Identification and Robust Inference for Multiple Treatments with Possibly Invalid Instruments
Abstract
The instrumental variable (IV) method is widely used to infer causal effects in observational studies with unmeasured confounding, but invalid instruments can compromise both population identification and finite-sample inference. This paper studies linear IV models with multiple endogenous treatments and possibly invalid instruments. Identification of multiple effects is more delicate than in the single-treatment setting because a single instrument no longer identifies a single candidate effect; instead, each relevant instrument defines a hyperplane in the multidimensional effect space. For identification of multiple treatment effects, we introduce generalized plurality and majority rules which require a sufficiently large number of IVs to be valid. For inference, data-dependent instrument selection may fail to separate certain invalid IVs from valid ones, leading to undercoverage of confidence intervals when these invalid instruments are mistakenly selected as valid. We propose a sampling confidence interval for each treatment effect, which is robust to IV selection errors. We establish asymptotic coverage and parametric-rate length of our sampling confidence interval under regularity conditions and illustrate this method in Monte Carlo simulations and a Mendelian randomization application.
Related
- Inference for Causal Path-Specific Effects in Longitudinal Studies with Repeated Treatments and Multiple Mediators
- Debiased Inference for AI-Generated Data without Gold-Standard Labels: Identification via Multiple Imperfect Measurements
- Inference with Granular Instruments in Large Panels
- Causal Inference with Generative Artificial Intelligence: Application to Texts as Treatments
- Nonlinear Sufficient Dimension Reduction for Causal Inference with Continuous Treatments
BibTeX
@misc{mei2026identification,
title = {{Identification and Robust Inference for Multiple Treatments with Possibly Invalid Instruments}},
author = {Ziwei Mei and Qingliang Fan and Zijian Guo},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2607.21481},
eprint = {2607.21481},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2607.21481},
}
RIS
TY - UNPB
TI - Identification and Robust Inference for Multiple Treatments with Possibly Invalid Instruments
AU - Ziwei Mei
AU - Qingliang Fan
AU - Zijian Guo
PY - 2026
DA - 2026/09/08
PB - arXiv
DO - 10.48550/arxiv.2607.21481
UR - https://arxiv.org/abs/2607.21481
AB - The instrumental variable (IV) method is widely used to infer causal effects in observational studies with unmeasured confounding, but invalid instruments can compromise both population identification and finite-sample inference. This paper studies linear IV models with multiple endogenous treatments and possibly invalid instruments. Identification of multiple effects is more delicate than in the single-treatment setting because a single instrument no longer identifies a single candidate effect; instead, each relevant instrument defines a hyperplane in the multidimensional effect space. For identification of multiple treatment effects, we introduce generalized plurality and majority rules which require a sufficiently large number of IVs to be valid. For inference, data-dependent instrument selection may fail to separate certain invalid IVs from valid ones, leading to undercoverage of confidence intervals when these invalid instruments are mistakenly selected as valid. We propose a sampling confidence interval for each treatment effect, which is robust to IV selection errors. We establish asymptotic coverage and parametric-rate length of our sampling confidence interval under regularity conditions and illustrate this method in Monte Carlo simulations and a Mendelian randomization application.
ER -