Embodied AI-Enabled Production Technology Intensity and Corporate Carbon Efficiency: Evidence from Chinese Listed Firms

Evolutionary and Institutional Economics Review · Springer Nature · 2026-11-17 · DOI: 10.1007/s10614-026-11489-3

Аннотация

This study examines whether embodied AI-enabled production technology is associated with higher corporate carbon efficiency, using a panel of Chinese A-share listed firms from 2013 to 2021. We measure its intensity by the natural logarithm of machinery and equipment book value per employee, a capital-based measure that also reflects conventional automation and capital deepening. To mitigate reverse causality, we instrument this measure with the leave-one-out mean among firms in the same broad industry, province, and year. The IV estimates imply that a 10% increase in machinery intensity is associated with an increase of about 2% in operating revenue per unit of carbon emissions. The relationship persists after conditioning on lagged productivity and financing constraints. Point estimates are larger for large firms and positive for non-state-owned firms. These subgroup differences are suggestive rather than conclusive. The evidence concerns the pre-2022 phase of embodied, production-side technology in Chinese listed firms, before generative and software-based AI applications diffused rapidly. It links the machinery-intensity proxy to the production–carbon trade-off that corporate climate risk governance seeks to improve.

Читать у издателя →

Похожие статьи