Modelling the circular economy–renewable integration nexus: a cross-national empirical analysis of smart-technology adoption, territorial CO₂ emissions, and the material footprint

Climate neutrality relies on two levers usually examined separately: circular-economy (CE) practices and renewable integration. Drawing on ecological modernization theory and the environmental Kuznets curve (EKC), we regress per-capita carbon dioxide (CO₂) emissions in 20 European Union member states over 2014–2023 ( N = 200) on the circular material use (CMU) rate and the renewable energy (REN) share, with smart-technology adoption [ST; internet penetration and research and development (R&D) intensity] as a moderator. The pooled ordinary least squares (OLS), two-way fixed-effects (FE), and random-effects models use cluster-robust standard errors and a wild cluster bootstrap. A 1-percentage-point rise in the renewable energy share is associated with approximately 0.9% lower per-capita emissions in the baseline FE model ( p = 0.095) and 0.95% in the moderated model ( p = 0.030; bootstrap p = 0.048). CMU has no significant unconditional association with emissions in the levels models, and income exhibits concave curvature, although the implied EKC turning point lies above the sample maximum. The CMU × ST interaction is positive and significant ( β = 0.0065, p = 0.009; bootstrap p = 0.033) and is robust across all 20 leave-one-country-out subsamples: the circularity–emissions slope becomes less favourable as adoption rises, with the estimated interaction coefficient larger for internet diffusion than for R&D intensity, although the two estimates cannot be statistically distinguished. Because no conditional slope is individually significant, this association is consistent with, but does not demonstrate, a digital-rebound mechanism. For the material footprint outcome, the CMU coefficient is negative, and the interaction changes sign, but neither is significant, so the pattern is specific to production-based CO₂ accounting.