Computation network and carbon neutrality: A quasi-natural experiment by externality of institutional support to energy security

Abstract

The computation network as new infrastructure holds strategic significance in ensuring energy security towards carbon neutrality. This study examines the impact of computation network construction on carbon emissions, addressing evidence through the institutional support of China’s National Computing Power Hub Policy (NCPHP) as a quasi-natural experiment. The policy seeks to transfer computational demand westward by establishing an integrated network, which serves as the backbone of the nation’s “Eastern Data, Western Computing” strategy. Empirical analysis employs a difference-in-difference-in-difference (DDD) approach to estimate the pro-environmental and spatial effects of NCPHP in eastern regions, using panel data from 30 provinces spanning 2006 to 2022. The findings indicate that the policy significantly reduces carbon emissions in eastern provinces and generates positive green externalities by alleviating environmental pressure through western computing supply, coordinated via the computation network. These results stand across a series of robustness checks and endogeneity adjustments. Additional computing social science methods are employed to deepen the discussion on empirical evidence. Bipartite matching and maximum flow programming are developed to identify key “trunk roads” and assess the performance of the integrated computation network, aligning with NCPHP’s design. This study provides valuable insights into fostering inclusive computational development from an energy security perspective within the carbon neutrality framework.