Predicting Conflicts

Jour Fixe talk by Sebastian Schutte on November 4, 2015

A temporary decline  in the number of ongoing conflicts around the world after the end of the cold war might be coming to an end: There are continuing episodes of conflict in Sub-Saharan Africa and South Asia. The Greater Middle East also has seen a major escalation: Both Afghanistan and Iraq have experienced more than a decade of civil war following the American-led invasions. Libya, Egypt, and Yemen are also experiencing civil conflicts following the 2011 Arab Spring; and finally the initially peaceful uprising against Bashar al-Assad in Syria has escalated into the most lethal conflict the region has seen in modern times.

But what keeps these conflicts going, what causes civilians to support the military actors? And could we have seen this coming? These are question that Sebastian Schutte tries to answer in his research. In his presentation on “Explaining and Predicting Large-scale Violence in Civil Conflicts” he gave some examples of how he approaches such questions.

“Scholars and practitioners agree that civilian loyalties play an important role in civil wars, particularly in insurgencies”, he stated. “Moreover, the counterinsurgency theory assumes that violence drives these loyalties. However, the way in which violence drives loyalties is contested.”

There are two general explanations: Deterrence-based explanations assume that higher quantities of violence against the insurgent weakens the insurgency by increasing the collective action problem for the insurgents. It is expected that if one applies more violence, fewer individuals will work with the adversary. On the other hand, alienation-based explanations assume that only selective violence against the insurgent weakens the insurgency while indiscriminate violence strengthens the insurgency by solving the collective action problem through selective incentives. In this case it is expected that if one applies more indiscriminate violence, more individuals will work with the adversary.

Empirical results show a weak but important effect: Civilians tend to collaborate more with the enemy of the perpetrator after heavy arms are used. This supports the alienation-based explanation and suggests that hawkish rhetoric might backfire.

In an ongoing collaborative work with his colleague Roos van der Haer, Sebastian Schutte tries to get time-variant individual-level data on experiences and hostilities in civil wars. To gather such data, researchers rely on surveys, although they have their own set of problems: face-to-face surveys are costly and dangerous to implement; web-based is often focused on the wealthy countries  and sms-based surveys have low response rates.

For this reason, the political scientist applies an innovative approach: He combines sms surveys with mobile payments (M-Pesa) and implemented a new computer system for “Reimbursed Mobile Surveys” (RMS). So far, the system has been used for a proof-of-principle application in India. “Based on a trial with over 1,000 participants, the system has produced higher response rates for reimbursed surveys than non-reimbursed surveys. Future deployments of the system in Kenya and Afghanistan are planned.