A special issue of The Lancet reports that gender inequality in the scientific community starts right at the beginning of the scientific process – when researchers petition for cash from funding bodies. But gender anonymity in applications could actually make things worse, argues Professor Sarah Kaplan, a Canadian academic.
When Kaplan, professor of gender and the economy at the University of Toronto, recently got invited to speak on a panel where she was the only woman, she immediately wrote back saying this was inexcusable. According to a Wired report, she got an apologetic answer from the organisers of the academic conference saying that they had tried, that they had invited six women, but that they hadn’t found anyone. “They said they had worked hard to find women speakers, and I am sure they had,” she says. “But they should have worked extra hard and put extra effort in this. That’s how you start making change.”
The report says it’s a mantra we all know: we need more women in science. But with female researchers making up only a third of the scientific workforce in North America and Europe, it’s clear we’re not doing very well when it comes to turning our well-meaning sentiments into hard actions.
Why exactly is that? The report says to try to answer that question, medical journal The Lancet launched a special issue on 8 February, 2019, dedicated to mapping the causes and consequences of gender inequality in the scientific community.
The report says that inequality starts right at the beginning of the scientific process – when researchers petition for cash from funding bodies. It had already been shown that men, for instance, had better odds than women of obtaining funding for their research – and one of the main studies published in this issue identifies where and when that gender gap originates. Things start going wrong, the issue explains, during the process of evaluating which research projects should be allocated public money. Evaluators tend to base their decision to invest on the scientist and not on the science proposed.
The research team, from a range of Canadian universities, analysed almost 24,000 applications submitted over five years at a national funding agency in Toronto. It divided funding applications between two grant programmes: one in which the review focused on the quality of the science proposed, and one where it focused on the calibre of the applicant.
Criteria for funding in the first programme, for example, included “importance” and “quality” of the idea; while the second one asked the candidate to demonstrate their leadership or productivity skills. In the first case scenario, the proportion of male and female-led proposals that received funding were roughly the same. But when the funding bodies focused on assessing the candidates, male applicants were 44% more likely to be granted public money than female applicants.
The report says the solution seems straightforward, then: reviews of research applications shouldn’t be accompanied by any information about their applicant’s gender. Not so fast, says Holly Witteman, from the University Laval in Quebec, and lead author of the study. “It is really important to diagnose the problem well,” she says. “If the problem is implicit or explicit bias against female researchers, then concealing their identity is a good way to fix that. But if the problem is systemic, then anonymity could actually make things worse.”
That’s because there are some inequalities within the system that gender-neutral pronouns or blind reviewing can’t tackle. Inequalities that range from the allocation of lab spaces to sexual harassment, and which all contribute to reducing the quality of the work that female researchers can produce. The report says Witteman calls that “cumulative disadvantage”.
It might be unfair, in this case, to review the work of female scientists based on the same criteria as that of male scientists – and, the report says, that’s where gender-neutral pronouns and blind reviewing become counter-productive.
Witteman herself holds a foundation grant. One quarter of the criteria to obtain it was based on her ability to demonstrate leadership in her career. “I’m in a medical school,” she says, “and we’ve never had a female Dean. The current Rector is a woman, but that’s a first in the 350 years that the institution has existed.”
The report says for many female researchers, arguing their case against male applicants is like racing against someone who has been given a head start. This is why one of the solutions Witteman suggests is to adjust review scores for women, to account for gender inequalities that exist within the system. “Funding groups like the National Institute of Health (NIH) in the US do those adjustments for specific groups that are disadvantaged, like early career investigators,” she says. “So that’s one potential avenue.”
The report says Kaplan advocates for similar solutions. The fix isn’t just about bringing more women – and minorities – to the scientific workforce to increase diversity, she says, it’s also crucial to bring them to a system that is built to give them the same opportunities as men. Including more people who have been historically underrepresented in certain fields – in this case, women and minorities in science – without then investing in mentoring and supporting them, will only backfire, she continues.
“That’s why just focusing on diversity doesn’t work,” she is quoted in the report as saying. “If you fail on the second part, which is to make people actually want to stay by giving them opportunities, they will eventually leave.”
A major obstacle standing in the way of inclusivity, however, is the nature of bias. The report says that is because at its very origin, bias is a process of categorisation that is central to the way we socialise. Because of the amount of information it receives, the brain simplifies things for us by attributing most of it to categories that our neocortical system has drawn throughout a lifetime of observation.
Those categories, such as race, age or gender, originate from patterns we perceive, and come to define our beliefs and expectations. For example, if we have rarely been exposed to female scientists, our brains will automatically perceive female scientists as being against the norm. And when something is embedded in the brain in such a way, changing it is extremely difficult, if not impossible.
If gender bias is endemic, then, does it mean that there is little hope for women who wish to have a scientific career? The report says Kaplan, for one, is far from admitting defeat. If we want people to change those categories, she argues, we have to start by fixing the procedures and practices of the system they are part of. “What we have been doing until now,” she says, “is just telling people that they are biased and hoping that they will change their behaviour. That doesn’t work. We have to give them the tools to do so.”
In evaluating research papers, for example, reviewers could have checklists – “scientists love checklists, why can’t they have one for this?” she asks – to make sure that criteria has been adjusted when reviewing papers submitted by women.
The report says this is not just about defending the democratic ideal of equality – not that there is anything wrong with that. But another study demonstrates that it is also the quality of scientific research that is strongly linked to the gender of the scientists leading it.
Studies led by two females, indeed, are 26% more likely to report on both male and female models than those that are not. And, the report says, this is important, because the same study, after analysing 11.5m research papers published between 1980 and 2016, found that almost 70% of them failed to report on outcomes for both men and women.
Sex accounts for many biological differences in vulnerability to heart disease or autoimmune issues, among many others. And, the report says, the consequences of science’s lack of concern for sexual variations are already well known. Consumer organisation DrugWatch estimates that women have twice as many chances of developing an adverse reaction to medication than men do, for example. That’s due to the fact that many drugs are developed based on male models; and now, this study demonstrates that this is more likely to happen when research is carried out by men.
While that can’t prove directly, of course, that including more female scientists in the workforce will directly lead to more sex-diverse research, it is a fair assumption to make that case, says its lead author Vincent Lariviere, from the University of Montreal in the report. More should be done, therefore, to increase funding for female-led research, or to impose the study of both sexes as a criteria for allocating grants. But that is not all. “We don’t just need more women in the system,” says Lariviere, “but more women in leadership positions, so that they have the power to decide what research is going to be done.”
And the implications go far behind science. For Witteman, getting scientific equality right is a matter of public health: “these are public dollars, and they are scarce,” she says. “It is important that they get allocated to research that will effectively improve public health – men’s and women’s.”