Abstract
A provincial analysis of Okun’s law in South Africa is provided in this article from 1996 to 2016. Empirically, we rely on the nonlinear autoregressive distributive lag (N-ARDL) model whilst the Corbae-Ouliaris filter is used to extract the ‘gap’ variables required for our regression estimates. Okun’s law is found to be significant hold in the long-run exclusively for the Western Cape and Kwa-Zulu Natal provinces whereas the remaining provinces partially display significant short-run effects. Our sensitivity analysis in which panel N-ARDL estimations for all provinces finds insignificant long-run Okun effects for the country as a whole, whilst validating the relationship only in the short-run. Our study hence advises that the epicenter of policy efforts in addressing the country’s high unemployment and low economic growth dilemma should be concentrated at a provincial level.