Off-Policy Evaluation with Strategic Agents via Local Disclosure
For decision makers evaluating policies in strategic settings (e.g., algorithmic fairness, personalized pricing), this provides a practical OPE method with partial knowledge of agent responses, addressing a key limitation of prior work.
The paper tackles off-policy evaluation under strategic behavior where agents modify covariates in response to policy. It proposes a doubly robust estimator using local disclosure to reveal pre-strategic covariates, achieving consistency under a log-normal cost sensitivity assumption.
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.