November 26, 2017
Update December 6, 2017: I’ll be presenting on this at AGU Tuesday, December 12! A link to the abstract here. Also, I’ll be looking at the data for atmospheric and ocean sciences (and updating the numbers through 2015) in early 2018 so stay tuned.
The thing that sucks about being in your (hopefully, theoretically, optimistically) last year of a PhD program, is knowing how close you are to being done timewise, but also how far you have to go workwise. I have a lot of things to do in order to finish by my Spring 2018 target graduation. It’s nice to fantasize about being done. About being a doctor.
This past January I was reading an article about a woman who, in 2010, became the 63rd African American woman to ever get a PhD in physics. While some commenters at the bottom of the article expressed shock that so few black women had earned PhD’s in physics (compared to 22,172 white men in the U.S.), all I could think was, “That seems high.”
That’s because earth sciences is the least diverse of all STEM disciplines. Worse than physics. Worse than math.
So naturally I (as someone who self-identifies as a black woman) wanted to find out what number black lady I’d be to earn a PhD in geology (once I finish, that is). Super easy, I thought, just download a spreadsheet from the internet the way that other lady in the article did.
It wasn’t that easy. The NSF Survey of Earned Doctorates (SED) puts out public statistical reports, but these are typically clustered in annoying ways. Like, you can see how many black people earned PhD’s, but it’s not broken down by gender, or if it is, it’s clustered with Atmospheric and Oceanic Sciences, or even with Physical Sciences. I wanted a year-by-year breakdown of nationality (Permanent and US residents only), race, ethnicity, gender for earth/geosciences only.
From 1973 (when the SED began recording race, gender and ethnicity) to 2006, you can download a master spreadsheet of demographics for every field of study from webCASPAR. Data collected after 2006 has to be obtained from something called the SED Tabulation Engine, which includes a “disclosure control mechanism” that protects the identity of respondents when using the gender, citizenship, and race/ethnicity variables. This basically means that any small numbers (like less than 5 or so) will be redacted to protect the identities of those people. So if you want to know how many black women got PhD’s in geosciences in 2008, you can’t, because that data entry would show up as a bunch of asterisks since that number is so small. Luckily, a really helpful program officer at NSF working with the SED sent me a spreadsheet with all the unredacted numbers through 2012 (I’ve since requested an updated spreadsheet for data through 2015, so stay tuned).
I sat on this spreadsheet for a long time; I never made time to look at it. But of course, now that I have a ton of stuff to do (applying for jobs, postdocs, prepping for AGU, PhD research), it seemed like a perfect time to play with the numbers. Here they are:
Let’s just start off with the total numbers (reminder: this is for US citizens and permanent residents only):
First thoughts: That is a lot of white men. Not even half the amount Physics has, but still. The other numbers are extremely small, the lowest being Black women and Native American women with 21 and 12 earning PhD’s, respectively. It will probably be impossible to know how many black women have earned PhD’s before me, as there were certainly a few to do so prior to 1973 (Marguerite Williams was first, in 1942) and more since 2012. But I’d feel confident saying it’s certainly less than 40. I was also interested in trends over time, so I made a bunch of other charts. Let’s look at number of people earning geoscience PhD’s in a given year, by race:
Nothing too shocking here, except for the spike in Asian/Pacific Islanders in the mid-1990s. Let’s break it down by gender too:
Now things are getting interesting. Before, it looked like white PhD recipients had remained constant, but from this you can see that males are decreasing and females are increasing. I was really surprised that more white men had earned PhD’s in 1973 than in 2012. We can also see from this chart that the Asian mid-1990s spike was only true for males, and not females. I also think it’s really surprising that so few Asian Americans earn PhD’s in geosciences (of course, this number is much higher if you include non-permanent residents aka international students). This group is not typically counted as “underrepresented,” but this data suggests they should be.
Here’s a plot of men vs. women (all races):
Nicely done, ladies. Again, unsure why (white) men are decreasing, as this is absolute number of PhD’s awarded, not percentages.
To see if minorities have improved over time, I zoomed in on the three smallest groups:
With the exception of maybe Hispanic/Latino women, it appears that there were no real improvements in diversity between 1973-2012. That sucks.
And now for the obligatory pie chart, for 2012, the most recent year I have data from:
Okay, that’s a lot of charts.
To summarize: Between the years 1973-2012, the number of geology PhDs earned has increased for white women, decreased for white men, and stayed depressingly low for all minorities (including Asian Americans, despite a brief jump in the mid-1990s) with no signs of improvement.
