Different impacts of democracy and income on carbon dioxide emissions: evidence from a panel quantile regression approach

The current research investigates the linear and non-linear effects of democracy and income levels on carbon dioxide emissions for 72 countries from 1993 to 2019. Considering the possibility that the impact of independent variables on carbon dioxide emissions may vary throughout the distribution of carbon dioxide emissions, the correlations between the variables are studied by the panel quantile regression method. The findings confirm the EKC hypothesis in countries having low and medium emissions and demonstrate an inverted U-shaped correlation between GDP and carbon dioxide emissions. In low- and medium-emission countries, a higher democracy level contributes to environmental quality, while it leads to increased carbon dioxide emissions in high-emission countries. On the contrary, it is remarkable that democracy causes a delay in the turning point of the EKC in countries with low and medium emissions. Furthermore, the results demonstrate an EKC correlation between democracy level and carbon dioxide emissions in countries with low and medium emissions. However, the correlation between democracy level and carbon dioxide emissions, which is inverted U-shaped in low- and medium-emission countries, turns into a U-shaped correlation in high-emission countries. These findings emphasize the necessity of considering the changing impact of democracy according to the emission structure of countries in policymaking.