Aggregation Trees

TL;DR

This paper introduces a nonparametric method for constructing nested subgroups to explore heterogeneous treatment effects, balancing interpretability and granularity, with valid inference enabled by honest and debiased machine learning.

Contribution

It proposes a flexible, nonparametric approach to subgroup discovery that integrates honest and debiased machine learning for valid inference.

Findings

Validated through Monte-Carlo simulations

Revealed heterogeneity in maternal smoking impact on birth weight

Demonstrated trade-off between interpretability and granularity

Abstract

Uncovering the heterogeneous effects of particular policies or "treatments" is a key concern for researchers and policymakers. A common approach is to report average treatment effects across subgroups based on observable covariates. However, the choice of subgroups is crucial as it poses the risk of $p$-hacking and requires balancing interpretability with granularity. This paper proposes a nonparametric approach to construct heterogeneous subgroups. The approach enables a flexible exploration of the trade-off between interpretability and the discovery of more granular heterogeneity by constructing a sequence of nested groupings, each with an optimality property. By integrating our approach with "honesty" and debiased machine learning, we provide valid inference about the average treatment effect of each group. We validate the proposed methodology through an empirical Monte-Carlo study and apply it to revisit the impact of maternal smoking on birth weight, revealing systematic heterogeneity driven by parental and birth-related characteristics.