Empirical Indicators of Crisis Phase in the Middle East, 1979-1995

The article uses several statistical techniques to identify and analyze phases in an event dataset measuring the political behavior among eight Middle Eastern actors -- Egypt, Israel, Jordan, Lebanon, the Palestinians, Syria, the United States, and USSR/Russia -- for the period July 1979 to June 1995. Consistent with narrative accounts, factor analysis identifies the Israeli-Palestinian conflict and the war in Lebanon as the two most important features of the data. Other factors reflect inter-Arab and major-power relationships. Discriminant analysis can distinguish a set of human-coded phases with about 90 percent accuracy. Stepwise discriminant achieves 70 percent accuracy using data from only 12 of the 54 directed dyads. Finally, K-means cluster analysis identifies five distinct phases that align fairly well with the human-coded phases, particularly in the first half of the time period. The dataset consists of WEIS-coded events generated from Reuters lead sentences downloaded from the NEXIS data service, then coded using the Kansas Event Data System (KEDS) machine coding system. One file provides the raw event data (approximately 110,000 events). A second tab-delimited file aggregates these events into dyad-months using the Goldstein (1992) scaling. The KEDS coding dictionaries and the program used to generate the data are also included in the ICPSR Publication-Related Archive release. Additional information can be obtained from the World Wide Web at http://www.ukans.edu/~keds.