Dynamic linkages between element input structure, global value chain, and carbon emissions: evidence from a heterogeneous panel cointegration model

Industrial CO2 constitutes a dominant component of global CO2 emissions. Therefore, modifications in industrial production processes are crucial. We analyse global cross-national panel datasets from 1995 to 2018 using the panel autoregressive distributed lag (ARDL) model to provide comprehensive evidence of the combined effects of input-side element structures and production-side global value chain (GVC) divisions on output-side CO2 emissions. The results show that: (i) Renewable energy and service elements in production factors can reduce CO2 emissions, though their abatement effects vary. Both can produce sustained carbon mitigation effects in the long term, but only renewable energy is effective in the short term. (ii) The GVC division pattern can induce a CO2 emission increase in both the short and long term, though these effects are heterogeneous depending on the differences in forward-backward GVC divisions. (iii) A path dependence analysis shows that input-side element structures can change production-side GVC divisions, thereby affecting output-side emissions. This study extends our understanding of the environmental effects of various production links and helps coordinate high-quality economic development and impending environmental constraints.