Causal Algorithmic Recourse: Foundations and Methods

2026-05-11

AI Summary

This work addresses a key limitation in existing algorithmic attribution methods, which treat counterfactuals as static outcomes and overlook the dynamic decision-making processes individuals undergo following an intervention. To overcome this, the paper introduces a causal framework that models attribution as the dynamic evolution of outcomes before and after intervention. Central to this approach is a novel “post-intervention stability” condition, which enables identification of attribution effects using only observational data. The proposed method integrates copula modeling with distribution-free learning algorithms, ensuring robustness under both model misspecification and correct specification. Empirical evaluations on real-world and semi-synthetic datasets demonstrate that the approach yields more accurate and reliable estimates of individual-level post-intervention decision changes compared to existing methods.

Abstract

The trustworthiness of AI decision-making systems is increasingly important. A key feature of such systems is the ability to provide recommendations for how an individual may reverse a negative decision, a problem known as algorithmic recourse. Existing approaches treat recourse outcomes as counterfactuals of a fixed unit, ignoring that real-world recourse involves repeated decisions on the same individual under possibly different latent conditions. We develop a causal framework that models recourse as a process over pre- and post-intervention outcomes, allowing for partial stability and resampling of latent variables. We introduce post-recourse stability conditions that enable reasoning about recourse from observational data alone, and develop a copula-based algorithm for inferring the effects of recourse under these conditions. For settings where paired observations of the same individual before and after intervention are available (called recourse data), we develop methods for inferring copula parameters and performing goodness-of-fit testing. When the copula model is rejected, we provide a distribution-free algorithm for learning recourse effects directly from recourse data. We demonstrate the value of the proposed methods on real and semi-synthetic datasets.

Problem

• algorithmic recourse
• causal inference
• counterfactual reasoning
• latent variables
• stability

Innovation

• causal algorithmic recourse
• post-recourse stability
• copula-based inference
• recourse data
• distribution-free learning