A Study of Crisis

Study of Crisis. By Michael Brecher and Jonathan Wilkenfeld. Ann Arbor: University of Michigan Press, 1997. 1,088p. $115.00.

The late Karl Deutsch was fond of describing data collection as "the Lord's work" in social science. Expensive, frustrating, tedious, and thankless, it nevertheless is the foundation on which cumulative findings are built. The material in this massive work is worthy of several hosannas, not only for the authors and their student assistants but also for the University of Michigan Press for publishing the more than 1,000 pages. The analysis and findings in this massive book account for only 140 pages, partly because the authors already have published a number of earlier studies based on the data. The bulk consists of a description of the data generated by the authors' International Crisis Behavior (ICB) project, including narrative accounts of 412 crises that occurred between 1929 and 1992.

Each brief narrative of a few paragraphs is followed by a list of sources, and a master table listing beginning and ending dates for each crisis, the participants (including third-party involvement), and the highest level of violence employed. The authors also provide estimates of several other variables, including which state initiated or "triggered" the crisis, the triggering event, the most salient threat perceived during the crisis, the principal conflict management technique employed by each participant, the type of outcome, and each state's satisfaction with the outcome.

Brecher and Wilkenfeld give a brief definition of each key variable, and they admit that some are not as operational as one might wish. Large-scale data collection efforts invariably involve trade-offs between validity and reliability. Often the variables of greatest interest, that is, those with the greatest...