Twitter
Facebook
YouTube
LinkedIn
RSS

Addressing Bias in Canadian Research Funding

November 3, 2016
By: 
Holly Witteman
Assistant Professor, Faculty of Medicine
Université Laval

 

In the last round of a major Canadian Institutes of Health Research (CIHR) funding competition, female mid-career and senior applicants had only half the success rate of their male peers. Eight percent (8%) of female applicants in this group received funding, compared to 16% of male applicants. These rates occurred in the 2015-2016 CIHR Foundation Scheme, the second round of a new program designed to fund, ‘people, not projects,’ and which was planned to consume approximately 45% of CIHR’s open operating grant budget. Disparities also occurred in the first round of this competition. Available data from the CIHR suggest that the gap occurs at Stage 1 of assessment, in which 75% of the score has to do with the researcher and her or his track record.

Meanwhile, in the CIHR’s Project Scheme—the ‘projects, not people’ counterpart to the Foundation Scheme—male and female applicants had approximately equivalent success rates across career stages. Grants awarded in the Foundation Scheme are mostly longer with higher annual budgets than those in the Project Scheme, meaning that in an era of historically low funding rates for health research, these imbalances in success rates have especially large impacts on the amount of research that female health researchers are able to do. Researchers who are women may be more likely to focus on topics relevant to women’s health, meaning this has implications well beyond the careers of the researchers themselves.

Other research funding programs whose review processes emphasize the reputation of the candidate show similar or worse imbalances. Vanier Canada Graduate Scholarships are awarded to PhD students; Banting Postdoctoral Fellowships are awarded to those doing additional training after a doctorate. Both programs give more funding and are more prestigious than other federal doctoral and postdoctoral awards. Both programs also had success rates consistently about 3% lower for female applicants than for male applicants over the past 5 years in a context of overall success rates around 15%. The prestigious Canada Research Chairs program, which fully or partially funds the salaries of awardees, continues to under-represent women, Indigenous peoples, people who are members of visible minority groups and people with disabilities. The even more prestigious Canadian Excellence in Research Chairs program, which awards $10 million to each chairholder, currently has 26 chairholders who are men (96%) and only one woman (4%).

Taken together, these data suggest that while women in Canada propose research that is deemed equally excellent as men’s research, female researchers are less likely to be perceived as leaders in their field. Evidence has shown that many academics of all genders have implicit biases against women. Students rated the same online instructors higher when they believed their instructor was a man, recommendation letters have been repeatedly shown to be gendered, the majority of faculty at an institution implicitly associated men’s names with leadership and women’s names with support, male scientists implicitly associated science with men, women and Black scholars are under-represented in fields believed to require brilliance, and both male and female scientists judged male candidates for a lab manager position as more competent & hireable. Possibly as a result, conference abstracts led by women have been accepted more often when reviewers don’t know who the authors are, papers led by women are cited less often, even when women publish in higher impact journals than men, and women may be invited to speak at conferences less often. Less data are available on other equity dimensions, but given findings in other settings, it is plausible that bias may also be affecting the funding levels of researchers in Canada who are Indigenous, who are members of a visible minority, or who have disabilities. Researchers who are members of more than one under-represented group (for example, women of colour, or Indigenous researchers with disabilities) may be especially disadvantaged.

When bias influences the results of a funding program, we fail to fund the best research. This must be rectified for the future of research in Canada. Given the barriers to the advancement of women and members of other under-represented groups in academia, it is especially urgent to correct the large gaps currently plaguing the most prestigious Canadian research programs.

 

5 Steps to Achieve Equity in Federal Funding

 

1. Measure & Report Equity Dimensions

The only equity dimension routinely reported by Tri-Council agencies is the sex (male, female, or not reported) of the nominated principal applicant. All federally-funded programs should collect data on applicants’ gender (man, woman, or other), age, career stage (measured by years in appointment), and whether or not the applicant identifies as an Indigenous person, a member of a visible minority group, or a person with a disability. Aggregate, non-identifying data should be transparently and rapidly reported so that Canadians know whether or not publicly-funded research is equitable.

 

2. Reflect on Program Design

When a program shows imbalances in award distribution by one or more equity dimensions, rather than wondering, “What’s wrong with applicants’ profiles such that they aren’t receiving these prestigious awards?” funding agency leadership should wonder, “What’s wrong with this program such that a particular group isn’t better represented among awardees?” Criteria such as Leadership, which represents 25% of the score for the CIHR Foundation Scheme’s Stage 1 assessment, bring bias to the fore. For example, the fact that women receive fewer leadership opportunities and are less likely to be perceived as leaders necessarily hampers women’s success by such a metric.

 

3. Automatically Equalize

For programs that award funding partially based on assessments of the researcher, the portion of the assessment focusing on the individual should be automatically equalized. For example, for the CIHR Foundation program, if 32% of mid-career applicants at Stage 1 are female, then 32% of mid-career applicants accepted to Stage 2 should be female. Automatic equalization at Stage 1—a stage at which no funding is allocated—is a low-risk, sensible policy. Such a policy would simply level the playing field by allowing equitable proportions of applicants to have their proposed research judged.

 

4. Train Reviewers

All reviewers should be required to complete anti-bias training that addresses implicit and explicit biases against women, Indigenous peoples, members of visible minorities and people with disabilities. Training must be evidence-based and must also be evaluated to assess its effectiveness and efficacy.

 

5. Reward Equity

Finally, federal funding policy should reward fairness on the part of programs and institutions. Rewarding fairness in programs would involve simply allocating a greater proportion of funds to programs demonstrated to be equitable. However, many of the current imbalances are rooted not in federal funding agency programs, but rather, in institutions. Institutions determine whom to nominate for prestigious scholarships, fellowships, and chairs. Federal funding agencies and programs have the power to change institutional behaviour. For example, allocation of Canada Research Chairs is currently determined by a formula largely based on the amount of funding that a given institution’s researchers are awarded from Tri-Council agencies. If this formula included additional points for equity targets, institutions would rapidly strive to achieve such targets. For example, additional points could be allocated for actions such as becoming signatories of and receiving awards from relevant charters (e.g., a gender equity charter such as the Athena SWAN charter, an Indigenous scholars equity charter, a race equity charter, a scholars with disabilities equity charter), addressing relevant Calls to Action from the Truth and Reconciliation Commission (which may include implementing Calls to Action 7, 16, 22, 24, 28 and others, and taking a role in 11, 62ii, and others) and contributing adequately to the very achievable equity targets set by the Canada Research Chairs program.

 

Funding Can Be Fair

We need to have the courage to implement steps towards equity. Not only because it’s 2016, but also because it is critical to ensuring that we are funding the best research and thus serving people in Canada.