Collective patterns of social diffusion are shaped by individual inertia and trend-seeking

Social conventions change when individuals collectively adopt an alternative over the status quo, in a process known as social diffusion. Our repeated trials of a multi-round experiment provided data that helped motivate the proposal of an agent-based model of social diffusion that incorporates inertia and trend-seeking, two behavioural mechanisms that are well documented in the social psychology literature. The former causes people to stick with their current decision, the latter creates sensitivity to population-level changes. We show that such inclusion resolves the contradictions of existing models, allowing to reproduce patterns of social diffusion which are consistent with our data and existing empirical observations at both the individual and population level. The model reveals how the emergent population-level diffusion pattern is critically shaped by the two individual-level mechanisms; trend-seeking guarantees the diffusion is explosive after the diffusion process takes off, but inertia can greatly delay the time to take-off. Social convention change due to diffusion is often described by agent-based models focusing on the role of social coordination. In this work the authors uncover two additional individual-level mechanisms, trend-seeking and inertia, that can critically shape the collective behavior of the population.