Biased Foundations of Macro-Social Order: Modelling How Micro-Level Cognitive Bias Shapes Aggregate Dynamics

This session addresses a core problem in micro-macro modelling: how systematic individual-level cognitive biases aggregate into macro-social dynamics, rather than cancelling out as random noise. While classical transformation mechanisms often assume idiosyncratic deviations from rationality wash away in aggregation, recent computational work shows that biases such as status quo bias and confirmation bias propagate through interaction and can be endogenously amplified — generating path dependence, lock-in, hysteresis, and persistent polarization that rational-baseline models cannot reproduce.

The session brings together theoretical, methodological, and empirical contributions that explicitly formalize the link from biased cognition to aggregate structure and dynamics. We welcome agent-based models, analytical and dynamical-systems approaches (including attractor-landscape and hysteresis-based treatments), and empirically calibrated simulations, alongside conceptual papers interrogating when bias amplifies versus cancels. Contributions that quantify the divergence between biased and rational baselines, that model bias as state-dependent rather than fixed, and that connect formal models to substantive domains — opinion dynamics, institutional inertia, inequality, technology adoption, collective decision-making — are especially encouraged.

Format: this is an open-call paper session. Following ISA convention, it will feature four to five presentations of roughly fifteen minutes each, organized to move from theoretical and methodological foundations toward applied and empirical cases. A discussant will offer integrative comments, followed by open floor discussion. Presenters will be invited to share materials in advance to deepen exchange. The session aims to build a cumulative conversation across modelling traditions, treating biased cognition as a generative micro-foundation for macro-social order and change rather than a residual nuisance.