Having read accounts of scarcity in academic research positions, some readers might expect this report to conclude that we are producing too many PhDs and should take immediate steps to cut back PhD programs in science and engineering. We are aware of the reports of unemployment and underemployment among new doctorate recipients, and survey data indicate that recent PhDs are finding it harder to make the transition from graduate school and postdoctoral study to career positions (see Chapter 2). The current situation probably results in part from the increase in annual PhD production to 25,000 in 1993 from the 18,000-19,000 per year in 1976-1986. In response, some graduate programs have begun to accept fewer new students. But forecasting demand for science and engineering PhDs is difficult, and, because it takes a long time for changes in graduate enrollment to manifest themselves in PhDs, past efforts either to increase production in response to perceived shortages (in the 1960s and 1980s) or to reduce production (in the 1970s) ended up not having beneficial effects, in that graduate students and programs had already made substantial adjustments. Because of the lag times between policy action and changes in the system and for reasons enumerated below, we do not think it possible to determine appropriate production targets. A better way to keep supply and demand in balance appears in the next chapter.
The committee is not convinced that the current low and stable unemployment
rates among science and engineering PhDs, even new ones, that are documented in
Chapter 2 prove that the system is working as well as it should. It is true
that science and engineering PhDs have prospered in an increasingly diverse
labor market. But as we illustrated in Chapter 2, there are indications of
employment difficulties, especially for recent graduates. For example, the
percentage of scientists and engineers looking for jobs in the first months
after PhD receipt has risen dramatically in some fields, and there is evidence
that an increasing percentage of those counted as "employed PhDs" have taken
temporary positions in either postdoctoral fellowships or short-term jobs. The
unemployment rate as of 1993 (the last year for which there were national data)
was still low at 1.6% but was increased from the roughly 1.0% of the 1980s and
1.4% in 1991 (Figure 2-3). Unemployment among new science and engineering PhDs
reached 2% in 1993, compared with the roughly 1.5% of the 1980s.
Nor do the available employment data take into account the nature of jobs held
by recent PhD recipients. Statistically, a PhD physicist working in a job
outside science and engineering is counted equally with a physicist on the
staff of AT&T Bell Laboratories or a tenure-track assistant professor at a
research university. Moreover, some PhDs who are finding good jobs in
nontraditional fields might be doing so regardless of their PhD training, not
because of it. The predominant view of the employers that we heard from during
the course of our study was that PhD work, including original research, made
students more effective employees. However, these graduates might be
attractive to some employers simply because they are members of a highly
qualified, hard-working, and carefully selected group of people. The time
spent in or the content of a PhD program might not be well matched to some
science and engineering graduates' jobs.
The committee cannot measure employment difficulties precisely, but the
evidence received from witnesses and other contributors is persuasive that
problems exist in at least some sectors. Some recent PhDs have indicated that
they regretted having spent time and money on doctoral work that turned out not
to be useful in their permanent jobs. Some even reported "hiding" their
doctorates so as not to appear overqualified, unbusinesslike, or too
theoretical in their approach to work.
We believe that the slow but steady shift 1 in
demand for doctoral scientists and engineers over the last 2 decades away from
academe and toward a greater variety of employment has accelerated somewhat in
the early 1990s at the same time that the number of new graduates (many of them
foreign students) has increased rapidly. For a variety of reasons, the number
of academic positions and traditional industrial research positions is steady
or shrinking, in accord with anecdotal reports that an unusually high number of
new PhDs had to change career plans on graduating or after several years in
postdoctoral positions.
To some extent, the science and engineering employment situation is cyclical,
and it might already be adjusting. The recent recession ended slowly, but
economic growth has resumed, and the demand for skilled people is increasing
even in some industries that have undergone substantial reduction and
restructuring. In addition, the high rate of increase in the number of PhDs
awarded to foreign citizens in the United States, which averaged more than 12%
per year in the late 1980s, began to fall after 1990 and was 0.3% in 1993
(calculated from Table 3 in NSF, 1994f). The number of doctorates awarded to
foreign citizens with temporary visas fell slightly in physics/astronomy,
chemistry, environmental sciences, and computer sciences from 1992 to 1993
(NRC, 1995: Appendix Table A-2).
However, we have already cited some indications of basic structural changes
that lower demand, including cuts in defense spending, industrial
restructuring, and reductions in growth of federal R&D spending. There is
no evidence that these trends of the last several years will end soon. Thus,
even if PhD production does fall in the near term, science and engineering
graduate students might do well to prepare themselves for an increasingly
diverse set of career paths.
Supply-demand models are not now adequate for predicting whether there will be
an undersupply or oversupply of trained scientists and engineers (Fechter,
1990; Leslie and Oaxaca, 1990; NSB, 1993; Vetter, 1993). That conclusion was
also expressed by the panel on estimation procedures of the Committee on
National Needs for Biomedical and Behavioral Research Personnel, which found
that previous supply-demand models for basic biomedical, behavioral, and
clinical research scientists had not proved accurate (NRC, 1994b).
At least two types of limitations of such models severely reduce their
reliability, especially over the 5- to 10-year periods needed to carry out
graduate-education plans. Internally, they are not based on an adequate
understanding of the behavior of the students, faculty, and other people whose
collective decisions affect the supply of new scientists and engineers;
externally, they cannot always predict the impact of major changes in key
variables outside the graduate system itself that affect demand for scientists
and engineers.
For example, predictions of a huge oversupply of scientists and engineers in
the early 1970s did not come true, because as a result of the predictions the
students changed plans, administrators reduced programs, and graduates found
new ways to use their training--all behavioral changes that were not included
in the models. More recent studies have forecast shortages of college and
university faculty, beginning in the middle 1990s. These shortages have not
occurred. The forecasters could not anticipate the behavioral effects of
recession and tight government budgets: fewer faculty have elected to retire,
and universities and colleges have begun to fill faculty openings temporarily
or leave them unfilled.
As an example of unanticipated external events affecting science and
engineering employment, the buildup of physical scientists and engineers in the
late 1980s stimulated by increased defense spending earlier in the decade was
followed by the end of the Cold War, which reduced demand for scientists and
engineers; similarly, no one could predict the immigration of experienced
scientists and engineers from the former Soviet Union and eastern Europe. The
intensified pressures of international economic competition have also had
unexpected effects, which have led some large high-technology companies to
reduce their research staffs and redirect those who remain toward more-applied
research with near-term payoffs.
With current techniques, it is not possible to forecast the future demand for
or supply of scientists and engineers. We can tell with some confidence
whether there are immediate mismatches between supply and demand; but in the
absence of reliable long-range models, we do not know whether a situation is
temporary and self-correcting or whether stronger action is required. In other
words, there is little basis for trying to control the production of new
science and engineering PhDs by limiting enrollments nationally through some
central control mechanism.
There are ways to improve the likelihood of a balance between supply and
demand that do not involve central planning and all the information
requirements on which such planning depends. We believe that a combination of
greater breadth and flexibility in graduate curriculums, better information and
guidance, and financial support mechanisms whose primary purpose is education
will provide scientists and engineers who can move more flexibly toward
employment demand. Our recommendations are presented in greater detail in
Chapter 5.
Meanwhile, efforts should continue to improve the collection and analysis of
employment-related information by the National Science Foundation (NSF), other
agencies, and the scientific societies and associations. Understanding the
dynamics of and trends in career paths of scientists and engineers with
advanced degrees in the various employment sectors is especially important.
The results should be disseminated to prospective graduate students, to
graduate students, to postdoctoral fellows, and to the faculty who advise them.
Better supply-demand modeling of PhD labor markets is also important. We offer
specific recommendations in Chapter 5.
4.2 THE ISSUE OF FOREIGN STUDENTS
As noted in Chapter 3, foreign-citizen students accounted for most of the
increase in the numbers of science and engineering graduate students and
numbers of PhDs since about 1986. In 1992, for example, foreign citizens were
nearly one-third of graduate students in science and engineering, up from less
than one-fourth in 1982. By 1993, 57% of the PhDs in engineering and more than
one-third in physics, computer science, and mathematics were awarded to
foreign-born scientists and engineers (Table 3 in NSF, 1994f). All together,
the increase in foreign graduate students with temporary visas accounted for
65.5% of the net increase in annual science and engineering PhD awards 1986 to
1993, and an increase in the number of foreign-citizen PhDs with permanent
visas contributed almost another 11% to the increase. Foreign citizens
achieved a majority of science and engineering postdoctoral appointments in the
United States in 1991.
Immigration laws have been changed to place some restrictions on foreign
citizens with temporary student visas who are enrolled in US graduate science
and engineering programs. They are required to be full-time students, and they
and their dependents are prohibited from taking jobs. They are prohibited from
taking most fellowships and traineeships or applying for federally guaranteed
loans and other forms of direct federal assistance. They can be employed as
research assistants on federally funded research projects. Many foreigners
receive support for the first year of their graduate study from their home
countries, but the universities usually support them after that, generally with
research assistantships and teaching assistantships (much of the support comes
from federal research grants) (CRS, 1992).
As a result, universities provide a greater degree of financial support to
foreign students than to US citizens. In 1992, for example, universities
provided support to 87% of the graduate students in the physical sciences with
temporary visas, 84% of those with permanent visas, and 72% of those who were
US citizens. In engineering, university support went to 76%, 73%, and 61%,
respectively. The pattern was similar in the life and social sciences.
It has been generally possible under the immigration laws for new PhDs of
foreign citizenship to find entry-level positions in the US labor force (NSF,
1990b). Historically, about half the foreign citizens with American doctorates
in science and engineering have left the United States after getting their PhDs
or later postdoctoral appointments (CRS, 1992).2
Opinions about the effects of an increasing number and percentage of
foreigners in American graduate science and engineering programs have been
mixed (see CRS, 1992, for review and citations). Some people say that the
United States benefits from high graduate enrollments of foreign students
because they help with research and teaching, counter the declining interest of
American students in science and engineering, and fill the employment needs of
industrial laboratories. They argue that in a global economy, US universities
and industries should be able to recruit the best talent available3 Some value the contribution of foreign students
to a multicultural educational environment. Others point out that US companies
later hire foreign students to help open new markets in their country of
origin.
Other people have begun to argue that the numbers of foreign students should
be limited, on several grounds. They charge that increasing numbers of
foreigners with US PhDs who remain in this country (many of whom become US
citizens) are competing with American graduates for jobs; that might explain
some part of the employment problems that recent PhDs have complained of in the
last several years. Meanwhile, some return home and work for our economic
competitors. Critics of increased graduate enrollment of foreigners also have
charged that cultural and language differences make many of them ineffective in
the classroom and limit their ability to succeed in the labor market, that
their graduate training has been unfairly subsidized by American taxpayers,
that they depress salaries and thus interfere with an important market signal
that would attract more American students, and that their presence discourages
defense-related research in industry and on campus (CRS, 1992). A bill was
introduced in 1992 in the second session of the 102nd Congress requiring
universities to give preference to US students in filling federally sponsored
research positions.4
The sharp jump in number of foreign-citizen graduate students in recent years,
as described in Chapter 3, has probably been caused in part by a set of
political events that are unlikely to recur, as well as changes in US
immigration laws. And many foreign students are in the United States because
their home nations lack adequate educational infrastructures. As the wealth of
developing nations grows, so will these infrastructures, providing more
attractive employment opportunities at home. Already, the aggregate of
undergraduate science and engineering enrollments in six economically important
Asian nations exceeds undergraduate science and engineering enrollment in the
United States (NSF, 1993c).
The number of American students entering science and engineering graduate
schools is not rising. There is no evidence that this situation would be
changed by limiting foreigners. In fact, artificial limits could have the
detrimental effect of disrupting the supply of scientists and engineers in key
fields.
To the extent that there is a limit on the number of departmental "slots" for
graduate students, we are inclined to believe that the real issue is the lack
of US students, rather than the increasing presence of foreign students in our
graduate science and engineering programs, but it is difficult to assess the
claim that the easy access to foreign students has prevented an adequate
response of the system to declining US student interest. If graduate programs
are filled with foreigners, the programs do not have to make adjustments in
enrollments or in content to make them more relevant to US students. Nor do
businesses have to increase salaries to increase their supply of American
students.
The committee suggests that the most appropriate response to flat or declining
graduate enrollments of American students is to implement the measures
advocated in this report, which should improve the functioning of the PhD labor
market, and to continue efforts to strengthen the teaching of precollege and
undergraduate science. Those measures, we believe, would make graduate
education in science and engineering more attractive, more effective, and
accessible to a larger group of qualified American applicants.
4.3 TIME TO EMPLOYMENT
For a variety of reasons that are not well understood, it has been taking
longer for PhDs in science and engineering to begin their careers with
"potentially permanent" jobs--i.e., post-PhD jobs that are not
postdoctoral fellowships and are not temporary. According to an NSF analysis
of the Survey of Doctorate Recipients, the median age of PhD recipients
entering their first permanent positions increased in all fields from 1971 to
1991--by more than a year for PhDs in engineering (from age 30 to 31), by 2
years in the physical and mathematical sciences (from 30 to 32), by 3 years in
the life sciences (from 30 to 33), and by nearly 4 years in the social and
behavioral sciences (from 30 to 34).
The committee is concerned about the longer time to first permanent job. The
prospect of many years of graduate study might discourage qualified candidates
from attempting a PhD. Also, extending the years of schooling burdens PhDs who
enter nonacademic employment with a disadvantage compared with their
contemporaries, who are years ahead in workplace experience and seniority.
Finally, long times to degree (TTDs) and more postdoctoral study increase the
time required for the supply of PhDs to respond to shifts in market demand;
this has both social and individual costs.
There are many possible reasons for the lengthening of time to first regular
position, some of them positive (e.g., time spent working between college and
graduate school, which adds experience and maturity), some negative (e.g.,
discouragement of graduation by faculty who need research assistants or
teaching assistants and an oversupply of PhDs relative to demand for academic
positions), and some unavoidable (more time devoted to child-care
responsibilities and a greater amount of material to learn in graduate school
and in postdoctoral appointments). Increasing TTDs of all types (registered,
elapsed, and total time) have been defined and documented in Chapter 3, and
entering graduate students on the average are probably older to begin with.
The median age of new PhDs has increased in all fields since 1971--by 1
year in physics and astronomy, by 1.5 years in engineering and chemistry, by 2
years in biology, by 2.5 years in mathematics, and by more than 3 years in the
social sciences (Table 4 in NSF, 1993b). A growing proportion of graduate
students come from groups that take longer to finish--women,
underrepresented minorities, and foreign citizens (Stricker, 1994:570).
The committee discussed the issue of longer time to first permanent position,
the possible causes, and the significance. We are concerned about the costs of
increased time to first permanent position and the role of increased TTD in it,
but we hesitate to recommend a particular time limit for completion of the PhD
or a particular length of postdoctoral study, partly in recognition of the
great diversity of graduate students, disciplinary requirements, and
educational institutions missions. Instead, we believe that clearly understood
quantitative guidelines for PhD completion times should be set by individual
institutions after discussions among students, faculty, and professional
societies. 5
The committee notes that many institutions already have guidelines intended to
limit time to degree. At the Georgia Institute of Technology, for example,
chemical engineers are expected to complete a PhD in 4 years, and flexibility
is granted as appropriate. Other institutions allow students to receive
teaching assistantships for only 4 years; still others limit the time that a
student can work on a single research project.
Whatever the institutional guidelines are, they must be implemented,
monitored, and enforced to ensure that graduate students are never used to
provide inexpensive labor on research projects or in teaching. As the report
of the Association of American Universities/Association of Graduate Schools
puts it, "policy changes alone are insufficient; the commitment to implement
them is crucial" (AAU, 1990). Each institution should adopt standards
appropriate to its mission and student body and should charge graduate schools
and their deans with oversight. That could be done at the departmental or
program level. Departmental rules should be developed with the active
participation of the faculty who carry out graduate education, and they must be
clearly communicated to students, faculty advisers, and dissertation
committees.
4.4 INFORMATION AND ANALYSIS NEEDS
It is characteristic of the issues described in this chapter, particularly
time to employment or first permanent job, that more information on and better
understanding of them is needed, despite the problems and sensitivities
involved in addressing the issues of employment, foreign students, and
institutional policies concerning time to degree. Accordingly, appropriate
recommendations aimed at NSF, which has the lead responsibility for gathering
and analyzing information about the science and engineering enterprise, are
included in the next chapter.
NAS Home Page |
NAP Home Page |
Reading Room |
Report Home Page
Previous Section |
HTML Home Page |
Next Section