Disaggregating the environmental effects of renewable and non-renewable energy consumption in South Africa: fresh evidence from the novel dynamic ARDL simulations approach

Previous studies have widely used the aggregate energy consumption in the energy–growth–CO 2 emissions nexus, which may not show the relative strength or explanatory power of several energy sources on CO 2 emissions. However, less explored in empirical literature are the effects of disaggregated levels of renewable and non-renewable energy sources on environmental quality. This study therefore contributes to fill this important gap for South Africa over the period 1960–2019. Our strategy is distinctively different from previous works in the following dimensions: we employ the recently developed novel dynamic autoregressive distributed lag (ARDL) simulations framework proposed by Jordan and Philips (Stand Genomic Sci 18(4):902–923, 2018) to examine the negative and positive changes in the disaggregated levels of renewable and non-renewable energy sources, trade openness, technique effect, and scale effect on CO 2 emissions. Second, we use an innovative measure of trade openness developed by Squalli and Wilson (World Econ 34(10):1745–770, 2011) to capture trade share in GDP as well as the size of trade relative to world trade for South Africa. Third, we use the frequency-domain causality (FDC) approach, the robust testing strategy suggested by Breitung and Candelon (J Econ 132(2):363–378, 2006) which enables us to explore permanent causality for medium-, short-, and long-term relationships among variables under review. Fourth, we employ the second-generation econometric procedures accounting robustly the multiple structural breaks which have been considerably ignored in earlier studies. For South Africa, the key findings are as follows: (i) hydroelectricity and nuclear energy consumptions contribute to lower CO 2 emissions in the long run; (ii) the scale effect increases CO 2 emissions whereas the technique effect improves it, validating the presence of an environmental Kuznets curve (EKC) hypothesis; and (iii) oil, coal, and natural gas consumptions deteriorate environmental quality. In the light of our empirical evidence, this paper suggests that South Africa’s government and policymakers should effectively study the optimal mix of all available energy resources to meet the increasing energy demands while improving the country’s environmental quality.