Socio-dynamic discrete choice: Theory and application

We find in this thesis that careful specification of both observed and unobserved heterogeneity matters critically for emergent temporal outcomes when there is sociodynamic feedback in the model, even when the feedback takes the simple form of an aggregate field variable. Agent heterogeneity impacts the magnitude of the mode shares, the speed of the transition to the steady state as well as very fundamentally the number of possible observable steady state solutions. Also the detailed effect of induced heterogeneity is important to understand in different network structures, including the speed of information flow across them. Misrepresentation of the appropriate scale at which social influence occurs and of the appropriate network structure can yield strongly flawed policy implications when studying social feedback.