South Africa's persistent inequality despite sustained economic growth and structural transformation contradicts the Kuznets curve hypothesis and raises question on whether other structural factors matter. This study revisits the Kuznets curve hypothesis in South Africa accounting for the roles of renewable energy and the health expenditure burden while controlling for urbanization. The study applies the Kernelized Regularized Least Squares (KRLS) machine learning approach over the period 2000Q1 to 2022Q4 to uncover both average and distribution-sensitive effects. Further, it applies the quantile-on-quantile KRLS (QQKRLS) method to test the full spectrum of the distributional effects and verify the robustness of the estimates. The average marginal effect results show that rising income levels in South Africa are associated with higher inequality in the early stage, while the inequality-reducing phase or the turning point predicted by Kuznets is yet to materialize. Similar findings are confirmed for the distributional effects. Further, both the marginal and the distributional effects of renewable energy adoption reveal consistent reduction in inequality, while the health expenditure burden in terms of out-of-pocket (OOP) health spending increases inequality. Moreover, the distributional effects demonstrate that the health expenditure burden shifts the Kuznets curve upward and delays any inequality-reducing turning point. Additionally, the results show that urbanization in South Africa exacerbates inequality. These findings suggest that combating inequality in South Africa requires policies that complement the already existing health financing protection in South Africa such as reducing OOP costs for definitive care and promoting inclusive energy access through policies such as renewable energy subsidies and incentives. • Applies machine learning methods to estimate nonlinear growth–inequality dynamics. • Out-of-pocket health expenditure exacerbates inequality. • Renewable energy adoption reduces inequality across income groups. • Results inform equitable health financing and resource allocation policy.
Iorember et al. (2026) studied this question.