This article has been reviewed according to Science X's and . have highlighted the following attributes while ensuring the content's credibility:
fact-checked
peer-reviewed publication
trusted source
proofread
Mapping our emotionally divided society: Mathematical model helps explain polarization

In our polarized political system, what's one thing that Democrats and Republicans have in common? Growing distrust and dislike for each other.
New research in PNAS Nexus from the USC Viterbi Information Sciences Institute (ISI), the University of Iowa, and Claremont Graduate University sheds light on how the presence of this growing emotional divide鈥攌nown as "affective polarization"鈥攕hapes individual decision-making across society. The work could help inform new strategies for reducing the disagreement on divisive but socially important issues like getting vaccinated.
"We became interested in this topic while trying to understand how quickly opinions diverge," explains ISI Principal Scientist Kristina Lerman, who oversaw the study. "We all know that issues get politicized, but why do people start believing completely opposite views?"
The research team, led by Buddhika Nettasinghe of the University of Iowa, developed a mathematical model to simulate how people make decisions in an emotionally divided society. In the model, individuals like and trust members of their own group, parameterized as "in-group love," and dislike and distrust members of the other group, parameterized as "out-group hate."
Importantly, this social and emotional dynamic drives their behavior: when faced with a binary choice, like whether to wear a mask or get vaccinated, people follow the actions of their in-group and reject the choices made by the out-group.
By experimenting with different parameter combinations, the model revealed several insights about polarization. "We can mathematically show that a system can reach complete consensus or split along party lines, where one group believes one thing and the other group believes the opposite," Lerman said.
Researchers found that the existence of out-group hate within a network is the key driver of polarization. When out-group hate is bigger than in-group love, there's no consensus possible, Lerman said. Additionally, while in-group love is necessary for consensus, it's not enough鈥攖he size of the in-group must be sufficiently larger than the out-group in order to avoid polarization.
Surprisingly, the model revealed that increasing cross-party connections often accelerates polarization rather than reducing it. This contradicts conventional wisdom that breaking out of echo chambers helps reduce division.
Instead, researchers found that when people from opposing groups interact more, their choices tend to be shaped less by personal preference and more by a desire to reject whatever the other side is doing. Even in cases where there was initial consensus between groups, the heightened awareness of partisan differences drove them apart.
The model shows that when people from opposing groups interact more, they often become more aware of their differences, which can trigger stronger negative reactions. Instead of bringing people together, exposing them to opposing views can actually push them further apart, especially when they already strongly dislike the other group.
"For every action, there's an opposite reaction," Lerman said. "As soon as you start seeing what your adversaries believe, you react against them."
These findings may help explain why seemingly apolitical decisions鈥攕uch as what to drink, which car to buy, and who to root for in the Super Bowl鈥攁re increasingly aligned with political identity in affectively polarized societies.
Looking forward, Lerman and team are focusing on how to mitigate corrosive emotional divides. They believe that news media and social platforms could play a role in reducing polarization by downplaying the emphasis on partisan divides. For example, adjusting network structures on social media to connect people from opposing parties through common interests rather than divisive topics could help reduce the reactionary mechanisms that drive polarization.
The study ultimately underscores the outsized role that negative emotions play in shaping social dynamics and decision-making across partisan lines. "Negativity is very powerful," Lerman said. "We are showing just how powerful it really is."
More information: Buddhika Nettasinghe et al, How out-group animosity can shape partisan divisions: A model of affective polarization, PNAS Nexus (2025).
Journal information: PNAS Nexus
Provided by University of Southern California