Improving Diversity in Science Journalism – A Case Study
It is hardly controversial these days to point out that science has a diversity problem. According to the National Science Foundation, women make up only 28 percent of employed scientists in the United States, and only one third of all scientists are non-white (compared to 40 percent of the total population). Black, Hispanic, and Native American professionals make up just 6 percent of the science workforce, and Black scientists received just 1 percent of National Institutes of Health research grants in 2018.
So far, so shameful — and we, as science journalists, have traditionally been part of the problem. According to the Global Media Monitoring Project, only 19 percent of the experts quoted in science and health stories are women. The percentage is even lower for underrepresented communities: Black, Indigenous and people of color (or BIPOC). Science reporting that underrepresents women and BIPOC reinforces the kind of biases that stop these groups from pursuing science careers or from rising to the highest levels in their fields.
It is a self-reinforcing cycle that hurts us all: a lack of diversity is not only morally and ethically problematic, it has also been shown to reduce the quality and relevance of scientific research.
Back in 2014, when we launched Gastropod, the podcast that explores food through the lens of science and history, we were determined to do better. From the very beginning, as we chose topics and decided which experts to reach out to and interview for the show, we constantly checked ourselves to make sure we didn’t just end up with a handful of white males discussing their research.
But we had no idea whether our efforts were good enough — how well (or badly) we were doing in trying to showcase diverse voices. Few of our peers knew either. While certainly many journalists make efforts to ensure a diversity of interviewees, only a handful have publicly shared their efforts to count the gender and race of the experts they quoted in print and on the air.
We wanted to take things a step further. Rather than simply talking about the need for diversity in science journalism, we wanted to create a clear, in-depth assessment of how we were doing.
When the Alfred P. Sloan Foundation invited us to apply for a grant to support our coverage of science and technology, we saw an opportunity. We requested an additional budget to conduct a thorough — and public — analysis of the backgrounds and participation levels of the guests on our show during a two-year period (October 2016 through September 2018.) While evaluating that data, we also set improvement goals and strategies for the project year, October 2018 to September 2019.
With Sloan Foundation support and the input of our advisory committee, particularly statistics wrangler-in-chief Kristi Lemm, we spent two years analyzing our track record and takings steps to try to improve it. We describe the process and results in our report “Tracking and Increasing the Representation of Diverse Voices,” published last fall.
The first big question was how to measure representation. The most straightforward approach would be to just count sources. But that can be misleading. One male scientist and one female scientist may be quoted in a story — gender parity achieved — but if the man is quoted or referenced 80 percent of the time, the actual representation is still skewed. Gastropod is a podcast, however, and we realized we could be more specific by adding up the amount of time each person spoke.
Our advisory committee helped us set our ambitious diversity goals: 40 percent of speaking time should be women’s voices as opposed to men’s; similarly, 40 percent of speaking time should be BIPOC voices. These goals were, we knew, a stretch. They overrepresent women’s and BIPOC voices based on their numbers in science. Yet, in terms of gender overall, our goal still underrepresents women based on their percentage — 51% — of the U.S. population.
We outlined our methodology in the report, and included templates for anyone interested in conducting a similar effort. We’ve also listed the strategies we used to try to achieve our goals: soliciting suggestions, searching through databases, reaching out to organizations that provide a network for women or BIPOC scientists, and more.
Overall, our efforts had a mixed result. Ultimately, we exceeded our goal that 40 percent of speaking time about science on Gastropod should be women’s voices as opposed to men’s, but we did not come close to our goal that 40 percent of speaking time should be BIPOC voices. As we analyzed what worked and didn’t, we came up with a number of explanations, and lessons learned, all listed in the report.
Our findings have reshaped our own working method. We found that one of our most successful strategies was to deliberately build an episode around a woman or BIPOC individual’s research, as opposed to picking a topic and then trying to find diverse sources.
This occasionally took some creative thinking to see how we could build an episode around work that maybe wasn’t at first glance obviously a Gastropod story — but it resulted in some amazing episodes. It also took a lot of time: for a show that puts out an episode exploring a single topic only once every two weeks, the ramp-up from launching this more concerted approach to finding diverse subjects and then creating episodes around them can be lengthy.
Perhaps the most surprising finding was that our focus on gender equity was, perhaps, a blinder. By prioritizing female scientists, we had likely been choosing to interview white women over their male BIPOC colleagues. As we more than met our goals for gender representation, this project helped us realize that it may at times make more sense for us to interview a BIPOC male scientist rather than a white female scientist. Of course, we are also continuing to work hard to find and interview female BIPOC scientists, such as the awesome soil biogeochemist Asmeret Asafaw Berhe and materials scientist Ainissa Ramirez, who have starred in recent episodes.
Improving the diversity of voices is not a quick fix. Undertaking this project was a huge investment for us, but we’ve realized it’s critical to continue not only monitoring how we’re doing, but also to identify strategies that are working best. We can and should do better — and the same applies to our colleagues at other podcasts and publications. We hope our report will help.