CONCLUSIONS
I had two aims in this chapter on case studies and conflict resolution: to summarize some of the recent literature on case study methodology, and to examine the role of case studies in the development of theory about crisis management.
With respect to case study methodology, I suggested a modification in conventional typologies of case studies, and focused on the theoretical purposes for which case studies are designed to serve. I also examined alternative research designs for facilitating use of case studies to test theories. I emphasized that different case study designs are more or less useful depending on the theoretical aims of the study, and that many designs are best conceived as stages in an overall research program. While I accept the conventional wisdom that for many theoretical purposes many cases are better than fewer, as long as the cases all satisfy the theoretical criteria guiding the study, I argued that for the purposes of hypothesis testing as well as hypothesis construction, a small number of cases or even a single case study can be extremely useful. With respect to hypothesis testing, this is particularly true for hypotheses that posit necessary or sufficient conditions or for situations in which cases satisfy “most-likely” or “least-likely” criteria based on theoretical priors.I then turned to case studies of crisis management, with particular attention to George's (1991a) provisional theory of crisis management and its application to the July 1914 crises and to the Cuban missile crisis.
I argued that World War I provides a useful reminder that crises can escalate to war not only because of the failure of crisis management, but also because the structure of preferences and international and domestic constraints sometimes create few incentives for leaders to try to manage the crisis. I argued that the Cuban missile crisis is a classic case of a crisis that could have escalated out of control but that was successfully managed by political leaders.