The American Astronomical Society's Examination of Graduate Education in Astronomy

The AAS Education Policy Board
Stephen Strom and Suzan Edwards, co-chairs
Alexei Filippenko, Andrew Fraknoi, Catherine Garmany
Eugene Levy, Lawrence Rudnick, Michal Simon

The Graduate Advisory Board
Frank Bash, Bernie Burke, David Schramm, Judy Pipher

2.1 Initiating the Dialog
2.2 Structured Community Workshops
2.3 Seeking Further Community Input
3.1 Charge to the Graduate Curriculum Working Group
3.2 Charge to Graduate Admissions and Support Modes Working Group
3.3 Master's Degrees
4.1 General Conclusions
4.2 Broadening Academic Options: Conclusions
4.3 Creative and Rewarding Teaching Experiences Should Be Provided
4.4 Graduate Student Support Modes: Conclusions
4.5 The Master's Degree: Conclusions
5.1 Action Items
5.2 Additional Recommendations
A. Participants in the AAS Regional Graduate Education Workshops
A.1Keynote Speakers
A.2 PhD Granting Graduate Departments
A.3 Others
B. Keynote Addresses
B.1 The Obvious and the Overlooked
B.2 Reshaping Graduate Education in Science and Engineering
B.3 Fostering Breadth in Graduate Education in Astronomy
B.4 The Changing Graduate Education and Research Environment
C. Career Trends for New Doctorates in Astronomy and Astrophysics
C.1 Statistical Trends in Astronomy and Physics
C.2 Summary
D. Additional Statistics
D.1 The AAS Job Register
D.2 Support of Graduate Students


Astronomy is entering an era which promises unprecedented discovery enabled by powerful new observational and computational tools undreamed of 20 years ago, and an enthusiastic cadre of recent PhDs eager to exploit these new capabilities. At the same time, long-term prospects for funding basic research are more uncertain than in recent decades. While it is dangerous to predict future funding patterns, there is near unanimous agreement that the growth astronomy and other sciences have enjoyed over the past 40 years cannot be sustained. As a consequence women and men who aspire to careers in basic scientific research will exceed the number of available long-term positions traditionally sought by young people following receipt of the PhD. These projections have led many to urge that the nation and its graduate schools take steps to decrease the PhD production rate.

Other participants in the national debate argue that "birth control'' is far too simplistic a response. Their major concern is that adopting this approach will discourage young people from pursuing graduate education in the sciences and consequently decrease the pool of technical talent vital to the nation's future. They view graduate schools as a national resource which for more than 50 years have trained the scientists and engineers who have developed ideas and technologies which have been central to the economic health and survival of the United States. Rather than decreasing the number of entering students, they urge that graduate schools be challenged to re-examine their goals, to devise curricula and cultures which continue to attract first rate students, and to prepare students not only for research in a particular discipline, but for a broader range of post-graduate careers.

In response to the unpredictable climate for support of basic research and the sometimes conflicting visions of the role of US graduate programs in society today, the scientific community has organized a number of studies aimed at re-examining the goals of graduate education in science. In 1995, the AAS Council charged its Astronomy Education Policy Board to structure a series of activities directed specifically at examining graduate education in astronomy.

This report summarizes the results of those activities, which included both society-wide discussions of the issues confronting graduate education in astronomy, and three focussed regional workshops attended by astronomy department chairs and graduate students from 59 PhD granting institutions, along with representatives from other professional societies, the industrial community, the science education community and the funding agencies which provide the bulk of support for astronomy.

The discussions were focussed on the following questions:

The discussions were both lively and thoughtful, and perhaps surprisingly, reflected or developed consensus on a number of points, among which the following are particularly noteworthy:

The AAS Education Policy Board and its Graduate Education Advisory Committee developed from the discussions a series of recommendations which they and the Society believe will enable the astronomical community to shape graduate programs which are responsive to the needs of students, the discipline, and the society which supports us. The full set of recommendations is included in the body of the report.

Key Recommendations

Three recommendations merit particular note:

Define and Support Experiments to Enrich Graduate Education

Federal and state resources should be allocated to develop pilot programs aimed at providing students with structured opportunities to gain broader experience and exposure to cultures outside academia during their careers. Example programs might include (1) developing new curriculum options involving partnerships between astronomy and other university departments, with industry, or with local schools; (2) developing enhanced MSc or specialized professional Master's programs which provide students with exposure to additional "hard'' (e.g. high speed computing and visualization techniques; instrumentation; teaching) and "soft'' skills (comunication; team problem solving) in addition to basic training in physics and astrophysics; and (3) providing fellowships/traineeships targeted at exposing students to instrumentation development, formal and informal science education and curriculum development, and the demands, expectations, and culture of industry and national laboratories.

This recommendation is made with the understanding that in times of constant or declining budgets, funding such experiments - even at a modest level - comes at the expense of other more traditional programs. The long-term gains, however, will likely outweigh potential short-term displacements. To ensure that limited resources are invested wisely, the community and the funding agencies must work together to subject any experiments to rigorous peer-review and evaluation.

Re-examine the Master's Degree in Astronomy

The astronomical community should carry out a national review of the merits and liabilities of (1) enhancing Master's degree programs to ensure they provide the broad and thorough training necessary for students either to pursue a PhD, or to leave graduate school armed with the skills necessary to embark successfully on a variety of career paths; and (2) requiring an MSc for admission to a PhD program. Central to this recommendation is a commitment to ensuring that a Master's degree would represent a desirable and "marketable'' outcome for many students, and would come to be recognized widely as a mark of significant academic achievement rather than a "failed PhD''.

Enhanced MSc degree programs would incorporate rigorous course work, research experience, a range of technical skills and exposure to the scientific world outside of academia, through internships or local partnerships. Such degrees - of duration 2 to 3 years - would provide a mix of academic and "real-world'' experience that would benefit both students who pursue an astronomy PhD, and those who choose other careers.

Several arguments persuade us to recommend further discussion of the possibility of admitting all entering students to an MSc rather than a PhD program: (1) it would allow institutions to recruit more widely and take greater risks in admitting potentially talented individuals; (2) it would provide both students and faculty with an opportunity to "pause and reflect'' prior to making a commitment to pursue a PhD or to choose a different career path; (3) it would allow institutions to be more selective in admitting students to a PhD program; and (4) it would allow graduate programs to grow in stimulating and potentially beneficial new directions, by forging links and partnerships with a wider scientific community.

Provide Students with the Information and Experience Necessary to Make Informed Career Decisions

Both graduate departments individually, and the AAS collectively, should (1) develop comprehensive databases which summarize employment histories of astronomy Master's and PhD recipients and (2) develop mechanisms to expose students to a variety of career options. Together these actions would provide prospective and current students with the background needed to make informed choices among graduate schools and realistic assessments of their post-graduate career options. These might include short-term activities, such as sponsoring career seminars featuring astronomy PhDs engaged in challenging and stimulating non-academic careers, and longer term investments, such as developing graduate curricula which incorporate "real-world'' skills and/or off-campus experiences.


Astronomy as a field and as a profession has never been more exciting. The discoveries of the past few decades - unseen dark matter which comprises more than 90% of the material in the universe; coherent filamentary structures traced by galaxies and extending for hundreds of millions of light years; massive black holes which power energetic outbursts during the lifetimes of most luminous galaxies; jets of exquisitely collimated hot plasma, extending over tens of light years and signalling the birth of stars and planetary systems; mature extra-solar planetary systems orbiting both sun-like stars and pulsars - have captured the imagination of scientists and the public alike. The next decades promise even greater change in our understanding of the universe as astronomers begin to exploit the power of a new generation of ground-based optical/infrared and radio telescopes, space-based observatories, and high-speed computing capability nearly unimaginable even a decade ago.

At the same time, the scientific community in the United States is entering a period of great uncertainty as the nation and its leaders debate the role of basic research in the post cold war era. While neither the ultimate outcome of this debate nor the resulting shape of a new "social contract'' between scientists and the society that supports them is yet clear, there is near-unanimous agreement that the growth in numbers of scientists and funding levels enjoyed by astronomy and other sciences over the past 40 years cannot be sustained.

Moreover, many universities are in the process of re-examining their mission and structure in response both to financial pressures and questions regarding their role in society. Federal laboratories are engaged in a similar process, while the industrial sector has already taken steps to alter or downsize major research laboratories. The anticipation of slower growth or a real decline in support for basic research, combined with the possibility of a decrease in the number of "traditional'' permanent research positions, has led some to argue that the nation should take steps to reduce the rate at which graduate programs produce PhDs.

Others express deep concern that taking such steps will discourage talented young people from pursuing graduate education in the sciences - leading both to a decrease in the pool of technically-trained individuals vital to the nation's economic future, and to erosion of the world leadership the US has enjoyed in most research areas over the past half century.

Still others add yet another element to the on-going national debate by arguing that many students currently emerging from PhD programs develop too narrow a range of skills, and as a result are unable to adapt readily to the rapidly changing needs of a technology-driven global economy. Moreover, because students are embedded within a culture which values academically-based fundamental research often to the exclusion of other, equally challenging pursuits, they reject or ignore career options outside academia. These participants in the national debate conclude that society can no longer afford generous support for "graduate education'' narrowly defined as "the process by which the next generation of research scientists is trained." Instead, they urge that graduate schools be challenged to re-examine their goals, revise their curricula and develop a "culture'' which broadens the definition of "graduate education'' and encourages students to explore and eventually pursue a much broader range of post-graduate careers.

The rapidly changing environment for support of basic research coupled with the multiple, sometimes conflicting political forces pulling on the nation's graduate schools, have combined to motivate the scientific community to organize a number of studies aimed at (1) examining and better understanding the goals of graduate education as we enter the next century; and (2) ensuring that they are aligned not only with discipline aspirations, but the long term needs of the nation and our graduate students. Two prominent recent examples develop a number of proposals aimed at effecting change in graduate education for a wide range of scientific and engineering disciplines (NAS/COSEPUP Report, Griffiths 1995; NSF/MPS Report, Armstrong 1996). Both reports urge that graduate programs in science be reshaped to (1) offer a broader range of academic options to enhance student versatility, (2) develop alternate methods for federal support of graduate students to encourage and enable students to gain broader experience during their graduate careers, and (3) re-examine the role of the Master's degree in light of both student aspirations and projected needs of the private sector.

In 1995, the AAS Council authorized the Astronomy Education Policy Board to structure a series of activities aimed at re-examining the goals of graduate training in astronomy. While informed and given initial focus by the seminal Griffiths and Armstrong reports, the activities were directed toward developing an understanding of whether and how changes in graduate education in astronomy might best be made in light of the following factors, many of which are specific to our community:

The outcome of the AAS examination of graduate education is reported here. The following sections (1) describe the format of the study, (2) summarize the key points raised in discussions held in the course of three regional workshops in November, 1996, and (3) set forth recommendations aimed at strengthening our ability to succeed in carrying out one of our fundamental missions as a community - to educate the next generation of scientists.


2.1 Initiating the Dialog

As a first step, the Society devoted a significant fraction of its January, 1996 meeting in San Antonio to plenary and special sessions designed to alert the community to the importance of re-examining the goals of graduate education, and to initiate discussion of these issues among a broad cross-section of the membership. These sessions included: addresses by NSF Director Neal Lane and National Academy of Sciences President Bruce Alberts; a summary by John Armstrong of the NSF/MPS study on graduate education; a panel discussion involving astronomy faculty from representative graduate programs from throughout the country; and an open dialog eliciting comments from a cross-section of the society.

2.2 Structured Community Workshops

In June, 1996 the AAS received financial support from the National Science Foundation, which in combination with Society funding enabled the AAS to continue this dialog via a series of three, 1 1/2 day regional workshops held during November, 1996. These workshops were designed by the Society's Astronomy Education Policy Board and its Graduate Education Advisory Committee to encourage detailed and structured discussion in three areas:

Workshop participants included representatives from 59 PhD granting institutions (most institutions sent both a faculty member - typically the department or graduate program chair - and a graduate student); and 31 representatives from the following communities: funding agencies, national observatories, institutions which grant only a Master's degree in astronomy, or a joint physics/astronomy MSc; 4 year colleges, community colleges, university administration, other scientific fields, industry, education. The full list of participants is included as Appendix A.

Each meeting was initiated by a keynote address aimed at framing issues from a national perspective. Keynote speakers at the three workshops were Neal Lane, Director of the National Science Foundation: Eastern/Southern Region; Phillip Griffiths, Director of the Institute for Advanced Study: Midwestern Region; and Marye Anne Fox, Vice President for Research the University of Texas at Austin: Western Region. The speeches from our first two keynote speakers, plus an invited talk at the western region workshop from Eugene Levy, Dean of Sciences at the University of Arizona, are included as Appendix B. An additional invited talk at the western region workshop by Judy Franz, Executive Director of the American Physical Society, summarizing the findings from a workshop on graduate education for physics department chairs, is not reproduced here (Franz, 1995).

In addition, statistical material was prepared and presented by representatives from the Office of Scientific and Engineering Personnel (OSEP) of the National Research Council and by the AAS Executive Office. Speakers from the OSEP presented data on PhD recipients in Astronomy and in Physics from 1973-1995, culled from the annual Doctoral Records File survey and from the longitudinal Survey of Doctoral Recipients. A manuscript summarizing this data, prepared by the OSEP, is included as Appendix C. To supplement the OSEP data, the AAS Executive Office examined the trends in availability of jobs advertised in the AAS Job Register and assembled statistics on federal support of graduate students in Astronomy and in Physics from the NSF data release of Selected Data on Graduate Students and Postdoctorates in Science and Engineering. These statistics are presented in Appendix D.

Discussions of graduate curricula, funding issues and Master's programs took place in 3 separate working groups, each of which comprised senior departmental representatives, graduate students and individuals from outside the astronomical community; all participated vigorously. In addition, the graduate students structured presentations summarizing their concerns to a plenary session at each meeting, while outside participants provided additional perspectives in the course of a panel discussion.

The organizers made a deliberate attempt not to inform participants in later workshops of the conclusions reached by earlier groups, in order that the degree of consensus (or lack thereof) on each of the key issues be assessed from the conclusions reached through independent discussions.

While each workshop developed its own distinctive "flavor," there was in the end considerable overlap regarding the merits and liabilities of various possible approaches to restructuring and funding graduate education.

2.3 Seeking Further Community Input

The results of the three community workshops were summarized in a presentation made at a special session at the January, 1997 Toronto AAS meeting which was organized with the goal of eliciting further community dialog and commentary. In addition, a summary of preliminary workshop conclusions was posted on the Web and members were encouraged to post comments using the AAS website.


In this section, we summarize the charges to the three workshop breakout sessions focussed on (1) broadening academic options, (2) graduate admissions and support modes, and (3) the role of the Master's degree in astronomy.

3.1 Charge to the Graduate Curriculum Working Group

The changing environment for support of scientific research has led the scientific community to re-examine the goals of graduate education. Two recent reports (see Introduction) urge significant reshaping of graduate curricula in order to provide students with a broader mix of skills and experience, and thus the training and background necessary for pursuing a variety of challenging post-graduate careers. They specifically recommend that graduate programs in science and engineering provide options:

Workshop participants were asked to discuss the relevance of these proposed new approaches to the astronomical community and assess their potential merits and liabilities as viewed from the perspective of: students; universities; the astronomical research enterprise and its goal of sustaining world-leading research; and the broader needs of US society in the United States which must develop the human resources to compete successfully in a rapidly changing, technology-based world economy. They were also asked to identify experiments which might stimulate desirable change.

3.2 Charge to Graduate Admissions and Support Modes Working Group

The anticipated downsizing of the scientific research enterprise has led some to suggest that admissions policies and the level and character of federal support for graduate students be re-structured with the goal of reducing the student population. If zero or negative population growth were the goal, departments might deliberately restrict admissions to graduate school; concommitently, federal research dollars devoted to graduate training might also be reduced.

Some suggest the possibility that a decreased number of graduate students and the aggregate federal investment in graduate student support might be balanced by an increase in the number of postdoctoral fellowships, thereby ameliorating the potentially negative effects of a decrease in the graduate student population on the health of the research enterprise.

A variant of this proposal would involve both reducing aggregate federal support for graduate students, and directing such support to the "best'' students, through (1) portable fellowships awarded at entry to graduate school (analogs of the Hubble Fellowships) or (2) "dissertation fellowships'' awarded following completion of pre-thesis PhD requirements. Such changes would in effect endorse a "birth-control'' policy and might alter significantly the balance of enrollment among institutions.

An alternative approach would be to view graduate education not simply as a means for producing the next generation of research scientists, but as part of a process essential to meeting a critical societal goal - preparing talented individuals for a range of careers which require creative approaches to solving challenging technical problems. Rather than attempting to tailor admissions to a fluctuating and uncertain job market, graduate admissions would be kept at near current levels, and advanced graduate training in science would be broadened in its intellectual content - to the benefit of both future researchers and those that pursue "alternative'' careers. Federal support for students would then be provided with the implicit understanding that these dollars represent an investment not only in the future of the discipline but in the future of society.

While following this path need not require a shift in federal funding patterns, some have suggested that the balance of federal support for graduate students should shift - away from traditional research assistantships supported through PI grants, and towards fellowships/traineeships or block grants.

Proponents of this approach argue that students supported via traditional PI-supported RAs carry with them obligations to carry out specific research tasks which in some cases can conflict with the goal of seeking knowledge beyond traditional discipline boundaries or working outside of their advisor's research specialty. Increasing the fraction of students supported via fellowships would, following this argument, "free'' such students to shape a more balanced research experience.

Others see block grants as a mechanism for providing strong incentives for graduate schools to develop a dual- rather than singular - focus to their programs. By providing funding to institutions which propose imaginative new approaches to graduate training - including a variety of broadening options - a strong message regarding national priorities would be sent, and the intellectual energies of some departments would almost certainly be invested in developing such approaches.

Participants were asked to assess merits and liabilities of (1) altering graduate admissions policies; (2) developing a dual focus to graduate education; and (3) proposed changes in the balance and character of federal funding for graduate students and graduate education. They were also asked to identify appropriate "experiments'' which might enable evaluation of different approaches or stimulate desirable change.

3.3 Charge to Master's Degrees Working Group

There is a widespread perception that a Master's Degree in Astronomy is a less than worthy achievement, offered to those who did not meet the criteria for admission to PhD candidacy. Are there advantages to restructuring the traditional path through graduate school so that a Master's degree is not a "failed'' PhD? Should Master's programs be designed to provide a broader more "applied'' approach to scientific training, with possible specialties in instrumentation, computing, or science education? An enhanced Master's degree programs of 2-3 years duration might incorporate some or all of the following:

An enhanced Master's degree would represent a desirable and "marketable'' outcome for many students, and would come to be recognized widely as a mark of significant academic achievement.

Workshop participants were asked to examine the merits and liabilities of (1) developing enhanced MSc programs; and (2) requiring an MSc for admission to PhD candidacy.


This section synthesizes the principal findings which emerged from the working group discussions at the regional meetings, those reached in plenary sessions by all participants at the three regional meetings, as well as those expressed in our open dialogs with Society members.

4.1 General Conclusions

4.1.1 Create a climate of experimentation to foster necessary changes

Workshop participants and AAS members recognize that we are entering an era when the assumptions underlying academic research which developed during the cold war decades are being challenged. As US society positions itself to compete successfully in the emerging global economy, funding levels and patterns, as well as the structure and size of the research enterprise, are bound to change.

Members recognize that adapting to change and shaping creative responses to political and economic realities will require flexibility and a willingness to experiment - with the attendent understanding that experiment entails risks and results in failures as well as successful new structures. They were nevertheless clear in their commitment (1) to undertake a critical re-examination of graduate education and the underlying assumptions of the academic culture in which it takes place; and (2) to urge the community and its funding agencies to develop and support experiments aimed at stimulating healthy change and finding new approaches to better serve both our students and society - fully recognizing that funds directed towards enabling such experiments, may decrease the number of dollars available to support other desirable programs.

Indeed, participants remarked that given the will to experiment expressed in the regional meetings, the small size of the astronomical community represents an advantage: lower community "inertia'' can enable relatively rapid response to a changing environment. Despite its modest numbers, astronomy could in principle play a leadership role in exploring new approaches to graduate education.

4.1.2 Retain primary focus on producing first rate research scientists

Participants urged that in selecting and funding experiments, great care be taken (1) to minimize risks to a structure which has proven enormously successful in training scientists who have carried out world-leading research in astronomy for more than 50 years; (2) to subject experiments in graduate education reform undertaken with federal support to the same rigorous standards of peer-reviewed evaluation that inform selection of first-rate research proposals.

4.1.3 Deliberate reduction of the population of graduate students or of graduate departments is not wise.

Past trends and current predictions suggest that of those presently enrolled in graduate school, no more than 30% will assume permanent positions in academia or national laboratories. Despite these predictions and the recent rise in astronomy PhD production, workshop participants - including students - felt strongly that neither graduate schools nor funding agencies should adopt practices aimed at matching graduate admissions to perceived "traditional'' job opportunities.

First, it was agreed that the admissions process is imperfect; identifying college seniors with the combination of intelligence and temperament matched to a research career is, with few exceptions, extremely difficult. Practicing "birth control'' at this stage would result in premature evaluations based more on "objective'' criteria than on assessment of a student's performance in a graduate research department. Moreover, "birth control'' at this early stage would almost certainly compromise the ability of graduate departments to meet their stated goal of enhancing diversity in the physical sciences.

Second, birth control via elimination of departments was also deemed both unwise and impractical both (1) because astronomers in small, recently-formed departments make significant and imaginative contributions to training productive MScs and PhDs; and (2) even were one to adopt implicit or explicit strategies to eliminate or curtail support for the "lower'' half of the 74 degree-granting institutions, the current PhD production rate would be decreased by only 20%.

Furthermore most participants believed that graduate training in astronomy provides students with a strong background in physics, and experience in posing and solving problems - two critical components of the skill mix needed for success in a wide range of careers. It was felt that if changes were to be made, they should be directed toward expanding the goals of graduate education. This would allow us to improve our preparation of individuals able to contribute creatively to the economic well-being of the nation rather than to retrench in order to avoid "overproducing'' PhDs for the academic market.

4.1.4 Provide students with accurate and up to date career statistics and clear information regarding alternative careers.

The mismatch between PhD and "traditional'' job production rates places a special obligation on graduate departments to provide students with accurate information regarding career trends and the career paths taken by past graduates. Participants agreed that departments should assume responsibility for longitudinal tracking of their graduates and for making such information available both to applicants and to the AAS. In turn, the AAS should use these data (with specific linkage between names, departments and outcomes omitted) along with corresponding NRC information for the physical sciences to compile an up-to-date statistical picture of short- and long- term job trends.

4.1.5 The combined median time-to-degree of 7 years followed by 2 postdoctoral fellowships is unreasonably long.

Participants were unanimous in the belief that the current norm - spending 7 years in graduate school and 6 or more years in post-doctoral positions before determining whether a career in astronomy represents a realistic option - is unacceptable. While some argued that placing a hard 4 to 5 year limit on time-to-degree would be a desirable first step in reducing the "time to career decision,'' there was no consensus, and in fact considerable reluctance, to commit to a specific set of guidelines or to mandatory federal funding caps aimed at reducing time-to-degree. Others argued that imposing mandatory statutes might have the undesirable goal of reducing the population of women and minorities in our field. Moreover, many participants remarked that a recommendation to broaden graduate education and to reduce the time to degree might well be inconsistent.

This critical area is one which would benefit from additional structured discussion both within the community and with our funding agencies, perhaps in the context of the next decadal survey. The discussion should have the goal of recommending community guidelines that would address reducing both the time-to-degree and the time young PhDs must wait to postpone long term career decisions.

4.1.6 Changing the academic culture

Participants acknowledged that a realistic re-examination of graduate education requires that departments and institutions look critically at the "cultural'' assumptions which inform their actions. In particular, the current system primarily (or exclusively) rewards those who pursue academically-based fundamental research. Often, pursuit of alternative intellectual challenges - for example, those focussed on the educational missions of the institutions - are discouraged either actively or passively, sending clear messages to young faculty and students alike.

Many participants lamented the constraints of the present system but were unable to develop consensus recommendations regarding mechanisms for escaping its "traps.'' As a first, relatively painless step, it was recognized that modest readjustments in attitude by graduate faculty could make a significant difference in the self esteem of students who wish to pursue broader educational goals. There was also agreement that in the long run institutional and federal incentives aimed at stimulating cultural change will be critical.

However, it was impossible in the time available to develop consensus regarding desirable changes, and the institutional and federal funding and review practices required to effect them. The subject is complex, and developing consensus regarding sound community guidelines for change will be difficult. However the issues are central to the future of the community and merit additional careful discussion, perhaps in the context of the next decadal survey.

4.2 Broadening Academic Options: Conclusions

4.2.1 Emphasize basic scientific training and problem solving skills

The regional working groups were unanimous in urging that astronomy graduate education remain focussed primarily on providing basic scientific training and problem solving skills, through coursework and research experience. Many argued that because astronomy is by nature highly interdisciplinary, this approach - thoughtfully applied - would ensure that students are prepared to address a broad range of problems in the physical sciences, even absent any additional curricular changes. Some participants expressed the belief that it might be timely to re-examine extant departmental offerings to ensure that courses continue to emphasize fundamental skills as opposed to increasingly specialized content.

There was unanimous agreement that curricula should stress early research experience to enable students to begin as soon as possible to develop the skills required to pose and solve problems, both independently and in groups.

4.2.2 Broaden graduate education

Despite the interdisciplinary training inherent in astronomy graduate programs, there was resonance with the Griffiths and Armstrong recommendations to complement sound basic training and problem solving skills with curricula that encourage increased exposure to both "hard'' (e.g. numerical simulation and visualization techniques, astronomical instrumentation, database management) and "soft'' (e.g. communication, team problem solving, management, teaching) skills during their graduate careers. Each group believed that broadening academic options in this way would benefit students both by enhancing their potential effectiveness as researchers in astronomy and their ability to secure rewarding, "non-traditional'' jobs.

Very modest levels of federal funding directed toward supporting curricular changes incorporating the above elements would, in our view, stimulate experiment and healthy change throughout the community. We strongly urge the funding agencies to work with the community to develop programs that would enable departments to propose innovative approaches aimed at providing opportunities for graduate students to develop broader skills. We believe that investments aimed at encouraging (1) partnerships with other departments within their institutions, with local industry, with local schools, and (2) development of enhanced MSc programs would be particularly efficacious.

4.2.3 Achieving broadening goals must not add new courses to graduate curricula

However, broadening the graduate experience was deemed best accomplished by restructuring extant courses and pre-dissertation research experience rather than by requiring students to take additional specific classes; current requirements have already driven the median time to degree to 7 years - a time which was deemed unacceptably long.

4.3 Creative and Rewarding Teaching Experiences Should Be Provided

Participants were unanimous in their belief that graduate curricula should be structured to provide mentored classroom teaching experience, and to encourage development of a wide range of teaching skills (e.g. communication of physical concepts to non-scientists, developing interactive course modules). Developing these skills will serve students well in a variety of careers: in traditional academic research positions, in teaching positions at all levels, and in a wide range of private sector jobs. Again, participants believed that integrating teaching experience within extant curricula would be far preferable to requiring students to enroll in specialized courses focussed on teaching methods. Developing approaches - including federal funding incentives aimed at encouraging curricular changes to enhance student teaching experiences (for example, through award of federally-funded "science education fellows'') - was deemed to have high priority.

4.4 Graduate Student Support Modes: Conclusions

4.4.1 Increase the opportunity for training in instrumentation.

There was near-universal agreement that NSF and NASA be urged to initiate or expand programs to identify, encourage and support individuals who have the inclination and aptitude to pursue careers related to astronomical instrumentation and technology development.

A number of participants urged initiation of short term internships of 2-4 months duration - modelled after the NSF REU programs and targeted at a broad cross-section of students just beginning their graduate careers. Such programs would be hosted by labs in academic, national observatory, or industrial settings, and would provide intensive exposure to astronomical instrumentation to students in their first or second year of graduate school. For some, such a program might act as a catalyst to choosing instrumentation development as a career path. For most, the experience of working in a laboratory environment would represent an important addition to their graduate training by providing exposure to a broad range of skills (instrumentation design and development; team research; project management) which have wide applicability in both the academic and private sectors. Indeed, observatory graduate internships may be essential for the future of astronomical research, in that they may be the only way that a sizable fraction of future researchers gain hands-on experience with the complex instrumentation of the next generation telescopes.

There was also considerable support for increasing the number of longer term internships or traineeships at industrial, national or major university laboratories, aimed at students desiring a career in instrumentation. This would require targeting additional NSF or NASA fellowships toward individuals with strong potential for careers in instrumentation development.

4.4.2 Increase the availablity of fellowships after 2-3 years.

A consensus view emerged from all 3 workshops: (1) competitively awarded fellowships represent an important mechanism for supporting first rate students and enabling them to structure successful graduate careers, and (2) that their number should be increased. Participants urged that more emphasis be given to awarding such fellowships not at entry to graduate school, but rather after 2 to 3 years when a student's competence, aptitude and temperament for research had been evaluated. Some were in favor of offering the fellowships after competitive review of dissertation proposals, citing as advantages: (i) developing early "focus'' for a dissertation through the discipline of preparing a structured proposal; (ii) providing "real world'' experience in competitive proposal writing; (iii) enabling students to chart and follow a brisk timetable for completing their research independently of the demands often placed by requirements to meet deadlines/requirements of funded research programs; (iv) the opportunity to structure mentoring relationships with a broader range of faculty, including those whose research support may already be commited to other students.

4.4.3 Do not abandon traditional research assistantships.

The traditional model in which most graduate students are supported via research assistantships administered by individual investigators is, however, considered basically sound and should remain the preferred mechanism for supporting PhD students. Participants argued strongly that this approach represents a demonstrated means of ensuring that graduate students are closely mentored, and work on cutting edge science selected for support via competitive peer review.

As a result, workshop participants and other members of the community urged caution in shifting the balance of federal support away from RAs and towards fellowships/traineeships - with the exception of urging "full speed ahead'' on the recommended programs to increase opportunities for training in instrumentation.

They argued instead that the community would be well served by a careful, structured community discussion regarding funding patterns and practices that will best meet the needs of students, the discipline and society. The upcoming decadal survey provides an appropriate and timely forum for developing such recommendations.

4.4.4 Do not adopt policies which disadvantage foreign nationals.

While endorsing experiments to somewhat alter the balance of funding support toward the fellowship/traineeship modes, participants at the same time expressed deep concern that effecting these experiments might have the unwanted consequence of decreasing support opportunities for foreign students (at present, federally-funded fellowships can be awarded only to US citizens). There was near-unanimous agreement that astronomers should avoid implementing practices which implicitly or explicitly erect barriers to access by foreign nationals to US graduate programs. It was emphasized that astronomical research in the US has benefited enormously from the vigor and imagination of foreign nationals, and that construction of large ground- and space- based research facilities increasingly depends on cooperative efforts with multiple international partners. The effectiveness of these partnerships depends significantly on establishing close personal collaborations and mutual understanding of cultural differences - both of which develop naturally in a system which is open and financially accessible to foreign nationals.

4.4.5 Do not move to block grants.

All three regional workshops considered the possible merits of block grants awarded competitively to departments who propose experiments aimed at reshaping their graduate programs to take advantage of unique circumstances (e.g. linking with industrial partners; linking with a school of education to develop a focussed program in science education). There was considerable skepticisim regarding both the mechanisms and review criteria that might be used for block grant evaluation. Moreover, participants felt that block grants historically have not typically met the same quality standards typical of peer-reviewed individual investigator grants.

However, the reluctance to adopt the block grant mechanism per se should not be interpreted as a lack of enthusiasm for providing a variety of opportunities for funding experiments aimed, for example, at comprehensive curricular reform, partnerships with industry, or imaginative Master's programs. There was, instead, near-unanimous support for encouraging such experiments, provided that they were designed by first rate investigators and subject to rigorous peer review.

4.5 The Master's Degree: Conclusions

4.5.1 Consider developing enhanced MSc programs

There was interest in encouraging departments to develop enhanced Master's degree programs designed to provide the broad and thorough training necessary for students to either continue with a PhD in astronomy or a related field, or to leave graduate school armed with the skills necessary to embark successfully on a variety of career paths.

Schools embarking on this path would offer a MSc program of 2-3 years duration with a core curriculum providing (1) basic scientific training via thorough introduction to the physical principles underlying astronomy and astrophysics; (2) mentored, creative teaching experience; and (3) mentored research experience. In addition to this curricular core, MSc students would also develop expertise in both "hard'' skills, such as high speed computing, visualization techniques, and instrumentation, and "soft'' skills, such as effective written and oral comunication and team problem solving through structured research projects, either in the classroom, in national or industrial laboratories, or with local facilities.

4.5.2 Consider the merits of requiring an MSc for entry to a PhD program

Perhaps the most surprising and potentially far-reaching outcome was the groundswell of support in advance of the workshops expressed for requiring all students to complete an enhanced MSc prior to admission to a PhD program.

While there was not uniform support for this idea, proponents argued that this path would (1) allow institutions to recruit more widely and take greater risks in admitting potentially talented individuals; (2) provide both students and faculty with an opportunity to "pause and reflect'' prior to making a commitment to pursue a PhD or to choose a different career path; (3) allow institutions to be more selective in admitting students to a PhD program; and (4) allow graduate programs to grow in stimulating and potentially benefical new directions, by forging links and partnerships with a wider scientific community.

The notion of an interlude to "pause and reflect'' before commiting to a PhD dissertation was particularly appealing to our graduate student participants. Many of them acknowledged that with no clear understanding of careers available to Master's Degree recipients, pursuing a PhD often seems the simplest path to follow. Some students even acknowledged that they continued graduate study past the Master's level, knowing that a research career in astronomy was not their ultimate goal.

Other participants were deeply wary of this approach because:

The issues raised by the proposal of an enhanced, required MSc require far more time for thoughtful dialog than was available during the workshops. Further discussion of this issue in the course of the upcoming decadal survey would allow the community to dicuss more fully the advantages and disadvantages of this proposal.

4.5.3 Potential niche for professional Master's degree in astronomy instrumentation

Participants noted the potential of a relatively small but still significant niche for specialized Master's degrees in astronomy-related instrumentation. Such programs might be tailored to develop the skills required to work on academic/industrial teams that will design and operate the sophisticated instruments and detectors required for present and future ground- and space- based facilities. There were felt to be advantages to having people with some astronomy background in these important positions.

4.5.4 Potential niche for professional Master's degree in astronomy education

A potentially larger role for Master's students might involve participation in formal or informal education programs (museums; planetaria; outreach programs linked to observatories or large projects). Some participants suggested the possibility of developing a joint Master's programs with schools of education, with the aim of bringing the excitement of astronomy to the classroom and to K-12 curricula. To effect such programs would realistically require developing incentives - in the form of funding opportunities - to overcome the barriers to cooperation between traditional academic and education departments.

4.5.5 Potential niche for professional Master's degree in astronomy and related fields

Several participants reported success in developing Master's programs involving collaboration with departments of engineering and computer sciences. In some cases, these programs involve internships with local industries. They appear to provide strong cross-disciplinary training for students and effective research collaborations for faculty, particularly at institutions with relatively small departments. Participants urged developing mechanisms - including seed funding opportunities - to encourage and support such programs where local conditions are favorable.


Recommendations stemming from discussions carried out during the past year fall into two general categories. The first comprises actions to be undertaken over the next two years by individual departments, the AAS, and the funding agencies. The second includes additional recommendations that should be considered carefully as part of a continuing effort to improve graduate education in astronomy.

5.1 Action Items

5.1.1 Support experiments to enrich graduate education

5.1.2 Provide students with information and opportunities for a broad range of career decisions

5.1.3 Re-examine the structure and role of the Master's degree in astronomy

5.2 Additional Recommendations

5.2.1 Provide students with better experience in real world skills

5.2.2 Provide students with creative and rewarding teaching opportunities

5.2.3 Expand opportunities for students to gain exposure to related disciplines and technologies

5.2.4 Critically examine the current "academic culture''


We deeply thank the invited speakers at each of our regional meetings: Neal Lane of the National Science Foundation, Phillip Griffiths of the Institute for Advanced Study, Marye Anne Fox of the University of Texas, and Judy Franz of the American Physical Society. We are grateful to the OSEP branch of the National Research Council, including Karen Greif, Charlotte Kuh, Don Shapero, and Jim Voytuk, for researching and presenting relevant statistical data. We thank John Armstrong for playing a pivotal role in the early stages of our study. We acknowledge useful conversations with Mary Gollady of the DSRS division of the National Science Foundation, and we acknowledge support of the National Science Foundation in carrying out this study through AST-9622599.


Armstrong, J. 1996, Graduate Education and Postdoctoral Training in the Mathematical and Physical Sciences, NSF/MPS Report (NSF 96-21)

Fox, M.A. 1997, Fostering Breadth in Graduate Education in Astronomy, in The American University: National Treasure or Endangered Species? ed. R.G. Ehrenberg Cornell University Press, 1997, p. 100.

Franz, J.R. 1995, Physics Graduate Education for Diverse Career Options, Department Chairs Conference, Special Report from the American Association of Physics Teachers and The American Physical Society

Griffiths, P.A. 1995, Reshaping the Graduate Education of Scientists and Engineers, Committee on Science, Engineering, and Public Policy (COSEPUP), National Academy Press


A. Participants in the AAS Regional Graduate Education Workshops

A.1Keynote Speakers

Marye Anne Fox, Vice President for Research, U. Texas at Austin
Phillip Griffiths, Director, Institute for Advanced Study
Neal Lane, Director, National Science Foundation

A.2 PhD Granting Graduate Departments

University of Alabama
Arizona State University
University of Arizona
Brigham Young University
University of British Columbia
Carnegie Mellon University
Case Western Reserve
University of Chicago
University of Colorado
Columbia University
Cornell University
UC Berkeley
UC Davis
UC Irvine
UC Los Angeles
UC Riverside
UC Santa Cruz
University of Delaware
University of Florida
Georgia State University
University of Georgia
Harvard University
University of Hawaii
University of Illinois
Indiana University
Iowa State University
Johns Hopkins University
University of Kentucky
Louisiana State University
University of Maryland
University of Massachusetts
Universidad Nacional Autonoma de Mexico
University of Michigan
Michigan State University
University of Minnesota
University of Nebraska, Lincoln
University of New Mexico
New Mexico State University
University of North Carolina
Ohio State University
Pennsylvania State University
University of Pittsburgh
Princeton University
Rice University
Rutgers University
University of Rochester
San Diego State University
Stanford University
Suny, Stony Brook
University of Texas
Texas Christian University
Vanderbilt University
University of Virginia
Virginia Tech
University of Washington
University of Wisconsin
University of Wyoming
Yale University

A.3 Others

Mark Adams, McDonald Observatory
Cindy Blaha, Carleton College
Marc Davis, CAA
Roger Davis, NM Science and Technology Company
Judy Franz, Executive Director, APS
Howard French, Honeywell Corporation
Richard Greene, NOAO
George Greene, Dean of Graduate School, UMN
Karen Greif, National Research Council
Bill Herbst, Wesleyan University
William Harris, Biosphere II Center
Patricia Keller, Education Department, UMN
Charlotte Kuh, National Research Council
Louis Lanzerotti, Lucent Technologies
Susan Lea, SFSU
Bernard McDonald, NSF
John Neithammer, MN Technology Inc.
Stephan O'Neil, Technology Transfer, U Colorado
Leonid Ozernoy, George Mason University
Ted Poehler, Provost, Johns Hopkins University
Jeffrey Rosendahl, NASA
Russ Ritenour, Radiology, UMN
Andrea Schweitzer, Honeywell Corporation
James Schuder, Sun Microsystems
Hugh Van Horn, NSF
Paul Vanden Bout, NRAO
Kim Venn, Macalaster College
Jim Voytuk, National Research Council
Sidney Wolff, NOAO
James Wright, NSF
Charlie Wu, STScI


Doug Duncan, Education Coordinator
Andrea Dupree, President
Suzan Edwards, AEPB
Andy Fraknoi, AEPB
Alex Filippenko, AEPB
Katy Garmany, AEPB
Len Kuhi, Treasurer
Gene Levy, AEPB
Bob Milkey, Executive Officer
Michal Simon, AEPB
Steve Strom, AEPB

B. Keynote Addresses

B.1 The Obvious and the Overlooked

Neal Lane, Director, National Science Foundation
1 November 1996
Washington, D.C.

Good evening. It is truly a pleasure and an honor to join you as keynote speaker for this first in a series of workshops on graduate education in Astronomy. My thanks go out to Andrea Dupree and Bob Milkey and the AAS's education policy committee for giving me this opportunity.

I appreciate that we have the full spectrum of the astronomy community represented here today. We know the success of our efforts hinges on hearing from everyone--students, faculty, recent graduates, industry employers, funding agencies, societies, and others. The fact that all of us are here and have taken time to participate in these workshops sows the seeds of success for our collective efforts.

For me personally, it is also especially gratifying to know that I share the marquee for this series with good friends and outstanding leaders like Marye Anne Fox and Phil Griffiths. I can't tell you how fortunate I feel that my talk comes first in this series. Following Phil and Marye Anne would be like being asked to apply a second coat after Michelangelo painted the ceiling!

As I was preparing for this talk, I stumbled across a story that underscored one of the ironies of life in today's world. It seems an assistant high school track coach had set up a make-shift office in the team's equipment shack. His job was to record the runners' times and distribute them to other coaches. The shack had no electricity, so computers and copiers weren't an option for data entry and communicating the results -- not unlike some observatories I've heard about.

For this reason, the coach ended up relying on the tried and true system of a manual typewriter and carbon paper to produce multiple copies of his tabulations. One of the student-athletes saw him lining up the carbon paper and pecking away and said, "Wow. That's amazing stuff. Soon we won't need copier machines anymore.''

This story, in addition to highlighting a modern-age generation gap, also reminds us of the astounding pace of progress in our world today -- which brings me to the subject of my talk today.

My remarks are entitled, The Obvious and the Overlooked. I want to focus on two topics -- one obvious, and one that we often overlook. The obvious and sometimes obsessive topic is change -- specifically changes in the so-called compact between science and society. We need to approach these changes in a thoughtful manner, recognize they are real and probably irreversible, and appreciate their implications for graduate education.

The second topic I will address -- the one we often overlook -- is the opportunity presented by this period of change, specifically the opportunities brought by the advent of a new age of information and knowledge, an age of advanced computation, communication, and information technologies. These have immense implications for all fields of science and engineering, astronomy in particular, and engineering and their role in our society.

My message today is that we must explicitly factor both of these topics into our thinking and our recommended course of action. Far too often, our instincts are to focus too narrowly on responding and adapting to change. As important as this is, it is not enough. The greater challenge -- less distinct but certainly more rewarding -- is to position ourselves to pursue the opportunities emerging at this, the dawn of the 21st Century.

I have a particular interest in these issues as they relate to astronomy. Throughout my career, in my own work and experience, I have seen time and again that astronomy is a source of inspiration, motivation, methods, and technologies, and insights for much of science and engineering. Astronomy is, without question, a key to progress for our entire enterprise. Our innate desire as humans to understand the cosmos has fired the creative genius of our brightest minds for generations. That will certainly always be true.

I once had the honor of writing a chapter on the contributions of my long-time mentor, Alex Dalgarno of the Harvard-Smithsonian Center, to atomic and molecular physics -- structure, collisions, and interactions with light. The experience was very pleasant but at the same time frustrating given the enormity of the contributions Alex has made and the paucity of pages I had been allotted to describe them.

As I was reviewing all this work and, more generally, the many connections between the atomic, molecular, and optical sciences and astrophysics, I was once again struck by the power of astronomy and astrophysics in motivating and stimulating scientific discovery across the broad sweep of time and virtually all fields.

Equally striking examples of these connections were highlighted just two weeks ago at a major event here in Washington. The five Americans awarded the 1996 Nobel prizes in physics and chemistry came here to spend a day in the media spotlight. It gave them the chance to showcase their achievement and discuss the value of science and the role of NSF and other Federal support in their success.

Hugh Van Horn, NSF's Division Director for Astronomy, reminded me of an interesting storyline that emerged from their presentations. The breakthroughs that earned them the Nobels were inspired in large part by questions emerging from astronomy.

Rick Smalley and Bob Curl of Rice University, who shared the Chemistry award with Harry Kroto of the University of Sussex, made this very point. They told of how Kroto's interest in the long-chained carbon molecules detected by radio astronomers first led them to apply their laser-supersonic cluster beam apparatus to carbon. That led them to the buckeyball, and the rest is history.

The physics laureates, David Lee and Robert Richardson of Cornell and Douglas Osheroff of Stanford had a similar story to tell. They described how their initial interest in the superfluidity of Helium 3 was inspired by questions about the behavior of Fermi liquids within neutron stars.

These examples underscore the importance of our gathering here today and of this series of workshops. It is vital that we sustain and nourish the astronomical and astrophysical sciences, and ensure that they remain rewarding and inspiring fields of scientific endeavor and education. As these examples show, this works to the benefit of the entire science and engineering enterprise and our nation as a whole.

Our workshop planners may not have known this, but it is highly appropriate that November 1 was chosen as the opening day for this session. Forty-four years ago today, the first H-Bomb test was conducted. That's an anniversary that may trouble some of us, while reassuring others. It nevertheless also provides an indelible reminder of the role science has played in our national security, and conversely of the role national security has played in justifying public support for science. That brings me to the subject of change.

All of us here today know better than most that the Federal government's significant role in the support of science grew from the contribution American scientists made during World War II -- not just in applications of nuclear science but in countless other areas as well. Although in many nations the scientific community has been called upon to help with a war effort, only the United States, through the articulation of Vannevar Bush, adopted a genuine pact between science and the rest of society.

The agreement outlined by Bush was reinforced, especially in our attitudes and rhetoric, by the significant demands of forty years of Cold War. It is debatable, and hotly debated, how important the Cold War was to Federal support of science. But, most would agree that Bush's agreement garnered much of its legitimacy and sense of irreversibility from the nation's overarching national security needs.

Since the end of the Cold War six years ago, many in our community have been debating the tenets of the compact between science and society. Does it need to be revisited, reformed, reconstructed? The implication being that a simple rewrite and update of the compact could usher in a new golden era.

Reality, of course, is not so simple. In a very real sense, this debate was eclipsed before it began. A more divergent perspective on science and technology has taken hold -- a perspective that I believe we need to understand and articulate. It seems to me something very different from the "compact'' has been emerging in practice over the last 10-15 years while many of us have remained wedded to the old rhetoric. The initial compact of the Vannevar Bush era has evolved, perhaps without our conscious attention.

What has taken shape, I believe, is a more expansive and multifaceted arrangement that encompasses the larger R&D enterprise in the nation, both public and private. This includes, of course, industry, small business, national laboratories, mission agencies, research universities, and state and local economic development councils. These, and others I have not named, are critical components of the new, and as yet unbaptized, arrangement that has grown from the original compact.

And so, as we continue to speak with reverence about the compact between science and the federal government, we are, in fact, thinking of the past more than the present. I want to emphasize strongly that I am not suggesting we lessen or eliminate the Federal role in support of fundamental research, primarily in our universities. I believe just the opposite. Rather, I am suggesting that this new perspective is a way to recharacterize the roles and positions of all participants in the national research and development system, as the system has evolved and matured. It is accurate to say that our once narrow compact has blossomed into a cornucopia.

And so for those who are threatened by imminent changes in the old compact, I would comfort them with reality -- the reality that those changes have been in play for some time. The current R&D enterprise that has evolved from the old compact is not now, and never will be, a cemented set of dogma. Rather, time and events have fashioned a dynamic and sometimes even spontaneous system. There will be no reversal of this momentum.

The reality of these changes -- and their irreversibility -- is most apparent in the focus of our discussion today: graduate education. There was a time, not very long ago, when newly minted PhDs were often considered disappointments if they had to venture beyond academe or a small set of eminent research labs to secure employment. Today, it is folly to try to turn back the clock and ask whether there will be enough faculty openings and traditional research positions for tomorrow's PhDs. That is too narrow a way to think about graduate education.

I noticed that the background materials prepared for these workshops included a number of references to what's commonly called the "COSEPUP report on graduate education.'' This report does deserve our attention, in part because it makes clear that a structural change has occurred. It notes that, "Among recent PhDs, there is a steady trend away from positions in education and toward applied research and more diverse, even non-research employment.''

We see this trend in virtually all of science and engineering. It is perhaps most acute in the mathematical and physical sciences, though no fields are immune. Where 20 years ago, the majority of PhDs embarked on academic careers after graduation, today a majority pursue careers outside the academy.

Our friends at the American Chemical Society and the American Mathematical Society tell us that a sizable fraction of new PhDs are encountering prolonged and increasingly frustrating job searches. Over 20 percent of PhD chemists report being unemployed up to six months after graduation. Over 10 percent of new PhDs in mathematics report being unemployed a full year after graduation.

The situation is virtually the same in astronomy. The Survey of Earned Doctorates sponsored by NSF's Division of Science Resources Studies testifies to this.

The bottom line is that 7 out of 10 recent PhDs in astronomy have not obtained what most of us would consider stable or satisfactory career positions.

We've all witnessed this feed a vicious cycle of frustration and even resentment for many of America's best and brightest. You may have seen the New York Times Magazine article that ran in late September entitled, "How to Make a PhD Matter.'' The author is a professor of English, but his opening observation matches much of what we are seeing in the physical sciences. He writes: "Getting a PhD today means spending your 20's in graduate school, plunging into debt, writing a dissertation no one will read -- and becoming more narrow and more bitter each step of the way.''

Most of us find more than a small measure of irony and incongruity in this state of affairs. We all recognize that the frustrations are real and the discouragement profound. But, we also know that the research itself remains a hotbed of excitement and amazement. Everything that attracted us and continues to attract young people to research remains as abundant as ever.

In just the past year, we have witnessed the discovery of planets orbiting stars beyond our solar system, gained new insights into dark matter, and viewed images of the birth of galaxies. We have even been tantalized by the prospect of ancient microbial life on another planet. While the jury is still out on that last finding, it nevertheless helps to fortify and elevate the sense of optimism and opportunity surrounding astronomy.

Indeed, I believe this sense of optimism and opportunity belongs at the center of our deliberations on the future of graduate education. We cannot afford to have it otherwise -- or to overlook these opportunities. We may be witnessing the emergence of a new era of discovery and progress for science and engineering in America. Leadership from the research community -- astronomy in particular -- will determine whether we realize the opportunities it offers.

Consider, for example, the data on U.S. economic growth since World War II. Our GDP has grown by a factor of six over the past five decades, thanks in large part to scientific and technological progress. Economists have concluded that innovations emerging from science and technology account for roughly one-third of all economic growth over the past half-century. That's an amazing accomplishment -- one in which we should all take great pride.

But I would argue that the best is yet to come for science and technology. Our role and our contribution will be even greater in the future. We have now entered what many are calling "the information age,'' and it promises to be an age where progress and success are shaped by scientific and technological endeavors.

This potential for progress emerges in large part from advanced computing, communications, and information technologies, such as those related to the World Wide Web. This has revolutionized physics, astronomy, and virtually every field of research. It's changed how we conduct research, post results, access data and images, develop and debug code, and search for jobs and explore career options. We can even read Ap.J. Letters weeks before it arrives in printed form.

All of this, as powerful and revolutionary as it is, represents just one small hint of the potential impact of these emerging technologies on our society. This impact is so profound that it may well redefine how we view terms like "school,'' "classroom,'' "laboratory,'' "university,'' and "career in science and engineering.''

In a trilogy of speeches delivered in February of this year, Vice President Gore suggested the metaphor, "distributed intelligence,'' to describe this new age of intelligent systems. It is a complicated metaphor, based on applying the principles of parallel processing and networking to social challenges and economic progress. It is noteworthy in and of itself for the Vice President of the United States to discuss things like massive parallelism and broad bandwidths. That's almost certainly a first in our history.

Distributed intelligence rests upon one key concept: shifting information and control away from centralized systems to the individual. It's a notion we've been discussing at NSF for some time, even before the Vice President introduced the metaphor.

One could say that this involves all of society getting wired, except that it won't always involve wires. This may yield an age in which the sharing of information is instantaneous and ubiquitous. We are already gaining a few glimpses of its potential impact on both science and society:

At NSF, we are exploring how to approach this concept in ways that maximize both its scientific potential and its benefit to society, and we are still grappling with some of the basic terminology. We are currently developing approaches that come under headings like "learning and intelligent systems'' and "knowledge networks.'' These include such activities as data mining, visualization, pattern recognition, improved human/computer interfaces, and learning technologies. We believe all of these hold immense benefit both from a scientific standpoint and as drivers of economic growth and social benefit.

When we view graduate education in the context of these activities, it points out the fundamental dilemma we face.

Fortunately for all of us, countless efforts are underway to remedy this dilemma. Most of these are department and discipline-based. At NSF, we have developed a number of mechanisms designed to encourage these grass-roots efforts.

None of these efforts provides the magic bullet for the challenges we face. We know that. It will take much more -- and most important of all it will take a commitment to working together and testing and exploring new approaches.

To conclude my remarks this evening, let me first say that I know I am asking a lot of you. I am asking you to do more than just respond to change, but also to exercise leadership that will help position the entire science and engineering community for new opportunities. That is a tall order, but I can't help but to recall the last time I spoke to an AAS gathering -- last January in San Antonio.

Those who were there caught the full throttle of my frustration over the government shutdown and prolonged budget battles. I called upon the astronomy community to lead and help spread word about the importance of stable funding for research. You delivered, and the entire science and engineering community followed your lead. I believe it is no coincidence that this year NSF's budget was in place prior to the start of the fiscal year.

Today, while I don't quite bring the same feelings of frustration or messages of immediate demise, we still face an uncertain future. The long-term threat to science is just as real as I described to you last January. We know that the drive to balance the budget will bring some rocky times. This gives all of us good reason to feel apprehensive about the future.

But as I've said throughout my talk today, we can't let this apprehension slow us down. I believe we will have a future golden age of science. It just won't be the same as in past years.

I believe that scientific research will continue to explore the most fundamental questions of nature. Clearly, this includes astronomy. And, doctoral-level graduate education will continue to prepare the nation's brightest and most capable students to make the major discoveries of tomorrow.

I nevertheless predict our system of research -- including university research -- and education will do much more than this. In a future golden age, research will also emphasize the integration and dissemination of knowledge beyond publishing in journals and presenting papers. We will rely on yet-to-be established networks of discovers and users. This new partnership will make the benefits of research more apparent and, at least some of the time, more immediate.

Graduate education at the doctoral and masters level will include valuable knowledge and skills, such as communication, teamwork, management, and leadership, that will enable scientists and engineers to excel in a wide range of professions. Why is this important? One reason is to ensure better jobs for science graduates.

But perhaps more important is the fact that future leaders in the world of business, law, medicine, and politics will need to understand science and technology to a degree society has never recognized and certainly not required before.

Will all of this come to pass? I don't know. But it will not come to pass unless we expand our views of research, of the university, of connections and partnerships involving the doers and users of science, and of graduate education.

This is why, as in January, I bring high expectations for leadership from all of you. Your efforts can once again show the way for the entire science, engineering, and technology community. We know that dramatic change is upon us, necessitating a thoughtful and appropriate response. We also know, but sometimes overlook, that even more dramatic and rewarding opportunities await us as well.

Your efforts through this series of workshops hold the key to success on both of these scores, and I once again have full confidence that you will deliver.

Thank you and thanks again to the AAS for inviting me to join you this evening.

B.2 Reshaping Graduate Education in Science and Engineering

Phillip Griffiths, Director, Institute for Advanced Study
8 November 1996
Minneapolis, MN

Today I'd like to speak to you about the way we prepare our students for careers in science and engineering, especially at the graduate level. This topic sounds somewhat mundane compared to thinking about quasars, dark matter, and the cosmos. And some of what I say may refer more to science in general than astronomy in particular. But most of you know how many young scientists -- including astronomers -- are having difficulty finding the secure careers they dreamed about in graduate school. And you also know that any scientific community needs a steady flow of bright graduate students. So I would suggest that the shape of graduate education has a powerful bearing on all of you, both as researchers who want the intellectual stimulation of younger colleagues and as educators who want to attract talented students to science.

This is a complex topic, and I owe a great deal to those who have helped to educate me. I've had the privilege of chairing COSEPUP, which is the National Academy of Sciences' Committee on Science, Engineering, and Public Policy, and over the past few years we've been talking with students, postdocs, faculty, and members of professional societies from all over the country about graduate education. As a result of these conversations and our own deliberations, COSEPUP released a report on graduate education about a year and a half ago, and this is a good time to reflect on what we learned in preparing the report and on what has taken place since then.

One thing we learned is how difficult it is to initiate the changes we want without causing changes we don't want. We certainly don't want to jeopardize the high quality of our present system. Also, my personal opinion and that of COSEPUP is that we don't want the level of graduate enrollment to drop off significantly. Any good research program needs a community of interactive faculty, postdocs, and graduate students; we count on the students to bring fresh perspective and to pose the original questions that are raised only by those who see a subject for the first time. We also need graduate students to staff research programs and to teach undergraduates. There is evidence that the influx of foreign students we saw in the late 80s and early 90s may be dropping off, and universities face a major challenge in attracting talented American students to science.

We've heard many informed people call for a reshaping of graduate education, and for good reasons. There are more graduates than there are academic positions and jobs in basic research, and recent graduates are finding the transition to other types of jobs difficult. More than half of new graduates with PhDs -- and much more than half in some fields, such as chemistry and engineering -- are now finding work in nonacademic settings. From this we can see that even PhDs who begin their careers in academia will probably make at least one career change -- a move to an adjacent field, to a position in academic or industrial management, or to a new career altogether. And I don't mean a new career in taxi driving. In spite of the anecdotes, there are very few PhDs in science working the graveyard shift at Yellow Cab. But there are many of them engaged in work that is not research. The job of our graduate schools is to equip them for challenging positions they have prepared for -- not just jobs they settle for. And today this preparation should include a range of options better aligned with the realities of employment.

One new reality of employment is that more graduates go to work in new kinds of scientific organizations. We're mainly familiar with two kinds of organizations: we might call them small science and big science. Small science operates on a kind of "Mom and Pop'' scale, with an individual investigator, a small number of researchers, and little need for management skills. Big science, which is typified by the national laboratories of defense and energy and large industrial facilities, employs many more researchers and promotes the development of its own managers to a degree not usually seen among university scientists.

Today we still have big and little science, but we're seeing growth in mid-sized organizations, such as engineering research centers, science and technology centers, and small to mid-sized companies. The most successful of these organizations are run by people who are closely tied to the science and who can also organize and run programs. This is a combination of abilities that is not addressed by the traditional graduate curriculum.

In preparing our report, COSEPUP talked with employers from both the academic world and industry, and asked them to critique the abilities of recent graduates. The employers praised their scientific ability, but they also reported difficulty in taking these talented graduates and developing the communication, leadership, and management skills that allow them to become administrators, directors, and managers.

COSEPUP also evaluated the Science and Technology Center programs that the NSF runs and found virtually the same condition. The level of science was almost uniformly high; the scientific community knows very well how to evaluate the quality of science in a proposal. But the selection process failed to evaluate the managerial aspects of proposals, and subsequent problems have been attributed to poor management. This is a challenge that is familiar to the astronomy community, which has traditionally managed telescopes and other facilities.

So when we think about the relationship between graduate education and careers in science, we see three things: a continuing tendency of departments to steer their students toward academic careers, a shortage of traditional academic research positions, and a feeling among employers that more non-research skills are needed. If we put these three things together, we see the need for a better alignment between the way scientists and engineers are educated and the way they are likely to spend their professional careers. So COSEPUP has recommended that universities offer a broader range of academic options to produce scientists and engineers who are more versatile and adaptable.

When we say "broader range,'' we don't mean that students should ignore the importance of the dissertation experience or chase after too many subjects. Instead, we envision a modified graduate program that is a richer variant of the current experience. We will always need the researchers who dive deeply into their subject and bring us the new discoveries and insights that underpin science and technology. Some people, especially those who are absolutely committed to academic research, will continue to best use their graduate years to focus deeply on their area of research, and then use postdoctoral time to broaden their understanding of the context of their work and perhaps of adjacent fields.

All students, however, should be offered more options. They should be able to prepare themselves to work across fields and to communicate and collaborate with people in other disciplines. Some of the most exciting problems in science today are found at the edges or interfaces of disciplines. We can think of problems in biophysics, neuroscience, computational chemistry, and of course astrophysics. A biologist working in biophysics needs not only expertise in biology, but also a true feel for physics. Getting a feel for another subject is hard to describe. It includes being able to communicate in the language of the other field, and most especially to gain a deep working awareness of how people in that field approach a problem and frame their questions.

For example, at the Institute for Advanced Study, we have a group of mathematicians and physicists working together on quantum field theory. In one very exciting development, the physicists, using ways of thinking that are more intuitive and less formal than those of traditional mathematics, have suggested that two models of quantum field theory should at a certain level be the same. Expressing this idea mathematically has led to certain mathematical predictions which seem to be true and that were unlike anything the mathematics community had ever seen. So this program is really being driven by the effort of the mathematics community to get inside the heads of the physicists and better understand how they think.

We can also think, for example, about a collaboration between mathematicians and geophysicists. Mathematicians know things about the relevant differential equations that can be helpful, but that's only part of the issue. They also have to know how one formulates geophysical problems and sets up the right equations. Differential equations are always approximate; they make certain assumptions. And those assumptions may give us an equation that doesn't really model what we're interested in. So the mathematician has to have an intuitive knowledge of the mind set of the geophysicist if the collaboration is to be a success.

One of the things that makes it difficult for students to gain insight into other disciplines is our tradition of organizing graduate departments by discipline. We do need to keep enough of this disciplinary organization so students have a firm foundation to work from. But many of the emerging areas of science lie between the traditional disciplines. So the challenge is to make modifications that allow graduate students to at least dip their toes into these emerging areas. We're not suggesting these modifications in response to a bureaucratic or organizational theory, but in response to the way science is being done today.

There are a number of steps departments can take to offer a wider range of options. They can offer academic options, such as minors, electives, and internships in adjacent or related fields. And they can help students sharpen more general career skills, such as the ability to express technical matters to nonspecialists and to work as members of a team. Faculty advisers can play an especially important role. By helping a student understand the shape of the scientific enterprise, they help the student make better career choices.

Another way to think about the need for breadth is suggested by Linda Wilson, president of Radcliffe College, who chairs the Office of Scientific and Engineering Personnel at the National Research Council. Linda points out that scientists are expected to perform many tasks during the course of their careers, including the following: synthesis of knowledge across disciplinary boundaries, follow-up of new discoveries, organization of new knowledge, transmission of new knowledge, transfer of technology, thinking about ethics, and educating the public about science. Students must learn to do these tasks just as they learn to do research.

There is some very good news about graduate education, and that is that faculty and students at many institutions have already initiated a variety of curricular modifications. Just a few examples: At Penn State there is an MBA/PhD concurrent degree program, and other programs that cross disciplinary boundaries. Ohio State has reorganized its departments and schools, and encouraged departments to broaden their master's programs and offer summer internships in industry. The University of Colorado has added multidisciplinary certificate programs in such areas as cognitive science, environmental policy, and biotechnology. Last June, COSEPUP hosted a National Convocation on Doctoral Education to bring together people who have tried strategies like these. A summary of this meeting is presented on the home page of the National Academy of Sciences. (

One way to encourage the kinds of modifications we've been talking about is to adjust the way the federal government supports graduate education. For the last two decades, about 85% of federal support for students has taken the form of research assistantships. These are awarded to a principal investigator to support a research program. Graduate students who receive support on that grant are expected to do research related to the grant. This apprenticeship method has, on the whole, worked marvelously. But as we go into the future, the overwhelming concentration of resources in research assistantships may not be the best method of support.

Let me suggest two alternatives. The first is to increase the proportion of traineeships, which have been used very successfully at the National Institutes of Health, especially for emerging or interdisciplinary areas such as cell biology and neuroscience. Training grants can be used in either emerging or traditional areas, where part of the grant's purpose is to encourage institutions to offer more curricular options and to provide more effective student guidance.

The other alternative is to give more student money in the form of graduate fellowships. My personal opinion is that fellowships offer a specific advantage: they allow students the maximum degree of choice. Students know which program is most attractive to them, and fellowships let them vote with their feet.

My own feeling is that we don't yet know which alternatives will work best. So instead of making radical reforms to the current system, we can experiment with incremental changes. Instead of putting 5/6 of our federal money into research assistantships, we could shift the proportion to 2/3 and see what happens. By making small shifts in the direction of fellowships and training grant mechanisms, we may get a feeling for what works best.

Another modification COSEPUP has looked at has to do with what we could call the "problem of the queue.'' Many new graduates are "lined up,'' waiting for tenure-track jobs, and this backup creates a tendency to spend more time in school and in postdoctoral positions. Over the last three decades, the time to degree has increased steadily -- in some fields by about 30%. According to the NSF, the median number of years between receipt of the bachelor's degree and the doctorate in science and engineering increased from 7.0 years during the 1960s to 8.7 years in 1991.

It may seem contradictory to suggest both a broader educational experience and a shorter time to degree, but I believe the key is to give students more of the information they need to plan their careers.

When I say information, there are several things I do not mean.

I do not mean we can predict where science is going to be in the future. We have a general sense of areas of science where activity is increasing and employment is growing -- such as information science, environmental science, and neuroscience. But that's not a prediction, or futurism; it's a projection based on what is known.

The second thing I do not mean by more information is trying to predict the job market. Many of us remember the famous prediction of a shortage of PhDs that the NSF put out in the late 1980s. No one could have known about the recession which followed that prediction and led to shrinking budgets for states and reduced numbers of tenured faculty at the state universities. There was no way to predict that, or many of the other factors that affect job markets.

What I mean by good information is information that can make students more aware of what their options might be once they have their PhD. One example of useful information is that everyone seems to agree that small to mid-size companies represent a growth area for the near to mid future. And these companies have certain employment needs that students should know about.

We had a number of people from these kinds of businesses talk to COSEPUP, and we asked them about the kind of people they like to hire. The first thing they're looking for is people who have a PhD -- people who have done an independent research project and know one area of science really well. The second thing they're looking for is people who are adaptable. They know that they can't hire a new scientist each time there's a breakthrough in a new technology; they have to use the people they already have. Graduate students who know this can see that they'll need to be versatile and flexible if they want such a job. So our second major recommendation is that universities and other institutions provide graduate students and their advisers with more up-to-date and accurate information about careers. Students need a much more realistic picture of what's out there.

Many universities have initiated public discussions of these issues and disseminated excellent career information. Brown University and Cal Tech's career service offices now assign a counselor specifically to graduate students. Johns Hopkins, Stanford, and other institutions hold career days, when students can learn how well their career goals match the needs of employers. The National Academy of Sciences has set up an on-line Career Planning Center for Beginning Scientists and Engineers. Earlier this year COSEPUP prepared a guidebook entitled "Careers in Science and Engineering: A Student Planning Guide for Grad School and Beyond.''

We hope that modifications like this will bring the supply of graduates most closely into balance with the demand of employers. Students with good guidance and information will be better prepared find the careers they want.

In the meantime, many institutions are grappling with the question of whether there should be limits on student populations. It's difficult to envision a good mechanism to limit numbers of students without being unfair or disrupting the educational process. And our country clearly benefits from attracting the best scientific talent, both from the U.S. and from abroad. However, some people have called for limits on international students, for two quite different reasons. One is that many of them remain in this country after graduation to compete for scarce jobs. The other is that many foreign graduates now return to their home countries, many of which offer rapidly improving opportunities for research. In Washington, this is beginning to raise the issue of whether we should be training and supporting citizens of other nations who eventually compete with us.

Whatever universities do about enrollment policy, the fiscal pressures for change are powerful. Some researchers believe that federal and state cuts in research funding over the next several years may be as high as 30%. Such enormous reductions would force institutions to go back to first principles and ask a basic question: How can we do excellent research and education in times of constricted external resources, and how can we treat students fairly in these times?

Too often we've lost sight of the educational aspect of our mission. As indicated earlier, one way to enhance this aspect is to shift resources from research assistantships to education and training grants. We can also provide better information so students can make wise decisions about their educational and professional careers. Certainly there are other ways to shift the balance, and this is where we rely on yourselves and others who have recognized the need for discussion and action.

I want to close with a thumbnail sketch of the kind of environment where more and more of our science graduates will be working. We've all read about the downsizing at Bell Labs, IBM, DuPont, and other major corporate laboratories. And we're generally aware of the shortening time horizons for R&D in industry. If industry is doing less of its own research, and doing it farther downstream, how will they retain access to the frontiers of science? The answer seems to be that they will have to rely on access to university research communities.

No one is sure how this will be done, but it will certainly include various kinds of consortia. Industries have a need to see what's coming down the road, and universities have a need to broaden their funding base as well as provide more educational options for their students. For example, we may see a "semiconductor research corporation'' which will give grants to campus-based researchers in much the same way NSF gives research grants today. We'll see more people moving back and forth between the universities and the research corporations, receiving and transferring new knowledge. Industry will put up money, as will universities, and the government will serve, among other roles, as advocate for long-term research and be an "honest broker'' between industry and the universities. The Frauenhoffer consortia are examples of this kind of effort.

There remain many problems of the structure and operation of consortia, because the objectives of universities and industries are quite different. Universities are interested in education and in knowledge for its own sake, while industries must apply the results of research to earn profits. For example, frequently the two parties disagree strongly over the control of patent rights to intellectual property. The universities would like to publish new results openly and to receive a portion of financial benefits, while industry would like to hold exclusive patent rights long enough to maximize their return on investment. Agreement on issues such as this will require a significant blending of two quite different cultures.

As these consortia become more common, the process of graduate education will have to take them into account. One way graduate students can prepare for them is to become familiar with industrial culture through internships and exchange of personnel. The graduate experience for a scientist who will work in a mixed research culture should be different from the experience of someone will work as a university professor.

The last few decades have been an extraordinary period in the development of science and in the public support for science. We can see now that the coming decades will be different, but they need not be less interesting, or less fruitful for science. It's important for the scientific community to see this now and to take positive steps to strengthen our system of education in appropriate ways.

We don't want to tamper with the essential soundness of graduate education, but we want to bring it into better alignment with the employment universe. This is especially important if we are going to attract the talented students we need to maintain our strength in science. With good cooperation among universities, industry, government, and the professional societies, we can maintain the level of excellence and world leadership we have all worked so hard to achieve.

Thank you very much.

B.3 Fostering Breadth in Graduate Education in Astronomy

Marye Anne Fox, Vice President for Research, U. Texas at Austin
22 November 1996
Tucson, AZ

Published in The American University: National Treasure or Endangered Species? ed. R.G. Ehrenberg Cornell University Press, 1997, p. 100.

B.4 The Changing Graduate Education and Research Environment

Eugene Levy, Dean of Sciences, University of Arizona
22 November 1996
Tucson, AZ

The changing shape of graduate education in science is a subject that is occupying a place of increasing national prominence. Signals abound: for example, over the past months prominent newspaper articles have appeared, even in the Arizona Daily Star, describing the allegedly shrinking career opportunities for scientists, and elaborating on the putative negative implications of this for today's cohort of students. This shrinking of opportunities is apparently driven by the reduction of support for basic research in the private sector -- especially in physical science -- and by the apparent petering out -- and threatened reversal -- of what had been, for a full half century, a sustained, high impulse thrust in growing governmental support for basic science.

It is my impression that the primary response of the academic community specifically -- has been to adopt a posture of studied denial...and thus I want to try to cast my remarks in a rather broad context, pertaining both to the changing research-support environment and to potential changes in the academic environment...even independently of the threats to research funding.

Beyond threatened reductions in support for scientific research, the situation in academic science is exacerbated by independent developments in university demographics and economics. For several decades in the middle part of the century, universities and colleges underwent an extraordinary expansion, as the bachelor's degree has displaced the high school diploma as the basic job certificate in much of the American workplace, and as the opportunity for college education was extended to a much larger fraction of the population than had previously been the case. This accelerated growth of universities seems more or less to be saturating out. It is not obvious that college will soon be the destination for a much larger fraction of the population than is already the case. While it is likely that the college population will continue to grow along with the general population growth, without continued increase in the fractional college enrollment, future expansion of the college/university population is likely to be much slower than it has been in recently past decades. I think we all are aware that the far-greater-than-replacement level of PhD production in the past was absorbed in significant measure by the expansion of higher education, as well as by expansion in the national laboratories and by a vigorous commitment to research and development in many private corporations.

On top of the change in demographics, significant public attention is being focused on the costs of universities and on the sources of funding for higher education and for research. During the past several years, cost consciousness and cost competition have started to become significant issues in the economics of higher education. I anticipate that this dynamic will accelerate, and that we will see increasing cost competition within the traditional community of colleges and universities. It seems clear, even as universities expand to accommodate a growing population, that expansion will not be funded as lavishly as has been the case over the past five decades.

Moreover -- and perhaps more importantly -- I anticipate the possibility of significant competition from alternative "education providers'' in an expanding and increasingly cost-competitive educational marketplace. Today we are confronted with the appearance of for-profit, private "education providers,'' named as universities...even with investors and stock-market listings...and already penetrating some constituencies that had previously been the exclusive province of traditional colleges and universities. It is as yet unclear how much of this competition will take place on the Internet, or whatever follows the Internet. This expanded competitive environment will likely continue to translate into an increasingly sharp focus on the ways in which funds are allocated and used within universities. The similarity between this developing situation and the dislocations currently occurring the health-care industry should not escape you. A substantial part of the U.S. academic research enterprise is supported by an implicit social contract that supports research -- largely through faculty salaries -- in an environment built largely for education. In the health-care industry, it has traditionally been the case that the cost of medical research was partly underwritten by hospital charges to patients and their insurance companies. In the past, health care was not a cost-competitive enterprise in the United States. In the current cost-competitive health-care environment, some of the biggest impacts are falling on medical research and education. It remains to be seen just how cost-conscious and cost-competitive higher education will become...but I believe that the potential threat to the current social/economic compact between education and research in the United States could be large and profound as they are becoming in medicine.

I want to remind you that I am not prescribing or advocating these changes and threats. I am merely describing what already has started to take place or what looms as potential threat to the prevailing social-economic-academic compact. I do not want to be shot as a messenger with grim tidings...nor do I want to play the role of Chicken Little crying about a falling sky. But rather I believe it is important for us to try to understand the changes that threaten to overtake us...and to consider, deliberately, how we preserve high value academic research environments and educational programs in a changing world.

These changes within universities and within the scholarly enterprise are being driven by social, economic and perceptual changes outside of universities. I have already mentioned the petering out of momentum in Federal support for basic research, and the possible development of a highly cost-competitive environment. But, at the same time, it is important to realize that there has been a substantial erosion in public confidence in major institutions of all kinds...not only universities, but governmental institutions, the Congress, legislatures and so on. While we in the university and research community feel mistreated by government institutions and funding sources, it is well keep in mind that the people in those institutions, who will most directly affect our future, are running scared themselves.

In thinking about the challenges we currently face, it is well to try to maintain a reasoned and balanced historical perspective. We are not at the end of history. The particular challenges that we face today are not going to last forever; history is somewhat cyclic. Good times are likely to roll again. In any case, research is not threatened with devastation, only with retrenchment. And even the retrenchment is not being visited on all disciplines equally.

However, the main reason I make this point about historical cyclicity is to preempt the issue. In some respects it is irrelevant. It is not enough to assert -- as I have heard many of my colleagues do -- that times will get better, and that we need only tough out the temporary downturn. Time scale is a central issue...some changes are indeed secular...and other changes are of sufficiently indeterminate and long time scale that they are indistinguishable from secular changes in terms of their impact on planning.

Indeed, I have already tried to make the case that at least part of the challenge we confront is of a secular nature -- the near saturation of the fraction of our population that will attend college. I think it is important realize that the world has changed in other ways as well.

Although we are focused here on astronomy, it is important to realize that similar discussions are going on more broadly, in other disciplines and in other forums. The recent studies of science graduate education carried out by the National Academy of Sciences and by the NSF Division of Mathematical and Physical Sciences mark the fact that there is growing anxiety about how our graduate programs fit into changing national needs and priorities. Much of the motivation for our most immediate anxieties derives from what we see as a discordancy between the supply of graduate students -- especially PhD students -- and the capacity of our disciplines, and our society more broadly, to absorb those students into jobs and careers that we -- and they -- regard as matching their skills, educations and ambitions in meaningful ways.

Not only is it the case that most, if not all, scientific disciplines are faced by similar challenges, but it is also important to keep in mind that changes important to the structure of the academic enterprise -- and to the assumptions and values that currently drive it -- are occurring broadly in our society. This past Monday (November 18), the front page of the New York Times carried an article worth glancing at. The article is entitled "Publisher's Squeeze Making Tenure Elusive.'' In a direct sense, this problem is hardly relevant to the scientific disciplines; but it is very important to academic disciplines in the social sciences and humanities. Despite the fact that these other disciplines have for a long time confronted problems similar to the ones that we are now beginning to grow alarmed about, and despite the fact that our colleagues in these disciplines feel a certain amount of unavoidable satisfaction over the potential of eroded support for science, these disciplines too are facing new challenges in the changing academic environment.

We all are convinced that astronomy departments constitute an undeniable social good...manifestly deserving of every resource society throws their way because of the spiritual reward that comes from delving into some of the deepest and oldest questions that humans have asked. And the reality is that astronomy, as a discipline, has become accustomed to a lot of resources. Astronomers are fond of citing how important an educational vehicle astronomy is for exciting the imaginations of students and for communicating scientific concepts...indeed, we astronomers frequently treat ourselves to the illusion that astronomy is unique in those respects. I want to tell you from my perspective having responsibility for the University of Arizona's College of Science that your colleagues in every one of the scientific disciplines believe their disciplines to be equally blessed with similar unique qualities. I am fond of an aphorism attributed to Richard Feynman...

The first principle is to not fool yourself, and you are

I think that we have too much tendency to overlook -- even to deny -- the magnitude of the change that is overtaking us, and the pervasiveness with which that change may affect all that we do...including life within our academic departments. The thrust of my remarks is that we are in need of much more substantial response to impending changes than to merely tune our graduate programs at the margins. We need to be concerned with how, as institutions, we adapt to changing circumstances. The issues involved are money, values and culture...within our disciplines and within our academic departments.

In my remarks at the AAS meeting in San Antonio last winter, I likened the scientific community (and particularly the astronomical community) to a person driving an open convertible on a beautifully maintained road, through a gorgeous landscape, and headed directly at a precipice where the road ends...kind of like a Road Runner cartoon. And in this cartoon, the driver of the car has not a care in the world he (or she) has one arm thrown over the seat and is happily proclaiming: "The world is great; everything is fine; look where I have come from; look at all that wonderful road and landscape behind me.''

Perhaps the biggest impediment that our universities confront in responding to changing circumstances is the pervasive denial -- among faculty and administrators -- that real change is actually upon us. The major manifestation of denial in the university research community is the fantasy that we are experiencing only a transient political and fiscal aberration, and that, if we just wait it out, the aberration will pass and "normalcy'' will return. One aspect of the putative "normalcy'' -- the return of which eagerly we await -- is a steady, year-after-year real growth in funding for higher education and for scientific research. Another aspect of the "normalcy'' for which we long is a relaxation of the tense, critical scrutiny that both higher education and research have been subjected to during the past several years. We long for the happy days of past decades, during which higher education was held in unquestioned esteem, world leadership in every area of scientific research was deemed to be of strategic importance and a national objective, and every dollar directed into scientific research programs -- all research, basic and applied, however esoteric -- was considered a dollar somehow well spent. It is not surprising, and probably healthy, that we long for the return of such circumstances. Indeed, scientific research is extraordinarily valuable to modern society, both in the way it drives the technological economy and for the spiritual expansion it provides by investigating deep questions of long-standing human interest. Such happy circumstances may one day return; if we wait long enough, they surely will. However -- given that our nation is feeling such an intense need for financial stringency, given the broadscale disillusionment with institutions that wracks our society, and given the political environment within which spending decisions are being made -- it may be unrealistic to expect a return to such happy circumstances on any temporal horizon useful for today's planning.

Especially since the end of World War II, scientific research has occupied a privileged position of declared strategic national importance. The current generation of scientists has lived its entire life in this environment of expansive support for scientific research. We have integrated into our personalities that this environment is the norm. I am not sure that it is. The past fifty years of post-Manhattan-Project, post-Cold-War-and-Sputnik scientific euphoria may be the anomaly...and we may be on the brink of relaxing back the more stringent and more retrenched norm.

To me, this is something of a paradoxical situation. Science and basic research are fundamental and at the heart of the technological economy in which we now live. One need only look around and ask how much of what we now take for granted on a daily basis and how much of what our economy (from plants growing in silicate soil, to "bits'' flying around silicon chips) is grounded in deep scientific understanding that did not exist as recently as 75 years ago. In my view, the 21st century is likely to see even greater advances in deep scientific technology.

Nonetheless, we are confronted with the possibility of serious retrenchment in scientific research. It is not rational situation, either in terms of the importance of the endeavor or in terms of the amount of money involved.

"Basic'' scientific research is hard to define uniquely, but...

In my view the fact that our society proclaims inability to support this small level of activity in the face of its manifest importance is indicative of non-rational thinking. But the nonrationality makes it harder to deal with the problem...not easier. As scientists, we too frequently dismiss that nonrational as being beneath attention. As citizens in the larger society, frequently it is the nonrational that must occupy us.

I want to describe two personal experiences, one of which is intended to emphasize that we are confronting secular -- or quasi-secular -- change in the research-support environment...and that this change has been gathering steam for a time. The second is intended to raise a flag -- in the spirit of Feynman's aphorism -- about the ability of the scientific community to fool itself.

As many of you know, I have been pretty active in space science policy and program planning for twenty years, from very early in my career...indeed starting in 1976 (so this is an anniversary year for me in that respect). Though I do it far less now, I have spent a great deal of time in Washington in connection with these issues...with nontrivial amounts of interaction both in the Congress and in Executive offices. I want to share a very well-founded perception with you: During the first part of those twenty years, it was an unquestioned article of faith in the Congress that the health and vigor of the U.S. university research enterprise was a matter of strategic national importance. In 1986, I was invited to testify as part of a series of science policy hearings in the House of Representatives Science, Technology and Space Committee, focused on ensuring the continued vigor of our nation's research enterprise in the universities. In the context of those hearings, in 1986, it still was an axiom that the health of the scientific enterprise in universities was of strategic national importance. The main question was to determine what was necessary to ensure that health. That article of underlying faith did not divorce the wrangling over science funding from the usual politics...but it modulated the politics in important ways.

By the early 1990s, to the best of my ability to perceive it in Washington, that notion, while not entirely dead, was extraordinarily diminished and diluted. I was told explicitly by key congressional staff in 1992 that the health of university research, per se, was no longer an explicit and widely accepted priority in the Congress. Events since then have only solidified the impression that times have changed...and that the next several decades are likely to be modulated by forces and perception different from those of the past five decades.

The second personal observation that I want to recount is addressed to what I worry is the level of self-fooling and denial of which the scientific community is capable. We are in a golden age in astronomy. The Hubble Space Telescope, AXAF (or rather its parts), gamma-ray observations, infrared astronomy, new ground-based telescopes and so on. And now we are talking seriously about the "next generation space telescope'' and ambitious interferometer arrays in space to discover and study other planetary systems. Looking at this conceptual agenda, and looking at what seems to be coming down the pike -- at least if the advisory studies and reports can serve as prophecy -- I detect no reason whatever to worry about the future.

In the late 1970s and early 1980s I served on the National Academy's Space Science Board, and chaired its Committee on Planetary and Lunar Exploration. I was first appointed to the committee -- a young assistant professor -- at almost exactly the time -- almost to the month -- of the 1976 Mars Viking landing...right in the midst of the rush of excitement and all of the expansive thinking about the future that the Viking success at Mars engendered.

In the committee we spent the next several years formulating an extraordinarily exciting terrestrial planets initiative...Venus, Earth and Mars, life in the Solar System, catastrophic planetary climate change...including a detailed, staged prescription for cooperating with the Soviet Union that was enthusiastically accepted by the Soviets as the basis for a possibly accelerated agenda of cooperation in planetary exploration. The intensive robotic presence on Mars, specifically planned to begin in 1984 was very exciting...the broader technological benefits of the planned robotics development were understood and laid out. It was a very satisfying, almost a monumentally sound, piece of scientific advice. The report was published by the NRC Press, and duly distributed and enthusiastically received by all the appropriate units of was very widely read and very well known. You can read it...but you have to find a is out of print now.

I do not have enough time to compare those plans, expectations and recommendations with the eventual realities...but, take my word for it, the plans, expectations and recommendations were not isomorphic with the eventualities.

The role of my remarks at the beginning of this conference is to try to provoke your thinking about the problems of astronomy in a broader context...a context that I believe is likely to have a profound affect on the overall academic research environment during the next couple of decades. Confining our attention narrowly within our own academic disciplines, and focusing solely on the details of this year's Federal budget battle (fooling ourselves into believing that what is "postponed'' this year will likely be funded next year) and not raising our eyes a little bit to look at the broader social scene...that narrow focus, I think, is hazardous and shortsighted.

I think that we confront some real changes in the next years to several decades...changes that will challenge the current structure of academic enterprise, both in teaching and research...and especially in what has been the implicit symbiotic social and economic contract between teaching and research.

According to a recently released Government Accounting Office report (hot off the Internet in Robert Park's weekly APS diatribe, so it must be true)...according to that report, while household incomes rose 82% (clearly that is in unadjusted dollars) in the years 1980-1994, tuition at public colleges and universities rose 234% in the same interval. Research expenditures rose -- also in the same interval -- 157%.

Predicting the future is hard -- I would never try to do it -- but you might ask yourself whether you think that trajectory is likely to be continued during the next decade and a half...and, as you contemplate that question, again keep in mind that the most recent specific discussions about the future of research support project a 30% real decline over a half dozen years. That 30% decline will almost certainly not come to pass...but the talk about it helps define the environment, and set the expectations.

I did not intend to come across negatively or pessimistically in these remarks. Indeed, I am very optimistic:

Indeed, I think there are sensible -- even exciting -- responses to the challenges that we confront. I have my own ideas about that, and some of you have heard me speak about those ideas. But my role at the beginning of this conference is not to hold forth on my ideas, but rather to challenge you on this subject, To find the answers, we will need to look within ourselves and examine some of the most basic assumptions that now shape the academic enterprise.

I hope this meeting can be a step in that direction. Thank you very much.

C. Career Trends for New Doctorates in Astronomy and Astrophysics

Karen F. Greif, Charlotte V. Kuh, and James A. Voytuk
Office of Scientific and Engineering Personnel
National Research Council
Washington, DC

When career trend data are reported for new PhD's, it is common to find that data for new doctorates in Astronomy and Astrophysics (henceforth referred to as Astronomy) are grouped with that for Physics. Although many graduate departments offer training in both Physics and Astronomy, it has been difficult to ascertain whether the employment situation for the two groups is the same. It is widely recognized that a crisis in overproduction of Physics PhDs occurred in the late 1970's, and that the persistence of a poor job market has reduced enrollments in graduate programs in Physics. What is not known is whether similar conditions also exist for new doctorates in Astronomy. The data presented in this report summarize trends in degree production and career paths for new doctorates in Astronomy, and demonstrate that significant differences exist between trends for Astronomy and Physics.

Data were derived from two sources collected by the National Research Council for the National Science Foundation (NSF) and other government agencies: the annual Survey of Earned Doctorates (whose records are summarized in the Doctoral Records File (DRF)) and the biennial Survey of Doctoral Recipients (SDR). A third source, the Survey of Graduate Students and Postdoctorals in Science and Engineering (GSESP), is collated by the NSF. All three databases examine different segments of the population, and data are constrained by differing survey methods.

The DRF is statistically the most complete, containing information on every individual earning a doctorate in a given year; however for this study its accuracy is limited by self-reporting of degree fields. For example a student receiving a PhD from a Department of Physics and Astronomy would have the option of selecting one of either Physics or Astrophysics or Astronomy for a field of study. Thus if a student considers his degree to be in physics he will be included in the physics DRF database, but if he selects either Astronomy or Astrophysics, he will be included as an astronomer.

The SDR is a longitudinal study which follows a sample population for an extended period and is the best source for employment trend data, but it is limited by the fact that only 8% of degree recipients in a given discipline are followed. The number of individuals tracked in the field of astronomy is low, and small numbers may compromise the results. However, the sample population in physics is statistically significant, and thus trends seen in both fields are likely to be real. In addition, since the SDR tracks only those who earned the doctorate in the United States, it will miss all PhDs who earned a degree abroad and who join the U.S. workforce after earning the degree.

The GSESP is an annual survey of graduate enrollments and postdoctoral appointments that is completed by academic departments, and includes those postdoctoral appointees with foreign degrees. Postdoctoral appointments outside of academia will not be counted by this survey, and postdoctoral appointees in centers and laboratories not directly connected to a department may not be known to the department providing the data. By piecing together data from these sources, it is possible to construct a picture of educational and employment patterns. We report here on major trends found for Astronomy and Physics for the decade 1985-1995.

[NOTE: For higher quality figures and tables print the PDF version of this report.]

C.1 Statistical Trends in Astronomy and Physics

C.1.1 Declining graduate enrollments after a period of growth

After a boom in graduate enrollments in the 1960's, enrollment plummeted in the 1970's in the physical sciences. Graduate enrollment in both Astronomy and Physics increased in the late 1980's but once again, graduate enrollments are on the decline. Enrollment has been dropping in Physics since 1992, but only began dropping in Astronomy in 1995 (Fig. 1. -2, Table 1). There were approximately 870 graduate students in Astronomy enrolled full-time in 1995, and 11,000 in Physics. The number of graduate students in both Astronomy and Physics declined by about 9% between 1994 and 1995.

C.1.2 PhD production is on the rise

Astronomy PhD degree production dipped during the 1970's and early 1980's, partially in response to reduced federal support for graduate education in the physical sciences but rose steadily since the late 1980's (Fig. 3, Table 2a). 173 degrees were awarded in 1995. After a large dip in the late 1970's and early 1980's, the number of degrees awarded in Physics increased since the mid-1980's, and has recovered to levels similar to those in the early 1970's. 1479 degrees were awarded in 1995 (Fig. 4, Table 2b). The recent declines in graduate enrollment have not yet been felt at the level of degree production, because of the long period of training required prior to degree.

Astronomy PhD production is unequally distributed across institutions and is dominated by a small number of institutions (Fig. 5). Of the 719 degrees awarded from 1991-95, 316 (or 44%) were awarded by 11 institutions. These institutions make up only 10% of the PhD-granting institutions awarding Astronomy degrees. However, of the 105 institutions that awarded at least one Astronomy degree during the period from 1973-1995, 20 did not award a degree from 1986 to 1995, and 26 did not award a degree from 1991 to 1995. There are a number of very small graduate programs that award the doctorate in Astronomy only rarely.

C.1.3 The demographics of new PhDs

Women represent about 16-17% of new doctorates in Astronomy, compared with 10- 12% for Physics (Fig. 6, Tables 2a, 2b for details). Little increase in the representation of women in Astronomy has occurred since the 1980's, although the fraction has varied from year to year. Efforts to encourage women to pursue advanced study in both Astronomy and Physics have apparently met with limited success.

There are frequently-voiced concerns about large influxes of non-US citizens into graduate programs in the US, and fears that these foreign students are depriving US citizens of graduate opportunities. This does not appear to be the case in Astronomy. The percentage of US citizens and permanent residents earning the doctorate is higher in Astronomy than in Physics. About 85% of degrees in Astronomy were earned by US citizens and permanent residents in 1995, as compared to 73% in Physics (Fig. 7). Unlike in Physics, there has not been a noticeable increase in the number of non-US citizens earning degrees in Astronomy. The fraction of recipients of degrees in Physics with temporary visas has risen steadily since the mid-1980's (Fig. 8).

C.1.4 Graduate students are taking longer to earn the PhD

The time to degree has lengthened in both Astronomy and Physics since the 1970's by approximately one year, reflecting a trend observed broadly across the sciences. However, Astronomy PhDs have a slightly shorter median time to degree (from the time of first enrollment in a graduate program to awarding of the PhD) than do Physics PhDs; 7.25 years compared to 7.5 years in 1995 (Fig. 9).

C.1.5 A growing pool of Postdoctoral Fellows

Disturbingly, the fraction of new PhDs in both Astronomy and Physics without definite commitments at the time of degree has increased since the early 1980's (Fig. 10, Fig. 11, Fig. 12, Fig. 13). However, the proportion with definite commitments for postdoctoral appointments at the time their degree was granted is increasing. This proportion is higher in Astronomy than in Physics, where a declining but still significant proportion plan to take permanent positions in industry. About 65% of Astronomy doctorates earning the degree from 1991-95 had definite plans at graduation to assume postdoctoral appointments compared to 38% of new Physics doctorates.

Not surprisingly, the number of individuals holding postdoctoral appointments has risen in both Physics and Astronomy (Fig. 14 - 15, Table 3). Over 260 Astronomy PhDs held postdocs in academia in 1995, double that of 1987. This is approximately 50% of those 1-4 years post-degree. The numbers also increased in Physics, but not as dramatically.

In both Astronomy and Physics, women are represented proportionately to their fraction in the PhD pool: about 20% in Astronomy and 10% in Physics in 1995 (Fig. 16).

In contrast, the fraction of non-US citizens in postdoctoral appointments in Physics rose from 47% in 1987 to close to 56% in 1995, while falling in Astronomy from 40% in 1987 to 33% in 1995 (Fig.17). These numbers do not include individuals holding postdocs in other sectors, such as industry or government labs, but do include individuals who earned their degrees outside of the US. Since the fraction of non-US citizens in the postdoc pool far exceeds their representation in the pool of new PhDs earned in the US, these data suggest that a significant number of foreign-trained individuals come to the US for further training as postdoctoral fellows. Whether they opt to stay in the US following completion of their postdoc is unknown.

Career opportunities for new PhDs in Astronomy are changing from those in the past. However, these changes are not as dramatic as that for new Physics PhDs, where uncertainties in employment impact on a much larger group. Because of the predominance of postdoctoral training for new PhDs in both disciplines, we examined the career paths of individuals 5-8 years post-degree, at a time when it seems reasonable to assume that most should have achieved stable employment. Because the numbers of individuals in a given employment sector are based on weighted estimates from the sample drawn in the SDR, and because the population of astronomers overall is small, changes in a small number of actual respondents may produce apparently large weighted changes from year to year. These variations should be viewed with caution. For detailed breakdowns, see Data Tables 4-8.

C.1.6 Recent growth in tenure-track faculty positions

By far the largest employer of PhD astronomers is academia, and 50-60% of astronomers (not including postdocs) are employed there. This fraction has remained relatively stable over the last 20 years. For Astronomy PhDs 5-8 yrs post-degree, both the number and fraction holding tenure-track positions declined until 1989, but has increased slightly since then (Fig. 18 - 19, Table 4). About one-third of the cohort held tenure-track faculty positions in 1995. As the number of tenure-track faculty positions declined there was a corresponding increase in those holding non-tenure-track positions in academia, and high numbers have persisted. This shift may reflect changes in university hiring patterns that are also seen in other disciplines.

The academic employment for physicists 5-8 years post-PhD is not as robust. Only 25% held tenure-track faculty positions in 1995, and the overall percentage holding positions in academia is somewhat lower (see Table 5).

C.1.7 Changing opportunities outside of Academia

The fraction of Astronomy PhDs 5-8 years post-degree holding positions outside of academia has remained relatively stable at around 35-40% over the past 20 years, although the number holding such positions rose in the 1980's (Fig. 20 - 21, Table 4). The most notable trend was the increase from the early 1980's until 1993 in positions held in federal labs and other government agencies. This growth might have been the result of increases in space- based astronomical research supported by the federal government. However, the number holding such positions dropped sharply in 1995. At the same time, the number of those employed in other positions in science and engineering (S&E) increased markedly. This sector includes, for example, those who are self-employed, work in museums or planetaria, or teach at the junior college or pre-college levels. Employment in industry represents only 10-12% of positions overall; this percentage has not changed markedly over the past 20 years.

In Physics, in contrast, the majority of those employed outside of academia hold positions in industry, although the fraction of those 5-8 years post-degree working in industry has fallen since 1985 (see Table 5). Comparatively fewer Physicists hold positions in federal labs or other employment in S&E. C.1.8 No evidence for growing unemployment

The SDR permits an estimate of the number of individuals who are unemployed and seeking positions, as well as those working outside of S&E. At least some of the latter might be viewed as under-employed, since their training and expertise are not fully utilized. Unemployment for the 5-8 year cohort in Astronomy appears to be very low (<1%), particularly in recent years. However, considerably variability is seen for other cohorts in across the survey period, probably as a result of very low reporting numbers. Similarly, low reporting numbers made it impossible to accurately estimate the population working outside of S&E. Examination of the entire Astronomy workforce, however, also indicates that both unemployment and employment outside of S&E remain low (see Table 8).

Unemployment in Physics, based on SDR data, ranges from 1-5 %, depending on the age cohort examined. In addition, a higher proportion of Physics PhDs report employment outside of S&E, particularly among older cohorts, where levels reach 8% in recent surveys.

C.2 Summary

D. Additional Statistics

R. Milkey, American Astronomical Society

This appendix contains charts with some additional statistical information concerning employment opportunities and methods of financial support for graduate students in astronomy, and a comparison of these with support for physics students.

D.1 The AAS Job Register

Since 1986 the AAS has operated its Job Register, for posting of open positions in astronomical research, education, and technology. While a large fraction of the jobs available in these fields are posted in this service, it is by no means a complete sample of the open positions, and one must be wary of conclusions drawn from an incomplete sampling, especially with regard to the type of positions available. Nevertheless, it is instructive to look at the trends in total job listings as compared to the total PhD production rates (as reported by the NAS).

Figure 22shows the total jobs listed each year since 1986 compared with the number of PhD degrees awarded in US institutions. A more detailed categorization of each job advertised was obtained beginning in 1991, and the number of postings for tenure-track faculty positions and post-doctoral research positions are indicated for those years. In a typical year between one-third and one-half of the total postings have been for post-doc positions. It is important to realize that many of the postings for post-doc positions are repostings of the same position as the incumbent's term runs out (typically two or three years) and the sponsor seeks another person to fill the position. The remaining jobs making up the total are in the areas of research support, management, and student positions.

Since 1991 the number of non-U.S. positions posted in the Job Register has averaged about 20% of the total number, and has been steadily growing. We have no way of knowing whether this indicates an increase in available positions overseas, or whether it reflects an increasing usage of the Job Register by foreign institutions.

The most notable trends from this data are (1) that the number of postdocs plus visiting professor positions advertised each year in the job register are well matched to the number of astronomy PhDs awarded each year, with 200 such jobs corresponding to 172 new PhDs in 1995, and (2) the number of tenure track positions advertised each year is approximately 30% of the number of PhDs awarded.

D.2 Support of Graduate Students

The data reported in this section were gathered by the NSF (National Science Foundation, Selected Data on Graduate Students and Postdoctorates in Science and Engineering: Fall 1995, Supplementary Data Release Number 3: by Source of Major Support) and are publically available on the World Wide Web.

Figures 23 and 24 show the sources of major support for Astronomy and Physics students respectively. It is interesting to note that Astronomy students are more dependent on funding by Federal funds than are Physics students, with Federal funding roughly equaling institutional funding for Astronomy students in recent years. It is also true that number of students funded for graduate study in physics has been declining since 1992, as has the total enrolment of Physics graduate students. The first year with a downturn in the number of supported Astronomy students was 1994, and it is too soon to tell if that indicates a trend.

Figures 25 and 26 show the types of Federal and non-Federal support received by Astronomy students in the years 1986 through 1994. Note that research assistantships dominate the Federal support while teaching assistantships dominate the non-Federal support.