Riot Networks and the Tullock Paradox: An Application to the Egyptian Arab Spring

If you are interested in joining the seminar, please send an email to Daniel Haerle or Sacha den Nijs.

Abstract:
We study a dynamic model of collective action – for concreteness, we speak of a riot – in which agents interact through, and learn from, a co-evolving social network. We consider two different scenarios. In one of them, conceived as a “benchmark”, agents changing their behavior are assumed to be completely informed of the prevailing state (action profile and network). Instead, in the alternative scenario, agents are assumed to shape their expectations about the state from a combination of local observation and social learning (modeled à la DeGroot). In both cases we provide a complete characterization of the long-run behavior of the system. While the first assumption of complete information is common, the second one is arguably more realistic. We show that the latter assumption leads to an increased probability of segmented network configurations, thus providing a plausible mechanism to understand what otherwise seems a puzzle, i.e. how do very large populations attain (“coordinate on”) collective action. Finally, we illustrate the empirical potential of the model by showing that it can be efficiently estimated for the so-called Egyptian Arab Spring by relying on large-scale cross sectional data on agents’ choices and their network of interactions. Our estimation results indicate that, in that instance of social unrest, both the local peer effect and the global conformity effect played a significant role in activating protest participation.