Bayesian Safe Policy Learning with Chance Constrained Optimization: Application to Military Security Assessment during the Vietnam War
This work addresses high-stakes algorithmic decision-making in military security, offering a method to safely update deterministic algorithms with transparent risk control, though it is incremental in applying existing Bayesian and optimization techniques to a specific historical case.
The authors tackled the problem of improving a deterministic security assessment algorithm used during the Vietnam War by developing a Bayesian policy learning framework with chance constraints to control risks of worse outcomes. Their analysis found that the learned algorithm assessed most regions as more secure and shifted emphasis from military to economic and political factors compared to the original algorithm.
Algorithmic decisions and recommendations are used in many high-stakes decision-making settings such as criminal justice, medicine, and public policy. We investigate whether it would have been possible to improve a security assessment algorithm employed during the Vietnam War, using outcomes measured immediately after its introduction in late 1969. This empirical application raises several methodological challenges that frequently arise in high-stakes algorithmic decision-making. First, before implementing a new algorithm, it is essential to characterize and control the risk of yielding worse outcomes than the existing algorithm. Second, the existing algorithm is deterministic, and learning a new algorithm requires transparent extrapolation. Third, the existing algorithm involves discrete decision tables that are difficult to optimize over. To address these challenges, we introduce the Average Conditional Risk (ACRisk), which first quantifies the risk that a new algorithmic policy leads to worse outcomes for subgroups of individual units and then averages this over the distribution of subgroups. We also propose a Bayesian policy learning framework that maximizes the posterior expected value while controlling the posterior expected ACRisk. This framework separates the estimation of heterogeneous treatment effects from policy optimization, enabling flexible estimation of effects and optimization over complex policy classes. We characterize the resulting chance-constrained optimization problem as a constrained linear programming problem. Our analysis shows that compared to the actual algorithm used during the Vietnam War, the learned algorithm assesses most regions as more secure and emphasizes economic and political factors over military factors.