You may have read in late September that the ratio of women receiving Royal Society funding has “plummeted from one in three in 2010 to one in 20 this year.” While the Society also awards the Dorothy Hodgkin Fellowships to early career women researchers, this award exists to boost women’s participation in science, not to augment or mask the issues in the Society’s mainstream Fellowship program.

The Royal Society was silent for a couple of days after its list of fellows list was made public, despite a large outcry by the scientific community on social media and opinion columns in the media. The Society President, Sir Paul Nurse, finally announced an investigation a couple of days after the fact. The question is: why did the Society wait until it was made public to assess their program?

I want to stress that while I’m using the Royal Academy’s Fellowship outcomes as a case study, the issue I am illustrating is the reactionary treatment of gender bias in all fields of Science, Technology, Engineering and Mathematics (STEM). The point here is to tease out institutional patterns and to make the case that institutional approaches are needed to address gender inequality. While this point may seem obvious, the fact is that inequality in science, as with other spheres of social life, is still treated as a surprise. This is because, on the whole, organisations (and society in general) remains reactionary to addressing gender inequality. Diversity is an afterthought, when it should be a proactive and ongoing project at the organisational and societal levels.

This is the first in a series of articles I’m writing on why the scientific community, inclusive of various disciplines, needs to re-examine its position on the problem of inequality in STEM. The picture I am building up is one of methodological rigour and interdisciplinary collaboration in order to better work towards gender inclusion.

Science Fellowships and Institutional Bias in STEM

Royal Society Fellowships

The Royal Society of London for Improving Natural Knowledge has made great contributions to gender equality in science, having hosted two annual Wikipedia Edit-a-thons, which encourage scientists to write Wikipedia entries about women scientists. This project has been adopted by other universities and science organisations such as the Australian Academy of Science’s Wikibomb. Yet when it comes to other organisational practices, there’s a long way to go.

No women have served as president for the Royal Society and only five percent of fellows at the beginning of the year are women. These facts feed into a broader problem of everyday sexism, where interpersonal and organisational practices validate broader gender inequalities at the societal level.

It’s great to host activities to promote gender diversity, but if an organisation’s structure and policies don’t actively reflect gender diversity, organisational commitment to gender equality has limited practical impact. Organisations like the Royal Society need to lead by example. Gender equality is obviously important given examples such as the Wiki hackathons and special awards such as the Dorothy Hodgkin Fellowship, but conflicting patterns within the Royal Academy reflect a broader problem in STEM.

Statistical Trends

As Zoology Professor Ben Sheldon noted on Twitter, having compiled data for the first two graphs below, the Royal Society did not respond to his public requests for a breakdown of Fellowship applicants. Sheldon’s graphs clearly show a worrying trend where less women are progressing through the Fellowship process (graphs via Katie Mac).

Royal Society Fellowships. By Ben Sheldon

Royal Society Fellowships application process. By Ben Sheldon

Some researchers defended the statistical trend arguing that the 2014 data are an anomaly. For example Chemistry PhD student Rebecca Murphy argues that the number of women fellows “has remained broadly stable over the last five years.” Physics Professor and Royal Society council member Athene Donald similarly argues, “as previous years’ data show, I don’t think the Royal Society can be said to be permanently showing deep prejudice towards women applicants.”

Swedish Professor of Mathematics David Sumpter, however, argued that the statistical trend was indeed worrying, and that “gender bias in academia can easily be created even when each career stage appears to be fair according to a statistical test.” He takes a long-term view of academic careers, focusing on different stages of recruitment using a binomial distribution (probability of success over a sequence of events). He shows that doing a statistical test of overall trends can obscure the statistical inequalities at each stage of academic application process. He writes:

“Doing a statistical test at one stage might show that that particular stage was fair, but it is more or less meaningless when done isolation. [Royal Society’s Edward] Hinds admits this at the end of his article, but does not spell out that each stage can appear ‘fair’ while producing an unfair outcome….

And that brings me to the Royal Society’s appointment of 2 out of 43 research fellows. Given the 21% female applicants each year, this was obviously statistically significant and I believe a poor decision. As Richard Mann argues in his blog post it might be reasonably considered as a statistical blip. Ben Sheldon provided statistics over the last 6 years giving the proportion of female appointments as 24%, 30%, 17.5%, 19.4%, 17.1% and 4.7%. The average is 18.78%, which given 21% female applicants gives a P value of about 0.4. Not biased against women, but hardly evidence that the Royal Society is helping women at this critical time in their career. More worrying is if we only take the last four years. Now the average is 14.68%. And the P value for this? About 0.2. Just around the level we might expect in my model for an institution trying to balance itself above statistical significance. While I am sure the Royal Society take this issue extremely seriously, I am afraid that these types of statistics do not look good.” [My emphasis]

So, to reiterate: taking the raw numbers, fluctuations this year may not look so bad if we compare them over a short period. Yet if we take a longitudinal approach, and examine inequality in stages, rather than simply from one year to the next, then gender bias becomes even more pronounced. What explains this? To get to the answer, we need to take into consideration various complex and interlocking institutional processes.

Institutional Bias

Girls in STEM. Photo by Argonne National Laboratory.

Elsewhere, I have made the point that competition for Fellowships mask institutional barriers that women face along way. This includes career and training opportunities that are not easily accessible to women due to life circumstance (such as family responsibilities), as well as other experiences of discrimination along a woman’s educational path. There are general issues that affect all women broadly, such as the effect of stereotypes and sexual harassment. There are even more issues that affect minority women specifically, such as sexism coupled with racism and homophobia.

The Royal Society would have known the winners well ahead of the announcement. It’s a competitive process but it is not a lucky draw. There are standard procedures and several steps that participants undertake. Why is the Society not monitoring its own procedures to ensure gender participation is more equitable at every stage of the process? As I will show, this means not only monitoring the Fellows’ application progression, but what it is doing to ensure that enough women are receiving adequate support leading up to the application.

The Society has publicly stated that they are disappointed by the 2014 figures as if they came as a surprise. This should not be the case. Like many gender issues in science, organisations do not critically examine their practices unless the public complains loudly enough. This is simply not good enough. Gender and equity issues require ongoing evaluation and commitment, and not simply reactions set against scandal.

Gender diversity cannot be taken for granted. In science, we continue to hold onto the myth of meritocracy. Some of the discussion on social media reflects this: perhaps not enough women applied; maybe the quality of women applicants was simply not high calibre this year; maybe women should be more proactive (read: “aggressive”) in pursuing a Fellowship. This is yet another example of the onus being placed on women to monitor their own progress. These types of arguments fail to see how women are being asked to participate in a structure that is biased against them at every stage of their education and career progression.

Science is not an even playing field. It never has been. The obstacles that women and minorities face are there not because they aren’t working hard enough, but because systems perpetuate inequality. Like the broader “leaky pipeline” problem, Fellowships need to be reviewed actively and routinely, not after the fact, but as an ongoing process. Policies may be inclusive on paper, but not in practice, especially if problems are not being actively monitored, evaluated and acted upon.

Male postgraduate students easily navigate academic bureaucracies with stronger support by their supervisors relative to women. Selection panels have been shown to be biased against women scientists even when they have the same qualifications and experience as men, and even when they put in the same number of work hours as men. Married academic men with children rise up quickly through academic ranks while their women counterparts are left behind. Even the public profile of White male scientists is more valued than women and minorities. These trends are not unrelated and they collude to prevent from women being treated as equal candidates in Fellowships and in the job market more broadly.

At every step of the way, before applicants have even started to look for Fellowships, women are already disadvantaged. Organisations need to own the responsibility of inequity. It’s not just one oversight one year; it’s a series of decisions that encourage men while obstructing women’s participation. The Royal Society is only one example where one “anomaly” actually reinforces a system of inequality. Training, education, policy and institutional reform are needed not just to lift numbers next year, but to put women on a stronger path of lifelong career success.

Moving Forward

I have a backlog of blog articles, so you’re going to read two from me over the next couple of days on this same issue and more in coming weeks. In part, I am motivated by moving beyond the cyclical arguments the sciences engage in when it comes to gender inequality. It seems that when we make some progress in one area, we take another step back elsewhere. This year alone, one science publication caused an uproar with sexist imagery, only to be followed by similar events not once but twice a few months later (with transphobia additionally thrown in the third instance). Important online discussions addressing sexual harassment have led to positive offline, cross-disciplinary engagement (such as with the study led by Professor Kate Clancy). Yet in between, high-profile science elders have engaged in incendiary sexist attacks on women, while ongoing issues in tech single women out for abuse.

From humanities to life sciences; from social to computer science; from pure to applied fields; gender inequalities are not easily overcome with a tunnel vision view. That is, we need to have a better grasp on why data collection on gender dynamics matter. We need to commit to working together, across disciplines, to address inequality. Above all else, we need to move away from individual-level explanations.

By thinking individual women can simply beef up their Fellowship applications, we miss the bigger picture – that is, how institutional processes affect different groups of women at different stages. We need to better understand the hurdles that stand in women’s way before they even consider applying for Fellowships (for example), and what happens to women at different points in their careers which make success so much harder.

It’s not just a numbers game – as Sumpter shows with the Royal Society, data can obscure patterns if we are not looking deeply enough. Even then, the data can tell you the outcome – it can show, yet again, that inequality exists, but statistical analyses alone cannot tell you why. As I will show in my next post, the way in which we collect, analyse and interpret data on gender dynamics in STEM needs to change.

We are all, in every science and tech field, complicit in gender inequality as long as we continue to allow the conversation to be shifted away to red herrings. The fact that gender inequality exists in STEM is a fact. What it means in different disciplines, and how to address the nuances of inequality is where we now need to focus our attention by also taking into consideration race, sexuality, class, disability and other socio-economic measures (this is the theory of intersectionality – more on this at a later time).

We are beyond the point of looking to individual women to raise their voices louder for help, because blaming women for inequality is a tactic that does not work. Whether its woeful outcomes in Fellowships, or pay discrepancies amongst faculty, or representation at conferences, or publications, or some other measure – these are all symptoms of institutional damage. We need to look to the causes and own them as a collective problem that requires collective effort. Enough with ducking our heads in the sand and hoping no one notices inequality. Enough with the surprise and the reactionary promise to do better next time. A stronger appraisal of longitudinal trends of inequality and long-term, proactive measures are the remedy. Let’s clean up this sick system in STEM, instead of making excuses for it.