The Ongoing Demographic Shift in the AAS
by Kevin Marvel
Kevin Marvel is the Executive Officer of the American Astronomical Society. Below is a transcript of his presentation to the 2003 Women in Astronomy conference that has been updated with 2008 statistics and an addendum.
AAS membership statistics indicate that a demographic shift is underway in our field. This demographic shift—if it continues—will result in gender parity in astronomy sometime in the next 30 years, likely sooner than later.
In 1973, the AAS undertook an initial survey of its membership. This survey was first called for by the Working Group on the Status of Women in Astronomy, which was founded in August of 1972. The survey had a sample size of 2,800 members and attained a 27 percent response rate. This was not a fantastic response, but looking at the survey, which was quite long, it is amazing we got this high of a response. It also appears that enough people responded that the survey does represent a statistically meaningful sample of the membership at that time.
Overall, the female membership was only 8 percent and there was some evidence for an increasing fraction of women in our youngest members. Furthermore, there were some disparities in women prizewinners and the number of women who served in the AAS leadership. Fewer women were being elected than what you might expect based on their membership fraction.
I have taken the statistics from this 1973 survey of our membership (which includes full, associate and junior members) as they were presented, and Figure 1A shows the men and the women. It shows five-year age bins horizontally and the percentage in that age band of the membership vertically. Note that the bin label represents the lowest age included in the bin. So you can see that we had quite a high percentage of senior women members and that the percentage of women was more or less flat across the whole society.
In Figure 2A, I show the percentage of the given gender at that age bin. There is certainly a youth peak. At any given age that you might pick, the women actually sometimes represented a higher percentage in that age bracket. For example, at age 33 there were about 30 percent women, so there was a marginal youth peak. The women members were young.
There is another complete survey that was done in 1990, and it was expensive, even in 1990 dollars (in Washington, D.C., we always scale our dollars). It cost about $25,000 to do this survey. The response rate was very good - about 42 percent. The overall female membership had increased percentage-wise to about 12.6 percent just between 1973 and 1990, perhaps because of the active work of the Committee on the Status of Women in Astronomy. There were more young female members than male, and the peak of the age distribution for women was about 30 and about 45 for men (see Figure 1B).
The total number of members at this time was roughly 4,000 to 4,400. Looking at this figure, you can see that in the youngest age bracket in 1990, roughly 30 percent of our membership in the society was women.
You can also see that for our middle-aged members, the flat percentage of women members is simply propagating forward. We do not gain many middle-aged or older members (this is true for both genders). What this indicates is that our membership is increasingly female, as a society we’re changing.
Figure 2B shows the percentage of the gender at a given age from the 1990 survey, and you can see for the younger age ranges that a larger fraction of women, are younger versus older. Most of our women members are younger.
There was another survey done in 1995. It was actually not a full membership survey but a partial membership survey. About 1,000 members were polled. This was done basically to save money in doing these sorts of surveys (they are expensive!) and to get good results.
Again, the overall percentage of women members had increased to about 16.5 percent (Figure 1C). And there was some evidence for episodic growth. I didn’t think the evidence was so strong, but it was mentioned in the report summary. At that time, Peter Boyce, who worked hard on surveys of this type, thought that it would be worthwhile to investigate the potential causes of what caused this episodic growth.
The percentage in the youngest age bracket had grown to about 40 percent and again, the earlier age bracket with 30 percent women, had moved forward in the age chart (we all get older!).
One comparison that I didn’t make, because the samples were done in different ways and cross-survey comparisons done incorrectly will lead to misleading conclusions, is to compare what’s happened to this cohort as it propagates through - has it grown or shrunk? That’s a difficult question and would require a different kind of survey to determine.
Looking at the percentage of the gender at a given age (Figure 2C) shows again an age peak in the youngest age brackets. At some level, this figure (and the others like it) are convolutions of a whole bunch of issues that affect astronomers, both men and women, as they enter astronomy as a profession and continue the aging process, some coming, some going, some, unfortunately, passing away’ also it is important to remember that figures like this represent many different things, not just the issue we are trying to understand, how women fare in our profession.
Figure 1. The age and gender distribution of AAS members as obtained from surveys in (A) 1973; (B) 1990; (C) 1995; (D) 2003; (E) 2008. Bin labels represent the lowest age in the bin.
Figure 3 is an interesting plot. This is a cumulative age band sample, so for everyone under 30, what fraction are women in the AAS? In 1995, it was about 30 percent. For those 40 or younger, it was about 24 percent, and so on. So, the number for the left-most bin more or less represents the fraction of women in the society as a whole. This is a little bit of a cheat, because it is a cumulative statistic, but, in fact, this parabola just gets deeper, if you actually did it for just those age bands. This figure conclusively shows that our Society has changed demographically in a very substantial way in the last 20 plus years.
On June 23, 2003 I took a snapshot of the AAS membership database and did a summation similar to the 1973, 1990 and 1995 data. I included every single membership type except publisher affiliates. That includes the very few lifetime members we have, the emeritus members, the full, associate and junior members, all of our divisions and all of our division affiliates. I lumped them all together. For a few years we have been collecting the birth date as an optional piece of information - a very helpful piece of information for producing plots of this type - and I would like to thank all of you, who have included your birth date in our database. It’s very worthwhile to do that, and it helps us produce graphs like the one I am going to show you. The birth date is optional, but I have a total of about 5,900 members that gave us that information out of 6,480 in total. Gender is also optional. We basically only require an individual to provide us with enough information to send them their membership materials. However, we strongly encourage everyone to provide us with as much information as possible. In a year or so, we will begin online updating of our membership database and each member can update or complete their records. We will have a big campaign to encourage this when the online tools become available.
Figure 1D shows the current situation. Fully 59.6 percent of our members from 18 to 23 year olds are women [at this point applause broke out in the lecture hall]. You can see in the next bin up that the fraction has also improved. We have more women from age 23 to 28 than in the previous 1995 data.
One thing to keep in mind is that there are only a few hundred people in the lowest age bracket, but, in my opinion, these represent people who will likely continue in astronomy as they have committed to our Society so early. Of course, they could just be joining to take advantage of our very favorable junior membership rate and the delight of their own electronic journal subscription. I’ll let my readers be the judge.
Figure 2D shows the percentage of the given gender at a given age, so of all the women in the society, something like 18 to 20 percent, are younger than 25. I should have mentioned it earlier, but the reason for the odd numbers for the bins is that the youngest age bin is meant to represent undergraduate students, the next graduate students, the next postdocs or early faculty and so on.
Another fact to keep in mind is that the various cohorts match up fairly well, meaning that as a cohort ages, the same fraction is retained later on, barring the small fluctuations from varying sample sizes. It is hard to use our membership data to gauge how many women might have left the field, mainly due to the varying samples. A full longitudinal study would have generated this data. I am happy to report that the AAS Council has funded just such a survey, which is being carried out by the AIP with input and help from our CSWA, CSMA and Employment Committee.
Figure 3 shows the same cumulative distribution plot as 1995 with the 2003 data included. The parabola seen in the earlier figure just gets bigger. So overall, as a society, we’re running just slightly under 20 percent women. As you go down in age, you can see that for members under the age of 35 there were about 35 percent women. This is a nice figure to remember.
The past AAS surveys have in large part come about based on membership needs and members stepping forward to help the society; both to organize it and to actually carry it forward.
Rachel Ivie (Statistical Research Center of the American Institute of Physics) reports on the status of a new jointly-sponsored project by the AAS and AIP, A Longitudinal Study of Astronomy Graduate Students:
Data collection was recently completed for the first phase of the AAS/AIP longitudinal survey of astronomy graduate students. The project, which began in early 2007, was the result of recommendations made at the 2003 Women in Astronomy Conference. Eventually, the study will track astronomy graduate students over the course of several years. The study has several purposes: to collect data on people who obtain graduate degrees in astronomy, to compare attrition rates for men and women, to collect data on people who leave the field of astronomy, and to collect data on astronomers who work outside the traditional employment sectors of academe and the observatories.
With support from AAS, the Statistical Research Center (SRC) of AIP is collecting the data. A working group composed of AAS members and an SRC staff member developed the questionnaire. Then, a list of current astronomy graduate students was compiled from various sources, including AAS junior members and lists provided by astronomy and physics departments to the SRC. The graduate students were contacted over a period of several months by e-mail and by postal mail. The SRC received about 1500 responses, and more than 800 of these volunteered to participate in future data collection efforts. Once preliminary results are available, the working group will seek additional support for the next round of data collection, which should begin in 2009 or 2010.
The Statistical Research Center of the AIP posts statistics and reports relating to physics and astronomy education, as well as of the profession at http://www.aip.org/statistics/
Figure 2. The percentage of a given gender as a function of age as obtained in the (A)1973; (B) 1990; (C) 1995; (D) 2003. Bin labels represent the lowest age in the bin.
Rachel mentioned her group at AIP has 12 staff. The AAS staff has 16 people in total to handle all the meetings and everything else we do, so in actuality we have only a fraction of a given individual to do any surveys at any given time (and that individual happens to be me). So, if anyone has an idea of doing a survey, we’re quite happy to listen to you and encourage you, but we would like to try and get help when we do these things.
So, I want to just take a little bit of a break here and share with you some comments we received from the individuals actually taking the 1990 survey to let you see the varied nature and attitudes of our membership.
These are the positive comments (about 50% of responses were similar in nature to these):
- “Please make it happen every five years.”
- “I’m happy that my dues are being used in such a concrete way.”
- “This questionnaire is the best I’ve seen.”
- “I usually hate questionnaires - this one is sensible.”
- “Reading this questionnaire has raised my level of awareness.”
The other 50 percent look like this:
- “How much did this questionnaire cost?”
- “Disappointed in this questionnaire, which seems to ignore all critical issues in astronomy.”
- “It’s a waste of time and money.”
- “Whose dues are paying for this questionnaire?”
- “Don’t waste funds on surveys.”
- “Ego expense is not a proper use of AAS funds.”
So, what has the AAS done in response to both the needs of the community and also in response to requests from the Committee on the Status of Women and so on? We’ve tried to implement at some level the report recommendations, which involve the creation of a list of women astronomers that could be used for prize committees, to find people who would be qualified to win prizes or to be nominated to various places. This service is in place and functioning, thanks to a volunteer effort and encouragement of women at the undergraduate and graduate levels. The Women In Astronomy meeting, for example, is along that level; employment
concerns and report dissemination was basically aimed at trying to make sure that women had a chance in jobs that they deserved and also to try and spread this report widely so that more people knew about the recommendations. The CSWA was formed in 1979 and moved out of a working group status; the AAS endorsed the Baltimore Charter in ‘94; we've done demographic surveys; and we continue to support CSWA activities. We're also a sponsor of the WIA meeting.
Figure 3. The cumulative distribution of the fraction of women AAS members as a function of age as of 1995 and 2003. The fraction is increasing at an accelerating rate amongst the youngest members.
I will now shift to speak about prize and leadership trends, basically to point out that the Baltimore Charter had a huge impact in some areas. Table 1 shows (as of 1990) the number of prize winners for our main prizes - the Russell, Warner and Pierce, Tinsley and Heineman prizes’ (not the division prizes) and these are just the raw numbers of men and women winning that prize previous to 1990. Since 1990, we have had two Russell winners who were women versus 17 men and so on. Obviously not all prizes have begun to reflect our changing demographics.
The AAS leadership trends are also interesting to review. As
of 1990, and considering all the years previous, the AAS had had only one female president. Between 1990 and 2003, four women
and four men served as president. For the vice-president positions, as of 1990, only about 9 percent of women served as members of the executive committee in aggregate - that means including treasurer and secretary (but those are somewhat special positions of the executive committee, because they serve for a longer period of time). Between 1990 and 2003, the AAS has had three female vice-presidents for an aggregate executive committee percentage
of about 28 percent. Since 2003, we had 3 presidents, all male. There have been 5 vice presidents, 4 male, 1 female. For Secretary and Treasurer there has been little change, except that we added one male male to each position. The bottom line on leadership of the AAS is that, we need more women nominees for these important positions.
This change has come about because of participation in the process. Change has come from within. It is imperative that women, and minorities more generally, continue to participate in the system as it stands as they work to change it for the longer term. Only through participation can change come about and I think these statistics speak directly to this inherent property of a democratic society and the AAS in particular.
My conclusions are that AAS membership demographics are changing and changing rapidly in my opinion. In 2003, having 60 percent women among our youngest members and 35% among our members under age 35 is very nice to see.
The Baltimore Charter had a large impact on the number of women serving in society leadership positions. I think we still have more room to go on the prize situation. Everyone should remember that nomination committees don’t work in a vacuum. They need nominations in order to select winners. If a small number of women are nominated, a small number of women will be selected as winners. If a large fraction of women are nominated, then a larger fraction of our winners will be women. Arlo Landolt and John Graham, our two most recent Society Secretaries, have made this point many times and very directly and I make it again here. The information on how to submit it is included with the form, so you all should consider making nominations if you can think of deserving individuals.
Addendum – material added since WIA meeting for publication in STATUS:
I have performed a similar demographic analysis of our membership for 2008 and the results show a continuing trend, with some fluctuation. The demographic ramp of women entering our Society is continuing and each five-year cohort appears to be marching forward over time into the next age bin. Shown as Figure 1E is the 2008 version of the standard demographic plot I described earlier. The lowest age bin now only has 40% women, but also only contains 111 people total, so the fluctuation we see is in line with expectations, but bears watching.
I continue to expect the demographics of the Society to continue changing over time, especially as the large fraction of mid-career women progress further. The AAS-AIP longitudinal survey will help us understand better how their careers progress and what difficulties they are facing. The AAS stands ready to take action, if needed, to ensure that all of our members receive fair treatment as outlined in the Baltimore Charter.
Back to January 2009 Contents
Back to STATUS Table of Contents