Merging algorithms and the human element, Missouri School of Journalism researchers are revolutionizing news desert research
COLUMBIA, Mo. (July 8, 2025) — One of the foremost challenges facing the news industry today is the spread of news deserts, which are areas of the country not adequately served by local journalism. While large-scale studies have helped identify the most obvious deserts with access to little or no news coverage, more nuanced but still-plentiful examples — communities that receive crime coverage, for example, but much less about community events or sports — often fall through the cracks.
Over the last two years, the Missouri School of Journalism has partnered with the MU Institute for Data Science and Informatics (MUIDSI) to develop an algorithmic approach that brings greater accuracy, depth and efficiency to the process of identifying news deserts. That work, led by Knight Chair in Journalism Innovation Damon Kiesow, has now entered a new phase: researchers are going out into the field to compare the data collected by the automated system with the firsthand thoughts, perceptions and preferences of news audiences.
Hannah Artman, a postdoctoral research fellow at the School of Journalism, spent this spring conducting a statewide survey of news consumers to examine their relationship with news. From in-person interviews to an online survey, the work is ongoing, but Artman is already gaining valuable insights into a hierarchy of news needs that can sometimes appear contradictory.
“It’s interesting because in news audience research, there is almost this cognitive dissonance where people say the news is too sensationalist and it’s too negative, but then they say, ‘I love to see what the arrests are’ and ‘I need to see which areas of town to avoid.’ We need to figure out how to ask the right questions — what are the news topics that help people get around in the day-to-day and affect them personally, and how do we find common ground in how to talk about those topics?”
Hannah Artman
“It’s interesting because in news audience research, there is almost this cognitive dissonance where people say the news is too sensationalist and it’s too negative, but then they say, ‘I love to see what the arrests are’ and ‘I need to see which areas of town to avoid,’” said Artman, who earned her doctorate from the University of Miami in 2024 and has professional experience in public opinion research. “We need to figure out how to ask the right questions — what are the news topics that help people get around in the day-to-day and affect them personally, and how do we find common ground in how to talk about those topics?”
Fittingly in a field that is all about communication, Artman said some apparent contradictions might not be contradictions at all, but could in fact be rooted in differences between how journalists and audiences group certain topics together.
For the past decade, one tool in journalism audience research has been the “critical needs framework,” which in this case organizes and prioritizes various types of news content based on what audiences consider most useful and important — much like the way a food pyramid organizes dietary needs. To use that framework effectively, researchers have to look beyond the surface-level meanings of words like “transportation” or “civics,” for instance, to understand what sort of information people actually value within a given category.
“People talk about traffic. They don’t talk about transportation systems,” Artman said. “So when they say transportation systems don’t apply to them, that doesn’t mean they don’t want to know about traffic.”
“There is some interest in thinking a bit differently about geography, about where people consume news and how that might shape their sense of what’s important. This project has been a really interesting opportunity to complicate the idea of locality and understand what that means in people’s day-to-day interactions with local news.”
Joy Jenkins
And if a critical needs framework is like a food pyramid, then determining whether a tomato is a fruit or a vegetable means little if the public isn’t on the same page.
“It’s not enough to know whether a place is a news desert or not,” Artman said. “It shouldn’t just be this binary classification. When we define communities, we’re usually using arbitrary guidelines like congressional districts, zip codes, media markets. But what happens, for example, to people who live in one area but work in another? We need to understand the more complex processes that make up news diets individually and collectively.”
Indeed, the influence of geography is yet another crucial aspect of the research. Joy Jenkins, an associate professor and collaborator on the project, said that physical proximity of one’s residence to a newspaper’s range is just one of many geographical considerations; work, leisure activities and political affiliations all create ties to locations that may or may not be where audiences actually reside, complicating their preferences when it comes to news.
“There is some interest in thinking a bit differently about geography, about where people consume news and how that might shape their sense of what’s important,” Jenkins said, noting that Artman presented preliminary findings at the annual International Communication Association conference in June and met with an enthusiastic response. “This project has been a really interesting opportunity to complicate the idea of locality and understand what that means in people’s day-to-day interactions with local news.”
“This is a feedback loop that helps us inform how we talk and listen to readers. We can feed what we hear back into the computational model, and the conversation between the two approaches refines the model until it’s rigorous enough to tell us the mix of news that audiences or the community both expect and need for a healthy news ecosystem.”
Damon Kiesow
Getting to the core of all these nuances is key to refining an algorithm-driven system that knows what audiences need and can accurately determine whether those needs are being met, providing researchers and the industry at large with a wealth of actionable data.
In partnership with MUIDSI, Kiesow and his team have built, trained and tested such a system, and Kiesow believes perfecting it comes down to uniting two distinct types of research data. With Artman’s help, algorithmically-collected results will be augmented with harder-to-measure aspects of public perception, bridging the gap between quantitative (numerical and measurable) and qualitative (descriptive and non-numerical) research methods and opening the door to news desert studies that capture a more holistic picture of how individual communities interact with news.
“This is a feedback loop that helps us inform how we talk and listen to readers,” Kiesow said. “We can feed what we hear back into the computational model, and the conversation between the two approaches refines the model until it’s rigorous enough to tell us the mix of news that audiences or the community both expect and need for a healthy news ecosystem.”
At a time when audiences continue to consider local news trustworthy even as trust in news overall sits at record lows, Artman, Jenkins and Kiesow believe the time is right to take an audience-centered approach and capture a sense of place that is more than a data point.
Updated: July 8, 2025