I completed my doctorate in 2001, when approximately 45 percent of biomedical doctoral students in the UK were women. At my current level — junior professor — this percentage drops to 30 percent, before falling to 15 percent at the top professorial level. There have been many attempts to explain why.
A commonly voiced hypothesis is that women rule themselves out of the race by deciding to have children (the average age of completion of postdoctoral training is between 28 and 35), and it is true there does seem to be a relatively high proportion of childless women in academic science. However, there are enough counter-examples (Nobel laureates such as Elizabeth Blackburn, Ada Yonath and Carol Greider to name just three) to make this scenario unlikely to be the whole explanation.
Then, of course, there is the suggestion that there are innate differences in the ability of men and women to do science, most famously put forward by former Harvard University president Larry Summers with his dribblings about “daddy and mummy trucks.”
In the fallout from that performance it was noted that the percentage of tenured job offers made to women had dropped dramatically since Summers took office, something he pledged to rectify.
Can it be that women are treated less fairly than men? A deceptively simple piece of research led by Jo Handelsman at Yale University has recently suggested that they are. The authors created application forms purporting to be from a recent science graduate wanting a laboratory manager job and asking for feedback. In total, 127 faculty members were asked to rank the candidate in terms of competence, starting salary they would offer, willingness to mentor the candidate, and likeability. The only difference in the applications was the name of the student — 63 were from “John” and 64 were from “Jennifer.”
The results were stark. Jennifer was ranked less competent than John and was offered a median starting salary almost US$4,000 lower than John. In addition, the faculty was less willing to mentor Jennifer, but, strangely, found her to be more likeable. All this from a piece of paper. I should point out here that there was no statistically significant difference between the responses from male or female faculty, nor were there differences between levels of faculty, suggesting this is not a hierarchical bias.
So, what does this mean? The study was nuanced — the curriculum vitae (CV) was deliberately designed to represent a good, but not stellar, candidate. This is a key point. When faced with a candidate who is clearly exceptional, gender rarely matters. This is because it doesn’t require thought to come to the conclusion of excellence. By the same token, a candidate who is really not good is also definable without much thought. However, for the large chunk of people who inhabit the “gray zone,” where subjective and objective evaluations matter, the outcome of this study indicates that men get the breaks where women do not.
The study, published in the Proceedings of the National Academy of Sciences of the United States of America last autumn, has prompted much debate and soul searching among my colleagues. Reactions have been quite eye-opening. Many people, including myself, were largely unsurprised. Several non-scientists found it hard to believe that the same CV could be evaluated so differently, and with such serious consequences in terms of pay and mentoring. Yet since the beginning of my career, I have always been acutely aware that I need to do better than a man to stand a chance of being hired ahead of him. Several people welcomed what they saw as concrete data to support their observations and Handelsman was listed as one of last year’s top 10 people by the journal Nature.
A less palatable reaction was one of irritation, and dismissal of the findings. Despite the fact that hard data is difficult to argue with, many scientists managed it. My own explanation for this reaction is that on a subconscious level, data like this support the implication that men in science didn’t necessarily get there on merit alone, but also because their female competitors were being discriminated against. That must be quite threatening and hence provoked a defensive response.
What are we going to do about it? Here’s where it gets tricky. The outcome of the study suggests that simply increasing the numbers of women on hiring panels is unlikely to solve the problem as the bias exists in all of us. The authors suggest educating scientists to be more aware of the potential biases that exist.
We should go further than this. Since this study shows that the discrimination occurs, at least in part, at the stage of the evaluation of the paper application, I would insist that hiring committees shortlist the “best” female applicants. Such positive discrimination is controversial. However, if 100 candidates apply, and six are shortlisted, how hard would it be to ensure the top two female candidates are also shortlisted? It would certainly allow the next step of the evaluation to be done in person, where many of the subconscious influences are no longer invoked.
On a final note, the persistent lack of women in the higher echelons of science is a constant source of lively debate. Several elder colleagues of mine have insisted that it is only a matter of time before things even out. However, I think this study shows that the discrimination occurs against women at the outset of their careers, and that it is systemic. Without a proactive approach to address the problem, we will be stuck at the current numbers for the next 100 years. This is not something I want to have to explain to either my son or my daughter.