SMU SOE Seminar Series (April 1, 2026): Genuinely Robust Inference for Clustered Data

TOPIC:

ABSTRACT

Conventional methods for cluster-robust inference are inconsistent when clusters of unignorably large size are present. We formalize this issue by deriving a necessary and sufficient condition for consistency, a condition frequently violated in empirical studies. Specifically, 77% of empirical research articles published in American Economic Review and Econometrica during 2020–2021 do not satisfy this condition. To address this limitation, we propose a new approach based on m-out-of-n bootstrap and establish its size control across broad classes of data-generating processes where conventional methods fail. Extensive simulation studies support our findings, demonstrating the reliability and effectiveness of the proposed approaches.

Keywords: Cluster-Robust Inference, Cluster Score Bootstrap, Unignorably Large Cluster, Domain of Attraction, Extreme Value Theory.

JEL: C12, C18, C46.

Click here to view the paper.