Causal Stability Selection

2026-05-09

AI Summary

Existing data-adaptive methods for identifying treatment effect modifiers often lack finite-sample control of false positives, leading to irreproducible findings. This work proposes a general framework that integrates cross-fitted conditional average treatment effect estimation with ensemble pathwise stability selection, accommodating arbitrary treatment effect estimators and variable selectors. For the first time, it provides a non-asymptotic upper bound on the expected number of false positives in modifier selection under finite samples. The method establishes a theoretical link between the accuracy of treatment effect estimation and the performance of variable discovery. Empirical validation on both a randomized oncology trial and an observational study of maternal smoking’s effect on infant birth weight demonstrates its ability to control the false discovery rate while achieving strong selection consistency.

Abstract

Identifying covariates that modify treatment effects is a central problem in causal inference. Yet existing data-adaptive procedures do not provide finite-sample control over the expected number of false discoveries, risking spurious findings that fail to replicate. We introduce causal stability selection, an algorithm that combines cross-fitted estimation of conditional average treatment effects with integrated path stability selection. The method accommodates arbitrary treatment effect estimators and arbitrary base selectors, and produces a selection set with an explicit, non-asymptotic bound on the expected number of false positives. Under standard causal identifying assumptions and regularity conditions on the base selector, we prove that the estimated selection probabilities converge to their oracle counterparts at the rate of the underlying treatment effect estimator. This establishes a direct connection between treatment effect estimation and effect modifier discovery. We illustrate the method on a randomized trial in oncology and on observational data on maternal smoking and infant birthweight.

Problem

• causal inference
• effect modification
• false discovery control
• treatment effect
• stability selection

Innovation

• causal stability selection
• conditional average treatment effect
• false discovery control
• cross-fitting
• effect modifier discovery