Ethnic movements – organized political campaigns by ethnic groups to influence state policy – take diverse forms from peaceful protest to large-scale violent mobilization. This excellent book tries to solve the puzzle of why some ethnic movements turn violent while others do not. It nicely bridges the literature on social movements with that on civil wars, arguing that social characteristics of the groups impact not only whether they will revolt but also how they will do it. Reminiscent of Horowitz's classical work on ethnic conflict, Vogt distinguishes different types of ethnic cleavages, defined by the degree of hierarchization of groups (the extent of political and socio-economic dominance) and the degree of social integration (the extent to which groups are embedded in the same socio-economic and cultural institutions). The main predictions are that violence is more likely in segmented unranked societies while non-violent conflict is more likely in stratified non-segmented societies. The breadth and depth of the empirical analyses is impressive: the book combines a rich global large-n dataset, regional large-n studies of Sub-Saharan Africa and Latin America, and qualitative case studies (Côte d'Ivoire and Gabon; Guatemala, and Ecuador). Although well-executed and thought-provoking, the book relies mostly on the demand side (aggrieved actors) and less on the supply side (state actors). Yet, the actions of state (for example, accommodation, exclusion, repression) are likely to have an impact on the outcomes. Also, violent and non-violent mobilization might not be substitutes but complements and they can be interrelated. For example, violence can emerge out of repressed non-violent contention. To observe this would take a more dynamic approach to conflict and escalation, which could complement the structural approach in this book.