Abstract for: A practical framework for using SD models to make allocation decisions in crisis situations: The Agropastoralist Policy Lab

System Dynamics models serve as tools for policy analysis. Ideally, a System Dynamics project begins with identifying reference modes of behavior. However, in cases where quality time-series data is lacking parameter estimation becomes challenging. This is particularly important when using a model to analyze policy tradeoffs, as model parameters impact system behavior, and accordingly policy responses. We provide a practical framework for overcoming parameter uncertainty to provide rapid, meaningful, and transparent policy insights from System Dynamics models. An equilibrium-based approach is proposed to circumvent the credibility issues associated with historical base-runs which cannot be externally and quantifiably verified. Model architecture, sensitivity analysis, and stochastic optimization are combined to manage parameter uncertainty and facilitate policy analysis for robustness rather than optimality. Our framework enables rapid scenario testing in case parameter assumptions are challenged by new data or expert advice, and prioritizes uncertain parameters for efficient optimization. Through a case study on drought impacts on Agropastoralist households of Somalia, we demonstrate the effectiveness of our approach in informing policy decisions. Insights are presented as preliminary, experimental, yet defensible and time-savvy. Our framework can help modelers efficienctly provide results under time-pressures of humanitarian crisis contexts, while further data collection and analysis is pending.