Research Information System
Energy Strategy Reviews, vol.65, 2026 (SCI-Expanded, Scopus)
Accurate assessment of embedded carbon emissions (ECE) is critical for designing effective climate governance strategies and promoting the low-carbon transformation of export-oriented industries. Unlike traditional multi-regional input–output approaches that focus on sectoral trade linkages, this study adopts a firm-level perspective to analyze the drivers of ECE in China's light industry. Using panel data for Chinese A-share listed light-industry firms from 2010 to 2024 , a two-way fixed-effects econometric model is employed to estimate the impacts of export intensity, energy structure, green innovation, firm size, profitability, and capital intensity on embedded carbon emissions. The empirical results show that export intensity significantly increases firms' ECE (β = 0.171, t = 4.55, p < 0.01) , indicating that expansion of export-oriented production remains associated with higher carbon burdens. Conversely, green innovation significantly reduces ECE (β = −0.065, t = −3.17, p < 0.01) , while improvements in clean energy utilization also exhibit a significant carbon-mitigation effect (β = −0.118, t = −4.07, p < 0.01) . Firm size and capital intensity are positively associated with emissions (β = 0.054 and 0.039, respectively; p < 0.05), reflecting the scale effect of industrial production, whereas profitability shows a modest negative relationship with ECE (β = −0.021, p < 0.10), suggesting that more efficient firms tend to achieve better carbon performance. Economically, the results imply that a 1% increase in export intensity leads to a 0.17% rise in embedded carbon emissions, whereas a 1% increase in green innovation reduces ECE by approximately 0.065% . Heterogeneity analysis further reveals that the export-driven carbon effect is stronger in heavy-polluting subsectors and non-high-tech firms, while the carbon-reduction effect of green innovation is more pronounced in large firms and state-owned enterprises. Moderating-effect estimations indicate that environmental regulation (β = −0.041, p < 0.05) and digital transformation (β = −0.036, p < 0.05) significantly weaken the positive relationship between exports and ECE. Robustness checks using alternative emission measures, lagged variables, and instrumental-variable estimation confirm the stability of the findings. Overall, this study provides new firm-level evidence on the carbon consequences of export-oriented growth in China's light industry and offers important policy implications for improving carbon efficiency, encouraging green innovation, and advancing sustainable export development under China's dual-carbon goals.