Analysis and forecast of donations at domestic hunger relief organizations impacted by natural disasters

Abstract

Food banks are non-profit, charitable organizations that distribute food and products to people in need. Food bank facilities become disaster relief centers after natural disasters. The uncertainty associated with the arrival of donations and demand make the planning and operations of food banks challenging during the disaster relief period. The goal of this research is to analyse and forecast the amount of donations received by food bank facilities in the U.S. when operating as disaster relief centers. This paper analyses the donations received by two food bank facilities affected by Hurricane Harvey in 2017. An extensive numerical study is performed that compares the donation behavior at each facility before and after the hurricane event. Multiple forecasting models are evaluated to determine their accuracy in predicting the observed behavior. The results show that under disaster operations, the best performing techniques for both food banks were smoothing techniques (i.e., CMA and Holt) and econometric models.