The Wire
South Africa’s jobs paradox: 2.3 million new jobs since 2008, none in the industries that pay for them
Statistics South Africa puts unemployment at 33.7%, 8.6 million people, while manufacturing output fell 3.5% in September. A University of Johannesburg economist argues the country has added jobs only where the money is spent, not where it is made; and sets out eight ways to rebuild the base.
Key points
- What: On 10 November 2026, Statistics South Africa released the Quarterly Labour Force Survey for the third quarter of 2026: the official unemployment rate rose to 33.7%, from 33.6% in the second quarter and 31.9% a year earlier, with 8.6 million people officially unemployed. 3
- Jobs: Employment rose by 101,000 in the quarter to 16.8 million, but the labour force grew faster, by 198,000 to 25.4 million; so the number of unemployed people still rose by 97,000. 8
- Broader measure: The combined rate of unemployment and the potential labour force reached 44.0%, and 3.7 million people are now classified as discouraged work-seekers. Youth unemployment stands at 47.3% for ages 15 to 34, and 63.6% for ages 15 to 24. 5
- Manufacturing: Factory output fell 3.5% year on year in September 2026, with motor vehicles, parts and transport equipment down 15% and wood, paper and printing down 11.2%. The sector still grew 0.9% over the third quarter as a whole. 6
- The argument: Prof. Daniel Meyer of the University of Johannesburg calculates that since 2008 South Africa has added 2.29 million jobs, yet the base sectors, agriculture, mining and manufacturing, have lost 209,000. Manufacturing employment fell from about 2.06 million to 1.60 million. 1
ONE EVENT. THREE READINGS.
In the Western reporting, the story is told as a political problem with a market footnote.
Reuters, reporting from Pretoria, leads with the voter: the figures landed days after municipal elections in which joblessness was a central grievance, and the governing party was punished for it.
What this reading highlights: The agency notes the rate has stayed above 30% for six years and ranks among the highest in the world, and that the bond market barely moved.
The Western lens reads unemployment as a verdict on a government; and the verdict, in this telling, has already been delivered at the ballot box.
Sources: Reuters via CNBC Africa (1), World Bank via Moneyweb (1)
In the Asian press, the story is a data release, not a political event.
Xinhua’s dispatch from Johannesburg reports the figures straight, 33.7%, 8.6 million unemployed, 16.8 million employed, without the electoral framing that dominates the Western accounts.
What this reading highlights: The Eastern reference point sits elsewhere: the World Bank’s report holds up China in the 1980s, Vietnam in the 1990s and Poland in the 2010s as cases where aligned reform rebuilt an economic base.
The question is whether South Africa can execute a sequence others have already run.
Sources: Xinhua (1)
In the Southern press, the story is about the people the headline rate leaves out.
The Daily Maverick, the Mail & Guardian, IOL and SABC all dwell on the same ledger: 3.7 million discouraged work-seekers, provinces where more people are unemployed than working, and young people locked out at 63.6%.
What this reading highlights: The trade union federation Saftu warned this week against blaming migrant workers for the crisis.
North-West University adds the long view: since 2008, the labour force has grown 35% while employment has grown 17%. The Southern lens counts the cost in households, not in basis points.
Sources: Daily Maverick (1), Statistics South Africa (1), Mail & Guardian (1), Business Day (1), IOL (1), SAnews (1)
Sources by lens: 2 West, 1 East, 6 Global South. See every source
THE POINT OF CONVERGENCE
On the facts. The Quarterly Labour Force Survey for the third quarter of 2026 puts official unemployment at 33.7%, with 8.6 million people unemployed and 16.8 million employed; the broader measure that counts discouraged work-seekers is 44.0%. Manufacturing output fell 3.5% year on year in September. Every independent outlet reporting the Stats SA release gives the same figures.
WHAT REMAINS CONTESTED
On the diagnosis. Meyer argues South Africa has an economic-base problem, not a labour-market problem, and that no amount of hiring policy fixes a shrinking productive core. Other economists quoted in the coverage stress weak demand, high input costs and low confidence instead, while the government points to its industrial strategy approved in June 2026 as the answer already under way.
The story
Two releases, nine days apart, tell one story about work in South Africa
On Tuesday 10 November 2026, Statistics South Africa released the Quarterly Labour Force Survey for the third quarter and manufacturing production for September. The official unemployment rate rose to 33.7%, up from 33.6% in the second quarter and 31.9% a year earlier, leaving 8.6 million people officially without work. Manufacturing output fell 3.5% compared with September 2025.
Nine days later, Prof. Daniel Meyer of the University of Johannesburg argued in the Daily Maverick that the figures expose something deeper than a bad quarter. Since 2008, he writes, the labour market has added 2.3 million jobs, yet not one net job in the sectors that earn the country’s living: agriculture, mining and manufacturing. Every net job since the global financial crisis, on his reading, sits in services that live off income earned elsewhere. His piece is the first systematic attempt to turn the new data into a programme, and that programme, as much as the numbers, is now being debated.
Why did the jobs go to services, not factories?
Between the second and third quarters, employment rose by 101,000 to 16.8 million; but the labour force grew by 198,000 to 25.4 million, so unemployment still rose by 97,000 to 8.6 million. North-West University’s analysis puts it plainly: the economy needed about 131,000 new jobs in the quarter just to keep the rate from rising. South Africa has now spent six years above 30% unemployment, among the highest rates in the world.
Inside the quarter, finance added 111,000 jobs, manufacturing 33,000 and mining 31,000, while private households, agriculture and construction shed workers. The broader measures complete the picture: the combined rate of unemployment and the potential labour force rose to 44.0%, discouraged work-seekers increased by 70,000 to 3.7 million, and youth unemployment stands at 47.3% for ages 15 to 34; and 63.6% for ages 15 to 24, the highest since 2022.
The provincial spread is brutal. The Eastern Cape records the highest official rate at 44.3%; Gauteng, the industrial heartland, is at 35.7%. In the North West and the Eastern Cape, the broader measure has passed 50%. In manufacturing, September’s 3.5% annual fall was driven by motor vehicles, down 15%, and wood, paper and printing, down 11.2%. The quarter as a whole still grew 0.9%, carried almost entirely by food and beverages.
Then comes Meyer’s long view, his analysis of the Stats SA series, not a Stats SA finding. In the third quarter of 2008, agriculture, mining and manufacturing employed 3.2 million people, 22% of all workers; today it is 3.0 million, or 17.8%. Manufacturing fell from about 2.06 million jobs to 1.60 million. Over the same period, finance added 1.29 million jobs and community and social services, mostly government; added 1.02 million, slightly more than the total net increase in employment. The ratio of base-sector jobs to the rest has deteriorated from 1:3.5 to 1:4.6.
What the 2008 comparison can and cannot prove
First, the 2008 comparison is Meyer’s own reading of eighteen years of survey data. The Bureau could not independently reproduce the full historical series in this cycle, and the detailed splits behind it do not appear in the public release. The figures are presented here as his analysis, and should be read that way until Stats SA or an independent replication confirms them.
Second, Stats SA warns that from the third quarter of 2025, formal and informal work is classified under a newer international standard, so that split is not comparable with earlier quarters. The totals are unaffected, but any claim about where exactly jobs sit must carry that caveat.
Third, nobody knows whether the eight proposals would work. They have never been tried as a package, and some depend on delivery by institutions that have missed deadlines before. Nor do we know what the fourth quarter will show: the next survey is due in February 2027, and the purchasing managers’ index showed factory employment falling for a third successive month in October.
One in three without work, and the young worst hit
One in three people in the labour force is officially unemployed; counting those who have stopped looking, it is closer to one in two. Among 15- to 24-year-olds, nearly two in three who are looking cannot find work, and 35.7% of that age group is not in employment, education or training. A generation is entering adult life with no foothold in the formal economy, and the practical skills that once carried people into work are thinning: Meyer records 121,000 fewer technician jobs and 68,000 fewer craft and trade jobs.
The politics follows the economics. Reuters tied the figures directly to voter anger at last week’s municipal elections, in which the African National Congress took what the agency called a drubbing. The new councils inherit the problem with thinner resources: community and social services shed 197,000 jobs over the past year, and the informal sector lost 485,000. Gauteng concentrates the damage: 800,000 manufacturing workers in 2008, 514,000 today, 63% of all manufacturing jobs lost nationally, and 117,000 jobs shed over the past year alone.
How can eight fixes rebuild the base?
Judge the programme fix by fix. Cutting the regulatory burden has heavyweight backing: the World Bank’s March 2025 report called the burden on business excessive and recommended wider use of equity-equivalent programmes instead of the complex conditions of broad-based black economic empowerment, noting the current system penalises small firms most. Protecting existing firms, knowing the largest employers, intervening early, costs little and has no ideological price.
Others are structural and slow. Cutting the cost of production means energy, logistics and water: faster roll-out of embedded and wheeled electricity for industrial parks, and more efficient, open-access rail and ports. Meyer argues this would do more for manufacturing than new incentive schemes, at a much lower cost to the fiscus. The government is, in principle, already moving this way: an industrial strategy approved in June 2026 is built around decarbonisation, diversification and digitalisation. The gap is between strategy and delivery; and the September figures suggest delivery is what has been missing.
The rest fill the gaps: municipalities as partners in production, with reliable services and approvals inside fixed time frames; processing mining and farm output at home rather than exporting it raw; tourism as a base sector in its own right; joint skills programmes between technical colleges, training bodies and employers; and measuring the base itself, with quarterly data per region and targets in municipal and provincial plans.
The strongest part is the test. A city that adds call-centre and security jobs while losing factories, Meyer writes, is not becoming economically stronger. By that test, the third quarter fails. The same arithmetic haunts other large economies on the continent; the Bureau’s country profiles for Nigeria and Morocco track similar pressures, in different forms. The question is whether Pretoria, and the new councils, will govern by that test or keep counting jobs without asking where they sit.
The second wave: skills, councils and the fiscus
If the base keeps shrinking, the effects arrive in a fixed order. The services that absorbed the jobs depend on income earned elsewhere, so the fiscus strains precisely as demand for support rises. Skills atrophy: technicians and artisans, once lost, are slow to rebuild. Gauteng’s concentrated losses hollow out municipal revenue where the bills are largest. And the politics hardens: new councils face thinner budgets and rising expectations, and the temptation to blame outsiders, already visible in the debate over migrant workers, grows. The base is not only an economic category. It is the floor under everything else.
What would prove the rebuild is working?
Three clocks are now running. The first is February 2027, when the fourth-quarter labour survey arrives: watch whether manufacturing employment and the broader 44.0% measure move together, and whether Gauteng stops shedding jobs. The second is the new councils’ first budgets; base compacts, protected industrial areas, approvals on a fixed clock. The third is delivery on energy and logistics, where announcements have outrun results for years.
The purchasing managers’ index is the early warning: factory employment fell for a third successive month in October, the sharpest shedding in six years. If that turns, and the next survey confirms it, the argument that the base can be rebuilt will have its first evidence. If it does not, the debate returns to Meyer’s question; whether South Africa has a jobs problem at all, or a production problem wearing a jobs problem’s clothes.
What we agree on, and what remains open
Established
- On the facts. The Quarterly Labour Force Survey for the third quarter of 2026 puts official unemployment at 33.7%, with 8.6 million people unemployed and 16.8 million employed.
- The broader measure that counts discouraged work-seekers is 44.0%. Manufacturing output fell 3.5% year on year in September. Every independent outlet reporting the Stats SA release gives the same figures.
Contested
- On the diagnosis. Meyer argues South Africa has an economic-base problem, not a labour-market problem, and that no amount of hiring policy fixes a shrinking productive core. Other economists quoted in the coverage stress weak demand, high input costs and low confidence instead, while the government points to its industrial strategy approved in June 2026 as the answer already under way.
Still open
- Whether the eight proposals would work. None has been tested as a package, and the long 2008 comparison behind the argument is the author’s analysis of the Stats SA series, not a Stats SA release itself. We do not know what the fourth quarter will show, that survey is due in February 2027, or how the new municipal councils will respond.
What to watch
Questions, answered
The official unemployment rate was 33.7% in the third quarter of 2026, according to Statistics South Africa’s Quarterly Labour Force Survey, released on 10 November 2026. That is up from 33.6% in the second quarter and 31.9% a year earlier. About 8.6 million people were officially unemployed, while 16.8 million had work.
Because the labour force grew faster than employment. Between the second and third quarters, 101,000 more people found work, but 198,000 more people entered the labour force, so unemployment still rose by 97,000. The economy needed about 131,000 new jobs in the quarter just to keep the rate steady, and it fell short.
The economic base is the set of sectors that sell outside the local economy and bring new money in: agriculture, mining, manufacturing and tourism. Other sectors, such as retail, personal services and much of government, circulate that money. The argument is that South Africa has added jobs only in the circulating sectors, while the base that pays for them has shrunk.
Prof. Daniel Meyer proposes cutting the regulatory burden, protecting existing firms, cutting production costs in energy, logistics and water, making municipalities partners in production, adding value to mining and farm output, growing tourism, rebuilding technical skills, and measuring base-sector employment and production per region with targets.
The next Quarterly Labour Force Survey, covering October to December 2026, is due in February 2027. It will be the first verdict on whether the third quarter’s pattern held, and on whether the new municipal councils’ first budgets changed anything for factories, mines and the workers who depend on them.
MAGNA BUREAU