Randomization Test for the Specification of Interference Structure

TL;DR

This paper introduces a new randomization test for verifying the structure of spillover effects in causal inference experiments, applicable to any null exposure hypothesis and effective without prior knowledge of the interference pattern.

Contribution

It develops a versatile randomization test leveraging hierarchical exposure relationships, enhancing testing power for interference structures in experimental settings.

Findings

Effective in social network experiments on farmers' insurance adoption

Applicable to diverse null exposure specifications

Produces powerful test statistics without prior interference knowledge

Abstract

This study considers testing the specification of spillover effects in causal inference. We focus on experimental settings in which the treatment assignment mechanism is known to researchers. We develop a new randomization test utilizing a hierarchical relationship between different exposures. Compared with existing approaches, our approach is essentially applicable to any null exposure specifications and produces powerful test statistics without a priori knowledge of the true interference structure. As empirical illustrations, we revisit two existing social network experiments: one on farmers' insurance adoption and the other on anti-conflict education programs.