Treatment Allocation with Strategic Agents
This addresses the challenge of strategic manipulation in personalized treatment allocation, such as in marketing or pricing, which is an incremental advance in mechanism design for causal inference.
The paper tackles the problem of optimal treatment allocation when individuals can strategically modify their behavior to receive better treatments, showing that the optimal rule may involve randomization rather than deterministic allocation. It proposes a sequential Bayesian optimization experiment that converges to this optimal rule without parametric assumptions on strategic behavior.
There is increasing interest in allocating treatments based on observed individual characteristics: examples include targeted marketing, individualized credit offers, and heterogeneous pricing. Treatment personalization introduces incentives for individuals to modify their behavior to obtain a better treatment. Strategic behavior shifts the joint distribution of covariates and potential outcomes. The optimal rule without strategic behavior allocates treatments only to those with a positive Conditional Average Treatment Effect. With strategic behavior, we show that the optimal rule can involve randomization, allocating treatments with less than 100% probability even to those who respond positively on average to the treatment. We propose a sequential experiment based on Bayesian Optimization that converges to the optimal treatment rule without parametric assumptions on individual strategic behavior.