In the context of rapid urbanization and climate change, evaluating urban forest development and the effectiveness of related policies is of great significance. This study takes Chinese prefecture-level cities as the research object and constructs an evaluation system for Urban Forest Development Effectiveness (UFDE), encompassing forest networks, forest health, ecological welfare, and development coordination. The analytic hierarchy process–entropy weight method is employed to measure UFDE. On this basis, leveraging the quasi-natural experiment formed by the staggered implementation of the National Forest City Policy (NFCP), this paper applies double machine learning (DML) to identify the causal effects of the policy. The results show that NFCP significantly improves UFDE, and this conclusion remains robust across various model specifications and robustness checks. Meanwhile, the policy effects exhibit significant heterogeneity, being more pronounced in eastern and central regions, as well as in humid climate zones, while being relatively weaker in western and arid regions. Methodologically, this study introduces DML to enhance the precision of causal identification, and in terms of measurement, it achieves a multidimensional, comprehensive evaluation. It provides a new analytical framework for assessing environmental policy effectiveness and offers empirical evidence for optimizing urban ecological governance and promoting green development.
Liu et al. (Sat,) studied this question.