Making Useful Conflict Predictions: Methods for Addressing Skewed Classes and Implementing Cost-Sensitive Learning in the Study of State Failure

One of the major issues in predicting state failure is the relatively rare occurrence of event onset. This class skew problem can cause difficulties in both estimating a model and selecting a decision boundary. Since the publication of King and Zeng’s (2001) study, scholars have utilized case-control methods to address this issue. This paper re-analyzes the landmark research of the Political Instability Task Force (Goldstone et al. 2010), comparing the casecontrol approach to several other methods from the machine learning field and some original to this study. Case-control methods are outperformed by almost all of the alternatives. A multilevel model on the raw data performs best. The article also introduces cost sensitive methods for determining a decision boundary. This explication reveals problems in the Task Force’s formulation of a decision boundary and suggests methods for making useful predictions for policy.