Empirical Global Games of Regime Change

2026-07-06

AI Summary

This study addresses the limitations in existing coup research, which often fails to adequately model coordination problems and suffers from a disconnect between multiple-equilibrium game-theoretic frameworks and empirical analysis. The paper develops an estimable global games model, uniquely embedding global games as an equilibrium selection mechanism within an empirical framework for coup analysis. It theoretically decomposes coup viability into two dimensions—“feasibility,” reflecting regime strength, and “attractiveness,” capturing the expected gains from rebellion—and links agents’ beliefs about these dimensions to observable variables. Employing simulated maximum likelihood estimation combined with a contraction mapping algorithm, the model identifies parameters and maps covariates onto regime strength and overthrow payoffs. This approach enables counterfactual analysis to quantitatively assess how coup payoffs, regime resilience, and information quality shape coup outcomes.

Abstract

Global games theory provides a tractable framework for analyzing coordination problems with multiple equilibria, with regime overthrow serving as a canonical application. A large empirical literature on coups d'état examines the relationship between country-level characteristics, coup occurrence, and coup success using reduced-form approaches that leave the underlying coordination problem implicit. Bridging these literatures, we develop an estimable global games model of coups d'état. The model incorporates strategic coordination into the empirical analysis of coups, employing the global games framework as an equilibrium selection device. The model distinguishes between the feasibility and desirability of regime overthrow, allowing observable fundamentals to enter separately into beliefs about regime strength and perceived gains from rebellion. The model therefore provides a theoretical basis for decomposing coup outcomes into feasibility and desirability components under maintained exclusion restrictions and equilibrium assumptions. If information on coup strength is available, the model also allows estimation of agents' uncertainty about regime strength. We show how the model can be estimated using simulated maximum likelihood coupled with a contraction mapping, and demonstrate how observable covariates map into regime strength and the perceived benefits of overthrow. We illustrate how the framework can be applied through counterfactuals varying coup benefits, regime strength, and information quality.

Problem

• global games
• regime change
• coordination problem
• coups d'état
• multiple equilibria

Innovation

• global games
• strategic coordination
• regime change
• equilibrium selection
• simulated maximum likelihood