South Africa’s private-household sector employed an estimated 1.024 million people in the third quarter of 2026, down 68,000 from the previous quarter, even as the number working in the formal sector rose by an estimated 268,000.
The household figure fell from 1.092 million in April to June, a quarterly decrease of 6.2 per cent. The official unemployment rate was estimated at 33.7 per cent in July to September, up from 33.6 per cent in the preceding quarter. The small rate change is not statistically conclusive, however, given the survey’s sampling uncertainty.
The results, released on 10 November by Statistics South Africa, show how different parts of the labour market can move in different directions. They are estimates of people employed, rather than a count of dismissal notices or a record of what happened to each worker.
Household sector and domestic work
Private-household employment covers people employed directly by households, including domestic workers, gardeners, cooks and childminders. It is not a measure of everyone who works from home. People working through independently run cleaning or gardening businesses are classified separately.
The survey also has an occupational category for domestic workers. On that measure, the estimated number fell from 831,000 to 800,000 between the second and third quarters. The 31,000 difference is not the same as the 68,000 change for the wider household sector: one figure describes an occupation and the other a sector.
There is also considerable uncertainty around the occupational change. The survey’s 95 per cent confidence interval runs from a decrease of 88,000 to an increase of 27,000, so it does not establish a quarterly fall in domestic-worker employment with conventional statistical confidence.
For the broader private-household sector, the corresponding interval runs from a decrease of 133,000 to a decrease of 3,000. That range is wide, although it points to a fall at the conventional 95 per cent confidence level.
Why the national picture differs
Formal-sector employment was estimated at 12.340 million in the third quarter, compared with 12.073 million in the previous three months. Employment in the informal sector was estimated to have fallen by 98,000 to 3.476 million. That informal-sector change is not statistically conclusive: its confidence interval spans both a decrease and an increase.
The sector labels describe the kind of economic unit in which someone works. They are different from the survey’s measure of informal employment, which can also include people working for formal-sector employers or in private households without certain basic benefits. A rise in the number working in the formal sector should not be treated as proof that every one of those workers has formal employment protections.
Across the economy, the number of employed people was estimated at 16.840 million, up 101,000 from the second quarter. The number counted as unemployed rose by an estimated 97,000 to 8.578 million. The labour force, which combines employed and unemployed people, grew by an estimated 198,000. Published totals and changes are rounded separately.
The reported changes in total employment and in unemployment also have confidence intervals that include movements in the opposite direction. They show the survey’s central estimates, not precise counts of new jobs or additional people out of work. The same caution applies to the 0.1 percentage-point rise in the official unemployment rate.
Different measures of joblessness
The official rate is the proportion of the labour force that is unemployed, not the proportion of the entire population without a job. Under the survey’s definition, unemployed people generally must be available for work and have actively looked for work or tried to start a business. The definition also includes some people with a definite future start who are available.
A broader measure combines unemployment with the potential labour force. It rose in the survey’s estimates from 43.8 per cent to 44.0 per cent. The potential labour force includes people who want and are available for work but are not actively looking, as well as some who are seeking work but are not currently available. This measure uses an expanded labour-force denominator and should not be read as a share of all residents. Its quarterly change, too, is not statistically conclusive.
The survey estimated that 3.735 million people were discouraged work-seekers, up 70,000 on the previous quarter. This category covers people who wanted and were available for work but had not recently looked because, for example, they believed no suitable jobs were available. The estimated quarterly change has a confidence interval spanning both a fall and a rise.
Pay and the limits of the figures
Domestic workers are covered by South Africa’s national minimum wage, which has been R30.23 for each ordinary hour worked since 1 March 2026. The amount earned over a week or month depends on the hours worked; an hourly minimum alone does not establish an individual worker’s actual income.
The survey does not explain why household employment changed. It cannot establish that wage rates, household bills or any other single factor caused the estimated decline. Nor does a net change identify whether the same people stayed in or left the sector between the two periods.
The figures come from a household sample, with estimates weighted to represent the population. The survey covers people aged 15 to 64, and the survey’s master sample is designed around approximately 33,000 dwelling units. Its estimates are subject to sampling and other errors, which is why the uncertainty around individual changes matters as well as the headline totals.
