Mapping Pathways of Ethnic Conflict Onset: Preferences and Enabling Conditions

Abstract

Recent quantitative studies of ethnic conflict tend to have characteristic strengths and weaknesses. Strengths are typical of sophisticated statistical work: the influence of many explanatory variables is examined across large datasets. Weaknesses are largely traceable to shortcomings of theory and measurement. Territorial and other ethnic conflicts are sometimes pooled with other types of internal conflict, which seem likely to have a different range of causes. Also, theory indicates that different conflict pathways have varying probabilities of giving rise to territorial ethnic conflict. These pathways involve different types of elite-level preferences, for either the outsider ethnic group or state side of a potential conflict, and for either incumbent or challenger elites. The pathways are also likely to be conditioned by different constraints. Yet little effort has been made to measure variation along these discrete pathways. More readily measurable variables, such as economic development and dependence on oil exports, often pick up predicted variation from multiple theoretical pathways. Regressions including such broad dependent and independent variables may deflate the estimated impact of discrete-pathway variables on ethnic conflict onset, and hence must be interpreted with care. There are significant practical barriers to measuring variation along the different conflict pathways; but further effort is necessary if the quantitative literature is to become more robust, easily interpreted and helpful in understanding specific cases.