Working Paper: NBER ID: w16152
Authors: Luke N. Condra; Joseph H. Felter; Radha K. Iyengar; Jacob N. Shapiro
Abstract: A central question in intrastate conflicts is how insurgents are able to mobilize supporters to participate in violent and risky activities. A common explanation is that violence committed by counterinsurgent forces mobilizes certain segments of the population through a range of mechanisms. We study the effects of civilian casualties in Iraq and Afghanistan to quantify the effect of such casualties on subsequent insurgent violence. By comparing uniquely detailed micro-data along temporal, spatial, and gender dimensions we can distinguish short-run 'information' and 'capacity' effects from the longer run 'propaganda' and 'revenge' effects. In Afghanistan we find strong evidence that local exposure to civilian casualties caused by international forces leads to increased insurgent violence over the long-run, what we term the 'revenge' effect. Matching districts with similar past trends in violence shows that counterinsurgent-generated civilian casualties from a typical incident are responsible for 1 additional violent incident in an average sized district in the following 6 weeks and lead to increased violence over the next 6 months. There is no evidence that out-of-area events--errant air strikes for example--lead to increased violence, nor is there evidence of short run effects, thus ruling out the propaganda, information, and capacity mechanisms. Critically, we find no evidence of a similar reaction to civilian casualties in Iraq, suggesting the constraints on insurgent production of violence may be quite conflict-specific. Our results imply that minimizing harm to civilians may indeed help counterinsurgent forces in Afghanistan to reduce insurgent recruitment.
Keywords: civilian casualties; insurgent violence; counterinsurgency; Afghanistan; Iraq
JEL Codes: F51; F52; H56; J22; K42; O53
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Civilian casualties caused by international forces in Afghanistan (F51) | Increased insurgent violence (F51) |
| Civilian casualties caused by international forces in Afghanistan (F51) | Increased insurgent violence over the next six months (D74) |
| Civilian casualties caused by international forces in Afghanistan (F51) | Revenge effect (G41) |
| No evidence of short-run effects from civilian casualties (H56) | Ruling out propaganda, information, and capacity mechanisms (D83) |
| Civilian casualties in Afghanistan (H56) | Reduced insurgent recruitment (H56) |
| Out-of-area events (errant airstrikes) (H84) | No increase in insurgent violence (F51) |
| Civilian casualties in Iraq (H56) | No similar reaction to insurgent violence (D74) |