Empirical Modeling of Food Demand and Characterizing Donor Behavior During a Natural Disaster

Humanitarian relief agencies have become very important to society today because of the role they play both during their normal operations and under emergency conditions. One challenge that these agencies face on a daily basis is to solicit donations amid severe competition from other agencies, all targeting a limited number of potential donors. Although studies have been done to characterize a donor behavior during normal operations, very little has been done to characterize the behavior during natural disasters. The first part of the thesis develops a model to predict whether a person will donate or not donate to a humanitarian agency during natural disasters. On the other hand, the nature of the service delivery by these agencies brings on board a lot of uncertainties in the supply and demand for their services. Inventory levels, maintenance activities, logistics, warehouse and storage facilities are all affected by these uncertainties, increasing the operating costs for the agency. However these uncertainties can be minimized if the demand can be forecasted with some level of certainty. This study identifies key factors that can be effectively used to predict the demand for the services of the Food bank of Central and Eastern North Carolina. The effect on the demand for their services when these indicators change is also investigated.