A study published in late January adds further weight to the refrain that diversity, equity and inclusion champions often share: DEI work is about more than just getting more minority voices in a room.

The study, published in State Politics and Policy Quarterly, evaluated how newsrooms with a good makeup of staff from traditionally underrepresented backgrounds and with an audience that was also diverse tended to cover political candidates — and whether differences emerged between the coverage of white vs. non-white candidates.

The study focused on an idea known as trait coverage, where candidates’ traits — such as socioeconomic status and other parts of a candidate’s social identity — are focus areas of news stories versus policies candidates may be pushing for.

Here’s how the authors of the study —  the University of Alabama’s Mingxiao Sui and the University of North Texas’ Newly Paul — actually did the study:

The authors used three main data sources: a database of political candidates running for offices in 2012, including their race, ethnicity, and gender; and Access World News, which contained a comprehensive collection of news articles that covered each state legislative candidate identified in the database (the researchers excluded op-eds and editorials from the search).

To determine the racial makeup of newsrooms, the researchers also used the 2012 American Society of Newspaper Editors newsroom census. To determine a legislative district’s racial makeup, Sui and Paul used data from the State Legislative Election Returns series.

The researchers narrowed their search to articles published between September 1, 2012, and November 6, 2012 (which was Election Day that year). The final dataset included nearly 1,000 news articles about almost 600 state legislative candidates across 13 states in the U.S.

Perhaps unsurprisingly, the authors found that having a diverse readership meant that coverage of minority candidates wasn’t also limited to just traits. The study found:

…the likelihood of news stories featuring positive traits dramatically increases when voting-age minority audiences constitute a large portion of the population in the newspaper’s area of circulation…

In terms of figures, the study found that when the market for a particular newspaper ranked in the 10th percentile in terms of diversity (in other words, if the paper had an overwhelmingly white readership), white candidates were almost 5% more likely to receive positive trait coverage than non-white candidates. However, if the readership placed that market in the 90th percentile in terms of diversity, then the tables flipped: non-white candidates were almost 5% more likely to get positive trait coverage.

However, things looked a little different when it came to the effect of newsroom diversity on trait coverage of political candidates. When the number of non-white staff in a newsroom increased, it didn’t correlate with a decrease in trait-based coverage of non-white political candidates.

Specifically, the researchers found that:

For newsrooms ranking among the 10th percentile, the predicted probability of receiving positive trait coverage for non-white candidates was about 3.12% higher than that for white candidates. For news media ranking among the 90th percentile, the predicted probability of receiving positive trait coverage was 4.86% lower for non-white candidates than for white candidates.

That was not what the authors expected, said Paul, who teaches at UNT’s Mayborn School of Journalism.

“We were thinking that if you have more minority journalists in the newsroom, then the trait coverage of minority candidates will be positive, that they will receive more coverage of positive traits and so on,” said Paul. “But surprisingly, we found that when you have more minority journalists in the newsroom, it is white candidates who get more positive, great coverage,” she added.

Paul and Sui also conducted a robustness check, where they looked to see if controlling for the competitiveness of a given race, the presence of a non-white supervisor in a newsroom or whether political candidates were incumbents influenced favorability of a candidate’s trait coverage. They found that these factors didn’t significantly affect whether white candidates were portrayed more favorably by more diverse newsrooms.

Joy Jenkins, an assistant professor of journalism and electronic media at the University of Tennessee who was not involved in the study, said that this last finding underscores what many DEI advocates already know: “Despite heightened conversations about the political, social, and economic benefits of ensuring newsrooms represent their local communities, internal barriers for minority journalists remain, as well as the persistence of biases in political reporting,” she shared.

Still, what was heartening, Paul said, was that there weren’t that many articles that included a lot of trait coverage of candidates. “I think it’s good that coverage is moving away from that trait-based coverage because we know that voters often just rely on shortcuts like that. They’ll read up on what a candidate is like or what their characteristics are like, and then they will use that as implicit cues to determine whom they should vote for,” she explained, versus using information about a candidate’s policies to make voting choices.

As far as takeaways, Paul said that the study reinforces the idea that just simply filling newsrooms with more minority reporters isn’t a simple solution to addressing systemic biases and related issues in journalism. “There’s a need to move away from that tokenism,” Paul said. It’s not enough to fill quotas for reporters from certain ethnic backgrounds, she said, because what’s important to consider is, “Are these journalists able to use their full potential to make the kind of changes that audiences would want and are they able to change the content as much as they need to?”

For the people who have been engaged in diversifying newsrooms in more ways than just hiring underrepresented talent, the results should be no surprise, Paul said. “The people who are really invested and trying to bring about a change in journalism understand that one person can’t change how journalism is practiced.”

Jenkins adds that another takeaway is perhaps that “Newsrooms should remain aware of and engaged with their communities, consider what approaches to political coverage will best serve them, and offer sufficient insights into candidates and their platforms to enable public engagement.”

Read the full study here.