Optimizing Policy Decisions in Interconnected Populations:  A Spatially-Dependent Framework for Dynamic Decision-Making

Abstract  Dynamic decision-making is crucial in policy formation. In reality, states or countries typically make these decisions, and their effects are interdependent due to spatial relationships. However, conventional policy decision-making strategies often overlook these spatial effects, assuming no interference among individuals. In this project, we propose a novel framework to measure these interdependent effects across multiple stages. We also propose the methods to estimate the optimal policy regime taking the spatial effects in consideration for a series of time spots, as well as constructing confidence intervals for the coefficients indexing the optimal policy regime based on multiplier Bootstrap methods. We rigorously demonstrate the validity of the proposed estimation and inference methods.We apply the proposed method on a case study on COVID-19 school closing policies. By applying our methodologies to real-world data on state-level interventions and outcomes, we illustrate how this approach can optimize policy decisions. Our findings have substantial implications for enhancing the effectiveness of personalized intervention strategies in interconnected populations. This approach has the potential to revolutionize precision policy-making in spatially dependent contexts while maintaining a balance between public health and economic considerations.

Slides

Photos

Participants: 16