Enrollment in U.S. graduate science and engineering (S&E) programs grew for almost 20 years, reached a peak of 436,000 students in 1993, and then began to shrink. From 1975-93, the overall number of students in graduate programs increased steadily at an average annual rate of 2 percent. Subsequent declining enrollment from 1993-97 has averaged 1.6 percent annually. Fewer students enrolling in engineering, mathematics, and computer sciences account for most of the decline. Engineering, mathematics, and computer science enrollments grew at a rate of almost 4 percent annually from 1975-92, but declined 3 percent annually from 1992-95. Engineering enrollment has continued to decline, while enrollment in mathematics and computer sciences increased slightly in 1996 and 1997. Trends differ when examining subfields: within the natural sciences, the physical sciences have decreasing graduate enrollment, while the biological sciences have increasing enrollment (NSF 1999a).
Graduate student enrollment in S&E, although shrinking, is becoming more diverse. In 1977, women represented only one-quarter of S&E graduate enrollment; by 1997, they represented 40 percent of enrollment. The increasing enrollment of minorities in graduate S&E programs partially stems from changing demographicsthe higher growth rate in the minority population relative to the white population. While women and minorities continued a decade-long trend of increased enrollment in graduate S&E programs, foreign students and U.S. citizen white males began a downward trend in their enrollment levels. (See appendix tables 1 and 2 and NSF 1999a.) The decline in foreign student enrollment in U.S. institutions is likely influenced by the increasing educational opportunities in other countries.
The overall trend in U.S. S&E programs at the master’s degree level shows rapidly increasing numbers of earned degrees throughout the 1980s and an even stronger growth in the 1990s. This growth is mainly accounted for by rising numbers of earned degrees in the social sciences and engineering, with relatively stable numbers in the natural sciences, mathematics, and computer sciences. (See appendix table 3.)
Over the 20-year period 1975-95, males accounted for the strong growth in master’s degrees in engineering, mathematics, and the computer sciences. Females were primarily responsible for the strong growth in social sciences; they also obtained a larger share of degrees in the natural sciences. The proportion of master’s degrees earned by females increased considerably in the last two decadesnot only in the natural sciences, but in engineering as well. In 1975, females earned 21 percent of the natural science degrees at the master’s level and almost 3 percent of the engineering degrees. By 1997, females accounted for 43 percent of the natural science degrees and 16 percent of engineering. (See appendix table 3.)
In the 1990s, minority groups in the United States earned, in most cases, increasing numbers as well as increasing shares of master’s degrees in S&E fields. The number of S&E degrees earned by Asian/Pacific Islanders consistently increased, especially in engineering, mathematics, and the computer sciences. The number of S&E master’s degrees obtained by blacks grew modestly in most fields, with strong growth in the social sciences. Hispanics earned a moderately increasing numberand proportionof degrees in the social sciences, as well as in engineering. White students showed modest growth in natural science and engineering degrees in the 1990s and strong growth in the social sciences. Notwithstanding these gains, the share of master’s degrees earned by white students in all fields declined during the 1977-97 period. (See appendix table 4.)
Analysis of master’s degrees by citizenship shows a trend toward a larger proportion of degrees going to foreign students in engineering, mathematics, and the computer sciences. In 1977, foreign students earned 22 percent of the engineering degrees and 11 percent of the mathematics and computer science degrees. By 1995, foreign representation at the master’s level was 34 percent in engineering and 35 percent in mathematics and computer sciences. The rate of growth of overall S&E master’s degrees obtained by foreign students slowed somewhat in the 1993-96 period, mainly due to a leveling off of their earned degrees in mathematics and the computer sciences. (See appendix table 4.) Engineering degrees awarded to foreign students declined in 1997, echoing the decline in foreign graduate enrollment in engineering from 1993-96. (See appendix table 2.)
A decade of relatively stable production of S&E doctoral degrees granted in the United States from 1975-85 was followed by a decade of increasing production of such degrees; in 1996, over 27,000 S&E doctorates were awarded. Large increases in the numbers of earned degrees were evident in engineering, mathematics, and the computer sciences. The number of degrees in these fields doubled from 1985-96. (See figure 1.) The natural science fieldsparticularly the biological sciencesalso contributed to the rising number of degrees during this period, increasing by 25 percent (NSF, 1999d.)
Male doctoral students accounted for much of the growth in engineering, mathematics, and the computer sciences; female doctoral recipients were largely responsible for the increasing number of natural science degrees. Within the past two decades, the share of S&E doctorates earned by women doubled, rising from almost 16 percent in 1975 to 33 percent in 1997. The proportion of increase has differed by field. By 1997, females earned half of the doctoral degrees in the social sciences and 40 percent in the biological sciences. Growth in the proportion of degrees awarded to women was greatest in engineering subfields. By 1997, women earned 12 percent of all engineering degrees, and 16 to 18 percent of doctoral degrees in chemical and material engineering. (See appendix table 5.)
Underrepresented minorities within U.S. universities received over 7 percent of all S&E doctorates awarded to U.S. citizens and permanent residents in 1995; this was up slightly from 4 percent in 1977. As a group, these minorities received 6 percent of earned degrees in the natural sciences, 4 percent in mathematics and the computer sciences, 10 percent in the social sciences, and 6 percent in engineering. For black Ph.D. recipients, the largest numerical increases in the past decade have been in the biological and social sciences. The largest percentage increases have been in the biological sciences and engineering. (See appendix table 6.)
The Committee on Science, Engineering, and Public Policy (COSEPUP) of the National Academy of Sciences recently reviewed U.S. graduate programs in S&E. The resulting report, Reshaping the Graduate Education of Scientists and Engineers (COSEPUP 1995), recommends broadening the education of doctoral students to better meet their actual career needs. The report noted that the current focus of doctoral programs on research training in a narrow discipline gradually evolved over previous decades when the demand for research was rising. U.S. R&D spending increased rapidly from the late 1970s to the latter part of the 1980s; consequently, doctoral R&D employment increased by almost 5 percent annually. Today, howeverthe report goes on to explainan even smaller minority than previously will enter academic research. Only one-third of future doctoral recipients in S&E will enter the tenured academic system; two-thirds will be employed in nonacademic settings. The report concludes that doctoral course offerings should be expanded to reflect the diversity and complexity of these employment options. What these options will all require is the ability to apply an advanced understanding of science and engineering to societal needs. Consequently, S&E doctoral students will need:
A variety of graduate reforms predated or stemmed from the recommendations of the COSEPUP report. These reforms focus on the education needs of students. Graduate programs are being expanded to include not only multidisciplinary coursework, but also to answer to students’ needs for business and teaching skills. The Council of Graduate Schools has held a series of national discussions with graduate deans about the need to prepare students more effectively for their roles as future faculty. Subsequently, the 1997 meeting of the National Science Board on the Federal Role in Graduate and Postdoctoral Programs recommended Federal encouragement to universities to increase diversity and the appropriate broad training of the S&E labor force (NISE 1998).Forces for Change
Underlying these policy studies are a variety of forces for graduate education reform. These include recent demographic, economic, technological and social changes, as well as the increasing complexity of viable solutions to real-world problems.
Among the demographic forces for change is a larger number of women and minorities earning bachelor’s degrees in S&E fields for potential recruitment into graduate S&E programs (along with a declining population and enrollment of whites and declining enrollments of foreign students). Emerging reforms that build on this demographic trend are graduate enhancement programs for underrepresented minority students and recruitment and retention programs for women in science and engineering. For example, Rice University initiated a graduate program for increasing diversity in computational sciences, and the University of Arizona and Notre Dame University promote the Graduate Education for Minorities Consortium (GEM) of industries, colleges, and universities to increase minority recruitment and retention (NISE 1998).
Economic and technological forces are combining to influence changes in graduate education. Spiraling education costswhich are increasing faster than the cost of livingare contributing to the growth of proprietary (for-profit) universities with cost-effective programs. The capital expense of major research programs is necessitating shared research facilities. Collaborative agreements among consortia of universities are being made to ensure efficient use of resources and expertise of graduate faculty. For example, in a new doctoral program in technology management, a consortium of nine universities across eight states links the top laboratories and faculty of key technical specializations (such as digital communication systems and industrial composite materials). This arrangement allows the participants to ensure the broad education needed to manage such advanced technologies (NISE 1998).
Another force for change is technology. Information technologies and distance learning technologies are changing how instruction can be given. For example, Engineering Research Centers supported by the National Science Foundation (NSF) are developing multidisciplinary engineering curricula through interactive instructional modules. (These centers are briefly described below under “Background: Federal Support for S&E.”) These modules can assist in teaching principles of diverse subjects using graphics, diagrams, and animation to convey key concepts, along with interactive exercises for practicing the principles’ application. Through alternative instructional delivery systems, both graduate students in university classrooms and researchers within private companies can use this software.[Skip Text Box]
|Background: Federal Support for S&E
Scientists played a key role in World War II within Federal defense research sites; following the war, policymakers chose to support scientists within universities. The Vannebar Bush Report stated that an increasing number of highly qualified scientists and engineers would be crucial to the U.S. economy, and recommended public support of advanced students in science and mathematics within universities. That policy produced significant Federal support for university-based S&T research and the training of scientists and engineers. These funds increased further following Sputnik, the Cold War, and the creation of the National Institutes of Health (NIH) and the National Science Foundation. By the early 1960s, NIH funding of university research exceeded total funding of university-based research by the Department of Defense.[*] This compact between the Federal Government and universities has continued to the present, with Federal academic R&D reaching $21 billion (in 1992 constant dollars) in 1996 (NSB 1998).
The growing demand for public accountability is driving the U.S. educational system to improve instruction in mathematics and science. At the graduate education level, this demand for accountability is focused on the improvement of teaching, with an increased focus on the educational and career needs of students rather than the research needs of faculty. Several universities have initiated efforts to improve both graduate and undergraduate instruction in science and engineering, such as Preparing Future Faculty programs and training for teaching assistants (NISE 1998).
Another dynamic for change is an emerging demand for broadly educated Ph.D. recipients who are able to address the complexity of real-world problems and contribute to their solution. For example, at a recent forum for graduate education reform, the director of research for the U.S. Department of Energy explained that the departmentwhich is one of the largest Federal supporters of basic research in the natural sciencesneeds an S&T workforce that can flexibly cross disciplines to solve complex problems in several mission areas. Issues that need to be addressed by the department include the security of existing nuclear stockpiles, the development and use of new energy technologies, the health and environmental effects of energy use, and structural genomics (which combines the disciplines of biology and informatics) in the human genome program (NISE 1998).
The above innovationsas well as new multidisciplinary programs and other efforts to broaden the preparation of graduate studentswere addressed at a recent National Institute for Science Education, University of Wisconsin at Madison, forum on graduate education. For more information, see NISE (1998).
During the course of their graduate careers, most S&E students are likely to be involved in some type of research activities. S&E graduate students thus play a unique role in the U.S. academic research system, in that they are both an input to and an output of this system. U.S. research universities have traditionally coupled advanced education with research, thereby generating new knowledge and producing advanced S&E talent. This complex, symbiotic relationship is exemplified by the variety of support mechanisms and sources through which financial resources are provided to S&E graduate students. Support mechanisms include fellowships, traineeships, research assistantships, and teaching assistantships. Sources of support include Federal agency; non-federal support (from academic institutions, state and local governments, foreign governments, nonprofit institutions, and industrial firms); and self-support (from loans or personal or family financial contributions). Most graduate students are supported by more than one source and mechanism during their time in graduate school; they also often receive support from several different sources and mechanisms in any given academic year.
The recent enrollment declines reported earlier for all S&E graduate students affected the number of full-time students in 1995. For the first time in almost two decades, enrollment of full-time S&E graduate students declined slightly in 1995. A 12-year trend of steady increases in enrollment of full-time graduate students whose primary source of support was the Federal Government also ended, as did an even longer upward trend in the number of graduate students whose primary source of support was from non-federal sources. For more information on Federal support, see sidebar on Background: Federal Support for S&E. The number of self-supported graduate students also declined for the first time since 1988. (See appendix table 7.)
Since 1980, there have been significant shifts in the relative usage of different types of primary support mechanisms. (See figure 2.) These shifts have been due more to rapid growth in some support mechanisms than to an absolute decline in the number of students supported by any of these mechanisms. The proportion of graduate students with research assistantships as their primary support mechanism increased from 22 to 27 percent between 1980 and 1995. This increase was offset by drops in the proportions of students supported by traineeships (from 7 to 5 percent) or by teaching assistantships (from 23 to 20 percent). Most of these changes had occurred by the late 1980s, with proportional shares being relatively stable during the first half of the 1990s. The proportion supported by fellowships fluctuated between 8 and 9 percent between 1980 and 1995; that with self-support as the primary mechanism fluctuated between 28 and 32 percent. These overall shifts in support mechanisms were evidenced for both students supported primarily by Federal sources and for those supported by non-federal sources. (See appendix table 7.)
The proportions of full-time S&E graduate students with primary support from various sources and mechanisms differ for private and public universities. (See figure 3.) A larger proportion of full-time graduate students rely primarily on self-support in private academic institutions as opposed to those in public institutions39 versus 30 percent in 1995.
Non-federal sources are the primary source of support for a larger proportion of students in public institutions (50 percent) than in private ones (41 percent). At both private and public institutions, about 20 percent of students receive their primary support from the Federal Government.
A larger proportion of students attending public academic institutions rely on research assistantships and teaching assistantships as their primary support mechanism (30 and 23 percent, respectively) than those attending private institutions (21 and 13 percent, respectively). This is balanced by greater reliance on fellowships and traineeships in private institutions (14 and 8 percent, respectively) than in public ones (7 and 4 percent, respectively).
Research Assistantships. Although research assistantships accounted for 27 percent of all primary support mechanisms in 1995, their role differed across S&E fields. They comprised more than 50 percent of the primary support mechanisms for graduate students in astronomy, atmospheric sciences, oceanography, agricultural sciences, chemical engineering, and materials engineering. They accounted for less than 20 percent in all the social sciences, mathematical sciences, and psychology. (See appendix table 8.)
Just as the significance of research assistantships differs across fields, so too does that of the Federal Government as the primary source of support for research assistantships. Overall, the Federal Government was the primary source of support for about half of graduate research assistants. However, it was the primary source of support for 75 percent of the research assistants in the physical sciences, just over 60 percent in both the environmental and computer sciences, but only 20 percent in the social sciences and 32 percent in psychology. (See appendix table 9.)
Teaching Assistantships. Teaching assistantships accounted for 20 percent of all primary support mechanisms in 1995. But they comprised more than 30 percent of the primary support mechanisms for graduate students in chemistry, physics, mathematics, and earth sciences; and less than 12 percent in the atmospheric sciences, oceanography, agricultural sciences, medical sciences, aeronautical engineering, and materials engineering. (See appendix table 8.) The Federal Government has an almost negligible role in supporting teaching assistantships.
Fellowships and Traineeships. Although fellowships accounted for only 9 percent of all primary support mechanisms in 1995, they are a much more important mechanism of primary support for students in the history of science, anthropology, and astronomy where they comprised 37, 20, and 17 percent of the primary support mechanisms, respectively. Students with traineeships as their primary support mechanism accounted for just under 5 percent of all full-time S&E graduate students in 1995. For students in the biological sciences, medical sciences, and other life sciences, however, traineeships accounted for between 11 and 14 percent of primary support. (See appendix table 8.)
The Federal Government was the primary source of support for about one-quarter of all graduate students with a fellowship as their primary mechanism of support and for about two-thirds of those with a traineeship as their primary mechanism of support. The Federal Government was a more important primary source for fellowships to graduate students in the atmospheric sciences, aeronautical engineering, and astronomy, providing 63, 56, and 50 percent, respectively, of the primary fellowship support. In contrast, it provided only 14 percent of primary fellowship support in the social sciences. The Federal Government provided almost 80 percent of primary support for traineeships in the life sciences, compared to 24 percent in computer sciences and 21 percent in the social sciences. (See appendix table 9.)
Self-Support. About one-third of full-time S&E graduate students were supported primarily by loans or from personal or family financial contributions. The importance of this type of support also differed across S&E fields. About 40 percent of students in the computer sciences, medical sciences, anthropology, and industrial engineeringand more than 50 percent of those in psychology and political sciencerelied on self-support as their primary support mechanism. Conversely, less than 10 percent of the students in astronomy, chemistry, physics, and the atmospheric sciences relied on self-support as their primary support. (See appendix table 8.)
There has long been great interest in whether the amount and type of financial support given to graduate students has an effect on degree completion rates, time to degree, and productivity and success in the labor market. How effective have the large investments in graduate education made by government, academia, and the private sector been? How do the various modes of supportteaching assistantships, research assistantships, fellowships, and subsidized loanscompare in terms of recipients’ educational and career outcomes?
Hypotheses of Relative Merits. The merits of various support mechanisms have been discussed and a number of hypotheses developed about the advantages and disadvantages of different mechanisms. In fact, some of the characteristics of a specific mechanism cited as disadvantages by some individuals are cited as advantages by others. For instance, the portability of fellowships and the independence they give to graduate students are seen by some as a distinct advantage because they provide these students with great freedom to pursue a wide variety of interests. Others argue that students with fellowships are more likely than those supported by traineeships or research assistantships to become isolated from their peers and from the faculty in their departments; they thus may either be less likely to complete their Ph.D. or to take longer to do so. Some argue that although having a fellowship at the beginning of one’s graduate career may be detrimental, having one when working on a dissertation is highly advantageous.
Similarly, some hold that since research assistantships are directed to the needs of funded research projects, doctoral students can become so involved on a specific project that they have little time for independent exploration or other educational activities, thus limiting the areas in which they acquire experience. A counter argument is that the research skills and experience students acquire by focusing on a specific project are indispensable to the high-quality, state-of-the-art research being conducted at U.S. universities and industrial laboratories; students with research assistantships thus may complete doctoral dissertations more frequently and faster than those with other forms of support. Some argue that strong reliance on research assistantships can bias research and graduate training toward those areas that have long track records rather than to new and innovate areas, and that they also may prevent beginning faculty from attracting graduate students. Others argue that it is the widespread availability of research grants that provides young faculty the opportunity to work closely with graduate students.
Lack of Quantifiable Data. Unfortunately, it is extremely difficult to examine many of these hypotheses analytically either because of the absence of data or the inability to capture the hypothesized outcomes quantitatively. In addition, most graduate students depend on multiple sources and mechanisms of support while in graduate school, and frequently on different sources and mechanisms in different phases of graduate work. This makes it quite difficult, if not impossible, to identify a one-to-one relationship between a student and a support source or mechanism.
Furthermore, there is a selection problem that is not easily overcome. Most external organizations and graduate institutions award financial support based on merit. In addition, the type of support that a student receives is affected by a graduate department’s view (and perhaps sometimes by the student’s own view) of the student’s relative ability to teach or to support research. If students receiving support have more ability or motivation than other students, the former are likely to be more successful than the latter irrespective of the effects of support mechanisms. To the extent that graduate support allocation decisions are successful in sorting students by merit and aptitude, it becomes more difficult to statistically isolate the effect of receiving graduate support from the effects of other student differences.
General Conclusions. Despite these difficulties, various studies have looked at some aspects of graduate support and student outcomes. A recent review of this literature summarized the results as follows (Bentley and Berger 1998):
Appendix table 10 shows the distribution of those in S&E occupations in the United States. Of the 11.5 million people with some kind of S&E degree, only 3.2 million are in jobs strictly labeled as science and engineering. Of these, nearly two-thirds are employed by private, for-profit employers. By this strict occupational measure of S&E workers, Ph.D. recipients make up 13 percent of the U.S. S&E workforce. If the definition were extended to include all workers with S&E degrees, the proportion of doctorate-holders would fall to 4 percent.
In the past decade, foreign students have accounted for the large growth in S&E doctoral degrees in U.S. universities. The number of foreign S&E doctoral recipients graduated from U.S. universities doubled from over 5,000 in 1986 to 10,000 in 1996. This doubling translates to an 8-percent average annual increase. In contrast, the rate of increase in doctoral degrees to U.S. citizens averaged less than 2 percent annually (NSB 1998).
Within natural science and engineering fields, the proportion of doctoral degrees earned in U.S. universities by foreign citizens climbed from 25 percent in 1985 to 33 percent in 1994; it has since begun to level off. In 1997, the share of natural science and engineering degrees earned by foreign students decreased slightly to 31 percent. This drop was mainly due to a decline in doctoral degrees earned by South Korean and Taiwanese students. Both of these economies (which are major contributors of foreign graduate students to the United States) have increased their internal capacity for graduate education in S&E, evidenced by the increasing number of in-country doctoral degrees in these fields (NSB 1998).
Even as Asian students entered U.S. graduate programs in record numbers, Asian universities were expanding their own doctoral degree programs in S&E fields. These two phenomena are related. The desire to increase in-country capacity to educate students through the doctoral level necessitated sending students abroad so as to prepare more S&E faculty for expanded graduate programs within Asian universities. For the period 1988-94, the Asian effort to receive doctoral training in U.S. universities was particularly intense, as evidenced by an increase from 2,872 earned degrees in 1989 to 6,229 in 1994. The annual rate of growth in S&E doctoral degrees earned by Asian students during this period was over 17 percent. However, this rate of growth has slowed considerably in the last few years, and in 1997, the number of degrees earned by Asian students within U.S. universities declined.
Although Ph.D. production in S&E fields is growing at a faster rate in Asian countries than in the United States, the Asian base is lower. In 1997, 18,513 S&E doctoral degrees were earned in five Asian countries. In that same year, U.S. universities produced almost 27,000 S&E doctorates; however, over 5,500 of these degrees were earned by foreign students from Asia. In 1997, the number of doctoral S&E degrees earned at universities within four Asian economies exceeded the number of such degrees earned by Asian foreign students at U.S. universities. Only for Taiwan do U.S.-earned doctoral degrees outnumber those earned within Taiwanese universities. (See figure 4 and text table 1.)
Technology transfer is often said to occur best through people. Thus, the mobility of foreign students throughout Europe, Asia, and the Americas is a significant source of diffusion of S&E knowledge in the world. NSF statistical data are limited to certain patterns of mobility to the United States. The Survey of Earned Doctorates captures the number of S&E doctoral degrees earned by foreign students, students’ planned location after completing their degrees, and any firm offers they've received of U.S. postdoctoral study or employment. The Scientists and Engineers Statistical Data System (SESTAT) captures the extent of the contribution of foreign-born scientists and engineers to the U.S. labor force. Little is known, however, of the return flow of foreign students and the contribution they make to build the S&T infrastructure in their home countries. Little is also known of those foreign graduate students who do not complete a doctoral degree. For example, Japanese industry sends its research personnel to top U.S. universities for 1 to 2 years of advanced study in particular fields (NSF 1997).
The diffusion of S&T knowledge may also occur through networking, without physical relocation of scientists and engineers for extended stays. Choi (1995) has shown extensive networking by Asian-born faculty and researchers working in the United States to advise, disseminate information, and assist in building their home country S&T infrastructure. This tendency is particularly true for foreign-born faculty in S&E departments. In 1993, foreign-born faculty in U.S. higher education accounted for 37 percent of engineering professors and over a quarter of mathematics and computer science teachers. More research is needed on the extent of this diffusion of S&E knowledge through exchange visits or electronic dissemination.
Cooperative research and information technologies are also diffusing S&T knowledge. International cooperative science programs often provide support for immigrant scientists and engineers to collaborate with home country scientists and to advise on building up a research area in a particular area of interest. For example, many of the grantees in the NSF U.S.-China Cooperative Science Program are Chinese American scientists and engineers who are most able to work effectively within the Chinese environment. Electronic dissemination through the Internet is allowing the dissemination of innovative teaching modules as well as specific information needed by home country S&T institutions.
Until 1992, around half of the foreign students who earned Ph.D.s in S&E in U.S. universities planned to locate in the United States after completing their degree. A significantly smaller proportion (one-third) received firm offers to remain in the United States for academic or industrial employment. The proportion of foreign doctoral recipients who plan to locate in the United States and accept firm offers differs considerably by country and region. Students from Asia, who are the most numerous, also represent the largest percentage who plan to locate in the United States. In contrast, students from North and South America, who are the least numerous, have a smaller proportion planning to locate in the United States.
For the period 1992-96, the proportions of foreign doctoral recipients planning to remain in the United States increased: over 68 percent planned to locate in the United States, and nearly 44 percent had firm offers to do so. This recent increase in stay rates, which may be temporary, is mainly accounted for by the sharp increase in the percentage of Chinese students with firm plans to stay in the United States. In 1990, 42 percent of the approximately 1,000 Chinese doctoral recipients in U.S. universities had firm plans to stay. By 1996, 57 percent of the nearly 3,000 Chinese doctoral recipients from U.S. universities had firm plans to remain in the United States. The underlying cause for this shift is the large number of Chinese students granted permanent residence status in the United States in 1992, following China’s response to student demonstrations. Selected countries in Europe (Eastern Europe) and the Americas (Canada), however, also increased their stay rates after completing advanced degrees from a U.S. university. Their numbers are small in comparison to Asia’s: 200 from Eastern Europe and 100 from Canada.
Among Asian countries, China and India apparently have a limited capacity to provide high-level employment to large numbers of returning S&E doctorate-holders. In 1996, 57 to 59 percent of the U.S. S&E doctoral recipients from these countries choose to accept further study or employment in the United States. In contrast, only a small percentage of 1996 doctoral recipients from South Korea and Taiwan (24 and 28 percent, respectively) accepted offers in the United States. The trend in the 1990s has been for relatively few doctoral recipients from these countries to remain in the United States; this is particularly true of South Korean engineering doctoral recipients (NSF 1998). (See figure 5.)
To a large extent, the definite plans of foreign doctoral recipients to remain in the United States revolve around postdoctoral study rather than employment. Among students born in those countries accounting for the largest numbers of foreign doctoral awards, the majority of definite plans to remain in the United States were for further study (58 percent on average between 1988 and 1996); followed by employment in R&D (27 percent); teaching (7 percent), or other professional employment (8 percent).
A recent study of foreign doctoral recipients working and earning wages in the United States (Finn 1997) shows that about 47 percent of the foreign students who earned doctorates in 1990 and 1991 were working in the United States in 1995. The percentages are higher in the physical sciences and engineering, and lower in the life and social sciences. These stay rates differ more by country of origin than by discipline, however. A very large percentage of the 1990-91 foreign doctoral recipients from India and China were still working in the United States in 1995. In contrast, only 10 percent of South Koreans who earned engineering doctorates from U.S. universities in 1990-91 were working in the United States in 1995.
Foreign doctoral recipients from 1970-72 were also examined in the same study. Finn estimated that 47 percent were working in the United States in 1995, and that the stay rate for that group had fluctuated around 50 percent during the 15 years leading up to 1995. There is no evidence of significant net return migration of these scientists and engineers after 10 or 20 years of work experience in the United States. This does not mean that there is not significant return migration: such migration is known to occur. However, the fairly constant stay rates indicate that any tendency of the 1970-72 cohorts to leave the United States after gaining work experience here has been largely offset by others from the same cohorts returning to the United States after going abroad.
In total, there were 135,000 foreign-born S&E doctoral recipients working in the United States in 1993. (See text table 2 and appendix table 12.) They accounted for 25.6 percent of all U.S.-employed S&E doctorate-holders. Academia is the largest sector of employment for foreign-born S&E doctorate-holders. In industry, however, they actually make up a larger proportion of total S&E doctoral recipients: nearly one-third.
Asia was the place of birth for over half of the foreign-born S&E doctorate-holders working in the United States76,000. Although this number is for the whole Asian continent, the two largest source countries combinedChina and Indiaprovided more S&E Ph.D. recipients to the U.S. labor force than all of Europe.
In 1995, at least 19,600 U.S. native-born naturalized citizen and permanent resident Ph.D. scientists and engineers lived outside the United States (text table 3). These included:
Not included are U.S. citizen Ph.D. scientists who held only a temporary student visa or work visa when they received their doctorate; it may be reasonable to assume that this group is as likely to work outside the United States as those who had already been naturalized by the time of degree.
The likelihood of foreign residence for U.S. natives is greatest for those with the most recent degreesranging from 2 percent of native-born doctorate-holders who received their Ph.D. between 1945 and 1954 to 3 percent of those who received their doctorate between 1985 and 1994. By field, the proportion of native-born Ph.D. recipients resident in foreign countries is greatest in the mathematical and computer sciences and in the social sciences (4 percent for each). It is lowest in the physical sciences.
Good estimates of the number of U.S. scientists and engineers who work abroad are not available, and the numbers presented here should be treated as lower bound estimates.
Bentley, Jerome, and Jacqueline Berger. 1998. The Effects of Graduate Support Mechanisms: A Literature Review. Prepared by Mathtech for the National Science Foundation under Contract #SRS97317954. Arlington, VA: National Science Foundation.
Burrelli, Joan S. 1998. "Graduate Enrollment of Women and Minorities in Science and Engineering Continues to Rise," Data Brief. Division of Science Resources Studies NSF 98-302. Feb. 23
Bush, Vannevar. 1945. “Science, the Endless Frontier,” Report to President Roosevelt.
Committee on Science, Engineering, and Public Policy (COSEPUP). 1995. Reshaping the Graduate Education of Scientists and Engineers. Washington, DC: National Academy Press.
Finn, M.G. 1997. Stay Rates of Foreign Doctorate Recipients from U.S. Universities, 1995. Oak Ridge, TN: Oak Ridge Institute for Science and Education.
National Center for Education Statistics. 1997. Earned Degrees and Completion Surveys. Washington, DC.
National Institute for Science Education (NISE). 1998. “Graduate Education Forum—Strengthening Graduate Education in Science and Engineering: Promising Practices and Strategies for Implementation.” Forum held June 29-30, Arlington, VA. http://nise.wcer.wisc.edu/GradForum.
National Science Board (NSB). 1998. Science & Engineering Indicators 1998. NSB 98-1. Arlington, VA: National Science Foundation.
National Science Foundation (NSF), Division of Science Resources Studies. 1997. The Science and Technology Resources of Japan: A Comparison with the United States. NSF 97-324. Arlington, VA.
______. 1998. Statistical Profiles of Foreign Doctoral Recipients in Science and Engineering: Plans to Stay in the United States. NSF 99-304. (Arlington, VA.)
______. 1999a. Graduate Students and Postdoctorates in Science and Engineering: Fall 1997. NSF 99-325. (Arlington, VA).
______. 1999b. Science and Engineering Degrees: 1966-95. NSF 97-335, by Susan Hill (Arlington,VA, 1997).
______. 1999c. Science and Engineering Degrees, by Race/Ethnicity of Recipients, annual series (Arlington, VA).
______. 1999d. Selected Data on Science and Engineering Doctorate Awards: 1997, NSF 99-323. (Arlington, VA).
Stanford Today. 1998. “The Cold War Era and the Modern University.” July-August: pp. 42-47.