The worker you dismissed in March is at the next gate in April: what a shared history catches

What a shared history catches, and what one page misses

Two Aprils, two farms, one worker. On the first farm the season opens and a picker you let go last March for stripping diesel is standing in your onboarding queue under a slightly different spelling of his name. You do not know it, because all you have is the ID copy in your hand. He is hired, and by June the diesel is walking again. On the second farm, twenty kilometres away, the same man reaches the gate and the system flags him: same ID number, worked at a neighbouring farm, left after a disciplinary for theft that was properly run. The supervisor asks a few more questions, and makes a decision with his eyes open.

That is the whole difference, and it is not about catching people out. It is about a farm owner in South Africa being able to see what he is actually hiring. A shared worker history turns a blind rehire into an informed one. One farm working alone sees a single page. A farm on a shared network sees the pattern.

Handled well, this is the one thing a payroll system can never give you, and the strongest reason to run labour on a network rather than a spreadsheet. Handled badly, it is an illegal blacklist that puts you in front of a defamation claim. The whole of this piece is about staying firmly on the right side of that line.

What one farm can see, and what it can't

Seasonal labour moves. A picker works citrus in Limpopo in winter, grapes in the Hex River valley in summer, and a third farm in between. Across a year he may pass through four or five gates, and each farm meets him fresh, with no memory of the last one.

What a single farm holds on a new worker is thin: an ID copy, whatever he chooses to tell you, and a gut feel from a five-minute conversation at a busy gate. It cannot see that he left the last farm mid-contract, or that he is already onboarded three farms over under a different spelling, drawing two advances. The information exists. It is just scattered across farms that never compare notes. That gap is exactly where a dismissed worker reappears, and where a duplicate slips through.

What a shared worker history actually holds

A shared worker history is not a rumour mill, and it is not a rating out of ten. To be useful, and to be legal, it holds facts and only facts.

The things worth sharing are the ones you could prove: which farms a worker has been registered on, and when; the role he held; whether a contract ran its term or ended early, and if it ended in a disciplinary, that a fair process was followed. And the plainest, most valuable flag of all, the duplicate: the same ID number trying to onboard twice, which is arithmetic, not opinion.

What does not belong in it matters just as much. Not “he seemed lazy”. Not “I heard he was trouble”. Not a dismissal that never went through a hearing. Opinion, hearsay, and unproven allegations are the raw material of a defamation claim, and they make the record worse, not better. A shared history earns its place by being narrower and harder than the gossip it replaces.

The decision stays yours

A flag is not a verdict. The point of a shared worker history is to put a fuller picture in front of the person doing the hiring, not to hire or reject for them.

When a worker comes up with history attached, the record tells you what happened; you decide what it means for your farm. A duplicate ID needs sorting out before anyone is onboarded. A contract that ended in a proven theft matters more for the fuel store than the packing line. An early exit might have a good explanation, and the worker is entitled to give it. Under South African data law you cannot leave that call to an automatic rule that bars a worker with no human judgement, a machine saying no on its own is a problem in itself. The owner sets the standard; a person applies it, case by case.

Keeping it legal: consent, facts, and the right of reply

This is the part farms get wrong, so it is worth being exact. A shared worker history is lawful in South Africa when it is built the right way, and unlawful the moment it is not.

Consent, at onboarding. Under the Protection of Personal Information Act, a worker's history is their personal information. The clean basis for sharing it is consent, captured when they onboard: they agree that their employment record can be checked against, and added to, the network, for the purpose of hiring. Explain it plainly, in a language they speak, and record it.

Facts, fairly established. Share only what is true and what you could stand behind. A dismissal counts only if it followed a fair process. Publishing “dismissed for theft” about a worker you fired without a hearing is both inaccurate and defamatory, and truth is the defence you would need if it is ever challenged.

The right to see and correct. POPIA gives every worker the right to see the information held on them and to correct what is wrong. A shared history has to allow that: a channel for the worker to say “that farm is not mine” or “that dismissal was overturned” and have it fixed. A record nobody can challenge is exactly the blacklist the law will not allow. The same discipline behind the three pre-hire checks every farm should run applies here: verify, document, and keep it to what you can prove. It is a different job from the shared incident record that tracks a suspect across farms; this is a worker's own employment history, shared with his consent.

A shared history earns its place by being narrower and harder than the gossip it replaces.

A rehire-vetting checklist for the season

The test for whether your shared history is a tool or a liability. If every line below is true, you are on the right side of it. Pin it where onboarding happens.

Labour Link's Workforce Control runs this as Background Verification across a network of farms at hr.labourlinksoftware.co.za: the duplicate-ID catch at the moment of onboarding, a factual shared history the worker consents to and can correct, and a flag the person at the gate acts on rather than a machine that bars them. It is the one thing a payroll tool cannot build, because it is not built from payslips. It is built from farms choosing to see the same worker clearly. The picker you let go in March becomes a decision, not a surprise.