A New Integrated Approach for Evaluating Sustainable Development in the Electric Vehicle Sector

Abstract

This study develops an innovative value creation process for the electric vehicle (EV) industry. First, this study conducts data envelopment analysis to measure the innovation, operation, and market efficiency performance of the EV industry. Second, this study conducts bootstrapped truncated regression to explore the impact of environmental, social, and governance (ESG) factors on the performance of the EV industry. Third, this study uses the classification & regression tree (CART), random forest, and eXtreme gradient boosting (XGBoost) algorithms to assist managers in identifying the key predictive variables for further classification and prediction. Results reveal significant differences in innovation performance across five industry sectors, among which the charging pile system sector exhibits the highest average value, and the battery system sector exhibits the lowest average value. The truncated regression analysis shows that innovation performance in Taiwan's EV industry is significantly influenced by energy management, data security, employee information statistics, and control over equity and board seats. Corporate governance transparency positively impacts operational performance, while energy and water management enhance market performance, with product quality and safety having a negative effect on market performance. This study identifies the relative importance of the classification attribute variables based on the classification rules of the target attributes by conducting further analysis with the CART decision model and constructs an optimal prediction model.