Globalization and Environmental Quality in Asia: New Evidence from Double Machine Learning and Heterogeneous Treatment Effect Estimation

Globalization is widely recognized as a key determinant of environmental sustainability; however, empirical evidence on its environmental effects remains inconclusive. Existing studies predominantly employ conventional econometric approaches that identify associations rather than causal relationships, leaving estimates vulnerable to confounding, reverse causality, and omitted variable bias. Although panel quantile regression has revealed distributional heterogeneity in globalization–environment relationships, it does not provide causal causal effects. Recent advances in causal machine learning, particularly Double/Debiased Machine Learning (DML), offer a rigorous framework for estimating causal effects in high-dimensional settings. Despite its growing application in environmental economics, DML has not been applied to the multidimensional globalization–environment nexus using the KOF Globalisation Index. This study proposes to employ the DML framework to estimate the causal effects of the seven KOF globalization dimensions on carbon footprint (CFP) and CO₂ emissions. Furthermore, generalized random forests will be used to estimate heterogeneous treatment effects and identify country-specific pathways through which globalization influences environmental outcomes. The proposed research is expected to advance both the methodological and empirical literature by providing robust, policy-relevant evidence on the causal environmental impacts of globalization.