Decomposing Discrimination: Causal Mediation Analysis for AI-Driven Credit Decisions

View PDF HTML (experimental)

Abstract:Statistical fairness metrics in AI-driven credit decisions conflate two causally distinct mechanisms: discrimination operating directly from a protected attribute to a credit outcome, and structural inequality propagating through legitimate financial features. We formalise this distinction using Pearl's framework of natural direct and indirect effects applied to the credit decision setting. Our primary theoretical contribution is an identification strategy for natural direct and indirect effects under treatment-induced confounding -- the prevalent setting in which protected attributes causally affect both financial mediators and the final decision, violating standard sequential ignorability. We show that interventional direct and indirect effects (IDE/IIE) are identified under the weaker Modified Sequential Ignorability assumption, and prove that IDE/IIE provide conservative bounds on the unidentified natural effects under monotone indirect treatment response. We propose a doubly-robust augmented inverse probability weighted (AIPW) estimator for IDE/IIE with semiparametric efficiency properties, implemented via cross-fitting. An E-value sensitivity analysis addresses residual confounding on the direct pathway. Empirical evaluation on 89,465 real HMDA conventional purchase mortgage applications from New York State (2022) demonstrates that approximately 77% of the observed 7.9 percentage-point racial denial disparity operates through financial mediators shaped by structural inequality, while the remaining 23% constitutes a conservative lower bound on direct discrimination. The open-source CausalFair Python package implements the full pipeline for deployment at resource-constrained financial institutions.

Comments: 22 pages, 6 figures, 2 tables. Open-source code at this https URL

Subjects:

Machine Learning (cs.LG)

MSC classes: 62P05, 62D20, 62H22, 91G70

ACM classes: I.2.6; J.4; K.4.1

Cite as: arXiv:2603.27510 [cs.LG]

(or arXiv:2603.27510v1 [cs.LG] for this version)

https://doi.org/10.48550/arXiv.2603.27510

arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Durai Rajamanickam [view email] [v1] Sun, 29 Mar 2026 04:20:09 UTC (71 KB)