The Ineffective Power of Assumptions: Why Structural Modeling Doesn’t Fly in IR
摘要
This paper critiques Kenkel and Ramsay’s (in J Politics. https://doi.org/10.1086/732953, 2024) “The Effective Power of Military Coalitions,” arguing that the study falls short of structural modeling standards despite its formal presentation. While adopting the aesthetics of game-theoretic analysis, their model simplifies coalitions to groups of independent actors and overlooks the interconnected nature of coalition warfare. This simplification commits what Tsebelis (in Am Polit Sci Rev 83(1):77–91, 1989) called the “Robinson Crusoe Fallacy”: assuming strategic independence when mutual responsiveness is essential. Their simulations do not re-optimize behavior as conditions change but instead reuse fixed parameters, resulting in calibrated, comparison-based results that are mistaken for structural inference. The model’s empirical fragility and theoretical rigidity weaken its explanatory power, especially when compared to historical examples of coordinated mobilization during World War I and World War II. As an alternative, I propose a theory-driven simulation framework that explicitly includes strategic interdependence. This approach preserves analytical clarity while recognizing the contingent and collective dynamics that shape military coalitions.
Graphical Abstract