The local approach to causal inference under network interference

• 成果类型: Article
• 署名作者: Auerbach, Eric; Guo, Hongchang; Tabord-Meehan, Max
• 署名单位: Northwestern University; University of Toronto
• 刊物名称: QUANTITATIVE ECONOMICS
• ISSN/ISSBN: 1759-7323
• DOI: 10.3982/QE2484
• 发表日期: 2026
• 关键词: Social networks; identification; tests

摘要

We propose a new nonparametric modeling framework for causal inference when outcomes depend on how agents are linked in a social or economic network. Such network interference describes a large literature on treatment spillovers, social interactions, social learning, information diffusion, disease and financial contagion, social capital formation, and more. Our approach works by first characterizing how an agent is linked in the network using the configuration of other agents and connections nearby as measured by path distance. The impact of a policy or treatment assignment is then learned by pooling outcome data across similarly configured agents. We demonstrate the approach by deriving finite-sample bounds on the mean-squared error of a k-nearest neighbor estimator for the average treatment response as well as proposing an asymptotically valid test for the hypothesis of policy irrelevance. We illustrate the empirical applicability of our method with simulations and an application to social capital formation.