Feasible Fusion: Constrained Joint Estimation under Structural Non-Overlap

TL;DR

This paper introduces a constrained joint estimation approach for causal inference in high-dimensional, multi-treatment systems with structural non-overlap, improving robustness and performance over existing methods.

Contribution

It formalizes treatment-induced non-overlap, identifies limitations of weighted fusion methods, and proposes a novel constrained estimation framework with a primal-dual algorithm.

Findings

Robust performance in synthetic experiments under nonoverlap.

Significant gains in a large-scale ridehailing application.

Matching RCT performance with less data.

Abstract

Causal inference in modern largescale systems faces growing challenges, including highdimensional covariates, multi-valued treatments, massive observational (OBS) data, and limited randomized controlled trial (RCT) samples due to cost constraints. We formalize treatment-induced structural non-overlap and show that, under this regime, commonly used weighted fusion methods provably fail to satisfy randomized identifying restrictions.To address this issue,we propose a constrained joint estimation framework that minimizes observational risk while enforcing causal validity through orthogonal experimental moment conditions. We further show that structural non-overlap creates a feasibility obstruction for moment enforcement in the original covariate space.We also derive a penalized primaldual algorithm that jointly learns representations and predictors, and establish oracle inequalities decomposing error into overlap recovery, moment violation, and statistical terms.Extensive synthetic experiments demonstrate robust performance under varying degrees of nonoverlap. A largescale ridehailing application shows that our method achieves substantial gains over existing baselines, matching the performance of models trained with significantly more RCT data.