Green Innovation's Heterogeneous Effects on Carbon Intensity: Spatial Spillovers in China

Abstract

Abstract:This study examines the effect of green technological innovation on carbon dioxide (CO2) intensity across Chinese provinces from 2000 to 2020. It distinguishes between invention patents and utility model patents to identify heterogeneous mechanisms across regional, regulatory and spatial contexts. Employing spatial Durbin models based on 630 province-year observations, the study reports three principal findings. First, green innovation is associated with lower CO2 intensity nationally, with an estimated elasticity of –0.111. This relationship remains robust across one- to three-year lag structures and dynamic panel (system-GMM) specifications, reflecting differential absorptive capacity and industrial structure constraints. Second, environmental regulation exhibits a threshold effect at –5.618, corresponding to 0.37 per cent of gross domestic product. Above this threshold, the effectiveness of green innovation in reducing CO2 intensity declines by approximately nine per cent, suggesting that excessive regulation is associated with diminished innovation productivity. Third, utility model patents generate substantially stronger spatial spillovers, with an indirect effect of –0.399, compared with 0.014 for invention patents. Moreover, the spillover effect of utility model patents exceed their direct effects by nearly threefold, indicating that incremental innovations diffuse more readily across provincial boundaries than breakthrough technologies. These findings suggest that optimal carbon mitigation policy should strengthen region-specific absorptive capacity-building, maintain moderate environmental regulation near the identified thresholds, and prioritise utility model patent development within regional cooperation frameworks to maximise interprovincial technology diffusion and collective emissions reduction.