Back to Eos Electronic Supplement
Peter Fiske, Lawrence Livermore Laboratory
"Have a good sense of what you want to do even if it means taking time off to work in your field (or even out of your field)? Early to mid 20s is a time when many of my colleagues gave serious thoughts to new career awareness. Two in particular left their field completely and have satisfying careers. ... Under no circumstances should you remain in a particular field if you find it unsatisfying (... I also know of a few people who have stayed on their original career paths and are miserable--hardly a way to spend the rest of your life, isn't it?). (28-year-old male postdoc in Atmospheric Sciences)"
In the past 5 years, concern about the job market for research scientists has grown tremendously. Once the topic of hushed hallway conversations among graduate students and postdocs, reports of a dismal job market are now emblazoned across the pages of newspapers, magazines, and even the nightly news [Browne, 1995; Greenberg, 1995]. Once viewed as an entree to an endless frontier, some young scientists describe their Ph.D. education as an arduous road to nowhere [Browne, 1997].
While there is a growing effort to gather more timely and precise data on the Ph.D. job supply, young scientists' perceptions of the job market and their graduate school experiences have received relatively little attention. For example, is job market pessimism pervasive or confined to a few hard-hit disciplines? What is the role of the job market in students' decisions to enroll in graduate school or to drop out? Most important, what advice does the current population of new Ph.D.s have for future students and for policymakers?
To explore these issues we surveyed job seekers at recent AGU meetings. Our findings confirm the general belief that job market pessimism is widespread in the community of young scientists and show that there is some variation in pessimism by subdiscipline. Survey respondents had many ideas for what they would have done differently in their graduate training; most relate to broadening their experiences in order to have more employment options.
With this research project, we strive to provide more discipline-specific data on the graduate school experiences and perceptions of geosciences students. We hope to provide information to students to help them make decisions about their graduate education and future careers, to faculty to help them become better advisors, and to institutional leaders to help them craft appropriate changes in graduate programs. It is our goal to continue to build a data set to monitor changing perceptions of the job market and experiences in graduate school in the geosciences. By surveying current job seekers, primarily graduate students and postdocs at the end of their training, we believe that we add an important voice to the policy debates.
Several studies have documented converging events that have created a substantial decline in the number of new employment opportunities in science in the United States [Goodstein, 1994; Tobias et al., 1995], though hard data on job supply are still lacking. In brief, financial restructuring both within academe and at the federal and state levels have curtailed research funding and reduced the number of available faculty positions. Concurrently, the number of doctorates granted in the physical sciences, including the Earth, atmospheric and marine sciences, has continued to rise [National Research Council, 1996]. Consequently, there is enormous competition for few academic research positions, which is reflected in the large numbers of people entering postdoctoral fellowships. The 1995 Survey of Earned Doctorates showed that over 50% of those earning Ph.D.s in the physical sciences were headed to postdocs, the highest proportion ever [National Research Council, 1996]. Many people choose to take a postdoc to broaden their skill base and to garner an edge in the job market. However, postdoc positions also serve as interim jobs in the absence of other options and effectively prolong the training period (marked by low pay and uncertainty about the future) for students.
Despite the dramatic changes in the economics of the research science job market, many decisionmakers are reluctant to tinker with the existing system, believing that students entering graduate school are making careful, reasoned decisions, based on the state of the job market. They argue that students will react to the paucity of research jobs by entering graduate school in smaller numbers and leave seeking nonresearch science employment [Committee on Science, 1995, p. 9]. There is evidence that this is broadly true on aggregated levels. The decline in the number of Ph.D.s produced from 1970 to 1976 is in part attributed to the poor academic job market, [Bowen and Rudenstine, 1992, p. 54]. The downturn in the number of humanities Ph.D.s currently being awarded is also attributed to the poor academic job market, although the decline has not been "proportional to the fall in demand" [Geiger, 1997, p. 244].
However, these market forces are only one of the factors affecting the broad, long-term trends in Ph.D. production. Emerging research suggests that students make their choices based not on a "free-market" rational utilitarian assessment of the job market but on a wide range of factors, many of which are personal and subjective. Golde [1996] found that many students enter doctoral programs with only a vague idea of what the academic life entails, little sense of how tight the academic job market is, and a lack of awareness of jobs outside of academia. This lack of information, combined with the pressures to find doctoral students to serve as research assistants [Goodstein, 1994; Massy and Goldman, 1995], buffers the system from quick responses to changes in the job market. In short, even drastic changes in the job market do not result in dramatic systemic changes in science graduate education.
Nonetheless, the changes in the research science environment are driving calls for two kinds of changes in graduate education. Some advocate reducing the size of doctoral training programs [Holden, 1995; Massy and Goldman, 1995], and some faculty members, departments and disciplines are responding (e.g., Washington University [Magner, 1997]). Others believe that doctoral programs must become broader and provide training that will help students find work in nonacademic positions; the most notable champion is the COSEPUP report [Committee on Science, 1995].1
Clearly, this debate needs to be informed by more than speculation about the future and hand wringing about the past. History shows that the research science job market is unstable and difficult to predict. The erroneous predictions in the late 1980s of a scarcity of scientists and engineers [Atkinson, 1990; Bowen and Sosa, 1989] are commonly cited by young scientists as one of the factors in their decision to pursue graduate education. Without reliable forecasts, students base their assessment of the job market on the experiences of their graduate school peers and on media reports. Many faculty advisors are unable (or unwilling) to provide a realistic assessment of the research job market and assist their students in shaping their career goals accordingly. Departments do not collect information that could inform students or could help reshape their programs. Indeed, there is evidence that departments are unwilling to do so [Lovitts, 1996].
Our two-page survey was given to users of the AGU Job Center during the Fall 1995, Spring 1996, and Fall 1996 AGU meetings when they registered. Overall, we received 538 usable surveys; at the first two meetings over 90% of the Job Center users returned surveys. The questions were designed to
Most of the users of the AGU job center are graduate students and recent Ph.D.s. To assess the employment concerns of recent Ph.D. graduates, we focused on the 72% of the survey respondents who had received or anticipated receipt of a Ph.D. between 1990 and 1997 (N = 387). Detailed information about this subgroup is shown in Table 1.
The broad range of disciplines represented by AGU is reflected in this population in roughly the same proportions as the overall membership. Members from the five largest sections (Hydrology; Ocean Sciences; Atmospheric Sciences; Volcanology, Geochemistry and Petrology (VGP); and Tectonophysics) make up 75% of our respondents. Seventy-six percent of those we surveyed were under the age of 36. The percentage of women (22%) is the same as the national average for recent Ph.D. recipients in the mathematical and physical sciences [National Research Council, 1996]. The percentage of non-U.S. citizens in this pool (38%) is lower than the national average (46%) [National Research Council, 1996].
Fifty percent of respondents from U.S. institutions come from geoscience or oceanography departments rated in the top 20 by the National Research Council [1995]; 31% come from top 10 departments. This high representation from top research departments is not surprising: attendance at AGU meetings is typically supported by research grant dollars, most of which flows to the departments with the best reputations. It is significant, because these findings represent the experiences and perceptions of the top young scholars in the country. It is in this pool of students that the federal government invests the most, with tuition support, research assistantships, fellowships and traineeships. Arguably, these students are likely to be the most successful in the job market and to have had access to the most financial and intellectual resources in graduate school. Their frustrations, then, bespeak of deep systemic problems.
Employment Goals. The young scientists in our survey reported varying levels of experience applying for jobs. Nearly 23% had applied to no jobs at all, 38% had applied to 1 5, 16% had applied to 6 10, 19% had applied to 10 50, and 5% had applied for more than 50 jobs. There was a general correlation between the years out from the Ph.D. and the number of jobs applied to.
In order to assess the employment goals of the respondents we asked them to rank order eight career opportunities, both as the job goal at the AGU meeting and the ultimate career goal. Many people filled out the question differently, checking a few boxes or ordering only a few options. Therefore we are reporting any check mark or ranking numbers 1-4 as a positive vote for this career option.
Not surprisingly, when asked what the ultimate goal of their job search was, recent Ph.D.s most frequent response was "College/University Academic, Predominantly Research" (72%), followed by "College/University Academic, Predominantly Teaching" (61%). "Industry/Private Sector Research/Technical" was the third choice (55%) and "Government Research/Technical" (50%) was close behind. In contrast, the jobs "Industry/Administration" (10%), "Research Post Doc/Contract Employee" (11%), and "Lecturer" (13%) were much less popular as ultimate goals.
Survey respondents distinguished between their meeting job goals and ultimate job goals. This was particularly evident in the "Post Doc" category. While only 11% were willing to consider this a permanent goal, fully 53% listed this as a goal for the meeting. On the other hand, the percentages for the Academic positions were nearly as high in the Meeting Goals lists as in the Ultimate Goals lists. This lends weight to the argument that people do not postdoc as a preferred choice (for skill development, for example) but that many believe that it is the only realistic option.
Only 2% listed "Other" as one of their preferred options. However, nearly 40% listed options they were considering. The most popular options specified were computer programmer/analyst, high school teacher, and writer. Among the unusual suggestions were architect, house husband, organic farmer, and multimedia artist. These data, taken together, show that top young scientists still desire to enter a traditional research science career path. Alternative careers options are just that, back-up plans if their preferred careers are unattainable. These data also suggest that graduate training has exposed students to only a narrow range of options. Other options may be invisible or discounted. Consequently, a traditional research science career is the track most students follow.
Overall, AGU Job Center users were not optimistic about the job market. The survey scale had five choices: "hopeless" (= 1), "bad," "neutral/fair," "good," and "excellent" (= 5). Respondents were asked to rate the "Current state of the 'traditional' job market for research trained geoscientists" both "in general" and "in your particular subdiscipline or AGU section." The results are shown in Table 2. The modal choice for both questions was "bad" (mean 2.45 and 2.50). The responses for both categories were quite similar. Over two thirds of the respondents marked the same choice for both job markets, while 17.9% thought the general job market was better, and 14.1% thought the subfield job market was better.
When these data are disaggregated, we see variance by subfield. Comparing the extremes, the contrast in job market perception between "healthy" fields, such as hydrology, and "stressed" fields, such as VGP, is quite noticeable (see Figure 1 Click here to see corrected Figure 1). Hydrology has seen considerable growth in the past few years [Eaton, 1995], while jobs in VGP have become far fewer in number. Correspondingly, hydrologists were more optimistic about the subfield job market than about the general job market, and the VGP respondents perceived the opposite.
Yet the differences in job market perceptions between these two groups were smaller than we expected. The mean subfield job perception score for volcanology was 2.22 and for hydrology was 2.85. In both fields these means lie between "bad" and "neutral/fair." We believe that this reflects in part that each AGU section represents a multitude of specific subsubdisciplines, and not all are experiencing the same amount of job growth or decline. We also suspect that general job market pessimism on the part of all young scientists may be permeating relatively "healthy" subdisciplines.
We predicted that job market pessimism would increase with the length of time young scientists spend looking for a job. Not only do long job searches take an emotional toll, but those who find permanent jobs disappear from the pool of respondents, leaving those for whom permanent positions are harder to acquire. While the trends were in the expected direction, the differences were small (see Figure 2). Those who have just graduated or who anticipate graduation in 1997 are more optimistic than their graduated counterparts. Similarly, we found a weak correlation between pessimism about the job market and the number of applications actually submitted. Between 50% and 60% of those who have applied to 0, 1 5, 6 10, and 11 50 jobs describe the general job market as "bad" or "hopeless." This percentage rises to 80% for those who have applied to more than 50 jobs.
Table 2 shows that postdoctoral fellows have the most discouraged and cynical perceptions of the job market. Since postdoc positions are relatively plentiful compared to the number of permanent positions, postdocs are arguably caught at the point the "career funnel" narrows most abruptly. Currently enrolled graduate students, on the other hand, are the most optimistic, reflecting the relatively little experience with the job market many of them have had. (This confirms the data on optimism by year of degree.) Those currently in permanent positions, while a small portion of our sample, fall in between. These people are in an enviable situation: having a permanent position, their need for a job may be lower than that of a postdoc, and they may also be more successful on the market.
Interestingly, perceptions of the job market have steadily risen over the three AGU conferences. The mean is rising, from 2.28 in fall 1995 to 2.58 in fall 1996. In the same time period the modal response moved up from "bad" to "neutral/fair." It remains to be seen whether this is an enduring trend. It is also possible people are getting used to a bad market and adjusting their sights accordingly.
Thirty-four percent of the survey respondents admitted that they had at some time considered dropping out of their doctoral program. Most of these selected the response "occasionally;" only nine selected "regularly," and two selected "constantly." This response rate was lower than we expected. Petrides [1996] found that nearly half (48.6%) of currently enrolled engineering doctoral students had thought about leaving their program. Golde [1996] found that the geology department she studied had a higher attrition rate than other science departments at the university under investigation. She also found that most of that attrition took place before candidacy. If this finding were true across institutions, it would explain the relatively low levels of reported dissatisfaction with graduate school we found. Those who were most unhappy have already left graduate school before making enough progress to be part of the AGU job placement service.
Women were more likely than men to consider leaving graduate school. Forty-seven percent of the women had considered dropping out, compared with 31% of their male colleagues. While this confirms other research that women in the sciences are more likely than men to attrite from graduate school, we find this disappointing. This confirms research showing that aspects of the culture of science discourage women from pursuing their goals [Brush, 1991; Etzkowitz et al., 1994; Widnall, 1988; Zuckerman et al., 1991].
We also found that of those thinking about leaving, only 27% were international students, although international students made up nearly half of the sample population. Again, this is consistent with other research; the explanation is that because it is more difficult for international students to enroll in graduate studies, they have made a greater intellectual and financial commitment to their studies [Girves and Wemmerus, 1988].
We also wanted to know what motivated students to consider leaving. Respondents could select as many of the list of responses that pertained, as shown in Table 3. We found that the job market plays an important role in student decision making: the top reason given was "poor job market." "Poor working relationship with advisor" was the second most common reason given. This finding is consistent with other research that highlights the importance of the advisor and the advising relationship in the sciences. In the sciences, unlike other disciplines, students are very dependent on the advisor for funding, and for the research project and, in many cases, most of the student's learning is in the advisor's lab. These dependencies make a good advising relationship critical for student success [Golde, 1996].
The survey respondents' observations and suggestions for changing their own graduate programs give great support to those advocating for reform of graduate education. Over half checked at least one aspect of their graduate training that, with the benefit of hindsight, they would have done differently. These results are shown in Table 4. The most common response was that students would have "selected a different field or subfield." Fourteen respondents specified that they would have chosen something in computers, seven each specifically mentioned engineering and environmental studies, and eight more listed professions outside of science. Once again the job market is uppermost in respondent's minds. Many indicated that they would select a field that was more applied and/or had better job prospects. Typical comments appended to this survey item included: "Possibly something with more crossover opportunities." "The job market is *&%#$." "Oriented more towards environmental geochem it's just as interesting and is marketable!" "More opportunities/better rewards."
The second most common response was "to take more technical classes in other departments." Nineteen respondents specifically listed computer science or computer programming classes as a top priority. Virtually all of the comments related to diversifying and enhancing skills. Eleven commented specifically on broadening their background, making comments like "While I feel that I have a broad background, the broader the better." "I would take more time, enjoy learning more subjects, rather than the minimum required." The acquisition of breadth has a price, however. As one respondent noted: "Math/Chem/Computers. This is what I would like to do, but practical concerns (time) may have made it impractical." Specific comments such as "to make myself more marketable" and "more flexibility" show that the desire for technical classes is often related to job market concerns.
Advisor-related issues emerged on this part of the survey as well. The third most common change respondents would make would be to change advisors. The comments portrayed a range of concerns and issues. Many respondents articulated characteristics that they would have liked their advisor to possess. Comments included: "One that will help with job search." "Need someone more politically savvy." "Personality conflicts." "I like my advisor, but he's too hands-off." Once again, this is consistent with other research on the importance of advisor selection. Unfortunately, many students select their advisor based on very little contact - perhaps one interview - which makes predicting a successful partnership difficult from the outset.
Seventeen percent of those who would have done something different would have attended another institution. The most common reason given was the recognition that comes from receiving a degree from a prestigious institution. Although most respondents were satisfied with their selection, the comments reveal how aware many students and employers are of institutional reputation. Sociologists have long commented on the "credentialling effect" and the way that a diploma is taken as a proxy for knowledge and skills. Survey comments included: "Chose my institution to work with several professors and specific projects. Now think it is more important to aim as high as possible (MIT, Stanford, etc.) because these are the people who get the jobs." "Need one with a better reputation and more money." "More prestige, better alumni connections." "When people I talk with about jobs ask me what school I am from, and I tell them, they look disappointed."
The most illuminating and thoughtful comments were offered in response to the open-ended request, "What advice would you give others entering, or in the early years of, graduate school?" About 28% of the survey respondents took the time to respond to this question. Overwhelmingly, the comments advised new students to be thoughtful and foresightful in the decisions that they make. Rather than leaping into graduate school, recent graduates urged those following them to thoroughly research graduate school and the job market. Furthermore, they provide sage words about the graduate school process, working with an advisor and the research project. Representative and particularly salient comments are shown in Table 5. The overwhelming message was that students should be directors of their own graduate experience, rather than passive recipients.
Several people also suggested that prospective students both research the job market and engage in introspection. These respondents linked success in graduate school with having clear plans based on self-knowledge and knowledge about the job process. Recent graduates say it best: "Don't just go to grad. school because you haven't checked out the job market. See where different degrees will get you and act accordingly." "Take time out to fully understand your strengths and which institution/lab/advisor will most likely lead to your working at full potential. Do not accept any Ph.D. offers out of insecurity or lack of choice. A Ph.D. is more than a job and demands full commitment." "Don't go just because it is the next logical step. THINK about what you want from life and how graduate school fits in."
We believe that this is the most important message to emerge from this research project. The job market is changing and fraught with uncertainty. Graduate school is an arduous process that requires significant financial, emotional, and intellectual commitment. It is not to be entered into unadvisedly or ill-informed.
While survey respondents might not have thought about the job market and the utilitarian outcomes of graduate education when they started graduate school, they are clearly urging their younger counterparts to do so. These findings may, of course, be skewed by reporting data from only job seekers, for whom the job market was a top concern as they were filling out the survey. Nonetheless, we believe that we have tapped important information, which is often hidden from new students, faculty, administrators, and policymakers.
The debate about the future directions for graduate education in the research sciences is now garnering considerable research and policy attention. With this survey we attempt to combine data on job market perceptions with information about graduate school experiences to promote informed discussion in departments in the geosciences.
We believe that the current pessimistic climate is in part attributable to the narrow definition of career tracks following from the Ph.D. in geosciences. We believe that the research science job market is hard to predict on the timescale that is relevant for graduate education (5 10 years in the future). In light of this uncertainty students may need more than just better data and reliable indicators of the research job market. They, and their faculty advisors, need to adopt more flexible and realistic training and employment goals.
Many survey respondents indicated dissatisfaction with their graduate school experience. If we take those concerns and suggestions seriously, they indicate that broadening the course work and emphases of graduate school would both make the experience more satisfying and prepare students better for the job market they will face. Recent graduates clearly would like graduate program faculty and administrators to be informed about the job environment and to make those realities (shifting though they may be) explicitly clear to students.
While our results are strictly applicable to only the research geosciences community, we believe that the phenomenon of job anxiety is widespread in the physical and life sciences. One interpretation of our findings, which may well apply across disciplines, is that young scientists are not being shown career options outside of traditional research science. When alternatives are visible, they are often described as lesser choices. These scientists, whether they become teachers, policy analysts, and writers, or computer systems experts, are often seen as exiles from the community of scientists. Rather than thinking about scientists as "lost" and the resources spent on their training as "waste," we would advocate a broadening of the purpose of graduate education.
Acknowledgments. We thank Beverly Saylor, who cast a critical eye over a draft, and Christine Maidl, who provided valuable assistance in the data entry, coding, analysis and presentation. We also thank Caroline Gilman and Karen Spaulding of AGU for their help in administering this survey and AGU's Committee on Education and Human Resources for their support and encouragement.
Atkinson, R. C., Supply and demand for scientists and engineers: A national crisis in the making, Science , 248, 1990.
Ausubel, J., Malthus and graduate students: Checks on burgeoning ranks of Ph.D.'s, The Scientist , Feb. 5, 11, 1996.
Bowen, W. G., and N. L. Rudenstine, In Pursuit of the PhD , 442 pp., Princeton Univ. Press, Princeton, N.J., 1992.
Bowen, W. G., and J. A. Sosa, Prospects for Faculty in the Arts and Sciences. A Study of Factors Affecting Demand and Supply , 1987 2012, Princeton Univ. Press, Princeton, N.J., 1989.
Browne, M. W., Supply Exceeds demand for Ph.D.'s in many science fields, N.Y. Times , 1995.
Browne, M. W., Comet's discoverer laments status of careers in science, N.Y. Times , p. A19, 1997.
Brush, S. G., Women in science and engineering, Am. Sci., 79 (5), 404 419, 1991.
Committee on Science, Engineering, and Public Policy, Reshaping the Graduate Education of Scientists and Engineers , National Academy Press, Washington, D. C., 1995.
Etzkowitz, H., C. Kemelgor, M. Neuschatz, and B. Uzzi, Barriers to women's participation in academic science and engineering, in Who Will Do Science? Educating the Next Generation , edited by W. Pearson Jr., and A. Fechter, 169 pp., Johns Hopkins Press, Baltimore, Md., 1994.
Geiger, R., Doctoral education: The Short-term crisis vs. long-term challenge, Rev. Higher Educ., 20 (3), 239 251, 1997.
Girves, J. E., and V. Wemmerus, Developing models of graduate student degree progress, J. Higher Educ., 59 (2), 163 189, 1988.
Golde, C. M., How departmental contextual factors shape doctoral student attrition, Ph.D. dissertation, Stanford Univ., Stanford, Calif., 1996.
Goodstein, D., The Big Crunch, paper presented at the NCAR 48 Symposium , California Institute of Technology, Portland, OR, 1994.
Greenberg, D. S., Surplus in science, Wash. Post , p. A25, 1995.
Holden, C., Is it time to begin Ph.D. population control?, Science, 270 , 123 128, 1995.
Lovitts, B., Leaving the Ivory tower: A sociological analysis of the causes of departure from doctoral study, Ph.D. dissertation, Univ. of Md., College Park, 1996.
Magner, D. K., Bucking Tradition, Washington U. Shrinks its Ph.D. programs, in Chron. Higher Educ., 1997.
Massy, W., and C. Goldman, The production and utilization of science and engineering doctorates in the United States, Stanford Univ., Stanford, Calif., 1995.
Menand, L., How to make a Ph.D. matter, in N.Y Times Mag., pp. 78 81, 1996.
National Research Council, Summary Report 1995. Doctorate Recipients from United States Universities , Nat. Acad. Press, Washington, D. C., 1996.
Tobias, S., D. E. Chubin, and K. Aylesworth, Rethinking science as a career: perceptions and realities in the physical sciences , 157 pp., Res. Corp., Tucson, 1995.
Widnall, S. E., AAAS presidential lecture: Voices from the pipeline, Science, 241 , 1740 1745, 1988.
Zuckerman, H., J. R. Cole, and J. T. Bruer, The Outer Circle: Women in the Scientific Community, 351 pp., Yale University Press, New Haven, Conn., 1991.
1 There are, of course, other suggestions for reform, but these two are the most widely discussed options. Geiger [1997] advocates expanding graduate education but also seeks commensurate expansion of the private sector market for Ph.D.s. Menand [1996] argues for sharply reducing time to degree by focusing training more precisely. Ausubel[1996] suggests "valorizing" the master's degree.