Kaldor's ‘new wars’ argument has stimulated a forceful debate over the last decade. Albeit providing important insights, this debate has messily conflated arguments, concepts and theories. As a result, when it comes to enhancing our understanding of contemporary armed conflicts, it is bringing diminishing results. In this article we suggest avenues for further research and point to current research programs that may help put the debate ‘back on track’ and push the discussion forward by examining systematically some of the aspects Kaldor described as defining features of ‘new wars’. Concretely, we stress the importance of undertaking a longer historical perspective for drawing inferences; call for bringing warfare in the study of civilian victimization; highlight the conceptual differences between ‘new wars’ and civil wars and emphasize the importance of taking their transnational dimension seriously; make a plea for disaggregation for capturing relevant temporal and spatial variation; and draw attention to new data sets and the opportunities they offer to statistically test the ‘new wars’ argument. In a concluding section, we broadly outline the significance of this discussion for policy making.
Cosmopolitanism, political authority, and human security, policy frames recommended by Kaldor, might be a good guide for national and international intervention. However, to be effective, these ‘master frames’ need to lead to more concrete and differentiated policy instruments that reflect the diverse realities on the ground.
If policy makers are to avoid interventions that unintentionally trigger violence against civilians, a robust mandate to protect civilians needs to be accompanied by an understanding of the diverse ways in which wars are fought, the strategic aims that armed groups pursue in each type of war, the role that civilians may play in the achievement of these ends, and the behavioural constraints and heterogeneous motivations that lead armed groups to victimize civilians.
Armed groups’ organizational structures and behavioural patterns exhibit great variation. Policy makers should take these variations into account in order to be able to distinguish among types of groups, e.g. according to their desire and ability to develop effective systems to govern civilians, and reflect these differences in their policies and interventions. Armed groups that take seriously the task of providing welfare to civilians should be targeted differently to those who exhibit a more coercive and predatory behaviour.