A0589
Title: Causal effect identification and inference with endogenous exposures and a light-tailed error
Authors: Ruoyu Wang - Harvard School of Public Health (United States) [presenting]
Abstract: Endogeneity poses significant challenges in causal inference across various research domains. The aim is to propose a novel approach to identify and estimate causal effects in the presence of endogeneity. A structural equation is considered with endogenous exposures and an additive error term. Assuming the light-tailedness of the error term, it is shown that the causal effect can be identified by contrasting extreme conditional quantiles of the outcome given the exposures. Unlike many existing results, the identification approach does not rely on additional parametric assumptions or auxiliary variables. Building on the identification result, an extreme-based causal effect learning (EXCEL) method is developed that estimates the causal effect using extreme quantile regression. The consistency of the EXCEL estimator is established under a general additive structural equation, and its asymptotic normality is demonstrated in the linear model setting. These results reveal that extreme quantile regression is invulnerable to endogeneity when the error term is light-tailed, which is not appreciated in the literature to our knowledge. The EXCEL method is applied to causal inference problems with invalid auxiliary variables, e.g., invalid instruments or invalid negative controls, to construct a valid confidence set for the causal effect. Simulations and data analysis of an automobile sale dataset show the effectiveness of our method in addressing endogeneity.
