Strategic Interaction and the Statistical Analysis of International Conflict

Abstract

Although strategic interaction is at the heart of most international relations theory, it has largely been missing from much empirical analysis in the field. Typical applications of logit and probit to theories of international conflict generally do not capture the structure of the strategic interdependence implied by those theories. I demonstrate how to derive statistical discrete choice models of international conflict that directly incorporate the theorized strategic interaction. I show this for a simple crisis interaction model and then use Monte Carlo analysis to show that logit provides estimates with incorrect substantive interpretations as well as fitted values that can be far from the true values. Finally, I reanalyze a well-known game-theoretic model of war, Bueno de Mesquita and Lalman's (1992) international interaction game. My results indicate that there is at best modest empirical support for their model.

Adverse selection and growth under IMF programs

“…This selection model incorporates the strategic interaction hypothesized by our theory into the likelihood function to be estimated, which should improve the efficiency of our results and remove any bias due to strategic misspecification (Signorino, 1999;Signorino and Yilmaz, 2003). We argue that a candidate government faces uncertainty about the IMF's response when deciding to seek an agreement with the IMF.…”

“…8 We use the agent error specification of Signorino's (1999) with an interested government. Thus, to model the overall probability of no program (P = 0), we need to employ a partial observability model that accounts for both possibilities.…”

Von Makro zu Mikro: Grundlagen und Perspektiven der Bürgerkriegsforschung

“…Eine rein statistische Analyse zu dieser Annahme kann aber nicht genügen, da eine zeitliche wie räumliche Streuung der Gewaltakte Konsequenz einer detaillierten Kriegsstrategie sein kann, die ein zentrales Kommandos entwickelt hat und bei deren Umsetzung es die möglichen und tatsächlichen Handlungen der Gegenseite berücksichtigt. Für die empirische Umsetzung eines solchen Modells bedeutet dies, dass neben der erwähnten Endogenität der geographischen Faktoren auch die strategische Interaktionen zu berücksichtigen wären, wie dies etwa für die Analyse von Krisenverhandlungsspielen vor allem durch die Arbeiten von Curtis S. Signorino (1999) (Schneider 2014). Angesichts der methodi-…”

Flexible Causal Inference for Political Science

“…A field that was once content to base its conclusions on garden-variety logit and probit results has created or imported methods for modeling strategic interaction (Signorino 1999;Smith 1999;Lewis 2003;c.f. Carrubba, Yuen and Zorn 2007), selection bias (Sartori, 2003;von Stein, 2005;Boehmke, Morey and Shannon, 2006;Chiba, Martin and Stevenson, 2014), split-population or partial-observability models (Xiang, 2010;Braumoeller and Carson, 2011), zero-inflated or rare events data (King and Zeng, 2001;Bagozzi, Hill, Moore and Mukherjee, 2015), network analysis (Dorussen and Ward, 2008;Hafner-Burton and Kahler, 2009;Maoz, 2009;Cranmer, Desmarais and Menninga, 2012), and more.…”