“…Two types of approaches are usually employed to address the selection bias issue of a binary treatment variable. The first is instrumental variable‐based (IV‐based) methods, including the endogenous treatment regression (ETR) model (Ma & Abdulai, 2017; Stata Press, 2019), endogenous switching regression (ESR) model (Ma, Zheng, & Yuan, 2021; Rodriguez, 2022), and control function approach (Dohmwirth & Liu, 2020; Wooldridge, 2015). The second refers to nonparametric methods, such as the propensity score matching (PSM) approach (Dohmwirth & Liu, 2020; Zheng & Lu, 2021), augmented inverse probability weighted (AIPW) estimator (Hossain & Samad, 2021), and inverse probability weighted regression adjustment (IPWRA) estimator (Zheng & Ma, 2021).…”