PrivATE: Differentially Private Average Treatment Effect Estimation for Observational Data

2025-12-16

AI Summary

Estimating the Average Treatment Effect (ATE) from observational data often compromises user privacy—particularly critical in sensitive domains such as healthcare and education, where heterogeneous privacy requirements pose significant challenges. Method: This paper proposes a dual-granularity differential privacy (DP) framework that jointly incorporates label-level and sample-level privacy protection. It introduces an adaptive matching boundary mechanism to jointly mitigate both DP-induced noise error and bias from propensity score matching. Additionally, it employs dynamic privacy budget allocation and error-tradeoff optimization. Contribution/Results: Extensive experiments on multiple real-world datasets demonstrate that the proposed method consistently outperforms existing baselines across diverse privacy budgets, achieving substantial improvements in ATE estimation accuracy while rigorously preserving privacy.

Technology Category

• Machine Learning: Privacy
• Computer Vision: Bias, Fairness & Privacy
• Natural Language Processing: Ethics — Bias, Fairness, Transparency & Privacy

Application Category

• Security and Privacy: Privacy-enhancing technologies
• User Modeling, Personalization and Recommendation: User privacy protection in personalized systems
• Responsible Web: Data and user privacy-enhancing technologies for the Web

Abstract

Causal inference plays a crucial role in scientific research across multiple disciplines. Estimating causal effects, particularly the average treatment effect (ATE), from observational data has garnered significant attention. However, computing the ATE from real-world observational data poses substantial privacy risks to users. Differential privacy, which offers strict theoretical guarantees, has emerged as a standard approach for privacy-preserving data analysis. However, existing differentially private ATE estimation works rely on specific assumptions, provide limited privacy protection, or fail to offer comprehensive information protection. To this end, we introduce PrivATE, a practical ATE estimation framework that ensures differential privacy. In fact, various scenarios require varying levels of privacy protection. For example, only test scores are generally sensitive information in education evaluation, while all types of medical record data are usually private. To accommodate different privacy requirements, we design two levels (i.e., label-level and sample-level) of privacy protection in PrivATE. By deriving an adaptive matching limit, PrivATE effectively balances noise-induced error and matching error, leading to a more accurate estimate of ATE. Our evaluation validates the effectiveness of PrivATE. PrivATE outperforms the baselines on all datasets and privacy budgets.

Problem

• Estimates ATE from observational data privately
• Provides flexible privacy levels for different scenarios
• Balances noise and matching errors for accuracy

Innovation

• Two-level differential privacy for ATE estimation
• Adaptive matching limit balancing noise and error
• Framework outperforms baselines across datasets and budgets