Enterprise Computing Power Deployment, Data as a Production Factor and Digital Innovation: Causal Inference Based on Double Machine Learning

Using a panel of Chinese A-share listed firms from 2017 to 2022, this paper estimates the causal effect of firm-level intelligent computing centers on digital technology innovation. We construct a novel firm-level measure of intelligent computing center from firms’ Internet Data Center (IDC) operating license records and apply a double machine learning (DML) framework to control for high-dimensional confounders. The results show that ownership of an ICC significantly increases firm-level digital technology innovation.Mechanism analysis indicates three primary channels: (1) converting raw data into certifiable, data assetization(monetizable data assets); (2) accelerating the accumulation and recombination of digital knowledge; and (3) enabling balanced cross-regional deployment of computing power that expands data flows and collaborative innovation. Heterogeneity tests show the effect is stronger in regions with more advanced digital infrastructure and policy support, and is concentrated among high-tech firms and non–asset-intensive industries. Although environmental regulation affects intelligent computing centers siting, it does not materially attenuate their innovation-promoting impact. By providing micro-level causal evidence on how enterprise intelligent computing centers deployment unlocks data as a production factor, the study offers actionable implications for infrastructure planning, data governance, and regional digital development.