Recovering Counterfactual Distributions via Wasserstein GANs

2026-01-24

AI Summary

This study addresses the limitations of traditional distributional synthetic control methods, which often produce structural artifacts and suffer from insufficient gradient information when support sets are mismatched or outcomes are multimodal. To overcome these issues, the authors propose a robust counterfactual distribution estimator grounded in optimal transport theory, introducing Wasserstein GAN into the distributional synthetic control framework for the first time. By replacing the conventional L2 distance with the Wasserstein-1 distance, the method enables effective matching between probability measures. Under affine independence conditions, it consistently identifies stable weights even in the presence of support misalignment or multimodality. Monte Carlo simulations demonstrate that the proposed approach accurately reconstructs counterfactual distributions under complex settings—such as heavy-tailed contamination and bimodal mixture distributions—and significantly outperforms existing methods.

Technology Category

• Search and Optimization: Distributed Search
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization
• Multiagent Systems: Adversarial Agents

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods

Abstract

Standard Distributional Synthetic Controls (DSC) estimate counterfactual distributions by minimizing the Euclidean $L_2$ distance between quantile functions. We demonstrate that this geometric reliance renders estimators fragile: they lack informative gradients under support mismatch and produce structural artifacts when outcomes are multimodal. This paper proposes a robust estimator grounded in Optimal Transport (OT). We construct the synthetic control by minimizing the Wasserstein-1 distance between probability measures, implemented via a Wasserstein Generative Adversarial Network (WGAN). We establish the formal point identification of synthetic weights under an affine independence condition on the donor pool. Monte Carlo simulations confirm that while standard estimators exhibit catastrophic variance explosions under heavy-tailed contamination and support mismatch, our WGAN-based approach remains consistent and stable. Furthermore, we show that our measure-based method correctly recovers complex bimodal mixtures where traditional quantile averaging fails structurally.

Problem

• counterfactual distributions
• distributional synthetic controls
• support mismatch
• multimodal outcomes
• Wasserstein distance

Innovation

• Wasserstein GAN
• Optimal Transport
• Counterfactual Distribution
• Synthetic Control
• Wasserstein Distance