Optimal Covariate Adjustment beyond the Average Treatment Effect: Treated-Population and Overlap-Weighted Estimands
Abstract
Graphical causal inference supplies a complete theory of efficient covariate adjustment for the average treatment effect: one adjustment set, computable from the graph, is optimal under every compatible distribution. We show that this is a property of the average treatment effect's inverse-prevalence weights, not of causal estimands in general. For the average treatment effect on the treated we index the efficiency bound by the adjustment set and derive exact identities for its change under treatment-side and outcome-side extensions of a valid set. Covariates that predict only the treated-arm outcome are exactly efficiency-neutral, and covariates that predict the control-arm outcome can strictly increase the bound when the propensity is below one half -- a reversal of the supplementation lemma whose source is an arithmetic-geometric-mean inequality that holds for the average treatment effect and fails for the treated-population estimand. A construction with two faithful distributions on one graph proves that no graphical optimality criterion exists for the treated-population estimand; under no effect modification the ATE-optimal set is nonetheless optimal among the graphically valid sets, with an exact expression for its advantage. The results extend to weighted average treatment effects with propensity-dependent weights, yielding symmetric thresholds for overlap weights, an estimand-drift phenomenon under instrument adjustment, and a characterization of constant weights as the only smooth positive weights for which outcome-side supplementation never increases the bound. Simulations and the LaLonde data provide illustrations.
Related
- Simple Covariate Adjustment for Many Estimands Using Stable Balancing Weights
- Automated, Efficient, and Model-Free Inference for Randomized Clinical Trials via Data-Driven Covariate Adjustment
- Covariate Adjustment in Randomized Experiments: A Unified Framework for Decision and Practice
- Covariate Adjustment for Time-Restricted Marginal Win Statistics in Randomized Trials: Assumption-Lean Inference and Guaranteed Efficiency Gain
- Beyond the Mean: A Weighted k-Sample Omnibus Variance-Ratio Statistic for Covariate Balance Diagnostics
BibTeX
@misc{okubo2026optimal,
title = {{Optimal Covariate Adjustment beyond the Average Treatment Effect: Treated-Population and Overlap-Weighted Estimands}},
author = {Shoki Okubo},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.11222},
eprint = {2609.11222},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.11222},
}
RIS
TY - UNPB
TI - Optimal Covariate Adjustment beyond the Average Treatment Effect: Treated-Population and Overlap-Weighted Estimands
AU - Shoki Okubo
PY - 2026
DA - 2026/09/13
PB - arXiv
DO - 10.48550/arxiv.2609.11222
UR - https://arxiv.org/abs/2609.11222
AB - Graphical causal inference supplies a complete theory of efficient covariate adjustment for the average treatment effect: one adjustment set, computable from the graph, is optimal under every compatible distribution. We show that this is a property of the average treatment effect's inverse-prevalence weights, not of causal estimands in general. For the average treatment effect on the treated we index the efficiency bound by the adjustment set and derive exact identities for its change under treatment-side and outcome-side extensions of a valid set. Covariates that predict only the treated-arm outcome are exactly efficiency-neutral, and covariates that predict the control-arm outcome can strictly increase the bound when the propensity is below one half -- a reversal of the supplementation lemma whose source is an arithmetic-geometric-mean inequality that holds for the average treatment effect and fails for the treated-population estimand. A construction with two faithful distributions on one graph proves that no graphical optimality criterion exists for the treated-population estimand; under no effect modification the ATE-optimal set is nonetheless optimal among the graphically valid sets, with an exact expression for its advantage. The results extend to weighted average treatment effects with propensity-dependent weights, yielding symmetric thresholds for overlap weights, an estimand-drift phenomenon under instrument adjustment, and a characterization of constant weights as the only smooth positive weights for which outcome-side supplementation never increases the bound. Simulations and the LaLonde data provide illustrations.
ER -