Our beliefs about political issues are not independent; they are embedded within interconnected belief systems. Models such as the Networks of Beliefs (NB) theory formalise how individuals adjust their beliefs to reduce dissonance, and study how this leads to macro-patterns like polarisation or consensus. In these models, the connections between beliefs are static, and the same belief network structure is assumed across individuals. This overlooks that perceived dissonance and the belief connections can differ across individuals. External events or belief interventions, such as economic shocks or political scandals, may trigger lasting changes in some individuals but only temporary or no change in others. We extend NB theory by allowing belief networks to evolve over time, shifting the focus from the dynamics of belief content to the co-evolution of content and structure. In our simulations of our model, individuals start with identical belief networks, but their structures diverge, producing stable disagreement between individuals. When we intervene on a single belief over a period of time, some individuals adapt their beliefs in line with the intervention, whereas others remain in opposition or conform only temporarily and revert to their prior beliefs after the intervention. Our model offers an illustrative and endogenous explanation of such asymmetries in dissonance and belief updating without requiring traits such as stubbornness, motivated cognition, and identity.

Please note that this talk is intended for internal participants only. If you are an external guest and would like to join, kindly send an email to .