Reads What has democracy meant for married women in the labour market? Married women are more actively seeking work but are still more likely to be unemployed than men. However, the expansion of education in the democratic era has significantly affected how she weighs up her potential wage and her husband’s earnings when deciding whether to look for work.
This month, as South Africa commemorates Freedom Day, it is worth pausing on what 1994 changed for women.
The end of apartheid brought with it a series of legislative changes that expanded women's freedoms, and the share of women actively seeking work began to rise. In the three decades since, the share of women aged 25 to 59 actively seeking work has risen from 40% to 65%, one of the more measurable shifts of the post-apartheid era. But more women looking for work has not meant more women in work. Women´s employment has not kept pace, and women remain more likely to be unemployed than men, a gap that has barely narrowed in a decade.
The reasons for any woman's decision to seek work are complex, but for married women, who pool time and money with their spouses, the trade-offs are particularly revealing. Far harder to answer, though, is what actually drives that decision, and whether it has changed as the democratic era has unfolded. There is a surprisingly thin evidence base on this question in South Africa. Answering it requires understanding two factors that influence a married woman's decision to seek work: what she herself could earn, and what her partner already brings home. If her own potential wage rises, does she become more likely to seek employment? And if her partner earns more, does she become less likely to? These are the own-price and cross-price elasticities of labour supply, measures of how sensitive a woman's decision to work is to her own potential wage on the one hand, and her partner's earnings on the other.
In the United States and Europe, both effects used to be strongly felt - a woman was more likely to seek work when her own potential wage was higher, and less likely to seek work when her partner earned more, reflecting a world in which the household ran on a single income and a woman's paid work was discretionary rather than necessary. Over the past half century, both effects have steadily weakened as women became more attached to the labour market in their own right. Whether the same shift has happened in South Africa, and what has driven it, is a different and far less settled question, and the picture that emerges from two decades of data is more complicated than expected.
A shift in the trend
Both wage effects weakened through the 2000s, then stabilised and in some cases began to strengthen again in the 2010s. Notably, a husband's higher earnings, once a stronger deterrent to his wife seeking work, lost force over the two decades studied. In the early 2000s, a rise in a husband's wage made it less likely that his wife would seek employment. By the 2010s, that same rise in his earnings had less bearing on her decision.
Covid-19 disrupted the trend sharply, and both effects weakened again. The stabilisation and partial reversal in the 2010s was most pronounced among groups with more labour market advantage: white women and those with tertiary education. Taken together, South Africa's wage sensitivity levels are broadly comparable to those found in high-income countries such as the United States, France, and the Netherlands.
Figure 1: Own- and cross-price labour supply elasticities of married or cohabiting women over time
Notes: Author's own calculations using The Post-Apartheid Socio-Economic Series (PASES.) Results are weighted using the StatsSA person weight. Sample is married or cohabiting women aged 25 to 59 years living with employed spouses or partners not older than 70 years. The left panel shows own-price elasticities, how sensitive a married woman's decision to seek work is to her own potential wage. The right panel shows cross-price elasticities, how sensitive that decision is to her partner's earnings. Each panel shows annual elasticity estimates with a linear fit, locally weighted regression fit, and 95% confidence intervals. The vertical red dashed line indicates 2020. The horizontal red dashed line indicates zero.
The 2000s pattern makes intuitive sense. When unemployment is high and job queues are long, landing any job matters more than what it pays, and so a woman's own potential wage becomes less of a deciding factor in whether she seeks work at all. But the 2010s tell a different story. Female unemployment had not eased, if anything, it worsened after the Global Financial Crisis of 2008. So why would married women become more responsive to wages precisely when their chances of finding work were no better?
Education as the turning point
Answering this question is complicated but one clue points to one of the post-apartheid era's most significant shifts: the expansion of access to education. The change in wage sensitivity tracks almost exactly to the moment when the median married woman crossed a specific threshold - completing matric. That threshold matters because it is the point at which employment prospects improve substantially. A woman with a matric certificate is significantly more likely to find work than one without it. As more women crossed that threshold through the 2010s, a direct consequence of post-apartheid investment in schooling, their sensitivity to what they could earn began to stabilise or rise .
In other developing countries it has not been uncommon for more education to go hand in hand with women opting out of the labour market. Once their husbands start earning enough, wives might prefer to not do menial jobs if they don’t have to, freeing them up to put in more time at home and, for some, avoid the stigma of manual work. But usually once women start attaining tertiary education and filling white-collar jobs, they opt to cash in on their greater earnings and career potential. So why in South Africa would tertiary educated and advantaged women be the same ones that have become more sensitive to the wage? This question is not settled yet, and South Africa’s pattern somewhat challenges conventional ideas about how labour markets develop.
Even as more women have entered paid employment, they continue to carry the overwhelming share of unpaid household and care work alongside it. Thirty-two years into democracy, married women’s model of when to work is dynamic and changing as they balance the burdens of paid and unpaid work. This change is closely connected to the expansion of freedoms afforded to women at the dawn of democracy as well as the broad-based expansion of their education levels.
