Wage inequality, firm characteristics, and firm wage premia in South Africa

This paper investigates the role of firm characteristics in driving wage inequality and firm wage premia in the South African labour market. The Abowd, Kramarz, and Margolis (AKM) and Kline, Saggio, and Sølvsten (KSS) regression-based decomposition methods are applied to matched employer–employee administrative tax data for the period 2011–19. Additionally, the Theil index is used as a comparative tool for estimating wage inequality, given that the variance of logarithms applied in the regression-based decomposition methods has been established as an imprecise measure of inequality. The results show significantly high dispersion in wages, as estimated by both the AKM and the KSS methods as well as the Theil index, reaffirming the extent of high inequality in the country. Worker and firm characteristics account for 35 per cent and 18 per cent of wage dispersion, respectively, with a positive worker–firm covariance accounting for 11 per cent. Firm size, industry, profits, geographical location, and whether firms are locally or foreign-owned are found to be important in driving firm wage premia.