Gender-union wage gap in South Africa: an unconditional quantile regression & coarsened exact matching approach

Journal Title
Journal ISSN
Volume Title
University of the Witwatersrand, Johannesburg
South African women have been fairing badly in the labour market compared to their male counterparts since pre-colonial times due to both cultural and legal restrictions (and discrimination). Despite the corrective measures (e.g. legislation changes) that were introduced by the South African government post-apartheid to improve women’s social status, labour market conditions remain unfavorable for women. Labour unions are actively advocating for gender equality as well as higher and more equitable salaries. However, the extant literature on unions and wages pay more attention on the impact of unions and wages. There is paucity of literature on the moderating role of unions on the gender-wage inequality. This study contributes to the literature by investigating the impact of labour unions on the gender-wage inequality in South Africa, using the National Income Dynamics Study (NIDS) data obtained from the DataFirst website for the 2008 to 2017 waves. Furthermore, this study uses more recent econometric techniques, viz. unconditional quantile regression (UQR) and coarsened exact matching (CEM) methods with interaction effects, on a panel of 12,881 individuals. These methods are superior than the ones used in the extant literature as they control individual heterogeneities, sample attrition and selection bias that may arise from individuals’ decisions to join unions. Both the CEM and UQR results suggest a strong positive relationship between unions and wages in South Africa. In addition, both models suggest that labour unions narrow down the existing gender-wage gap in South Africa. However, this impact is weakened at higher-ends of the income distribution as the magnitude by which unions reduce the gender-wage gap reduces at higher percentiles (i.e. the 75th and 90th percentiles). This is possibly due to high-incomeearners being individuals with higher productive abilities (such as higher levels of education and additional skills) and therefore having more bargaining power to negotiate their own wages in the absence of unions. Another possible explanation could be that high productive abilities are a signal to employers that an individual will be able to do the job better than individuals with low productive abilities. This leads to employers being able to better estimate wages for such individuals outside the bargaining power of unions. These findings are important as they indicate that either unions have a weakened bargaining power for high-income earners, or that the high-income earners experience less gender-wage discriminations. It is therefore worthwhile for policymakers to analyse such trends before implementing “one-for-all” union policies and other related policies aimed at reducing gender inequality
A research report submitted in partial fulfillment of the Degree of Master of Economic Science in the School of Economics and Finance at the University of the Witwatersrand, 2023
South African women, Labour market, Post-apartheid, Women’s social status, Gender equality, UCTD
Hlapisi, Nthabiseng. (2023). Gender-union wage gap in South Africa: an unconditional quantile regression & coarsened exact matching approach [Master’s dissertation, University of the Witwatersrand, Johannesburg]. WireDSpace.