1993b). Graduate students in the physical sciences have shorter-than-average overall completion timesâabout 7 yearsâand social scientists have longer-than-average completion timesâabout 11 years (see Figure 3â2). The remaining science and engineering fields average between 8 and 9 years. The median time registered in doctorate programs is shorter than total TTD (the interval from receipt of a bachelor's degree to receipt of a PhD) because many graduate students take some time between college and graduate school to work, and some take time off during graduate school. Because time out between college and graduate school can be valuable for gaining work experience and more mature decision-making about careers, an increase in years from bachelor's
THE EDUCATION OF GRADUATE SCIENTISTS AND ENGINEERS 52 FIGURE 3â2 Median years to degree for doctorate recipients, by broad field, 1993. SOURCE: Appendix Table B-29. NOTES: Total time is the number of years between receipt of the bachelor's degree and receipt of the PhD. Registered time is the amount of time actually enrolled in graduate school (thus, it might be less than the time elapsed from entry into graduate school and completion of the PhD).
THE EDUCATION OF GRADUATE SCIENTISTS AND ENGINEERS 53 degree to doctorate is not a problem. But registered time to degree (RTTD)1 has also increased steadily over the last 30 years. The median RTTD for engineering PhDs increased from 5.0 years in 1962 to 6.2 years in 1992. In 1992, it was 6.7 years for PhDs in the life sciences, 6.5 years in the physical sciences, and 7.5 for the social sciences (Table 6 in NRC, 1993). Our understanding of factors that affect TTD is incomplete. One finding, reported for psychology, is that TTDs are longer when there are many students per faculty member or many students overall (Striker, 1994). The National Research Council's Office of Scientific and Engineering Personnel in 1990 tested a five-variable model over 11 fields of science and could find no cause or set of causes to explain the trend (Striker, 1994; Tuckman et al., 1990). Some researchers explain the increase in TTDs by pointing to the increasing complexity and quantity of knowledge required for expertise in a given field. Another possible explanation is the tendency of some faculty to extend the time that students spend on research projects beyond what is necessary to meet appropriate requirements for a dissertation. The Council of Graduate Schools (CGS) reports that lack of financial support during the dissertation phase substantially extends TTD, as do difficulties in topic selection, unrealistic expectations for the amount of work that can be completed in a dissertation, and inadequate guidance by advisers. Still other reasons are poor undergraduate preparation, student reluctance to leave the congenial life of academe, and postponement of graduation in the face of uncertain employment prospects (CGS, 1990). There has been little research on how students spend the extra time that they take to earn a degreeâwhether in class work, studying for general examinations, doing thesis research, working as teaching assistants or research assistants, etc. In a tight labor market, students might hope that the extra time might provide them with a better thesis and thus a better chance at a research position, but information on this is not readily available. 3.5 MECHANISMS OF ASSISTANCE FOR GRADUATE EDUCATION Research grants, whose primary purpose is to support research, exert a powerful influence on the format of graduate education. Table 3â1 provides an overview of the sources of graduate school support for doctorate recipients by broad field in 1993. Master's-degree students are mainly self-supporting (and often hold full-time jobs while studying), but most PhD students offset the cost of graduate education with grants and other forms of support from state and federal governments, industries, universities, nonprofit groups, and others. The amount and kind of support vary widely by field (see Appendix B, Table B-7). 1 Registered time is the amount of time actually enrolled in graduate school (thus, it might be less than the time elapsed from entry into graduate school and completion of the PhD).
THE EDUCATION OF GRADUATE SCIENTISTS AND ENGINEERS 54
THE EDUCATION OF GRADUATE SCIENTISTS AND ENGINEERS 55 In 1992, according to a survey of graduate departments, 41% of the 126,000 full-time graduate science and engineering students received their primary support from their institutions, 31% provided most of their own funds (including funds from federally guaranteed loans), and 20% depended primarily on federal sources, primarily in the form of research assistantships, graduate fellowships, and training-grant positions (Table 12 in NSF, 1994a). However, federal support for students in the biological and physical sciences was higher (34% and 36%, respectively). One-fourth of those with institutional support received it in the form of research assistantships, half received teaching assistantships, and the remaining one-fourth were supported by a mix of fellowships, traineeships, and other forms of support. The preceding discussion underestimates the importance of federal support, especially to RAs, because they were measured at one time (1992). Typically, graduate students depend on different sources of support in different phases of graduate workâperhaps as teaching assistants (TAs) in the first 2 years and then as RAs while doing dissertation research. By the time students receive the doctorate, nearly two-thirds have been RAs and half TAs (see Figure 3â3). The students reporting this information are not always sure of the ultimate source of their RA funds, and the reported data do not distinguish between federal and institutional RAs (Table A-5 in NRC, 1993). But we believe that most RAs are supported by federal research grants and contracts. Since the early 1970s, virtually all growth in federal support of scientists and engineers in academe has been in the form of grants, contracts, and cooperative agreements for R&D projects (Figure 3â4). Federal fellowship and traineeship programs were cut back substantially in the early 1970s. The research-assistantship mechanism began to grow in importance as faculty used their research grants to support graduate students. Federal support of graduate fellowships and traineeships fell steadily as a percentage of overall federal funding (Figure 3â5). As a result, the federal government has supported graduate education for the last 2 decades primarily through its support of faculty research projects, rather than direct support of graduate students. There are no clear guidelines for distributing the various types of federal support. The research assistantship has become dominant, but not as a result of an education policy. The number of PhDs produced now reflects more closely the availability of research funds than the employment demand for PhDs. There are several drawbacks to this dependence on research grants. One is that the pressure to produce new research results extends to graduate students, who easily gain the impression that hard, goal-oriented work on a specific project is the most important aspect of graduate education. As already noted, PhD students can become so involved in the work of the faculty investigators under whose grants they conduct their dissertation research that little time is left for independent exploration or other educational activities. Even the best-intentioned professors might lack the time to impart a broad appreciation of their discipline or to encourage their RAs to investigate the discipline thoroughly or plan their careers. Efforts should continue to be made to make this experience as profitable and broadening as possible so that graduate scientists and engineers are prepared for all kinds of careers. In addition, the peer-review process, effective as it is at judging the research ability of academic researchers, does not try to evaluate the educational value of the research projects that can constitute the central activities of graduate students (although contribution to education is
THE EDUCATION OF GRADUATE SCIENTISTS AND ENGINEERS 56 FIGURE 3â3 Incidence of research assistantships and teaching assistantships among US PhDs, by broad field, 1993. SOURCE: NRC 1995 Table A-5. NOTES: 1993 doctoral recipients also reported many other sources of support (see Table 3â1).
THE EDUCATION OF GRADUATE SCIENTISTS AND ENGINEERS 57 FIGURE 3â4 Types of support for academic R&D, 1966â1992 (billions of 1987 dollars). SOURCE: NSF, 1994c: Table 5. NOTES: Research assistantships are included as part of R&D projects. Other includes R&D plant, scientists and engineers facilities, general scientists and engineers support, and other scientists and engineers activities.
THE EDUCATION OF GRADUATE SCIENTISTS AND ENGINEERS 58 FIGURE 3â5 Mix of federal support for academic scientists and engineers, 1966â1992 SOURCE: NSF, 1994c: Table 5 NOTES: Research assistantships are included as part of R&D projects. Other includes R&D plant, scientists and engineers facilities, general scientists and engineers support, and other scientists and engineers activities.
THE EDUCATION OF GRADUATE SCIENTISTS AND ENGINEERS 59 technically one of four criteria used to judge National Science Foundation grants). And a project or research environment of high educational merit will not necessarily impress a peer-review committee charged with judging the scientific merit of a proposed research topic and the ability of a principal investigator to carry it out. 3.6 CAREER INFORMATION AND GUIDANCE Graduate students do not routinely receive accurate, timely, and complete information on the array of available careers in science and engineering. Several government agencies and private organizations collect and publish information relevant to the careers of graduate students, including the National Science Foundation, the Bureau of Labor Statistics, and the National Research Council. Those data are of interest to three more or less distinct communities: â¢ The professionals who generate the data, including universities, professors, students, and professional societies. â¢ The National Science Foundation, which arranges and presents data to be used by others. â¢ A small number of people who study and use human-resources data. In general, the data that are available are not presented in formats designed for use by students or faculty advisers in choosing and planning careers in science and engineering. Moreover, in most cases, there is a lag of several years between the gathering of data and their publication. As a result, graduate students lack adequate information to New employment trends are already obliging some universities to pay more attention to PhD placement. Theodore Poehler, vice provost for research at Johns Hopkins University, told the committee that his university used to pay little attention to placement of graduate scientists and engineers. However, now that they are paying more attention, they are finding that one incentive for doing so is that increasing numbers of students are considering unconventional careers. For example, of the six new PhDs in one graduate program at Hopkins last year (three US women, three foreign men), two went to small companies, two went to postdoctoral positions, one had numerous offers from around the world, and one became a NASA program manager. Because of that experience, Johns Hopkins is trying to provide graduate students with more educational options to prepare them for a wider range of career opportunities. For example, the university offers more faculty and graduate students more opportunities for interdisciplinary research and education and for âlife- long learning.â In addition, when funding is available, the university encourages graduate students to travel to national meetings where they can present their research results and to workshops where they can meet representatives of small companies and other potential employers. Box 3â2: Enhancing Graduate Student Career Opportunities
THE EDUCATION OF GRADUATE SCIENTISTS AND ENGINEERS 60 â¢ Design their own education and career-development strategies. â¢ Gain a realistic understanding of employment prospects. â¢ Recognize likely future demand for scientists and engineers, by field. â¢ Understand the dynamics, structure, and evolution of the scientific-research system. More-effective guidance is clearly required. A prevailing belief in higher education is that faculty members ânaturallyâ know how to be dissertation advisers through their own experience as students and teachers. That might be true when it comes to advising students who will enter academic careers, but many (if not most) faculty members have little experience with or awareness of nonacademic job opportunities and so cannot be effective advisers for other students. Feniosky Pena, a doctoral candidate in engineering at MIT who is performing an internship with industry, told the committee that he experienced a troublesome culture gap when he began his studies. As a native of the Dominican Republic, he had been taught to respect authority. At MIT, he was reluctant to question his adviser, who in turn thought that Mr. Pena lacked a grasp of his subject. Furthermore, the adviser used a technique of persistent interrogation, which Mr. Pena found humiliating. He heard this difficulty described by others at minority-group conferences, where students told him that they felt âstupidâ when dealing with their advisers, classmates, or teachers. He suggested that if faculty were familiar with other cultures, such misunderstandings could be avoided. He said that minority-group students need more ânurturingâ if they are to reach a good understanding of the education environment in the United States. Mr. Pena added that the racial diversity that minority-group members bring to campuses is not valued by everyone. He suggested training for both the minority and the majority so that each gains a better understanding of the other's culture. Box 3â3: Minority Issues: the Culture Gap 3.7 THE GRADUATE EDUCATION OF WOMEN AND MINORITY-GROUP STUDENTS The presence of women and minority-group students, although increasing, is still small relative to the population as a whole in nearly all science and engineering fields. In the long run, it is in the interest of all to recruit a fair share of the most-able members of society into science and engineering. Meanwhile, efforts should continue to ensure that all people with talent have an equal opportunity to enter science and engineering careers. Women and minorities are underrepresented as graduate students and particularly as
THE EDUCATION OF GRADUATE SCIENTISTS AND ENGINEERS 61 faculty, researchers, academic officers, administrators, and policy-makers. The proportion of new entrants into the workforce who are minority-group members and women has risen and will continue to rise, and the quality and extent of their education should have high national priority. Linda Wilson, president of Radcliffe College, was asked by the committee to comment on issues pertaining to women in graduate science and engineering education. Dr. Wilson chairs the National Research Council's Office of Scientific and Engineering Personnel. Dr. Wilson said that the unsatisfactory position of women in graduate education indicates the need to change the system for both men and women, both minority and majority. In her view, the key elements requiring improvement are access, including expectations and encouragement; mentoring and career guidance; recognition and respect; and accurate information about career paths. She recommends changing the university into a more supportive culture, moving toward a âcontinuous learning system,â and maximizing our âhuman capital investmentâ by including more women and minority- group members throughout the science and engineering enterprise. Dr. Wilson said that key assumptions about graduate school are seldom tested, such as the notion that scientists do their best work when young and that independent work is more important than collaborative work. She suggested that more careful scrutiny of such assumptions might lead to constructive policy changes. Box 3â4: Improving access for women and minority-group students Statistically, the position of women students in advanced science and engineering is improving, in part because of special efforts. From 1982 to 1992, the total number of women in graduate schools rose by about 3% a year, compared with a rise of 1% a year for men. In 1982, women received 23.7% of science and engineering doctorates; in 1992, they received 28.5% (see Appendix B, Table B-22). In 1993, women constituted 33% of all full-time faculty (and 37% of combined full-time and part-time faculty) but only 6% of the full-time faculty in engineering, 20% in the natural sciences, and 27% in the social sciences (Table 6 in NCES, 1994). Women have been most successful at entering the social and life sciences. In 1992, 54% of graduate students in the social sciences and 44% of those in the life sciences were women (see Appendix B, Table B-3). Fewer women enroll in the physical sciences or engineering. In 1992, 15% of engineering graduate students and 27% of those in the natural sciences were women, but their percentage gains over the preceding decade have been greatest in those fields. Entry into science and engineering graduate schools is lowest among minority-group students. The percentage of science and engineering doctorates awarded to members of underrepresented minorities increased from only 4.1% in 1982 to only 5.5% in 1992 (see Appendix B, Table B-24). In 1992, fewer than 29,000 (9%) of science and engineering graduate students were US citizens who belonged to underrepresented minorities (black, Hispanic, and American Indian) (see Appendix B, Table B-4). That is related to their low representation on college faculties: 8% of full-time faculty in 1993â6% in engineering, 7% in the natural sciences, and more than 9% in the social science (Table 6 in NCES, 1994). By comparison, in 1991, 22% of Americans were black, Hispanic, or American Indian. Committee witnesses indicated that a âcritical massâ of students is particularly important for minority-group members, who as students often suffer from a âone and onlyâ syndrome.
THE EDUCATION OF GRADUATE SCIENTISTS AND ENGINEERS 62 As the demographics of the workplace shift rapidly, it is clearly in the national interest to encourage and facilitate the entry of women and minority-group members, with white men, into science and engineering fields. Overall enrollment of scientists and engineers grew by just over 2% a year from 1982 to 1992, but foreign enrollment grew by more than 5% a year. Foreign participation varies widely by field: non-US citizens make up 46% of all full-time graduate students in engineering, 39% of those in the physical and mathematical sciences, 27% of those in the life sciences, and 17% of those in the social sciences. In 1992, foreign students earned more than half the new PhDs in engineering (up from 39% 10 years earlier), more than one-third of the PhDs in physical and mathematical sciences, and one-fourth of those in the life sciences. Box 3â5: Distribution of Foreign Graduate Students 3.8 FOREIGN GRADUATE STUDENTS The number of foreign science and engineering students enrolled in US graduate schools and the number receiving PhDs have both risen more rapidly than the comparable numbers of US citizens. The number of science and engineering doctorates earned annually by people who are not US citizens and have temporary visas increased sharply from 3,400 in 1983 to almost 8,100 in 1993. This group received 18.5% of the doctorates in 1982 and 32% in 1992 (see Appendix B, Table B-25) and accounted for most of the net increase in the number of doctorates awarded since 1986 (see Figure 1â2). One reason for the marked increase has been a series of political events that have encouraged in immigration. The Immigration Reform Act of 1990 gave visa preference to applicants with science and engineering skills (NSB, 1993). The arrival of many of those students results from one-time political events, but American universities continue to attract students for whom comparable education is not available at home. The issues raised by the increase in the number of foreign students in American graduate schools and earning US doctorates are discussed in Chapter 4 (Section 4.2). As discussed in Chapter 4, the decision of an increasing number of those students to seek permanent jobs in the United States increases the talent available to our country, although it adds to the employment-related pressures on advanced scientists and engineers. 3.9 POSTDOCTORAL EDUCATION The postdoctoral population has increased faster than the graduate-student population. Some of the increase might be due to employment difficulties.
THE EDUCATION OF GRADUATE SCIENTISTS AND ENGINEERS 63 According to the latest National Science Foundation (NSF) survey of science and engineering graduate departments (unpublished), there were 24,024 science and engineering postdoctoral appointees2 in doctorate- granting institutions in the fall of 1992, compared with 14,672 in 1982âa 63.7% increase, compared with a 26.7% increase in the number of graduate students. Part of the growth can be assumed to reflect the legitimate need for postdoctoral study and exploration to prepare for the increased complexity of modern science; in biology, chemistry, and physics, for example, postdoctoral study has become the norm. But committee testimony and other anecdotal evidence indicates that many postdoctoral appointees are extending their studies because permanent positions in academic or industrial research are not available. An important additional factor is the increasing percentage of postdoctoral appointees who are foreign studentsâ53% in 1992, compared with 42% in 1985 (NSF, unpublished). More foreign citizens than American citizens have had postdoctoral appointments in US universities since 1991 (Tables C-29 and C-30 in NSF, 1993a). However, surveys do not determine the extent to which young scientists and engineers take postdoctoral positions because they cannot find regular employment. One measure of the impact of employment-market problems on the growth of the postdoctoral pool would be an increase in the length of postdoctoral time before a permanent position is found or an increase in the percentage of scientists and engineers who take second or third postdoctoral positions. Another indication would be an increasing percentage of scientists and engineers taking postdoctoral appointments at the institutions where they received their doctorates; this would indicate that professors are retaining their former students as RAs when they cannot find regular jobs. The Survey of Doctorate Recipients can be analyzed to address the question. The comparative analysis of cohorts of scientists and engineers 5â8 years after receipt of their PhDs, done for this report, indicated that the percentage still in postdoctoral positions grew from 2% in 1977 to 3% in 1989; the increase was greater and smaller in specific fields (see Appendix C, Table C-2).3 In 1979, more than 600 (4.9%) of the biologists who received PhDs in 1971â1974 held postdoctoral appointments; in 1989, nearly 1,300 (9.2%) of those with PhDs from 1981â 1984 were in postdoctoral positions. The percentage of each cohort in the faculty tenure system fell from 40% in 1979 to 25% in 1989. The above changes might partially explain the finding that the percentage of young biologists (aged 36 and younger) who applied to the National Institutes of Health for individual-investigator research grants fell by 54% from 1985 to 1993 (NRC, 1994a); clearly, fewer of them were in a position of independent investigator, from which they are permitted to apply for research grants. 2 Both numbers include foreign citizens, but the postdoctoral total includes doctorates in science or engineering from foreign universities. Part of the larger increase in the number of postdoctoral fellows over the last decade, therefore, might be ascribed to a greater propensity of foreign scientists and engineers to immigrate at the postdoctoral than the predoctoral stage, rather than to an increase in the pool of postdoctoral fellows who cannot find a job. 3 The 1989 data from the Survey of Doctorate Recipients are used because, owing to a change in the timing of the survey, the 1991 data on postdoctoral appointments are not comparable.
THE EDUCATION OF GRADUATE SCIENTISTS AND ENGINEERS 64 Regardless of the proportion of postdoctoral appointees who are in a vocational âholding pattern,â their numbers are rising, and each year they vie with the new class of graduating PhDs for available positions. The postdoctoral appointees have an advantage in being able to offer more research experience and publications in competing for available research positions. That competition, in turn, increases the trends among new PhDs toward postdoctoral study and nontraditional jobs.