Standard practice in IR conflict research segregates analyses of civil war from those of interstate war. This is reflected in fundamental research design issues (state-years as the units of analysis in civil war studies, dyad-years in interstate war studies) and also in the frequent assertions by researchers that civil wars are profoundly different from interstate wars. We recognize that civil and interstate wars may differ fundamentally in their causes, courses, and consequences. But since there is no pre-existing theoretical reason to anticipate these differences, it is probably the case that data availability (specifically the absence of data about rebels in most intrastate wars, and particularly about rebels during years of peace) causes the distinction between civil and interstate war studies. We discussion justifications for and against distinct analyses of civil and interstate wars. In the course of this discussion we show how the very claims used to justify separate analyses of civil from interstate wars can be used to justify separate analyses of “types” of interstate wars. We investigate the inferential consequences of making arbitrary distinctions among interstate wars. A second empirical section combines civil and interstate wars in unified analyses of war duration and outcome, looking for distinct patterns between the two types of war. We find in those analyses that variables representing the bargaining conditions across all wars account for a substantial amount of the variation in duration and settlement patterns between civil and interstate wars. We conclude that there is substantial evidence that the division of wars into civil and interstate categories is a distinction without a difference.
Download from SSRN