South Africa’s gender pay gap starts long before women sit down to negotiate a salary - IOL
South Africa’s gender pay gap is about more than equal pay. Research shows women are also less likely to reach the higher-paying firms and industries where the biggest salaries are found.
The pay gap in South Africa is not decided only when two people sit across a desk and negotiate a salary. By the time that conversation happens, the industry, the occupation and the company have already done much of the work.
On Friday, 18 September, the world marks International Equal Pay Day. In South Africa, the day highlights a stubborn fact: in the formal sector, women earn, on average, 12% less than men. Almost half of that difference is not about what a woman is paid inside a firm. It is about which firms she is able to reach.
That finding comes from an eight-year study by economists Ihsaan Bassier and Leila Gautham, published in the Journal of Development Economics. Using administrative tax records covering the universe of formal workers between 2010 and 2018, they tracked millions of employees and isolated what each company pays its workers, regardless of gender. About 45% of the 12% gap, or 5.5 percentage points, is explained by women being concentrated in lower-paying companies.
The remaining 55% cannot be read as a clean measure of “same job, different pay”. The study did not compare women and men doing identical work. Researchers say the remainder can include unmeasured differences between workers and bias that operates across the labour market rather than inside a single payroll. Either way, the implication is the same: an explanation that stops at the office door is incomplete.
The 12% figure is the formal-sector average from tax data. Broader household surveys paint a harsher picture because they include informal work, shorter hours and private households.
A UN Women analysis of labour-force data found an unadjusted gap of 20.1% on an hourly basis and 32.5% monthly, with the wider monthly gap reflecting that women, on average, work fewer paid hours. After accounting for education, occupation, sector and other observable characteristics, an adjusted gap of about 7.9% remained.
Those numbers sit on top of a more basic exclusion. In the second quarter of 2026, Statistics South Africa counted 21.3 million working-age women. Only 7.3 million of them were employed, little more than one in three. Women’s labour-force participation was 54.9%, against 64.4% for men. Women’s official unemployment rate was 37.5%, compared with 30.3% for men. Their absorption rate – the share of working-age people actually in a job – was 34.3%, against 44.9% for men.
Women already make up just over half of the working-age population. They hold only 44 of every 100 jobs.
Equal pay for the same work is protected by law in South Africa. Equal access to the work that pays is not yet a lived reality.
The usual suspects – hours, experience and qualifications – explain some of the difference. They do not explain why women and men, moving through the same economy, land in companies that pay at different levels.
Bassier and Gautham measured firm pay premiums by following workers who switched jobs and seeing how their pay changed. That method holds the worker roughly constant and asks what the company adds. The answer is large. In high-income countries such as Portugal and Italy, firm sorting explains about 20–25% of the gender pay gap. In South Africa, it explains nearly half.
Why the difference? Formal jobs are scarce. High unemployment gives firms more power to set wages. Women enter that hierarchy lower down, and the hierarchy itself is steeper. Women entering formal work are more concentrated in lower-paying industries, while men are over-represented in higher-paying sectors such as construction, mining and manufacturing. In regions with less formal employment, the firm-pay gender gap is wider still.
The study finds that within-firm differences play a smaller role than sorting across firms. Women and men change employers at similar rates. Men are more likely to move to better-paying firms. Women returning from unemployment or informal work tend to re-enter at lower-paying firms than men do. The penalty compounds.
The divide does not sit in one place. It runs through industry, occupation and employer, and the three layers stack.
Sectors do not offer the same pay premium. Women entering formal work are concentrated in education, retail and personal care. Men are over-represented in construction, mining and manufacturing – industries that pay more across the workforce.
A Stats SA analysis published in August 2026 confirmed that pattern: women form more than half of employment in personal services, accommodation, food and beverages, and retail, while their representation is lowest in construction, mining, transport and storage, and the motor trade. Construction is male-dominated across job types, not only at the coalface. Women occupy more than 70% of jobs in social work and hospital, medical and dental activities. They account for more than 60% of textile and clothing jobs inside manufacturing, while remaining under-represented in the sector as a whole.
Even inside the same industry, the work is not the same. In Q2 2026, women were more heavily represented in clerical, sales and service and domestic occupations, while men were more represented in managerial roles and several technical occupations.
Women remain less represented in the jobs that carry authority and, in many cases, the steepest pay ladders.
Two people can do comparable work in the same sector and still earn different amounts because their companies sit at different points on the pay distribution. That is the layer the tax study isolates and the layer that accounts for nearly half the formal-sector gap.
“These divisions do not operate in isolation. Layered together, sector, occupation and employer can place workers on sharply different earnings paths, allowing disadvantage to accumulate from one level to the next,” says Prof Linda Meyer, managing director of Rosebank International and visiting professor at Nelson Mandela University.
The public sector is an instructive exception, showing that the structure can be changed. It employs a much higher share of women than men and offers relatively high pay premiums. The researchers link this in part to the state’s active pursuit of gender equity in hiring. Access to better-paying firms is not a law of nature. It can be shaped by policy and institutional choices.
Women remain scarce in high-premium sectors such as mining even as some have broken in. Representation there is still among the lowest in the economy.
The pay gap is measured among people who already have a job. A large part of the economic divide never reaches a payslip.
Of the 21.3 million working-age women in Q2 2026, 9.6 million were outside the labour force. About 2.5 million of them formed part of the potential labour force – available but not actively seeking work, or seeking work but not immediately available. Discouraged work-seekers made up the bulk of that group. Young women face a wider gap still: 38.2% of females aged 15–24 were not in employment, education or training, against 34.6% of young men.
Unpaid care is the quiet twin of that statistic. The latest available Stats SA figures show that 2.1 million women, or 88.2% of the 2.4 million people outside the labour force because they were homemakers, were women. For every man in that position, more than eight women were.
That unpaid work keeps households running and frees other adults for paid jobs. It can also narrow the set of jobs a woman can realistically take: predictable hours, flexibility and proximity to home. Those constraints do not show up as a line item called “discrimination”. They can show up in the type of firm and job a worker enters.
The firm-pay gap is negligible among the youngest workers. It widens sharply from the mid-20s through the child-rearing years before narrowing among older workers. The timing overlaps closely with the years in which many women are raising children, although the study does not establish that care responsibilities cause the divergence.
Women who stay continuously employed still lose ground: they switch firms as often as men, but less often to better-paying ones. Women who leave and return face an entry penalty. In a labour market with high churn and thin formal opportunity, each re-entry can become a second lottery – and the tickets are not evenly distributed.
Education is supposed to be the great equaliser. Women are now the majority of students in South Africa’s post-school system and the majority of graduates. That has not erased the occupational pipeline. Department of Higher Education and Training statistics still show men outnumbering women among graduates in physical and computer sciences, manufacturing and engineering, the fields that feed higher-premium industries. Women dominate education, health and the humanities. Industry figures show that only 16% of registered engineering professionals in South Africa are women.
“Education alone doesn’t determine pay, but different fields of study open different occupational pathways. The labour-market divide is shaped by a series of connected pressures, not one free and isolated decision about which job to accept,” Meyer says.
None of those steps requires a manager to write two different numbers on two identical contracts. Together they can produce two different careers.
South African law already prohibits unfair discrimination in pay for work that is the same, substantially the same or of equal value. That protection matters. It is also limited. Inequality that is hidden cannot easily be challenged, while inequality produced before two people share a workplace cannot be audited inside one company.
Stellenbosch Business School research, led by Prof Anita Bosch, has long argued that employers must examine remuneration practices, identify unjustified differences and correct them, and that transparency, regular pay audits and enforcement make inequality harder to defend. Bosch and colleagues have also warned that South Africa’s median gender pay gap has been stagnant for years in the broader 23–35% range cited in governance debates, and that mandatory public disclosure of gender pay data remains weaker than in jurisdictions that already require it. Most JSE Top 40 companies still do not publish comparable gender pay-gap figures.
A pay audit can catch the woman paid less than the man at the next desk. It cannot put her into mining, manufacturing, a better-paying professional services firm or a managerial grade. Closing the wider gap also requires:
The researchers who measured the firm-pay gap put it plainly: reducing the gender pay gap takes more than placing women into male-dominated jobs or promoting equal pay for equal work. It means tackling the structures that steer women into lower-paying companies. Left alone, “sorting into low-paying firms” will keep reproducing the gender pay gap, one job move at a time.
Women can do the work that pays. They already do in classrooms, hospitals, shops, offices, mines and plants. The evidence is not that South African women lack skill. It is that women and men do not have equal access to all parts of the labour market, and that this unequal distribution contributes to the earnings gap.
Equal pay for work of equal value is non-negotiable. Women should receive equal pay for the same, substantially the same or equal-value work. That principle is written into South African law. It should be visible on every payroll.
But a country cannot close its gender earnings gap if only a minority of women ever reach the jobs and firms where the higher salaries are paid. The test is double. First: when a woman and a man stand side by side, is the pay fair? Second: did she have a fair chance to stand there at all?
Until South Africa answers both, women can arrive at salary negotiations with their earning potential already marked down – not necessarily because they asked for less, but because the path to better-paying work may have been narrower from the start.
