This project contains code for running a coordination game simulation based on David Siegel (2009)'s "Social Networks and Collective Action" in the American Journal of Political Science.
All relevant code is contained in coordination_model.py
, and simulations are
started using the run_coordination_simulation
function.
Returns pandas Series where each entry is share of people coordinated on candidate 1 at end of each simulation run.
The one potential divergence from Siegel's model is that I do not update individual nodes simultaneously. Siegel's writing suggests that all nodes examine behavior in their neighborhood, then update simultaneously. This, I discovered, leads to what I feel is a pathological dynamic that is knife-edge dependent on this simultaneity. The simplest illustration of this is when a network has only two nodes,