Primary Submission Category: Heterogeneous Treatment Effects
Finding Valid Adjustments under Non-ignorability with Minimal DAG Knowledge
Authors: Abhin Shah, Karthikeyan Shanmugam, Kartik Ahuja,
Presenting Author: Abhin Shah*
There are two very different schools of thought for treatment effect estimation from observational data. On one hand, the Pearlian framework commonly assumes structural knowledge (provided by an expert) in the form of directed acyclic graphs and provides graphical criteria, e.g., the back-door criterion, to identify valid adjustment sets. On the other hand, the potential outcomes (PO) framework commonly assumes that all the observed features satisfy ignorability (i.e., no hidden confounding), which in general is untestable. In prior works that attempted to bridge these frameworks, there is an observational criteria to identify an anchor variable and if a subset of covariates (not involving the anchor variable) passes a suitable conditional independence criteria, then that subset is a valid back-door. Our main result strengthens these prior results by showing that under a different expert-driven structural knowledge — that one variable is a causal parent of the treatment variable — remarkably, testing for subsets (not involving the known parent variable) that are valid back-doors is equivalent to an invariance test. Importantly, we cover the non-trivial case where the entire set of observed features is not ignorable (generalizing the PO framework) without requiring the knowledge of all the parents of the treatment variable. We also leverage Invariant Risk Minimization to connect finding valid adjustments (in non-ignorable observational settings) to representation learning.