South Africa entered the second quarter of 2026 with an unemployment crisis deep enough to dominate almost every economic debate, yet agriculture continued to support about 944,000 jobs and remained ahead of its employment level a year earlier.
At the same time, farmers in some regions and commodities have reported real difficulty securing labour at peak periods. The revealing question is not whether one side is right and the other wrong, but how both realities can exist in the same agricultural economy — and what happens if policy mistakes a local, seasonal pressure point for a national collapse, or dismisses it because the national numbers still look healthy.
A crop does not negotiate with a labour market. It reaches maturity when weather, cultivar, soil, water and the season say it is ready. A table-grape block cannot be told to wait for a policy meeting, citrus cannot be left indefinitely on the tree while paperwork catches up, and a labour-intensive harvest can turn from orderly to expensive remarkably quickly if too few hands are available at the moment they are needed. That biological reality makes the current farm-labour debate far more instructive than the political slogans surrounding it.
South Africa is being asked to hold two facts in its head at the same time. Statistics South Africa reported an official unemployment rate of 33.6% in the second quarter of 2026, with the number of unemployed people rising by 345,000 to 8.5 million. Total employment slipped by 16,000 to 16.7 million, while agriculture itself lost about 15,000 jobs compared with the first quarter. Yet agricultural employment still stood at about 944,000, which Wandile Sihlobo of the Agricultural Business Chamber of South Africa (Agbiz) notes was 4% higher than a year earlier. The revealing picture is therefore not one of a sector either booming or collapsing; it is one of a large labour market that can be robust nationally while remaining exposed locally and seasonally.
That distinction matters because public discussion has increasingly presented agricultural labour as a binary choice: either foreign workers are indispensable and crops will rot without them, or South Africa’s mass unemployment proves there can be no genuine shortage of farm workers. Neither position survives close inspection. The most revealing evidence points toward a more complicated middle ground — a sector employing hundreds of thousands of South Africans, supplemented in some areas and seasons by foreign workers, with labour pressure concentrated in certain provinces, commodities and harvesting windows rather than spread evenly across the country.
The numbers do not tell a story of collapse
Stats SA’s Q2 2026 Quarterly Labour Force Survey provides the hardest national reference point. The unemployment rate rose from 32.7% in Q1 to 33.6% in Q2. Unemployment increased to 8.5 million people, while total employment fell marginally to 16.7 million. Agriculture was among the industries recording a quarterly decline, shedding about 15,000 jobs. Those are not numbers to wave away. They show agriculture did experience a softer quarter even while its year-on-year position remained stronger.
Sihlobo’s interpretation adds the second half of the picture. He places agricultural employment at 944,000 in Q2, up 4% from the same period in 2025, and describes the broader agricultural labour market as stable. He also notes that primary agriculture employs close to one million people and that more than 300,000 additional people work across agricultural value chains. The important point is that the sector’s employment footprint extends well beyond the farm gate — into packing, processing, logistics, inputs, services and the wider chain that turns production into food and exports.
A quarterly decline and a year-on-year increase are not contradictory. They answer different questions. Quarter-on-quarter data tell us whether employment rose or fell from the immediately preceding three-month period; year-on-year data tell us whether the sector is stronger or weaker than at the same point in the previous year. Agriculture can lose 15,000 positions in one quarter and still employ more people than it did a year earlier.
That is why the phrase “farm labour crisis” requires precision. A national employment collapse would imply broad deterioration across the sector. The evidence Sihlobo cites from Agbiz and Agri SA surveys does not point to that. Instead, he says labour difficulties have been isolated to a few provinces and industries. The challenge is therefore not to deny the farmers who are struggling, but to locate the pressure accurately enough to understand what is actually failing.
A million unemployed people cannot simply be moved to a harvest
South Africa’s unemployment numbers invite an obvious question: if 8.5 million people are officially unemployed, how can any farmer struggle to find workers? It is a fair question, but it treats labour as if it were stored in a national warehouse and could be dispatched anywhere on demand. Agriculture does not work that way. Fields, orchards and vineyards are fixed in place; harvest dates are governed by production cycles; and labour demand can rise sharply for a relatively short period before easing again.
The national QLFS cannot tell us that every unemployed person is located within practical reach of every farm, has access to transport or accommodation, is available for temporary seasonal work, or is suited to every task a farming operation may require. Nor does Sihlobo’s note quantify those factors. What it does identify is the revealing importance of geography and seasonality: foreign workers are particularly present in provinces bordering neighbouring countries and in certain value chains, including sugar cane and fruit, while seasonal workers become more important during busy harvesting periods.
This distinction changes the debate. The relevant question is not whether South Africa has unemployed people; plainly it does. The revealing policy question is whether available workers and temporary agricultural vacancies can be matched quickly enough, close enough to production areas and for long enough to carry a crop through its peak labour requirement. A national unemployment rate can describe the depth of economic distress without proving that the labour market clears efficiently at farm level.
There is also a difference between a job that exists throughout the year and a labour requirement that expands for several weeks or months. Permanent employment can support a household with some predictability. Seasonal work, by definition, is tied to the agricultural calendar. That does not make it unimportant — on the contrary, it can be decisive to the crop — but it does mean the incentives, logistics and recruitment dynamics are different.
The harvest window is where a small shortage becomes a large problem
Agriculture’s exposure to labour shortages is not measured only by the number of workers missing. It is measured by what those missing workers were supposed to do, and when. In a labour-intensive harvest, being short of workers at the wrong moment can affect picking speed, product condition, packhouse flow and the producer’s ability to move a perishable crop through the chain. Sihlobo makes the revealing point that timeliness matters precisely because agricultural products are perishable and delayed harvesting can reduce quality, income and long-term sustainability.
This helps explain why national stability and local urgency can coexist. Suppose a shortage affects one fruit-producing district for a few weeks. It may barely move national employment data. Yet for the farms involved, the consequences can be severe because the commercial value of the crop is concentrated in a narrow window. Macroeconomic data are essential for judging the national picture, but they are too blunt to measure every operational crisis occurring below it.
That is also why claims that “crops are rotting across South Africa” deserve scrutiny rather than repetition. Sihlobo explicitly warns that such a description may exaggerate the national situation. He says Agbiz and Agri SA survey work found the problem concentrated in a few provinces and industries. This matters because it allows two statements to be true at once: some producers may have experienced serious labour disruption, while the evidence does not support portraying the entire agricultural sector as unable to harvest its crops.
Good agricultural journalism has to preserve both truths. Minimising an affected farmer’s loss because the national number looks healthy is lazy. Extrapolating an affected farm or district into a nationwide agricultural breakdown is equally lazy. The journalistic task is to measure scale, identify commodities and regions, and distinguish a real operational bottleneck from a national narrative that has outrun the evidence.
Foreign labour matters — but nobody has produced a credible national share
The presence of foreign workers in South African agriculture is not in dispute. Sihlobo specifically notes their role in border provinces and in value chains such as sugar cane and fruit. Some join the sector during peak harvesting periods through legal processes. What remains uncertain is their share of total agricultural employment. Sihlobo says the data are difficult to assess, and that gap is one of the most revealing weaknesses in the public argument.
Without a credible national denominator, sweeping percentages about agriculture’s “dependence” on foreign workers should be treated cautiously. The available evidence supports neither the idea that foreign workers are irrelevant nor the claim that South African farming as a whole would cease functioning without them. The revealing reality is likely much more uneven: foreign labour can be operationally significant in specific places and seasons without representing the majority of the national agricultural workforce.
This matters because language shapes policy. If foreign workers are portrayed as the only people willing or able to work on farms, South African farm workers are erased from the story and agriculture risks reputational damage. If foreign workers are portrayed as unnecessary everywhere because the national unemployment rate is high, producers with legitimate seasonal requirements may be pushed into an administrative dead end. The revealing middle ground is to recruit South Africans wherever possible while maintaining lawful, efficient routes where seasonal foreign labour is genuinely required.
Sihlobo is careful on this point. He argues that agriculture has a responsibility to continue recruiting as many South Africans as possible, particularly because the sector is viewed as having job-creation potential in a country with extreme unemployment. At the same time, he says legal processes that have enabled foreign workers to participate where needed should continue to function. The revealing strength of that position is that it refuses the false choice between local employment and lawful seasonal migration.
Home Affairs is closer to the farm gate than it looks
Immigration administration is usually discussed as a border-control or documentation issue. For agriculture, it can also become a production-timing issue. If a farmer has a legitimate seasonal labour requirement and the lawful process for meeting it moves more slowly than the crop, the cost of administrative delay is not abstract. The crop continues to mature while forms, permits and approvals move through systems designed on a different clock.
Sihlobo therefore places a responsibility on the Department of Home Affairs, organised agriculture and other stakeholders to manage the process carefully and efficiently. That is an important agricultural point. It does not argue for bypassing immigration law; it argues for a legal system capable of functioning at the speed required by a seasonal industry.
The same responsibility falls on employers. A rushed harvest is not an excuse for unlawful employment, weak record-keeping or poor labour practices. South African agriculture cannot defend its reputation abroad while treating compliance at home as optional. The revealing policy challenge is to build a system in which lawful recruitment is workable enough that farmers are not encouraged toward informality by administrative dysfunction, while enforcement remains credible enough that compliance has meaning.
This is where the debate should become practical rather than ideological. Which districts experience recurring seasonal shortages? Which commodities are most exposed? How many workers are required, for how long, and at what point in the production cycle? How many positions are filled locally? Where legal foreign workers are needed, how long does the process take? The revealing absence of consistent, public answers to those questions is itself part of the problem.
The missing dataset may be the biggest story
South Africa has a national labour-force survey capable of showing that agriculture employed about 944,000 people in Q2 2026. What it does not give the public debate, at least not in the headline figures currently being used, is a detailed real-time map of seasonal vacancies, worker nationality, district-level shortages, unfilled harvesting positions and crop losses attributable specifically to labour constraints. The revealing gap between national employment data and farm-level operational data is where speculation flourishes.
That gap allows dramatic anecdotes to become statistics in the public imagination. A video of unpicked produce can be genuine and still not prove a national trend. A farmer unable to recruit enough pickers can be telling the truth and still not represent the entire sector. Conversely, a national employment figure can be robust and still fail to reveal a serious local shortage. The revealing answer is not to choose which anecdote we prefer; it is to improve the data connecting them.
Organised agriculture is well placed to help. Commodity organisations, producer associations and regional structures already sit closer to the seasonal realities of farming than a national household survey can. A consistent framework recording labour demand, vacancies, recruitment outcomes, legal migrant participation and harvest disruption could turn the current argument into something measurable. The revealing opportunity is to move from claims about “farmers” and “foreigners” toward evidence showing exactly where labour constraints exist and how large they are.
Government would benefit as well. A labour department trying to connect unemployed South Africans with vacancies needs more than a national unemployment rate. Home Affairs needs to understand when and where legitimate seasonal migration pressure is likely to emerge. Provincial departments need to know which production districts face recurring constraints. The revealing value of better data is that it could allow policy to respond before a harvest reaches crisis point rather than after images of lost produce have already shaped the narrative.
Agricultural technology does not make labour irrelevant
Sihlobo expects agricultural expansion, alongside technological advancement, to create more jobs. That may sound counter-intuitive to those who assume mechanisation simply replaces workers. In practice, the relationship between technology and employment differs sharply by commodity, farming system and stage of the value chain. The sources used for this article do not quantify how many jobs future technologies will add or displace, so any precise forecast would go beyond the evidence.
What can be said is that the revealing labour question is likely to change rather than disappear. Mechanisation can reduce the number of workers needed for some repetitive or physically demanding tasks, while expansion in production can create work elsewhere in the chain. Horticulture, packing, quality control, logistics, technical maintenance, irrigation management and processing all have different labour profiles. A sector can become more technologically sophisticated without becoming labour-free.
This makes training and mobility part of the longer-term discussion, although the current Q2 data do not tell us how large those mismatches are. If agriculture is expected to absorb more South Africans, policy cannot stop at telling unemployed people that jobs exist somewhere in the sector. The revealing challenge is to ensure people can reach those opportunities, meet their requirements and remain economically connected to them beyond a short production spike.
Export agriculture has more at stake than getting the crop off the land
The labour debate does not end when the last bin leaves the orchard. Sihlobo warns that South Africa must manage immigration and farm-labour issues without creating unnecessary reputational risk for export-oriented agriculture because key export markets place considerable value on farm labour practices. That is a revealing warning in a country whose high-value agricultural industries depend heavily on international buyers.
This should not be exaggerated into a claim that every labour dispute threatens immediate market closure. It does mean, however, that labour practices are part of the commercial identity of an export supply chain. Buyers, retailers and markets increasingly scrutinise how products are produced, and allegations of unlawful or abusive labour practices can travel far beyond the district where they originate. The revealing consequence is that a poorly handled local labour problem can become a national reputational problem even when it is not a national employment problem.
For producers, that changes the calculation. Labour compliance is not merely about avoiding penalties. It protects traceability, buyer confidence and the broader credibility of South African agriculture. For government, efficient legal processes are likewise not simply administrative housekeeping. They help protect continuity in sectors earning foreign revenue and carrying South African produce into competitive global markets.
The strongest defence is evidence. If South Africa can quantify where legal seasonal foreign labour is used, demonstrate that recruitment processes are lawful and document where genuine shortages occur, the sector is in a far stronger position than it is when public debate runs on assertion. The revealing lesson is that transparency is not agriculture’s enemy; uncertainty is.
Why labour-related risks at Transnet matter for SA agricultural and food trade
Five questions farmers and policymakers should be asking
1. Is South Africa experiencing a nationwide farm-labour crisis?
The available evidence does not support that conclusion. Agricultural employment stood at about 944,000 in Q2 2026 and was 4% higher than a year earlier, even after falling by about 15,000 jobs from Q1. Sihlobo also reports that Agbiz and Agri SA surveys found labour difficulties concentrated in a few provinces and industries rather than across the entire sector. The revealing answer is therefore that South Africa appears to have specific seasonal and regional labour pressures, not a uniform national collapse.
2. How can farmers report shortages when 8.5 million South Africans are unemployed?
A national unemployment total measures the scale of joblessness; it does not prove that every unemployed person is geographically close to seasonal agricultural vacancies or available at the precise time a harvest requires labour. The sources show that seasonality and location matter, particularly in fruit, sugar cane and border provinces, but they do not quantify transport, skills or mobility as causes of every shortage. The revealing distinction is between having a large pool of unemployed people nationally and successfully matching workers to temporary vacancies locally. South Africa needs better evidence on that matching process before simplistic conclusions are drawn.
3. Does South African agriculture depend on foreign workers?
Foreign workers are clearly part of the agricultural workforce, particularly in certain border regions, commodities and peak harvesting periods, but no credible figure in the sources establishes their share of total national farm employment. Sihlobo explicitly says that share is difficult to assess. The revealing conclusion is that foreign labour can be important in specific farming systems without proving that the entire agricultural sector is dependent on it. Claims of either total dependence or total irrelevance go beyond the evidence currently available.
4. Are claims that crops are rotting because workers have disappeared false?
They should not be dismissed outright, because isolated farms or regions can experience genuine labour disruption and perishable crops can suffer quickly when harvesting is delayed. What is not supported by the evidence cited by Sihlobo is the leap from individual cases to a picture of crops rotting countrywide. The revealing journalistic test is scale: where did losses occur, which crops were affected, how many farms were involved and what proportion of the harvest was actually lost because labour was unavailable? Until those questions are answered systematically, dramatic individual cases should not be promoted into national statistics.
5. What should South Africa do next?
The immediate answer is not to choose between South African workers and legal foreign seasonal workers as if one must exclude the other. Agriculture should recruit South Africans as widely as possible, while Home Affairs and organised agriculture ensure that lawful processes remain efficient where seasonal foreign labour is genuinely needed. Government and industry should also build better district- and commodity-level data on vacancies, labour demand and harvest disruptions. The revealing objective should be a labour system that protects jobs, keeps farms compliant, gets crops harvested on time and gives export markets confidence in how South African food is produced.
Conclusion: The labour crisis is not where the loudest argument says it is
South Africa does not need another agricultural argument in which facts are selected according to politics. It needs to confront the revealing paradox in full. The country has 8.5 million officially unemployed people and an official unemployment rate of 33.6%, yet agriculture still employs about 944,000 people and can experience real labour shortages in specific crops, districts and periods. Those facts do not cancel one another; together they describe a labour market that is large, stressed, seasonal and imperfectly connected.
The evidence does not show a farming sector collapsing because foreign workers have vanished. Nor does it justify dismissing farmers who face a genuine shortage when fruit must be picked, cane must be cut or a perishable crop must move. The revealing failure would be to turn either experience into an ideological weapon instead of asking what the data can actually support.
What South Africa appears to have is not one national farm-labour emergency but a series of narrower pressure points with the potential to become expensive when they collide with harvest timing. That is a very different problem, and it demands a very different response. National employment data should tell us whether agriculture is expanding or contracting. Commodity and district data should tell us where labour is short. Recruitment systems should connect South Africans to real opportunities. Home Affairs should make lawful seasonal processes workable where they are needed. Employers must remain compliant. Organised agriculture should document the gaps before anecdotes harden into accepted fact.
