The relationship between renewable energy production and CO₂ emissions in 27 OECD countries: A panel cointegration and Granger non-causality approach

Human-caused CO₂ emissions are the primary cause of global warming. In this regard, determining the most effective approach for lowering CO₂ emissions and the collateral risk of catastrophic natural disasters is crucial. This study examines the long-run relationship between disaggregated renewable energy production and carbon dioxide (CO₂) emissions per capita for a panel of 27 OECD countries from 1965 to 2020. The panel-autoregressive distributed lag (ARDL) models of the pooled mean group (PMG), mean group (MG), and dynamic fixed effect (DFE) were used to evaluate the relationship between CO₂ emissions and energy production from biofuel, aggregated geothermal and biomass (GEOB), hydropower, nuclear, solar, and wind. As robustness checks, fully modified ordinary least squares (FMOLS), dynamic ordinary least squares (DOLS), and common correlated effects mean group (CCEMG) estimators were used. Then, using a generalized method of moment (GMM) framework for panel vector autoregression (PVAR), the Granger non-causality between CO₂ emissions and renewable energy production was investigated. GEOB, hydropower, nuclear, solar, and wind were found to be negatively and significantly correlated with CO₂ emissions. GEOB, hydropower, and solar were the most effective renewable resources in reducing CO₂ emissions. Granger non-causality approach showed unidirectional causation from hydropower, solar, and wind to CO₂ emissions, bidirectional causation between CO₂, and biofuel and GEOB, and unidirectional causation from CO₂ emissions to nuclear. The findings were consistent across different model specifications and suggested a faster transition to GEOB, hydropower, and solar energy in OECD countries to reduce CO₂ emissions and enhance environmental sustainability.