Thinking about Graduate School in Astronomy?
Another View.

Author: Don Barry  (

Written 9 February 1999, revised 18 February 1999


As a former graduate student at GSU, I feel an obligation to provide information to others considering a graduate career in astronomy (at GSU or elsewhere) representing views not likely to be presented by faculty and departments with successful recruitment as their primary agenda. I would like to stress that the views presented here are mine and in no way represent any official views of the GSU astronomy program or any other. However, they have been endorsed and amended by a number of current astronomy graduate students at GSU and at UT/Austin, and therefore represent not just the views of one person, but of a segment of a graduate population. We feel that it is important that we share some of the facts and resources we have come across in our exploration of the current job situation. We invite opposing commentary from interested parties and will gladly provide links in the interest of scholarly dialogue.


There are several primary facts which students entering astronomy graduate programs must consider:
  1. Currently (1999), approximately 190 astronomy Ph.D.s per year are granted by American universities.
  2. Currently (1999), approximately 20 research astronomy faculty positions at American universities become free.
  3. Currently (1999), approximately 30 teaching astronomy faculty positions at American universities become free.
  4. An influx of eastern-bloc senior scientists has diluted hiring in this country for the last decade. It is likely to continue for the next few years.
  5. The average length of time to a Ph.D. in astronomy is 6.5 years nationally. A Ph.D. program in the natural sciences is a major undertaking, requiring complete immersion in the subject matter for a period of years, and usually results, regardless of intent, in a strong identification with the field and the work.
  6. Unemployment for recent Ph.D. graduates is low, but employment outside one's field of interest and identity is high. This percentage increases with time since Ph.D. Whereas employment rates within one's field for most engineering majors are high, even long term (56% of civil engineers are still employed within their field after 20 years), the corresponding rate for astronomy Ph.D.s is closer to 25% (and dropping).
  7. The market for those attempting to find astronomy jobs is brutal. Faculty openings typically receive 200+ applications. This is the size of the pool of those still trying to find permanent positions in the field. Even with no new production of Ph.D.s and support of the current pool with temporary positions, it would take years to consume the surplus.
  8. History has shown the professional bodies (AAS, NSF) cannot be trusted in predicting job demand in the future. As late as 1989 the AAS was predicting huge shortages of astronomy Ph.D.s for the early to late 1990s. Although they now admit the field is tight, they have undertaken no efforts to restrict or discourage production of new Ph.D.s. (An excellent investigation of NSF studies touting future shortages that were prepared by political appointees who bypassed NSF's own actuarial divisions has been prepared by MIT mathematician Eric Weinstein using documents released after Freedom of Information Act inquiries).
  9. As many as half of graduate students cannot find postdoctoral work after graduation. Fewer than half of postdocs find academic work after completion of their (often multiple) appointments. Postdoc salaries average in the mid 20K range. Starting salaries for faculty astronomers at research universities are typically in the upper 30K range, with full professors making typically 70K in 9-month salary (they can boost this by 33% with summer salary paid from grant support).
  10. Few "industry" jobs exist for astronomers doing astronomy.
  11. Departments do not do their duty by hiring numbers of postdocs necessary to sustain the field in steady state. Graduate students are cheaper. The GSU astronomy program carries 19 graduate students at the beginning of 1999 (source: GSU Astronomy Graduate Information Page) but only supports one three-year postdoc.
  12. Astronomy is a field in a period of fiscal contraction and stagnant hiring. National budgets have fallen by 30% in constant dollars for ground-based astronomy during the past 15 years.

GSU's Alumni History: A Countervailing View

The GSU Graduate Student Information page analysis of post-graduate career paths paints a fairly rosy picture in its list of employed, established graduates. But a few structural adjustments are in order to the claims.

The GSU program, first off, has a larger than average graduate student to faculty member ratio. Seven graduate faculty currently supervise 19 graduate students (a number of whom are part time), for a ratio of 2.7. The ratios for two other, representative programs are UT/Austin's (10th ranked) at 1.75 (35 vs. 20) and 1.19 (25 vs. 20) for U/Toronto. At current rates, this is what these figures imply: for a faculty career of 25 years, each GSU graduate faculty member will produce approximately 10.3 Ph.D.s, of which one will find a job in the stable marketplace. For every 8 Ph.D.s that graduate (48 person years), GSU will support one three-year postdoc.

The career paths of 15 GSU Ph.D. graduates are tracked on their graduate info page. However, of the 5 graduates who are now in permanent academic jobs, 3 were foreign students who had been guaranteed professorships in their home countries upon successful completion of their degrees. Discounting this cohort, only 3 of the remaining 12 have achieved permanent positions. Depending on definition, 5 of the 12 can be said to have left the field (none have more than 8-years post-Ph.D. under their belt). Depressing as this is, it is, in fact, a somewhat better performance than many departments can claim, and the faculty is justifiably proud of this record.

GSU prospective students are assured that their teaching responsibilities will generally end by their third year, freeing them to devote full time to research and dissertation work. Yet these same promises were delivered to students in the late 80s and 90s and went unfilled. A number of current students still teach well into their graduate programs. In addition, graduate students fulfill a number of expected support roles in the department including maintaining the computer systems and network and performing maintenance at the research observatory.

The "training in pedagogy" received by graduate students is more or less pro-forma. GSU does not permit graduate students to entirely supervise classes. Yet teaching academic jobs now usually request teaching "beyond lab supervision" in their descriptions. Students will need to look elsewhere for this experience. Graduate students should be warned that outside adjunct teaching appointments demand a great deal of work and pay very little (they have a captive audience of people who must gain teaching experience to even compete in the tight market). A full semester of teaching introductory astronomy at Austin Community College, for example, pays the instructor $1500.

Prospective students should be informed of the community's ranking of their intended programs before committing to them. Rankings, despite their somewhat arbitrary nature, do reflect the perception of graduates and their parent programs in the community. The 1993 National Research Council ranking of Ph.D. granting astronomy programs placed GSU's 33rd: the bottommost listed entry (several schools have programs which were not ranked). Although I am convinced the program is somewhat better than this, it is unrealistic to compare the program through rose-colored glasses to top-10 or even top-20 programs. The physics department is unranked by the NRC in 1993 (having recently received the right to grant the Ph.D. at that time). However, most would agree it is a program not in competition with Georgia Tech's, which is ranked 61st, or even UGA's (this is now debatable), which is ranked 75th.

On the positive side, it is clear that the GSU has a program on the rise, with an active, ambitious faculty and a number of interesting projects underway. A fairer ranking of the program would be in the mid 20s at present, with the possibility of a top-20 assignment within the next few years, particularly in the unlikely event that the CHARA Array succeeds.

Whither the Ph.D.?

Let us reiterate: unemployment rates are low for natural science Ph.D.s, though unemployment within one's area of specialty and identification is very high. Yet these aspects are seldom discussed with bright-eyed students contemplating a graduate education. The Scientist in a recent issue quotes Arthur Sowers, a membrane biologist at the University of Maryland, Baltimore, who reports, "The nasty facts of life are really not discussed very often, People in the know and institutions in the know are not openly disclosing up front the realities of what people are getting into. I certainly have never been pulled into an office and told all the bad things that can happen." Sowers further states, "Young people have to somehow get the idea that science is a field that's in contraction. I think they have to first learn that there are better pastures someplace else. I think the failure rate in achieving career goals is just far higher than worth the effort to find out if you're going to succeed."

Why this doom and gloom? The fact is that Ph.D. programs require typically 6-7 years of concentrated effort, pay apprenticeship wages, and offer no prospects of employment with one's, by then, career identification. It would not be so bad if, say, those years were filled with a reasonable wage so that one could begin a family or a life (like most other contingent jobs in this era of "flexibility" in hiring); or if the apprenticeship period was followed by a relative guarantee of success in the field (as medical doctors and lawyers enjoy). Some faculty criticize student expectations, and would have, as Sandra Gilbert has stated, students be grateful for being permitted to participate in intensive study which "animates their soul." They encourage students to think of a Ph.D. education as being rather like a stint in the Peace Corps. "Make the world a better place, and then get on with the rest of your life."

This would not be so bad for terminal M.S. (two-year) programs, which probably could be expanded to fulfill some of the needs of the academy and industry. But having dramatically lopsided numbers of Ph.D. graduates with clear career expectations versus few postdocs and fewer faculty show that academic planning is not ruled by consideration of a healthy discipline which looks out for its members, but by simple economic greed for cheap labor.

The Reality of the Job Search

Research faculty positions routinely receive 200+ applications. Yet only 20 or so of these open per given year. Over 120 students now receive Ph.D. degrees annually in this country. In addition, the last decade has (as will likely the next) seen an exodus of top-quality, senior researchers from eastern-bloc countries vying for all positions here. The numbers are clear: it is impossible for more than 20% of currently graduating Ph.D.s to find long-term employment in their field, and much of that will go to graduates of the "prestige" institutions. More than half will eventually leave their field entirely, the others will find teaching positions, or a succession of low-paid, low-security adjunct or postdoctoral roles. Some have been playing this soft-money game for 25 years or more.

The nature of the field changes qualitatively with this oversupply. Candidates who would once have transitioned directly from graduate school to secure faculty appointments now find that they must endure a number of postdoctoral relocations, assuming they can find them. Tenure itself is now harder to obtain, often requiring 7 years instead of 5 or 3. Job security may not come until one reaches the mid 40s.

Hiring changes as well. It is impossible to quantitatively make reliable decisions out of a hiring pool of 200+. Any conceivable reason to disqualify candidates must be found. The ones that endure tend to be careful, incremental researchers who have planned several lines of inquiry guaranteed to produce a number of research papers in the short term. They have made no "career errors." Speculative or longer-term research, especially in the younger creative and productive years, suffers. Any glance at the typical qualitative nature of papers in the major journals from the 1960s and today will reveal this shift. Overproduction breeds scientific conservatism and scientific stagnation.

Exits to Industry

Astronomers, with their training, find ample employment in high technology areas. It is not that astronomers are unemployable, it is merely that they have great difficulty finding employment in astronomy. Yet graduate astronomy programs simultaneously preach the versatility of astronomy graduates and yet practice a narrow focus into supporting the needs of the individual research group. One GSU faculty member told me at the 1999 Winter AAS meeting that they needed to find another graduate student to work on an instrumentation project in support of the department. Yet this attitude precisely states the problem: graduate students are seen as the means to an end (fulfilling department needs) rather than an end of themselves (adding peers to a community). They represent primarily cheap labor, and also in some cases the hope for an "intellectual heir" for an advisor. But they are a disposable commodity, several times cheaper than postdocs or especially faculty members. Larger graduate groups enhance the reputation of an individual department, grow the prestige of a single research group. But when everyone practices this policy, the result is disastrous for the field. It is the crisis of the commons.

Students exit the field in a variety of ways. Some simply cut short their graduate programs with M.S. degrees. Many are unable to find an initial postdoc after the Ph.D. and are forced to immediately leave astronomy. Others seek permanent positions for long periods, moving every few years, living on the lower salary of a contingent or temporary employee. It is not unknown to have astronomers search for the cherished faculty position into their 40s (and sometimes 50s) before changing direction. Yet most postdoctoral salaries are below 30K (and are easily filled at this). Industry salaries are almost twice this for similar training. Clearly an attachment to research is cultivated and bred with great strength in the course of a graduate career.

What Cost a Graduate Career?

Faculty members, especially those facing tenure, are often poor sources for perspective on the job issue. They, after all, represent those who have found permanent positions, so naturally their view tends to be less negative. In addition, the prospect of cheap research labor, the offloading of some teaching responsibilities, and the potential to build large, prestigious reseach teams often blinds faculty to the long-term damage their unrealistic visions create. This is manifested in sometimes astounding statements of dissociation from academic reality. Chris Sneden, full professor (and current department chair) at UT/Austin, once stated in a student discussion (1997), "There are jobs for each and every good astronomer. We take care of our students." (when the Texas state legislature mandated universities cut support to graduate students after 5 years of Ph.D. work, his response to a student inquiring what the department would do in response was, "What are you, a bunch of wimps? You can't finish in five years?") Hal McAlister, Regents' professor at GSU, touched on the exploitive angle (1992) in a lunchtime conversation at which several students were present, saying, "The secret to the success of the American economy is the cheapness of graduate student labor." (At a 2004 salary of $152,400, he is GSU's highest-paid astronomer. [source, Georgia Department of Audits and Accounts])

The American Astronomical Society is positively mendacious in their misrepresentation of the job issue. Their current Career Education Page (intended to lure people into the field) has the audacity to misrepresent the current case as follows:

On the other hand, there is a small turnover of positions each year and, therefore, strong competition for positions. In recent years, there have been about 150 job openings for astronomers in North America, while the number of Ph.D.s conferred annually in recent years has averaged about 125. It is common for astronomers to spend from three to six years in postdoctoral positions before finding a steady position in a university department, national facility, or government lab.
But most of these 150 positions are temporary, and attract competition not only from the new Ph.D.s but from recycled temps of the previous two decades. Their implied "apprenticeship period in a postdoc" before full entry into the field applies to only a distinct minority of Ph.D. graduates. They circularly argue in this brochure that 100% of astronomers work in astronomy, defining, of course, an astronomer as one who practices astronomy. They do not mention that most people are squeezed out of the field after a succession of increasingly frustrated job searches. Only when the number of permanent jobs created annually equals the Ph.D. production rate will equilibrium be obtained. Not quite ready to sing "Happy Days are Here Again," the AAS leadership contents itself with offstage renditions of "Keep Your Sunny Side Up!"

But at what cost is this empire building constructed? Wishful thinking, writer and academic Cary Nelson claims, is often academia's major contribution to the public sphere. (Manifesto of a Tenured Radical, NYU Press, 1997. He also produced the delightful metaphor at the end of the previous paragraph.) Graduate students, he says, "acquire an identity they do not have at the outset. These commitments can become very deep. We are, after all, talking about six to eight years of teaching and research while working for the Ph.D. ... Then suddenly, in their thirties, they are cut off from a field they have inhabited for a decade or more. And they have to invent an alternative future they have not even imagined, let alone one for which they have trained." Yet departments and faculties often feel their mild disclaimers exculpate them from responsibility. "As for real recruitment brochures, suffice it to say that it is disingenuous and dishonorable to claim that warnings to prospective graduate students about the job market are sufficient, that such warnings take faculty and institutions morally off the hook. The undergraduate senior can easily dismiss warnings about a career to which he or she is still quite uncommitted. The commitment comes later, the career seduction comes later, identity formation comes later. It is with a certain corrupt relief that faculty and administrators note they can post job warnings and still lure applicants into the labor pool ... The package is appealing, its eventual psychological cost often at once large and unimaginable." He concludes, "...when apprenticeship leads to no future it becomes not only unethical but pathological. Apprenticeship with no future is servitude."

A Personal Statement

I began graduate school in 1988 amidst predictions of a dire shortage of Ph.D.s in astronomy to be manifest in the mid and late 1990s. This has not only failed to emerge, but the job situation is even more desperate than it was in the late 70s and 80s. The American Astronomical Society has taken no substantive measures to rein in the size of graduate student populations, despite the ominous signs of stagnant funding and the astoundingly clear evidence of the crisis. In 1989, I even picked up an AAS-commissioned poster at the AAS winter meeting (with the crisis already at full steam) headlined, "Wanted: Astronomers."

Despite some doses of the usual petty academic politics, I found graduate school a congenial place, the work fascinating, the colleagues stimulating and generally friendly. I quickly bred a strong attachment to the field, which continues to this day. Though I am currently between positions, I consider my own prospects better than that of most in the field, and have high hopes for my own future.

Yet if I were to make the decision to enter graduate school today under the same circumstances as I did in 1988 (but with more complete information), I would have refused. At the very least, I would have demanded of departments that they provide a substantially higher stipend (at the 25K+ level) to compensate me for the work I would perform in the field with little prospect for being further ahead in a career at the end. I would have also involved myself less in support roles (according to the pay), or demanded in advance a guaranteed quid pro quo for the work not directly toward my dissertation. (In fact, I did this, requiring a principal investigator to sign a document assuring me a 20% observing share on an instrument which I built--a commitment which I have not yet tested, and which may very well be fought bitterly by the department in its realization.) As it was, I myself, like many of my cohort, were enablers for a number of dysfunctional expectations of my department. But this is universal in the field. I do not criticize GSU uniquely for this, but only point it out as an example of that which a potential student should beware.

Links to other resources

Don Barry
Last modified: Thu Feb 18 17:26:50 1999