Аңдатпа
China’s “East-Data West-Computing" strategy and carbon-neutrality targets are jointly reshaping the risk structure of the data center industry chain. Spatial reallocation creates new cross-regional dependencies, while carbon constraints impose heterogeneous compliance, financing, and energy-transition pressures on electricity-intensive firms. Existing single-layer network or financial-risk frameworks cannot fully capture how transaction ties, equity links, and carbon-exposure dependence jointly transmit systemic stress. This study constructs a directed and weighted three-layer network of transaction, equity, and carbon-exposure relationships among data-center-related enterprises, and develops a carbon-constrained SIRS-type contagion model. In this economic setting, the infected state denotes firm-level distress, and firm-level carbon emission intensity enters the infection-rate function as a susceptibility amplifier. The main innovation is to treat carbon exposure not merely as an external regulatory background, but as both a structural contagion layer and a source of heterogeneous firm vulnerability. Using a 2018–2023 panel of 1243 enterprises, with 2018–2021 as the pre-rollout baseline period and 2022–2023 as the early post-rollout period, we compare carbon pricing, renewable subsidies, technology mandates, and their combined package. The results show that stand-alone carbon pricing reduces emissions but raises the short-run peak infection rate, interpreted as peak distress incidence. By contrast, technology-oriented instruments reduce contagion pressure by lowering firm-level carbon intensity, and the combined package achieves the best joint outcome in systemic-risk reduction and emission abatement. The findings extend industry-chain contagion research by embedding carbon-exposure dependence into both network structure and firm-level distress susceptibility. They also support a stability-aware low-carbon transition governance framework for digital infrastructure systems.