Measuring the determinants of CO2 emissions in Türkiye: An instrumental variable regression analysis

This research study investigates the determinants of CO2 emissions in Türkiye using instrumental variable (IV) regression analysis. The research focuses on the influence of energy use (ENU), population density (PD), foreign direct investment (FDI), and GDP growth on CO2 emissions. This study Employ a two-stage least squares (2SLS) regression approach, take GDP as endogenous variable, military expenditure, and ICT goods export as instrument variables. The results suggest varying impacts of different variables on CO2 emissions. While foreign direct investment (FDI) demonstrates a statistically insignificant effect, energy use (ENU) emerges as a significant predictor of CO2 emissions (p <0.005), indicating that an increase in ENU leads to a corresponding rise in CO2 emissions. Similarly, Population density (PD) also demonstrates a significant impact on CO2 emissions (p <0.002). Conversely, while GDP is found to be a significant predictor (p <0.047), it shows a negative association with CO2 emissions. However, it's noteworthy that endogeneity tests reveal significant endogeneity of the variables. Additionally, the robustness check using the Generalized Method of Moments (GMM) regression confirms these results. These findings highlight the importance of further investigating the strength of instrumental variables to ensure the reliability of the conclusions.