Findings

The empirical results of this study will be presented in several stages. First, a descriptive analysis of the data will be provided through a system and geographic distribution for the scale points of the indicators of severity.³⁶ Second, the indicators of severity and importance will be correlated within each index. These exercises are intended to demonstrate that the indices are streamlined and valid, a prerequisite to any analyses that include either overall severity or importance. Third, the overall severity of all 278 crises from 1929 to 1979 will be examined in different system-periods and geographic regions. Fourth, the top 10 percent of the data set along the dimension of severity will be isolated for intensive study. Consisting of 26 cases at the apex of severity, the subset to be examined obviously in interesting (and manageable) for purposes of comparison. This stage of the analysis will describe the distribution of cases in terms of geography, system-period, power discrepancy, and each of the six components of severity. Thereafter the link between severity and importance for this subset of crises is discussed. This is followed by in-depth discussion of two case studies of severity and importance, the Middle East Crisis, 1973-1974, and the Angola Crisis, 1975-1976. The final stage of the analysis will concern findings in the aggregate. A contingency table and ordinary least squares (OLS) regression will be used to test the predictive power of severity with respect to importance.