Of midterms, mayors, and complex political identity

As American midterm elections quickly approach, conventional narratives no longer fully explain who votes how. “The phenomena of Obama voters who switched to Trump in 2016, or the mayoral race in New York City, which was won by an underdog, demonstrate a deeply complexifying identity landscape that can lead to political upset,” says SFI Resident Faculty member Brandon Ogbunu, also an associate professor of ecology and evolutionary biology at Yale University. The unpredictable effects of identity on voting have real consequences for societal inequality. Elected politicians make decisions about who receives benefits and redress in critical areas like healthcare, education, and criminal justice. “The fact that people change who they are and change how they feel manifests in a real outcome that dictates the shape of democracy,” says Ogbunu. As a first step to developing a new theory of “Complex Political Identity,” Ogbunu co-organized a working group at SFI from August 10–11, 2026 with SFI Research Fellow Andrew Stier. Stier and Ogbunu met during SFI’s 2024 Complexity Global School for Emerging Political Economies in Bogota, Colombia. Soon they were mapping Stier’s data about implicit racial bias in 140 American cities onto voting records. This revealed patterns that didn’t match conventional political wisdom: places with low racial bias that leaned conservative, and high racial bias that leaned progressive. The findings were published in 2025 and inspired the working group. Ogbunu and Stier, who recently concluded a Complexity Postdoctoral Fellowship at SFI and is now an assistant professor of behavioral science at the University of Chicago Booth School of Business, convened experts from fields like genetics, complex systems, psychology, anthropology, political science, and neuroscience. These working group participants discussed questions including how we think about groups and their members, how coalitional groups form, and how institutions and identities shape each other. “What became clear over the meeting is that there’s a lot of work on identity, spread across many different fields. But more often than not, all that work stays separate and disconnected. We don’t know how to reconcile mechanisms at different scales — the micro-psychological all the way up to the institutional — to make better sense of the multi-dimensional facets of identity,” says Stier. Speaker presentations particularly challenged the narrative of a highly polarized binary political identity of either Democrat or Republican. “It’s rare that you don’t have a multidimensional identity,” Stier says. “We discussed how tangible kitchen-table considerations — things that you’d talk about over the kitchen table with your family that have an impact on your day-to-day life, rather than policy or ideological concerns — complicate the two-party picture in the U.S.” Eventually, Ogbunu and Stier hope to expand beyond politics to develop an empirically testable general theory for all complex social identity. This working group received support from SFI’s Emergent Engineering theme (which is sponsored by the Robert Wood Johnson Foundation Grant #81366) and the Emergent Political Economies program.