Interpretable Almost-Matching-Exactly With Instrumental Variables
This addresses the challenge of causal inference in observational studies for researchers and practitioners, offering a more interpretable and scalable method, though it is incremental as it builds on existing IV matching approaches.
The paper tackles the problem of estimating causal effects in observational studies with unmeasured confounding by proposing a matching framework using Instrumental Variables (IV) that avoids strong parametric assumptions and scales well, showing it constructs better matches on simulated datasets and produces interesting results in a political canvassing application.
Uncertainty in the estimation of the causal effect in observational studies is often due to unmeasured confounding, i.e., the presence of unobserved covariates linking treatments and outcomes. Instrumental Variables (IV) are commonly used to reduce the effects of unmeasured confounding. Existing methods for IV estimation either require strong parametric assumptions, use arbitrary distance metrics, or do not scale well to large datasets. We propose a matching framework for IV in the presence of observed categorical confounders that addresses these weaknesses. Our method first matches units exactly, and then consecutively drops variables to approximately match the remaining units on as many variables as possible. We show that our algorithm constructs better matches than other existing methods on simulated datasets, and we produce interesting results in an application to political canvassing.