'Natural' disaster, conflict and aid allocation

This paper looks into aid allocation in the response to multiple crises, focusing more specifically on the cases of concomitance between so-called 'natural' hazard/disaster and conflict situations. Over 150 natural disasters have occurred alongside complex political crises in the past seven years alone. Yet, the fields of conflict and disaster research remain largely isolated from one another, and in fact, no aid related research has addressed the issue of the concomitance of conflict and disaster. We exploit a large panel dataset that includes official development aid, and information about the victims from natural disasters and conflicts for 112 developing countries over a period of 35 years. For eight different donor countries and groups of donor countries we find that while conflict does not affect their aid allocation patterns, the occurrence of natural disasters does. The econometric analysis demonstrates that aid allocation needs to be analyzed in a disaggregated fashion - for each donor individually - as donors clearly have different agendas. Applying GMM techniques we account for the endogenous nature of the control variables such as per capita GDP. In addition we use the relative size of the youth cohort as exogenous instrument for conflict. -- Disaster ; Conflict ; Aid Allocation ; Longitudinal Panel Methods ; GMM