The New Rivalry Dataset: Procedures and Patterns

Abstract

The 1990s saw the emergence of a new research agenda focused on enduring rivalries, longstanding competitions between the same pair of states. The original Diehl & Goertz dataset on international rivalries has been perhaps the most widely used collection to study those rivalries. Here, that dataset is extended through 2001, and additional criteria beyond the time-density approach are used to define a population of rivalries. In the first half of the article, the conceptual and operational bases on which the original rivalry collection was based are described. The article explores each of the dimensions of the rivalry concept and the associated operational criteria. The ‘linked conflict’ dimension of the rivalry concept is made more explicit in the discussion of rivalry dimensions. The article then presents and discusses all the major changes made vis-‡-vis the earlier rivalry collection. In the second half of the article, empirical analyses highlight the conceptual dimensions of rivalry. Particular attention is devoted to the issue of rivalry symmetry, with an investigation of rival power capabilities. In an analysis on the linked conflict dimension, the article examines war occurrence and sequence in rivalry (most of which occurs at or near the outset of the rivalry) as well as the outcome and waiting times between disputes. The article concludes with a comparison of this dataset to another prominent rivalry collection.