A Decision-Support Framework for Food Bank Operations: Overcoming Waste and Uncertainty in Food Distribution

Problem definition: Food banks face the complex operational challenge of allocating uncertain and per- ishable food donations to maximize community benefit. This requires navigating difficult trade-offs between effectiveness (delivering nutritious and fresh food relative to need), efficiency (maximizing food distributed and minimizing waste), and equity (ensuring fair distribution), a task for which traditional myopic poli- cies are often ill-equipped. Our research addresses how food banks can design practical, forward-looking allocation strategies in this dynamic and challenging environment. Methodology/results: We develop a comprehensive dynamic programming framework to model the food bank allocation problem. From this, we derive two practical and implementable policies: a simple, static Scaled Fluid (SF) policy and an adaptive Re-Optimized (ReOPT) policy. We provide rigorous theoretical proofs that both policies are asymptotically optimal. Furthermore, extensive simulations using both synthetic and real data show that our proposed strategies yield significant utility gains. Managerial implications: This research provides food bank man- agers and decision-makers with theoretically sound and empirically validated tools to improve their opera- tions. Our framework offers a clear path to enhance the impact and sustainability of food distribution efforts, enabling organizations to better serve their communities by making more efficient, effective, and equitable decisions with the resources they have.