Nearly 40% of U.S. cities analyzed in a new study in NPJ Complexity diverge from the common narrative that Republican-dominated areas have high levels of implicit racial bias while Democratic strongholds are more tolerant.
Led by Santa Fe Institute Complexity Postdoctoral Fellow Andrew Stier , the study combines prior research on implicit racial bias — people's unconscious attitudes and beliefs about race — with voting data from the 2016 and 2020 presidential elections across 146 U.S. cities.
Instead of neatly dividing along party lines, the study paints a patchwork picture, revealing 58 cities where political affiliation and racial attitudes reflect local culture and lived experience more than partisan expectation.
"We found many exceptions to the common narrative," Stier says. "This reflects what many people experience. In some Southern cities, you might meet people who are kind to individual people from different racial or ethnic groups, yet still vote for discriminatory policies. Meanwhile, there are places that feel more interpersonally exclusionary but vote for inclusive policies."
Cities like Knoxville, Tennessee, and Spokane, Washington, voted Republican but showed lower-than-expected levels of implicit racial bias, once factors like diversity, segregation, and city population were accounted for. In contrast, Chicago, Illinois, Albany, New York, and Minneapolis, Minnesota, all Democratic strongholds, exhibited higher-than-expected bias levels.
The idea for the study began with an informal conversation between Stier and co-author Brandon Ogbunu , an SFI professor and Yale University evolutionary biologist, at SFI's Compexity Global School last summer program in Bogotá, Colombia. They noticed that many patterns in the data didn't match the common narrative told about the link between racial attitudes and political identity.
"What stood out to me was how closely the data mirrored things I've experienced firsthand — things that don't always show up in national narratives," Ogbunu says. "They reflect the kind of social dynamics that we often sense but rarely see captured in research."
Rather than looking at overall bias levels, the researchers adjusted for city size, diversity, and segregation. By removing these structural effects, they focused on what remained — subtle signals that may reflect a city's cultural norms and everyday social dynamics.
"There's a temptation to tell simple stories about political identity and race," Stier says. "But the truth is much more complicated — and that complexity matters if we want to really understand how attitudes and behaviors shape our democracy."
Next, the researchers plan to extend their analysis using 2024 election data and explore other forms of bias, including gender and age.
"Cities are complex systems," Ogbunu says. "If we want to understand what shapes a place's political and cultural climate, we have to look beyond easy assumptions and start embracing that complexity."
