EconBase
← All papers

Data-Automated Policy Learning for Nonlinear Welfare

Chunrong Ai, Zeqi Wu, Zheng Zhang

arXiv 1 Jun 2026 · Econometrics

arXiv:2606.01659 · PDF · DOI · OpenAlex · Extracted main text

Abstract

This paper explores policy learning from observational data, focusing on a nonlinear welfare criterion in a binary treatment setting. The nonlinear criterion is inspired by scenarios where policymakers prioritize specific population segments. We model this criterion using a utility function that encompasses potential outcomes and intermediate parameters, with the latter capturing higher moments of the outcome distributions. When formulated in the context of observational data, both the intermediate parameters and the welfare criterion depend on the propensity score, which we estimate using machine-learning techniques. To address bias in machine learning estimates, we introduce a novel reweighting-based debiasing approach that offers a promising alternative to traditional orthogonality-based methods. To tackle the complexities of infinite-dimensional policy spaces, we employ sieve approximations and $K$-fold cross-validation for model selection, thereby fully automating the policy-learning process. Despite these complexities, we demonstrate that both the welfare regret and the average welfare regret of our proposed policy learning method satisfy an oracle inequality, thereby providing theoretical guarantees on the performance of the estimated policy relative to the best possible policy. This finding extends the existing results from linear to nonlinear welfare criteria, from finite-dimensional to infinite-dimensional policy spaces, and from a known propensity score to a machine-learned one.

Citation extraction

63
references
118
in-text mentions
63
distinct cited
3
self-citations
10,217
main-text words

appendix boundary found by appendix_command · 24% of the source is main text. Read the extracted text to check this.

Most heavily cited references

The works this paper leans on most, across its whole bibliography — not restricted to papers in our corpus. Ranked by composite intensity, which combines how often a work is mentioned, how many sections mention it, and how much of that falls in the main text rather than the appendix.

ReferenceIntensityMentionsSectionsMain text
1Athey, Susan and Wager, Stefan (2021) Policy Learning With Observational Data1.00053100%
2Mbakop, Eric and Tabord-Meehan, Max (2021) Model Selection for Treatment Choice: Penalized Welfare Maximization0.93717582%
3Kitagawa, Toru and Tetenov, Aleksey (2018) Who Should Be Treated? Empirical Welfare Maximization Methods for Treatment Choice0.92844100%
4Crippa, Federico (2025) Regret Analysis in Threshold Policy Design0.64422100%
5Liu, Nan and Liu, Yanbo and Sasaki, Yuya and Wan, Yuanyuan (2025) Nonparametric Uniform Inference in Binary Classification and Policy Values0.64422100%
6Boyd, Stephen P. and Vandenberghe, Lieven (2004) Convex Optimization0.5112250%
7Fan, Yanqin and Qi, Yuan and Xu, Gaoqian (2025) Policy Learning with $ $-Expected Welfare0.51121100%
8Farrell, Max H. and Liang, Tengyuan and Misra, Sanjog (2021) Deep Neural Networks for Estimation and Inference0.51121100%
9Jiao, Yuling and Shen, Guohao and Lin, Yuanyuan and Huang, Jian (2023) Deep Nonparametric Regression on Approximate Manifolds: Nonasymptotic Error Bounds with Polynomial Prefactors0.51121100%
10Lecué, Guillaume and Mitchell, Charles (2012) Oracle Inequalities for Cross-Validation Type Procedures0.51121100%

Showing the top 10 of 63 scored citations.