In 2020, the National Institutes of Health (NIH) has invested $41.7 billion in research. And many Black scientists appear to be missing out on those awards.
More than 80% of that budget is handed out each year through almost 50,000 grants to researchers at universities and other institutions. But as research by economist Donna Ginther, PhD, at the University of Kansas, and colleagues found, Black scientists were 10 percentage points less likely than white applicants to be awarded NIH research funding from 2000-2006 In other words, the award probability for Black applicants was roughly 55% of that found for white applicants (Scienceopens in new window, Vol. 333, No. 6045, 2011). In the latest in a series of analyses to find out why, psychologist Richard Nakamura, PhD, former director of the NIH’s Center for Scientific Review and former deputy director of the National Institute of Mental Health, has co-authored a paper exploring at which stage of the peer-review process disparities emerged. Nakamura and colleagues found that racial disparities appeared in the very earliest stage of peer review (Erosheva, E.A., et. al., Science Advancesopens in new window, Vol. 6, No. 23, 2020).
Nakamura, who retired from NIH and spoke solely in his role as a volunteer for the organization, spoke with APA about these findings.
What does your paper reveal about the causes for the funding gap?
With colleagues including Carole J. Lee, PhD, and Elena Erosheva, PhD, both of the University of Washington, we assessed data from 2014 to 2016 and found the funding gap identified in the 2011 paper remains, despite changes to the peer-review process instituted in 2009. We found that the disparities arise at the earliest stage of peer review, when reviewers give grant applications preliminary criterion scores in five areas: significance, investigator, innovation, approach and environment.
Importantly, this paper suggests that later stages of the review process, such as when reviewers engage in discussion about the grant application or assign a final score, do not change the racial disparities in scores. That doesn’t mean we couldn’t somehow compensate or correct for bias at later stages. But in our search for the source of the bias, we can focus on whatever happens at or before the first stage of peer review.
What might be happening to disadvantage Black scientists?
The causes are complex, and there are almost certainly multiple factors. Implicit bias among reviewers is one possibility. Reviewers aren’t given information about the identity of the investigator. But because reviewers are usually well acquainted with other researchers in their domain of expertise, it wouldn’t be a surprise if some knew the race of the applicant.
Another likely possibility is that there is some qualitative difference between applications from Black scientists and white scientists. Research has shown that compared to white investigators who are applying for grants, Black investigators tend to have smaller publication records and lower citation rates, and they are more likely to publish in journals with lower journal impact factors. Those differences may be due to years of cumulative discrimination. We know there is racial discrimination in all parts of society. That cumulative discrimination may be working against Black applicants, such as by contributing to differences in career mentorship opportunities or collaborator networks. And that cumulative bias could work in the other direction, too, by creating the conditions that favor white applicants [For more on this, see Ginther, D.K., et. al., PLOS Oneopens in new window, Vol. 13, No. 11, 2018].
How might the review process change to correct these disparities?
Figuring this out is a priority for the NIH. I never had a meeting with the leadership that didn’t discuss the issue. With others at NIH, we’re working on an anonymization study to determine whether further concealing the identity and demographics of the principal investigator would eliminate or change the scoring disparities. Obviously, a scientist’s track record is an important indicator of their future productivity. Yet if it turned out that such a step eliminated the difference in scoring, we might consider making changes to the peer-review process. One idea that has been suggested is a two-stage review: First, reviewers would assess the research idea without any information about the investigator. Additional information, like the scientist’s track record, could be considered during a second level of review.
What more can the research community do to address racial disparities in science?
We have to take a look at the environment to make science feel welcoming to diverse perspectives and personalities and lifestyles. Diverse teams are more creative; they bring different ideas to the table. But the scientific subculture can be unwelcoming to a lot of people, including women and transgender and gender-nonconforming individuals, as well as Black scientists. That may be why Black scientists’ professional networks are smaller, or why they have less citation activity.
To make it more welcoming, scientists need to look deeply to figure out how and why the culture may be toxic to people from different backgrounds.
Institutions should certainly be paying attention to this and working to increase diversity. We should also encourage senior scientists to improve the diversity of their labs and better assist their minority students. Senior scientists with strong track records are the ones who get their students into the best graduate programs or postdoctoral positions. Those senior scientists help steer their students and put them in touch with more collaborators, so their efforts can snowball.
Disparities have been built into the structure of this country for centuries. Our recent struggles remind us how little progress we've made as a nation. We need to dig deep to make it possible for Black Americans of every kind to succeed in our society. Improving diversity isn’t just good for individual scientists, it’s good for science and for society.
