Religion and Conflict: Explaining the Puzzling Case of “Islamic Violence”

Despite considerable interest in the relationship between Islam and political violence, there is little systematic empirical research that explores the intrastate conflict proneness of Muslim countries, and existing studies provide mixed results. This article examines the causal factors that explain the prevalence of intrastate conflict in Muslim-plurality states and the conditions under which Islam may influence civil war onset. Further, following Ward, Greenhill, and Bakke’s (2010) suggestion, the effects of Islam and other socioeconomic and political factors in actually predicting civil wars are examined by utilizing ROC curves and cross-validation exercises. Utilizing the Uppsala Conflict Data Program’s data for the 1981--2009 period, the findings indicate that Muslim-plurality countries are indeed disproportionately involved in intrastate conflicts, but these countries are also characterized by lower GDP per capita, oil dependency, state repression, autocracy, and youth bulges, all of which correlate strongly with intrastate conflict onset. The significance of Islam disappears when controls for such factors are included in the statistical model. The variable of Islam does not make any significant contribution to either the in-sample or out-of-sample predictive power of models. Among the factors that increase the risk of intrastate conflict, the presence of a youth bulge has the greatest impact in Muslim-plurality countries.