Determinants of Carbon Dioxide Emissions: New Empirical Evidence from MENA Countries

Abstract

The paper emplyos the Pooled Mean Group (PMG) estimation to investigate empirically the relationship among carbon dioxide (CO2) emissions and potential determinants for a panel of 18 MENA countries during the period 1980-2018. The good properties in terms of consistency and efficiency of the coefficient estimates make the PMG approach very useful for examining the determinants of pollution emissions in the framework of dynamic heterogeneous panel data models over both the long- and short-run. Unlike the extant literature on MENA economies, many determinants are included in the analysis to avoid the bias problem of omitted variables. Three energy sources and two classes of sub-panels according to regional proximities and oil wealth are considered in order to provide a sensitivity check on the findings and to make the analysis more homogeneous. The results reveal long-run relationships between pollution emissions and the selected variables. All determinants are found to be statistically significant for all panels and energy sources over the short-run. However, some variables are not significant determinants over the long-run. The Environment Kuznet's Curve (EKC) hypothesis is supported only for the panel of non-oil countries, which has meaningful implications and reveals the importance of splitting the global panel in order to appropriately examine the EKC hypothesis and conduct policy debates according to the findings of each panel. The obtained results provide important policy implications.