Difference-in-Differences under Local Dependence on Networks

2026-02-02

AI Summary

This study addresses causal inference under violations of the stable unit treatment value assumption in the presence of local network interference. The authors propose a nonparametric identification strategy that leverages a neighborhood-treatment-vector-based conditional parallel trends assumption to identify and estimate both direct and indirect average treatment effects. Innovatively, they define a novel form of indirect effect that captures the total spillover impact of an intervention without requiring a pre-specified exposure mapping, thereby overcoming limitations of conventional interference models. The proposed method integrates inverse probability weighting with a doubly robust estimator and establishes asymptotic consistency even under misspecification of the interference model. Simulations and empirical analyses demonstrate that the approach effectively and accurately estimates causal effects within locally dependent network structures.

Technology Category

• Machine Learning: Causal Learning
• Reasoning under Uncertainty: Causality
• Knowledge Representation and Reasoning: Action, Change, and Causality

Application Category

• Social Networks and Social Media: Influence propagation, information diffusion, and the prediction on networks
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs
• Economics, Online Markets and Human Computation: Incentives in network design for Web infrastructures and ecosystems

Abstract

Estimating causal effects under interference, where the stable unit treatment value assumption is violated, is critical in fields such as regional and public economics. Much of the existing research on causal inference under interference relies on a pre-specified"exposure mapping". This paper focuses on difference-in-difference and proposes a nonparametric identification strategy for direct and indirect average treatment effects under local interference on an observed network. In particular, we proposed a new concept of an indirect effect measuring the total outward influence of the intervension. Based on parallel trends assumption conditional on the neighborhood treatment vector, we develop inverse probability weighted and doubly robust estimators. We establish their asymptotic properties, including consistency under misspecification of nuisance models under some regularity conditions. Simulation studies and an empirical application demonstrate the effectiveness of the proposed method.

Problem

• interference
• difference-in-differences
• network
• causal inference
• indirect effects

Innovation

• difference-in-differences
• interference
• network spillovers
• doubly robust estimation
• nonparametric identification