The impact of energy prices and socio-economic factors on CO2 emissions in OECD countries: A STIRPAT and machine learning analysis

Climate change presents a major hurdle for OECD countries, which have accounted for substantial historical greenhouse gas emissions and now face the task of reducing carbon dioxide (CO 2 ) emissions while sustaining economic growth. This research investigates the influence of oil price dynamics, Brent, OPEC, and West Texas Intermediate, and socio-economic factors on CO 2 emissions in OECD countries from 1990 to 2020. Using the STIRPAT framework, we estimate static and dynamic panel regression models to assess the effects of oil price benchmarks alongside variables such as Gross Domestic Product (GDP), urbanization, and education. Granger causality tests evaluate directionality, while artificial neural networks serve as robustness checks. Results show a significant inverse relationship between oil prices and CO 2 emissions, signifying that increased oil prices are associated with lower emissions, as they encourage conservation, efficiency, and cleaner energy transitions. Socioeconomic factors are also essential, with GDP growth and urbanization contributing to variability in emissions across countries. These findings highlight the significance of designing differentiated, context-sensitive mitigation strategies. Policy recommendations include adopting energy pricing reforms, such as carbon taxes, to internalize environmental costs, while also supporting education, urban planning, and the adoption of renewable energy to strengthen long-term emission reduction efforts. • Investigates the impact of energy prices and socio-economic factors on emissions. • Employing machine learning techniques. • Reveals an inverse relationship between energy prices and CO 2 emissions. • Artificial neural networks indicate the critical role of education in shaping emissions.