Beyond digital investment: How value embedding modes shape ESG performance in energy-intensive firms

There is no consensus in existing literature on how digital transformation facilitates sustainable development in energy-intensive enterprises. Based on the Resource-Based View and Dynamic Capabilities Theory, this paper proposes the “Value Embedding Mode” analytical framework, which categorizes the value conversion pathways of digital resources into asset-embedded and information-embedded pathways, and empirically examines this framework using panel data from listed companies in China's energy-intensive industries from 2011 to 2024. The study finds a positive association between digital transformation and ESG performance; however, this relationship does not simply stem from the scale of digital investment but depends on whether digital resources can realize value conversion through effective embedding mechanisms. Specifically, the asset-embedded pathway uses green innovation as a mediating channel; however, due to structural factors such as the capital-intensive nature of energy-intensive industries and long asset renewal cycles, its mediating effect is not robustly supported. In contrast, the information-embedded pathway, which relies on improvements in the external information environment as its channel, exhibits more stable effects on value conversion. This paper reveals the underlying logic by which the same digital resources generate differentiated ESG effects depending on the embedding objects and value realization approaches, offering a new theoretical perspective on the heterogeneous outcomes of digital transformation in energy-intensive enterprises.