From text to effectiveness: Quantifying green industrial policies in China

Abstract

The evolution of green industrial policy in China is deeply embedded within a unique political, economic, cultural, and social milieu. The intricacies and complexities inherent in this context present challenges for quantitative policy research. In the existing literature, the study of thematic clustering of green industry policy instruments and their effectiveness measures in large-scale policy text repositories remains a relatively unexplored area. To fill this research gap, this study provides an analysis of China's green industrial policies from 1994 to 2022. By adopting the Latent Dirichlet Allocation (LDA) topic model, implicit and optimal themes are identified, and key policy instruments are systematically and comprehensively parsed. Based on this, the cutting-edge Policy Modeling Consistency-Text Encoder (PMC-TE) index model was used to conduct an in-depth quantitative evaluation of various policy tools. The findings reveal 15 policy instruments from 7891 Chinese green industry policy documents. Notably, technical support, information resource demonstration project policies emerged as the most frequently deployed and efficacious. Conversely, government procurement and service outsourcing policies, while showing incremental growth year-on-year, lagged in utilization and effectiveness. These policy instruments can be broadly categorized into three primary archetypes: supply-type, demand-type, and environment-type. The supply-type instruments dominate, while environment-type instruments remain underutilized. Additionally, this study delineates the four chronological phases of China's green industrial policy evolution, illuminating China's strategic pivots-transitioning from resource allocation, to market dynamics, and ultimately to environmental governance trajectories.