Abstract
Actor-level variations in the amounts of uncertainty have been widely ignored in the growing literature on statistical models of strategic interaction in international relations. In this article, I provide a tool for testing theories about the level of uncertainty in strategic interactions. I show that ignoring potential variations in levels of uncertainty across different cases can be a source of bias for empirical analyses. I propose a method to incorporate this form of heteroskedasticity into existing estimators and show that this method can improve inferences. With a series of Monte Carlo experiments, I evaluate the magnitude and the severity of the bias and inconsistency in estimators that ignore heteroskedasticity. More importantly, the tools developed in this article have many interesting substantive application areas. Examples considered include measuring speculators’ suboptimal behavior tendencies in international currency crises, and capturing varying levels of signaling and Bayesian updating behavior in the recent strategic models of signaling.
Highlights
This paper aims to contribute to the burgeoning literature on strategic models and complement the existing estimators in three ways: first, I propose a method, heteroskedastic strategic probit (HSP), that helps capture heteroskedasticity in strategic models and correct the potential bias and inconsistency when the data generating process violates the homoskedasticity assumption
What happens if heteroskedasticity is not present in the data generating process and the analyst fits a heteroskedastic model? Would this be a source of bias or inefficiency in the estimates? This Monte Carlo experiment assesses the effects of running a HSP model, when the underlying model is homoskedastic with private information specification
The Monte Carlo analyses presented in this paper have shown that, if not controlled for, the presence of heteroskedasticity in the uncertainty parameters of such models can be a source of bias and inconsistency in estimates
Summary
In recent years, increased attention has been given in the international relations literature to testing hypotheses about strategic interaction (Fearon, 1994b; Smith, 1999; Signorino, 1999; Schultz, 1999; Leblang, 2003; Signorino and Tarar, 2006; Bas, Signorino and Walker, 2008; McLean and Whang, 2010). In the international conflict literature, for instance, there is a big debate about the effect of regime type on the amount of private information states have in international conflict (Fearon, 1994a; Schultz, 1998, 1999, 2001) By limiting their focus to homoskedastic models, empirical international relations scholars cannot incorporate these possibilities into their research. Scholars have previously noted a lack of attention given to theorizing the variance parameter – the “second moment” – in the empirical international relations literature (Braumoeller, 2006) They emphasized the importance and the potential value of substantively interpreting heteroskedasticity instead of treating it just a nuisance (Downs and Rocke, 1979). I first replicate Leblang’s (2003) analysis of speculative currency attacks using the HSP estimator, and discuss the potential extension of the technique to the recent statistical strategic models of signaling and Bayesian updating
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
