Using Chyrons To Understand Diversity In Television News

When television news channels turn to outside experts to interview on the major stories of the moment, who do they turn to? In a time of increasing awareness of the lack of diversity on television news, of particular concern are questions of diversity and representativeness of who television news channels give a voice to. Who are the voices that are given the opportunity to tell their stories, what communities have their voices heard in how major stories are affecting them and who are the outside experts channels turn to to offer expert commentary and to interpret major events?

In the opening months of the Covid-19 pandemic, we demonstrated how the chyron ("lower thirds") text of television news offers an exceptionally powerful lens onto these questions by giving the name, title and affiliation of each guest they interview. With just a single SQL query it is possible to compile a daily chronology of all of the medical doctors that have been interviewed across ABC, CBS and NBC evening news shows over the past decade or CNN, MSNBC and Fox News since the start of the pandemic, for example.

While the chyron text only gives the guest's name, title and affiliation, this can then be cross-walked against external self-reported demographic data, such as recorded in datasets like Wikidata. Using chyron text to identify the guest and then looking up their self-reported demographic information moves past the fraught ethical challenges of using highly biased AI models to "estimate" the degree to which a speaker's name, visual or spoken presentation conforms to historical demographic "norms" which typically yields extremely wrong and highly offensive results that reinforces harmful stereotypes and instead makes it possible to study critical diversity questions by utilizing the self-reported identities of speakers.

We would love to see further work in this space and would love to hear from those with an interest in these kinds of questions!