Supercompliers

TL;DR

This paper introduces the concept of supercompliers in instrumental variable analysis, identifying subpopulations that benefit from treatment eligibility and demonstrating methods to characterize and estimate them, with applications to policy evaluation.

Contribution

It defines supercompliers within a binary-treatment IV framework and provides tools for their characterization and estimation under testable assumptions, advancing policy-relevant analysis.

Findings

Supercompliers are identified as the subpopulation benefiting from treatment eligibility.

Estimation methods using IV regression are developed for supercompliers.

Application to job-training experiments shows practical utility in policy analysis.

Abstract

In a binary-treatment instrumental variable framework, we define supercompliers as the subpopulation whose treatment take-up positively responds to eligibility and whose outcome positively responds to take-up. Supercompliers are the only subpopulation to benefit from treatment eligibility and, hence, are important for policy. We provide tools to characterize supercompliers under a set of jointly testable assumptions. Specifically, we require standard assumptions from the local average treatment effect literature plus an outcome monotonicity assumption. Estimation and inference can be conducted with instrumental variable regression. In two job-training experiments, we demonstrate our machinery's utility, particularly in incorporating social welfare weights into marginal-value-of-public-funds analysis.