Double Machine Learning for Panel Data: Identification, Specification, and a Bank Regulation Effect
Abstract
This paper examines robustness and causal interpretation of panel estimates using double/debiased machine learning. The prototype treatment setup is the 2018 US banking regulation. Under common treatment timing, we characterize identification and nuisance restrictions and derive exact bank-level representations of specified "within" and "first-difference" scores. Simulations assess panel model transformation-induced specification bias and inference under weaker overlap. Empirically, estimates vary with asset-size classification and prior growth. Prior-growth adjustment reduces overlap-weighted estimates across four Machine Learners (ML); doubly robust estimates depend on the fitted propensity and concentrated weights. Both bank groups received relief, so we estimate differential exposure effects. The findings highlight specification issues and the relation between functional-form robustness and causal identification.
Related
- Measurement Error and Nonlinear Confounding: Structural Equation Modeling Versus Double Machine Learning
- Causal Mediation Analysis for Panel Data with Double/Debiased Machine Learning
- Covariate Selection for Doubly Robust Double/debiased Machine Learning Estimators for Causal Inference
- Correcting Bartlett factor-score error for latent covariates in partially linear double machine learning
- On the Asymptotic Inadmissibility of Double Machine Learning Estimators Under Structure-Agnostic Models
BibTeX
@misc{habibnia2026double,
title = {{Double Machine Learning for Panel Data: Identification, Specification, and a Bank Regulation Effect}},
author = {Ali Habibnia and Esfandiar Maasoumi and Gaurav Shah},
howpublished = {arXiv},
year = {2026},
month = nov,
doi = {10.48550/arxiv.2611.07249},
eprint = {2611.07249},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2611.07249},
}
RIS
TY - UNPB
TI - Double Machine Learning for Panel Data: Identification, Specification, and a Bank Regulation Effect
AU - Ali Habibnia
AU - Esfandiar Maasoumi
AU - Gaurav Shah
PY - 2026
DA - 2026/11/04
PB - arXiv
DO - 10.48550/arxiv.2611.07249
UR - https://arxiv.org/abs/2611.07249
AB - This paper examines robustness and causal interpretation of panel estimates using double/debiased machine learning. The prototype treatment setup is the 2018 US banking regulation. Under common treatment timing, we characterize identification and nuisance restrictions and derive exact bank-level representations of specified "within" and "first-difference" scores. Simulations assess panel model transformation-induced specification bias and inference under weaker overlap. Empirically, estimates vary with asset-size classification and prior growth. Prior-growth adjustment reduces overlap-weighted estimates across four Machine Learners (ML); doubly robust estimates depend on the fitted propensity and concentrated weights. Both bank groups received relief, so we estimate differential exposure effects. The findings highlight specification issues and the relation between functional-form robustness and causal identification.
ER -