The impact of energy intensity, urbanisation, industrialisation, and income on CO2 emissions in South Africa: an ARDL bounds testing approach

We examine the dynamic short and long-run relationship between urbanisation, industrialisation, energy intensity, per capita GDP, and CO2 emissions in South Africa from 1970 to 2016. The Autoregressive Distributed Bounds Testing Approach (ARDL) is employed to calculate short-run and long-run relationships in the presence of structural breaks and the vector error correction model is applied to determine the direction of causality. The bounds tests suggest that the five variables are bound together in the long run when CO2 emission is the dependent variable and that urbanisation has the largest impact on CO2 emissions. Granger causality tests show a strong bidirectional causality among all the variables, with the exception of GDP and energy intensity where there is a strong unidirectional relationship running from energy intensity to GDP per capita. Our findings suggest that policymakers must develop comprehensive policies for mitigating CO2 emissions, particularly those that focus on managing rapid urbanisation.