Focusing on Africans and coloureds, this study found that households headed by women are more likely than those headed by men to experience a situation where no household member is employed. Moreover, since having no household member employed is associated with an increase in the likelihood of household poverty, it implies that female-headed households (FHHs) are more likely than male-headed households (MHHs) to be poor. But among households where at least a member is employed, there is no difference in the likelihood of poverty between FHHs and MHHs in general. This general picture however, masks important nuances based on gender-based employment patterns in such households. FHHs are about four times more likely to have only women employed in the household relative to MHHs. On the other hand, MHHs are about five times more likely to have only men employed than FHHs. Consequently, we found that FHHs where only women are employed are significantly more likely to be poor relative to MHHs where only men are employed. Our analysis suggests that this difference may be due to labour market discrimination faced by women, rather than FHHs working for less amounts of time, having fewer household members employed, or holding lower skilled jobs in the work place.
We also found that FHHs are not homogenous. Comparing FHHs where only women are employed with fellow FHHs where at least one man is employed (irrespective of the number of female household members also employed), we found that the former are significantly more likely to be poor relative to the latter. However, FHHs where at least one man is employed are not more likely to be poor relative to similar MHHs, while FHHs where only men are employed are significantly less likely to be poor relative to similar MHHs.
These results indicate that one way to fight gender-based poverty in South Africa is to understand and address the root causes of higher total household non-employment among FHHs. This does not imply neglecting MHHs, given that they also experienced high rates of complete household non-employment. Moreover, the heterogeneity of FHHs indicates that while a significant part of households headed by women are economically vulnerable, some of them are relatively well off. Therefore, using the gender of the household head as a poverty-targeting metric will result in an inefficient allocation of resources.
Another way to address the relatively higher poverty prevalence among FHHs may be to tackle gender discrimination in the labour market. As earlier indicated, among households with someone employed, FHHs with only women employed are likely to be economically vulnerable. While our study did not identify any specific method to be used in tackling the perceived gender discrimination in the labour market, we call on policy makers to strengthen the implementation of existing anti-discrimination labour laws in the country. Some studies highlight the difficulties facing the realization of gender equality in South Africa despite constitutional provisions. Perhaps, a clearer understanding of gender equality (especially among employers), coupled with the strengthening of judicial procedures and legal consequences for gender-based (wage) discrimination may ameliorate poverty in FHHs.