Does a motherhood penalty exist in the post-apartheid South African labour market

Type Working Paper - SALDRU Working Paper Number 247 Version 1/ NIDS Discussion Paper 2019/14
Title Does a motherhood penalty exist in the post-apartheid South African labour market
Author(s)
Publication (Day/Month/Year) 2019
URL http://www.opensaldru.uct.ac.za/bitstream/handle/11090/962/2019_247_Saldruwp.pdf?sequence=1
Abstract
Do working mothers earn less than non-mothers in the South African labour market? This study examines whether there exists a motherhood (or child) penalty for Black African female employees in post-apartheid South Africa using data from wave 5 of the National Income Dynamics Study (NIDS), from 2017. NIDS is the first nationally representative survey in South African to include comprehensive child birth history. Restricting analysis to women aged 20 to 49, the Mincerian regression model results from the analysis indicate that a motherhood penalty does exist, ceteris paribus. Moreover, the study uses unconditional quantile regressions (RIF-OLS) to examine the wage returns of mothers versus non-mothers along the wage distribution. The study finds that, when controlling for relevant observable characteristics, there exists a motherhood wage penalty at lower wage levels, but this effect wanes in prominence at higher wage quantiles. At higher wage levels, mothers earn higher hourly wages than their child-free counterparts, especially if they are married and work part-time. This result indicates the effect of a part-time hourly wage premium. The study then applies OaxacaBlinder type decompositions within the RIF framework to decompose changes in the motherhood wage gap along the distribution into explained and unexplained contributions related to a range of factors. The decomposition results indicate that only at the hourly wages of mothers minus wages of non-mothers are negative only at the 10th quantile, but positive everywhere else. Moreover, even though most of the wage differential between mothers and non-mothers is due to explained characteristics, at the lower levels unobservable traits have an impact on the wage gap. This implies that there are additional relevant factors such as societal norms, selection effects into employment and behavioural characteristics which should be considered when analysing women’s wage outcomes. Labour market policy needs to accommodate women with children, particularly if they are the main breadwinners at lower wage levels. Workplaces should consider embracing flexible work hours and provide the option for staff to work remotely.

Related studies

»