The Impact of Digital Transformation on Electricity Trading Volume: A Double Machine Learning Approach

This study focuses on the effect of prefecture-level digital transformation initiatives on electricity trading volume in China, employing both a traditional Difference-in-Differences (DID) model and a Double Machine Learning (DML) approach. The study initially applies the DID model to ascertain the policy effect of digital transformation on power trading. The application of DML has been instrumental in isolating the policy effect of digital transformation, offering a robust framework for causal inference in the presence of high-dimensional and nonlinear data. Our findings suggest that digital transformation at the prefecture level results in a significant reduction in electricity trading volume, and this finding remains robust even after adding control variable quadratic terms or altering the sample splitting ratio and machine learning approach, implying a potential enhancement in energy efficiency and sustainability. This paper enriches the existing literatures by providing a nuanced analysis of the intersection between digital transformation and energy markets, offering insights into the operational efficiencies gained through technological advancements. The results have implications for policy-making and the strategic allocation of resources within the energy sector.