Proximal Causal Inference without Uniqueness Assumptions
Authors:
- Jeffrey ZhangJZJeffrey Zhang
- Wei LiWLWei Li
- Wang MiaoWMWang Miao
- Eric Tchetgen TchetgenETEric Tchetgen Tchetgen
Preprint
Abstract
We consider identification and inference about a counterfactual outcome mean when there is unmeasured confounding using tools from proximal causal inference (Miao et al. [2018], Tchetgen Tchetgen et al. [2020]). Proximal causal inference requires existence of solutions to at least one of two integral equations. We motivate the existence of solutions to the integral equations from proximal causal inference by demonstrating that, assuming the existence of a solution to one of the integral equations, $\sqrt{n}$-estimability of a linear functional (such as its mean) of that solution requires the existence of a solution to the other integral equation. Solutions to the integral equations may not be unique, which complicates estimation and inference. We construct a consistent estimator for the solution set for one of the integral equations and then adapt the theory of extremum estimators to find from the estimated set a consistent estimator for a uniquely defined solution. A debiased estimator for the counterfactual mean is shown to be root-$n$ consistent, regular, and asymptotically semiparametrically locally efficient under additional regularity conditions.
Keywords
All Related Versions
This publication has 10 references indexed in Scilit:
- Inference on strongly identified functionals of weakly identified functionsJournal of the Royal Statistical Society Series B: Statistical Methodology, 2025
- Semiparametric Proximal Causal InferenceJournal of the American Statistical Association, 2023
- Automatic Debiased Machine Learning of Causal and Structural EffectsEconometrica, 2022
- A Selective Review of Negative Control Methods in EpidemiologyCurrent Epidemiology Reports, 2020
- Multiply Robust Causal Inference with Double-Negative Control Adjustment for Categorical Unmeasured ConfoundingJournal of the Royal Statistical Society Series B: Statistical Methodology, 2020
- Identifying causal effects with proxy variables of an unmeasured confounderBiometrika, 2018
- On varieties of doubly robust estimators under missingness not at random with a shadow variableBiometrika, 2016
- Efficiency bounds for estimating linear functionals of nonparametric regression models with endogenous regressorsJournal of Econometrics, 2012
- Estimation of Nonparametric Conditional Moment Models With Possibly Nonsmooth Generalized ResidualsEconometrica, 2012
- Instrumental variable methods for recovering continuous linear functionalsJournal of Econometrics, 2011