Abstract/Summary

What shapes the level of civilian targeting in counterinsurgency campaigns abroad? While practitioners do not explicitly address this important question, political scientists produced competing explanations pointing to independent variables like regime type, military organizational interests, military culture, force structure affecting intelligence collection, the image of the enemy, and the military threat posed by insurgents. After assessing the explanatory power of current scholarly explanations, in this work I build a typological model that can account for spatial and temporal variations in the level of civilian targeting in counterinsurgency abroad. My model includes two causal factors: the external threat environment as perceived by the incumbent state and the local alliance strategy of the incumbent itself. Since each of my causal factors can assume two values, my model includes four possible combinations of independent variable values. Each combination – or type – is a scenario in which the incumbent is likely to use a specific level of indiscriminate violence by way of a specific causal logic. As a result, my work does not produce an overarching explanation indistinctly applicable to the whole population of cases; indeed, my model includes as many explanations and causal pathways as the number of types; each type can only explain a group of cases in the population. In this way, my typological model acknowledges the complexity of a phenomenon like civilian targeting, while still allowing for contingent generalizations. On the empirical level, I show and test my model against competing arguments by studying four cases: British counterinsurgency in Kenya and Cyprus, French counterinsurgency in Algeria, and German counterinsurgency in South-West Africa (SWA). I conclude by highlighting the policy prescriptions stemming from my model.