Proximal Causal Learning of Conditional Average Treatment Effects
This addresses causal inference limitations in fields like medicine and marketing where proxy variables must replace implausible exchangeability assumptions.
The paper tackles the problem of estimating heterogeneous treatment effects when standard conditional exchangeability assumptions are violated, proposing the P-learner method that achieves an oracle error bound with kernel regression when nuisance components are well-estimated.
Efficiently and flexibly estimating treatment effect heterogeneity is an important task in a wide variety of settings ranging from medicine to marketing, and there are a considerable number of promising conditional average treatment effect estimators currently available. These, however, typically rely on the assumption that the measured covariates are enough to justify conditional exchangeability. We propose the P-learner, motivated by the R- and DR-learner, a tailored two-stage loss function for learning heterogeneous treatment effects in settings where exchangeability given observed covariates is an implausible assumption, and we wish to rely on proxy variables for causal inference. Our proposed estimator can be implemented by off-the-shelf loss-minimizing machine learning methods, which in the case of kernel regression satisfies an oracle bound on the estimated error as long as the nuisance components are estimated reasonably well.