The Status of Women in Astronomy

By Meg Urry

June 2000

Meg Urry heads the Science Program Selection Office at STScI, and does research on active galaxies, notably multiwavelength studies of blazars. She was the chief organizer of the 1992 STScI conference on Women in Astronomy which led to the Baltimore Charter, is currently a member of the Committees
on the Status of Women in Astronomy (AAS) and in Physics (of the American Physical Society), and has been co-editor of STATUS since 1998.

Surveying the Playing Field

ASTRONOMY is a highly competitive profession, and to succeed requires brains, dedication, energy, imagination, and luck. It is hard for almost everyone to get a faculty job, to get tenure, to rise to leadership
roles in the profession. Is it harder for women? Or easier? The answer lies not in anecdotes —
which abound to support either view — but in an objective assessment of the data. If women are
being given an unfair advantage, we should see that they are being hired in greater numbers than
their percentage in the talent pool. If vice-versa, perhaps greater measures are needed to ensure
their fair access to the profession.

Accordingly, we looked at how many astronomers are women, and how this changes
with professional level. Because astronomy is a relatively small profession, it is usually combined
with physics (which is 10 times bigger) for statistical purposes. The only available statistics for
astronomy alone, spanning graduate school through the full professor level, come from the
following three surveys: the 1992 and 1999 STScI surveys of four observatories and 32 universities
with astronomy graduate programs (100% response, ~1,300 Ph.D. astronomers),
and the 1999 AAS survey of ~300 institutions (~60% response, ~1,600 Ph.D. astronomers).
The STScI surveys were done by Ethan Schreier in 1992 (published in the Proceedings of the
Conference on Women in Astronomy, online at www.stsci.edu/stsci/meetings/WiA/) and by this
author last year. The AAS survey was initiated in early 1999 by consensus of the Chairs of the
Committees on Employment, Education, Women, and Minorities, and was carried out by
Kevin Marvel and AAS Executive Office staff; Brett Blacker (STScI) and I analyzed the results
(BAAS 31, 1552, #121.01).

Evaluating the Survey Data

The raw data are shown as bar graphs in Figures 1-3. The picture that emerges from
these surveys is a consistent one: roughly 1/4 of graduate students are women, ~1/6 of the post
docs, assistant, and associate professors, and only ~5% of the full professors are women.
Most men in astronomy are full professors (65%), compared to only 1/3 of the women (the
largest group of women are postdocs). Interestingly, the percentages of women are
slightly higher in the STScI sample, which includes the four observatories not in the AAS
data base (STScI, CfA, NOAO, NRAO) and institutions that are generally the largest and
most prominent.

The AAS survey will be repeated regularly, so there will be new data to show how these numbers
evolve. Ideally, the bulge of women at the young end of the profession will propagate smoothly up
the hierarchy. That is, 25% of the Ph.D.s will go to women, 25% of the new postdoc hires (now) will
be women, 25% of assistant professors hired (in a few years) will be women, and so on. This would
indicate a gender neutral system. In the meantime, we can use the present, somewhat limited, data to
assess the current situation.

In fact, there are disturbing signs that the advancement of women lags behind that of men.
The clearest disparity occurs at the entry level — the transition from graduate school to postdoc.
Statistically, 43% (+/– 2%) of the men in graduate school can expect to obtain postdocs, while
only 26% (+/– 3%) of the women will. (This discrepancy is significant at the >5 sigma level.)
These percentages follow from a comparison of the numbers of men and women in graduate
school and postdoc positions at the surveyed institutions, under the assumption that the gender
compositions of those groups change little over the time scale for transition from one to the next.

At later transitions, the statistics are too poor to distinguish between the advancement rates for
women and men astronomers; the raw numbers for women are still lower but only at the ~1
sigma level. (This is a “Catch-22” situation: there are few enough women astronomers that the
error bars are large, thus it is hard to establish with high statistical significance that women are
falling behind.) For combined physics and astronomy, where the statistics are more robust, the
progress of women lags behind at all levels. Women are less likely to be hired, are less likely
to be given tenure, and spend longer at lower levels (e.g., as associate professors) than their
male colleagues. (See articles by Gerhard Sonnert and Virginia Valian, STATUS, January 1999, and
references therein.)

It is noteworthy that this lesser progress for women occurs during a veritable explosion in
national astronomy faculty. Between the two STScI surveys, in 1992 and 1999, the number of
professional astronomers increased by 1/3, assistant professor positions increased by more than
50%, associate professor positions by nearly as much (43%), full professors by 1/4, and postdocs
by 1/5. It is still a tough job market for new Ph.D.s, certainly, but it is much better than it
would be in a steady-state situation. If women fall behind even now, when and how can we
expect to attain the gender neutral state?

The Statistics of Invited Speakers: Rough Parity?

Several other statistics are important, if more specific, indicators of the status of women
in astronomy. The percentage of women invited to speak at meetings is one measure of the gender
neutrality of the field. If women are invited in the proportion appropriate to the particular
sub-field (at a seniority level comparable to the male invited speakers), then one would conclude
no gender bias is present, at least on average. The speaker-invitation process also has an
important feedback effect: evaluation of astronomers for hiring, promotion, tenure, or
prizes usually includes an assessment of the frequency with which the candidate is invited to
give talks at major meetings. Thus underrepresentation would not only indicate unfairness, it
could help perpetuate it.

A random survey of about 25 topical astronomy meetings (submitted to this author, roughly
equally, by people outraged at the exclusion of women and by others demonstrating how effectively
women are included) shows that 9% (+/–2%) of invited talks were given by women,
87% (+/–7%) by men, and 4% by people from whose names gender could not be determined.
This is roughly consistent with the percentage of women Ph.D.s over all astronomy, and so is gender-
neutral, at least in an average sense. That meeting rosters so often anger women and make
them feel excluded may simply be because the numbers are very low — there are still very few
women in astronomy.

However, there may be more to the story: this author noted a number of rosters that lacked
any women, despite many who have contributed extensively to the particular sub-field. Obviously
other rosters must have over-represented women, for the average to end up close to the national
average. We would have to evaluate the second moment of the overall distribution to quantify
whether this perceived bifurcation is actually non-Gaussian. In the meantime we can conclude
that, if meeting organizers make a conscious effort, they should be able to achieve
the appropriate 10% representation of women (up to 25% if there are many young speakers).

Inequality in Honors and Prizes

Another statistic is the percentage of women given prizes or high honors. One example is the percentage
of women elected to the National Academy of Sciences. In the physical sciences, about 5% of
the new members elected over the last 15 years are women, and this is also the percentage in astronomy at present (4 women of 78 astronomy members). This is comparable to the percentage of women full professors across all of astronomy but lower than the percentage of women full professors at the dominant astronomical institutions (8%).

We can also ask what percentage of AAS prizes in the last decade went to women (see table, page 4). Of 96 science prizes, seven went to women (or 7% +/– 2%). The Warner, Pierce, and Urey prizes, by design, go to young astronomers; of the 29 recipients, five were women, less than, but comparable to, the percentage of women postdocs averaged over the past decade. Excluding the planetary award, however, only two of 19 (11% +/– 8%) were given to women, while 17-20% of the postdocs over this period were women. For the more senior science prizes over the past decade (including division prizes), two of 67 were given to women, whereas, based on the percentage of women full professors at top universities and observatories, five to six would be expected. Perhaps most striking, none of the 16 intermediate-age prize winners (Heinemann and Tinsley) have been women; given the ~10-14% women in associate professor positions, the average expectation is about two (13% probability that the absence of women would happen by chance).

The expectation is that roughly 12 of the 96 science awards “should” have gone to women if
there were no dependence on gender. This comes from assuming 20% of the young winners (5.8),
12% of the intermediateage winners (1.9), and 8% of the senior winners (4.1) should be women if
drawn randomly (with respect to gender) from the appropriate age pool in the past decade. The
probability of seven women winning awards when the expectation is 11.8 is only 4%. For service
and education or public outreach, the percentage going to women (11% +/–7%) is slightly higher
than for science and still below the gender-free expectation (though with large uncertainty, and
here the probability of this happening by chance, independent of gender, is 27%).

In summary, women have been winning AAS prizes at a rate significantly below their percentage
in the pool of candidates. Certainly women are not winning a disproportionately high share of
awards — as is sometimes the claim — with the possible exception of young planetary astronomers (a
20% random probability to have gotten three, rather than two, of the 10 awards).

Conclusions

The bottom line is that there are still very few women in astronomy, particularly at the senior levels
of the hierarchy. The data show clearly that the relatively large numbers of women astronomers at entry levels are not achieving the same success as their male peers. Although at least 10% of the Ph.D.s in astronomy have been awarded to women for more than 100 years — and for the last 20 years, the number has been closer to 20% — the number of women full professors of astronomy is still well below 10%. Women astronomers are not making it to the full professor level at the same rate as their male peers, nor to the National Academy, nor are they receiving a fair share of AAS prizes. And this lack of equal progress is happening right now.

This article describes the objective situation of women in astronomy. It does not speak to individual
cases — to the hiring of this or that person, to the awarding of a particular prize in a particular year, to invitations to speak at particular conferences — usually the statistics are too limited in any one instance (e.g., 1 +/–1!) and there are always rational reasons for whatever actions occur. But overall, the data show women doing less well than men in astronomy, most obviously at the first-postdoc stage. At higher levels, the statistics in astronomy alone are too sparse to say, but in astronomy plus physics, the differential attrition continues. We can at least dispel the myth that women astronomers are being hired and promoted and rewarded in preference to men — it simply is not happening. Or rather, if it is, there has to be a “cosmic conspiracy” such that as many women are being discriminated against as are being given preferences.

This unequal situation persists despite the fact that most institutions have affirmative action
plans, the intent of which is to identify qualified women and minorities in hiring situations and to
make sure they are considered fully. Some universities that feel particularly behind the curve have
targeted searches for women and/or minorities, often competing them across several departments.
Some view this as reverse discrimination, making it harder for a young man to succeed
than a young women. However, the data clearly falsify this perception, at least in a global sense.

Some may ask, what is the reason for this gender difference? (Sometimes the implication of this question is that, if we cannot identify the cause, it is not a real effect, or at least, nothing can be done about it.) Some may conclude that women are less able, although there is certainly no objective evidence supporting this notion. Indeed, many women (and men) perceive just the opposite, that women need to be better to succeed. Another possibility is that (as an NSF program director once suggested to me) women choose preferentially not to advance in the profession. Or there may be subtle barriers, the socalled “micro-disadvantages” that Virginia Valian talks about (STATUS June 1999). We can see that overt discrimination has almost vanished.
Faculty search committees today rarely discuss gender explicitly, and never to exclude women candidates. Few of us consider ourselves prejudiced, and few would advocate the promotion of men above women simply because of gender.

There are probably many reasons for the dependence of success on gender, different ones applying in different places and at different times. To “fix” the situation may require diverse small actions, many of which will improve the situation for all astronomers, not just women. But make no mistake: we do not now have a perfect system, we are failing to capitalize on the talents of women who have demonstrated strong interest in our field by pursuing advanced degrees, and we are not attracting and retaining and fostering success among the best minds in astronomy.

Back to June 2000 Contents

Back to STATUS Table of Contents