State Characteristics and Crisis Outcomes: A Test of the Spatial Model

Abstract

Two recent books, Interstate Crisis Behavior, 1816–1980 by Russell Leng and Untying the Knot of War by Cliff Morgan, have focused on the study of international crises. The two studies follow fundamentally different approaches, with Leng following an inductive strategy and Morgan developing a deductive theory, but there appears to be a great deal of overlap in the questions studied and in the variables examined. This research attempts to evaluate each of these previous studies in light of the other. By using Leng's data to test Morgan's hypotheses we hope to discover areas in which the theory requires additional development and areas in which the current data may be an inadequate basis from which to draw inferences. Our results suggest that the theory does need to be extended in certain directions; e.g., to incorporate misperceptions regarding the balance of military capabilities. Furthermore, the results suggest that the current version of the data may not allow valid inferences regarding the impact of crisis structure on crisis outcomes. In particular, the sample size and the biases in the non-random sample selection constitute threats to the validity of any such analysis. The bottom line is that each work contributes to our understanding of international crises, but additional theoretical and empirical work is warranted. The existence of each study enhances our understanding of the other, however, and this exercise produces specific suggestions for the directions of future research.