Off-Policy Evaluation with Strategic Agents via Local Disclosure

2026-06-05

AI Summary

This study addresses covariate shift induced by strategic agents’ responsive behaviors in single-round interactions, challenging the exogeneity assumption commonly adopted in offline policy evaluation. Under partial knowledge of the agents’ response mechanism, the authors propose a novel local information disclosure framework that leverages post-hoc interpretability techniques to reconstruct pre-adjustment features. They further develop a cost-sensitive response model based on a conditional log-normal distribution and introduce a doubly robust policy value estimator that simultaneously ensures consistency and corrects for covariate shift. Theoretical analysis establishes the estimator’s consistency, while empirical experiments demonstrate its effectiveness and accuracy in evaluating policy performance under strategic agent behavior.

Abstract

We study off-policy evaluation (OPE) under strategic behavior where decision subjects (or agents) respond to a decision maker's policy by strategically modifying their covariates. Such behavior induces a policy-dependent covariate shift, breaking the standard assumption in existing methods that covariates are exogenous to the policy. Related work addresses this challenge by imposing strong assumptions such as repeated interactions or full knowledge of agents' response behavior, substantially limiting its applicability to OPE. In contrast, we consider a one-shot OPE setting where the decision maker has only partial knowledge of the agents' response behavior. Our key insight is that disclosing local information through post-hoc explanations reveals agents' pre-strategic covariates prior to adaptation, mitigating the information loss induced by strategic behavior. Leveraging this structure, we estimate a statistical model for the agents' responses and construct a doubly robust estimator for policy value. By assuming that the agents' cost sensitivity follows a conditional log-normal distribution, we establish consistency of the proposed estimator and validate our approach empirically. More broadly, our results highlight how interaction design can mitigate information asymmetry by revealing otherwise hidden structure in agents' strategic responses.

Problem

• off-policy evaluation
• strategic agents
• covariate shift
• information asymmetry
• policy-dependent behavior

Innovation

• off-policy evaluation
• strategic agents
• local disclosure
• doubly robust estimator
• covariate shift