How can new energy vehicle adoption affect urban ecological resilience? Evidence from a double/debiased machine learning method

New energy vehicles (NEVs) are increasingly regarded as a driver of energy transition. However, few scholars have conducted quantitative research on the impact of the large-scale adoption of NEVs on urban ecological resilience (UER). Based on the Pressure-State-Response framework adopted to assess UER, this study utilizes panel data from 294 cities in China from 2017 to 2023 and applies a double/debiased machine learning approach to analyze the impact of NEV adoption on UER and the mechanisms driving this effect. The findings yield four main conclusions. First, the adoption of NEVs significantly increases UER. A 1% rise in NEV stock leads to a 0.0132% increase in UER, and this outcome remains stable through multiple robustness tests. Second, the mechanism exploration demonstrates that the adoption of NEVs improves UER through three channels: industrial structure upgrading, boosting green technological innovation capabilities, and raising public environmental awareness. Third, heterogeneity analysis reveals that NEV adoption has a stronger effect on UER in western cities, non-resource-based cities, and cities with stricter environmental regulations. Fourth, sub-dimensional analysis indicates that the adoption of NEVs mainly improves the state criterion and response criterion of UER, while improvements in the pressure criterion occur only in non-resource-based cities and cities with stricter environmental regulations. This study offers empirical evidence demonstrating how NEV adoption contributes to UER and drives sustainable development, emphasizing that this process requires participation from multiple stakeholders including governments, enterprises, and the public.