Structural Factors and International Crisis Behavior

We began this analysis with the point that the structural characteristics of collectivities of states have been largely ignored in studies of conflict and crisis. The analysis was aimed at identifying, operationalizing, and examining the effects of a set of such structural factors on process and outcome characteristics of crises. While structural factors alone can hardly be expected to provide full explanations for the behavior of states in crises, our analysis has shown that in several key areas, such factors warrant greater attention on the part of the research community. At this point, it is appropriate to review the key findings for the three structural factors considered here. (1) Triggers of Crises: Our expectation was that the trigger for the entire crisis would set the tone for the manner in which the participants would pursue their goals during the course of the crisis. Specifically, we find that crises which were triggered by violence had a far greater propensity to exhibit violent crisis management techniques than did crises triggered by non-violent acts. Also, crises with violent triggers were more likely than non-violent crises to result in an escalation of tensions, defined in terms of the recurrence of crisis among the parties during the subsequent five-year period. Ten of the international crises with violent triggers fit our profile for the first structural factor. That is, these cases exhibit a violent trigger, violent crisis management technique, and escalation of tensions. The cases are: Israel Independence 1948, Chinese Civil War 1948, Korean War (June) 1950, Congo 1960, Yemen 1964, Gulf of Tonkin 1964, Pleiku 1965, El Samu 1966, War of Attrition 1970, Angola 1975. (2) Group Size: Consistent with the public choice literature on the effects of group size on the outcomes of bargaining and negotiation, it was our expectation that international crises with two actors would exhibit substantially different patterns than would larger groups of crisis actors. More specifically, we found that the larger international crises (in terms of number of crisis actors) tended to be associated with longer duration. These larger crises also exhibited a heightened probability of achieving agreement among the parties. Finally, the larger crisis clusters tended to exhibit a lower tendency toward satisfaction with the outcome. Seventeen international crises involving three or more actors met the profile for the second structural factor: group size of three crisis actors or larger, long duration, agreement; and lack of satisfaction. These cases are: Iran Hegemony 1945, Berlin 1948, Israel Independence 1948, Korea October 1950, Taiwan Straits 1954, Suez 1956, Yemen (1962, May 1964, December 1964, 1966), Greece/Turkey 1963, Czechoslovakia 1968, Pakistan/India/Bangladesh 1971, Vietnam Port Mining 1972, Vietnam Christmas Bombing 1972, Middle East War 1973, Angola 1975. (3) Homogeneity of World View: It was our expectation that Western states would differ markedly from non-Western states in the way they view crises. Accordingly, we felt that international crises involving heterogeneous groups of states (mixed Western and non-Western) would exhibit different outcome patterns than either non-Western homogeneous clusters or Western homogeneous clusters. In fact, we found that the homogeneous Western crises were quite uniform in their characteristics, and differed markedly from the other two types. Further, the heterogeneous crises and the homogeneous non-Western crises exhibited strikingly similar patterns. More specifically, the Western homogeneous crises were less likely than others to exhibit violence as the principal crisis management technique, and were more likely to lead to tension reduction. Seven crisis clusters fit this homogeneous Western profile: USSR/Hungary 1947, USSR/Czechoslovakia 1948, Berlin 1948, Yugoslavia/Italy 1953, Poland/USSR 1956, Berlin Wall 1961, Cod War 1973. Clearly, our work constitutes only the beginning of what should be serious and systematic attention to the identification of the conditions under which structural characteristics of collectivities of interacting states impact on international behavior. Our exclusive focus on international crises makes problematical any generalization of the findings to the entire range of international behavior, or even to the more focused phenomenon of international conflict. Nevertheless, our discovery that the structural factors of crisis trigger, group size, and homogeneity of world view are variously related to such process and outcome characteristics of crises as the use of violence, duration, substance, form, satisfaction with, and durability of outcomes should encourage a wider examination of the role of these and other structural factors.