A network-based microfoundation of Granovetter's threshold model for social tipping

TL;DR

This paper refines Granovetter's threshold model using network theory to better understand social tipping points, highlighting how social interactions can lead to large-scale collective action and addressing previous modeling limitations.

Contribution

It introduces a network-based refinement of Granovetter's model that predicts realistic collective behavior and explains social tipping phenomena through analytical derivations.

Findings

Broad threshold distributions emerge from social interactions.

Social tipping occurs as saddle-node bifurcations and hysteresis.

Refined model aligns with real-world social movement observations.

Abstract

Social tipping, where minorities trigger larger populations to engage in collective action, has been suggested as one key aspect in addressing contemporary global challenges. Here, we refine Granovetter's widely acknowledged theoretical threshold model of collective behavior as a numerical modelling tool for understanding social tipping processes and resolve issues that so far have hindered such applications. Based on real-world observations and social movement theory, we group the population into certain or potential actors, such that -- in contrast to its original formulation -- the model predicts non-trivial final shares of acting individuals. Then, we use a network cascade model to explain and analytically derive that previously hypothesized broad threshold distributions emerge if individuals become active via social interaction. Thus, through intuitive parameters and low dimensionality our refined model is adaptable to explain the likelihood of engaging in collective behavior where social tipping like processes emerge as saddle-node bifurcations and hysteresis.