Abstract: [Purpose/Significance] From the perspective of reproduction of policy content, this paper extracts the referential relationships of local provincial policies to national policies in digital transformation of China's government. This is beneficial to reveal the patterns of the provincial-level governments in implementing, adapting, and expanding national policies, providing support for optimizing digital transformation policies. [Design/Methodology] Taking the policy texts of government digital transformation as the analysis objects, this paper constructs a dataset of the referential relationships of provincial-level policies to national policies, calculates the reproduction coefficients of national policies content based on the reference relationship, depicts the spatio-temporal characteristics and laws of the reproduction of policy content, and quantitatively reveals the degree of refinement and updating of provincial-level governments in implementing national policies. [Findings/Conclusion] The local governments and the central government within each phase of the development of digital transformation policies in China are closely synergy, and national policies are diffused and innovated by provincial-level governments. All provincial-level governments can form blueprints with local characteristics relatively well. The degree of policy content reproduction increases over time, with the eastern region exhibiting a higher degree than the central and western regions. [Originality/Value] This study explores the strategy for measuring the reproduction of content in government digital transformation policy by integrating the policy reference relationships. It reveals the intrinsic policy diffusion patterns at the content level, thereby facilitating for uncovering the experience of policy synergy in government digital transformation and promoting the high-quality development of government digital transformation in China.

Keywords: Government digital transformation, Reproduction of policy content, Policy reference relationship, Policy text, Similarity calculation