Assessing Racial Disparities in Healthcare Expenditures Using Causal Path-Specific Effects

2025-04-30

AI Summary

This paper quantifies the causal path-specific effects of socioeconomic status, insurance access, health behaviors, and health status—mediating factors underlying racial disparities in healthcare expenditures. To address the zero-inflated and right-skewed nature of expenditure data, we propose the first machine learning–enhanced framework that integrates causal path-specific effect estimation with influence function–based inference. Our approach combines super learners, two-part models, and robust influence function estimators, ensuring both model misspecification robustness and asymptotic linearity. It enables precise decomposition of each mediator’s independent contribution to expenditure disparities, yielding interpretable, high-precision, and doubly robust causal effect estimates. The method advances the identification of key drivers of health inequities and informs targeted policy interventions.

Technology Category

• Machine Learning: Causal Learning
• Reasoning under Uncertainty: Causality
• Intelligent Robots: State Estimation

Application Category

• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Economics, Online Markets and Human Computation: Trust and reliance of crowd workers and data experts on GenAI
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Racial disparities in healthcare expenditures are well-documented, yet the underlying drivers remain complex and require further investigation. This study employs causal and counterfactual path-specific effects to quantify how various factors, including socioeconomic status, insurance access, health behaviors, and health status, mediate these disparities. Using data from the Medical Expenditures Panel Survey, we estimate how expenditures would differ under counterfactual scenarios in which the values of specific mediators were aligned across racial groups along selected causal pathways. A key challenge in this analysis is ensuring robustness against model misspecification while addressing the zero-inflation and right-skewness of healthcare expenditures. For reliable inference, we derive asymptotically linear estimators by integrating influence function-based techniques with flexible machine learning methods, including super learners and a two-part model tailored to the zero-inflated, right-skewed nature of healthcare expenditures.

Problem

• Quantify racial disparities in healthcare expenditure drivers
• Assess mediator impact via causal path-specific effects
• Address zero-inflation and skewness in expenditure data

Innovation

• Uses causal path-specific effects for disparity analysis
• Employs super learners for flexible machine learning
• Applies two-part model for zero-inflated skewed data