Abstract:Causal inference is the process of using assumptions, study designs, and estimation strategies to draw conclusions about the causal relationships between variables based on data. This allows researchers to better understand the underlying mechanisms at work in complex systems and make more informed decisions. In many settings, we may not fully observe all the confounders that affect both the treatment and outcome variables, complicating the estimation of causal effects. To address this problem, a growing literature in both causal inference and machine learning proposes to use Instrumental Variables (IV). This paper serves as the first effort to systematically and comprehensively introduce and discuss the IV methods and their applications in both causal inference and machine learning. First, we provide the formal definition of IVs and discuss the identification problem of IV regression methods under different assumptions. Second, we categorize the existing work on IV methods into three streams according to the focus on the proposed methods, including two-stage least squares with IVs, control function with IVs, and evaluation of IVs. For each stream, we present both the classical causal inference methods, and recent developments in the machine learning literature. Then, we introduce a variety of applications of IV methods in real-world scenarios and provide a summary of the available datasets and algorithms. Finally, we summarize the literature, discuss the open problems and suggest promising future research directions for IV methods and their applications. We also develop a toolkit of IVs methods reviewed in this survey at https://github.com/causal-machine-learning-lab/mliv.

Causal Inference with the Instrumental Variable Approach and Bayesian Nonparametric Machine Learning

Bayesian Nonparametric Modeling for Causal Inference

Bayesian variable selection in linear regression models with instrumental variables

Heterogeneous causal effects with imperfect compliance: a Bayesian machine learning approach

Valid Causal Inference with (Some) Invalid Instruments

Flexible Instrumental Variable Models With Bayesian Additive Regression Trees

Counterfactual Prediction with Deep Instrumental Variables Networks

Instrumental Variable Model Average With Applications in Nonlinear Causal Inference

Mostly Harmless Machine Learning: Learning Optimal Instruments in Linear IV Models

Estimation and inference for high-dimensional nonparametric additive instrumental-variables regression

Scalable Quasi-Bayesian Inference for Instrumental Variable Regression

Quasi-Bayesian Dual Instrumental Variable Regression

Non-parametric causal inference for bivariate time series

Instrumental Variables in Causal Inference and Machine Learning: A Survey

High-dimensional Model-assisted Inference for Local Average Treatment Effects with Instrumental Variables

Causal Bounds and Instruments

Selective Machine Learning of the Average Treatment Effect with an Invalid Instrumental Variable

Strategic Instrumental Variable Regression: Recovering Causal Relationships From Strategic Responses

Robust causal inference with continuous instruments using the local instrumental variable curve

Control Function Instrumental Variable Estimation of Nonlinear Causal Effect Models

A Non-Parametric Bayesian Method for Inferring Hidden Causes