Policy effects of belt and road initiative on corporate green transformation: empirical tests based on dual machine learning model

The Belt and Road Initiative (BRI) aims to foster the harmonious coexistence of man and nature with a strong emphasis on sustainable development. However, it remains uncertain whether and how this initiative can enhance corporate green transformation (CGT). This research builds a comprehensive evaluation index system and a combination empowerment-TOPSIS model based on game theory to measure the level of CGT. Then, we regard the BRI as a quasi-natural experiment (QNE) and use dual machine learning (DML) methods to assess its impact and mechanisms. The findings reveal that the BRI has significantly enhanced CGT, and this relationship is consistently supported by multiple robustness tests. The mechanism analyses demonstrate that the BRI has advanced CGT through technology, configuration, and structural effects. Heterogeneity analyses find that the BRI more substantially facilitates the green transformation of non-heavy pollution, non-state-owned corporations, and firms with low environmental information disclosure. Further analysis reveals that the BRI and the National Big Data Comprehensive Pilot Zone (NBDCPZ) exhibit a synergetic effect in advancing green transformation among participating firms. These findings offer insights for optimizing China’s open-door policy and fostering sustainable corporate growth.