"A Strategic Logic of Attacking Aid Workers: Evidence from Violence in Afghanistan," coauthored with Neil Narang, International Studies Quarterly, vol.61, no. 1 (2017), 38-51.
Why do armed groups ever direct violent attacks against humanitarian organizations? While scholars have analyzed wartime violence against civilians, little research exists on violence against other noncombatants, like humanitarian organizations. Violence against aid workers, however, is common in wartime, with devastating consequences for civilians, who suffer when aid organizations respond by reducing services. This article argues that much of the violence against humanitarian organizations is strategic. By serving as substitute providers of public goods, aid organizations can bolster the government. Insurgents thus target aid workers in an effort to force them out of particular regions, undermining government support. To test this argument, we analyze variation in violence across space and time using an original panel dataset on attacks against aid workers in Afghanistan, 2008–2012. Despite aid organizations’ attempts to remain neutral, we find evidence that insurgents strategically target aid workers in areas where their services likely strengthen government support.
"Regulating Militias: Governments, Militias, and Civilian Targeting in Civil War," Journal of Conflict Resolution, vol.59, no.5 (2015), 899-923.
In nearly two-thirds of civil wars since 1989, governments have received support in their counterinsurgency operations from militias. Many scholars predict higher levels of violence in conflicts involving pro-government militias because governments are either unable or unwilling to control militias. This article challenges this view, arguing that governments can and do often control militia behavior in civil war. Governments make strategic decisions about whether to use violence against civilians, encouraging both regular military forces and militia forces to target civilians or restraining regular military forces and militia forces from attacking civilians. In some cases, however, government and militia behavior differs. When a militia recruits its members from the same constituency as the insurgents, the militia is less likely to target civilians, as doing so would mean attacking their own community. Statistical analyses, using new data on pro-government militia violence in civil wars from 1989 to 2010, support these arguments.
Although scholars have focused primarily on transnational terrorism, much of the terrorism occurring worldwide is domestic terrorism carried out by rebel groups fighting in civil wars. This article examines variation in terrorism across civil wars, asking why some rebel groups use terrorism, while others do not. Rebel groups make strategic calculations, assessing how their government opponents and their own civilian constituencies will react to terrorism. Rebel groups challenging democratic governments are more likely to use terrorism, believing that their opponents will be sensitive to civilian losses and, therefore, likely to make concessions in response to violence. Rebel groups also consider the costs of violence, which depend on characteristics of the rebel group’s civilian constituency. Rebel groups with broad civilian constituencies select lower-casualty civilian targets to minimize public backlash. Evidence from a new data set on rebel-group terrorism in civil wars from 1989 to 2010 provides support for these arguments