National innovation capacity and the decoupling of economic growth from carbon emissions evidence from dynamic panel analysis

Abstract This paper examines whether national innovation capacity accelerates the decoupling of economic growth from carbon dioxide (CO₂) emissions and how institutional quality and policy context condition this relationship. Moving beyond emissions-level analyses and Environmental Kuznets Curve reasoning, the study adopts a decoupling-index framework that captures changes in carbon intensity rather than emissions alone. Using a balanced panel of 47 countries over the period 2005–2024, the analysis employs fixed-effects estimation, dynamic System GMM, and panel quantile regression to account for unobserved heterogeneity, endogeneity, nonlinearity, and distributional differences in decoupling performance. The results show that innovation contributes to growth–emissions decoupling within countries, but its effects are neither uniform nor automatic. Innovation exhibits clear nonlinearities, with diminishing and, in some contexts, adverse marginal effects at higher levels of innovation capacity. Institutional quality improves decoupling in some settings but operates primarily as a complementary rather than mediating factor. Distributional estimates reveal substantial heterogeneity, with innovation exerting weak effects in low-decoupling regimes and increasingly negative marginal effects in high-decoupling regimes. Evidence from the post-2015 period further suggests that innovation becomes relatively more effective when embedded within a coordinated international climate policy framework. The findings challenge simplistic innovation-led decarbonization narratives and highlight the importance of policy alignment, institutional context, and innovation direction in achieving environmentally sustainable growth. JEL Classification O33, Q53, Q55, C23, Q01, O43