The migrant labour system that built South Africa’s gold economy was a machine for moving value in one direction. It pulled men into the Witwatersrand goldfields from the Eastern Cape, Lesotho, Mozambique, Malawi and Zimbabwe, took their working years underground and sent them home when their lungs failed. Wages left as remittances. Silica dust stayed in their chests. And the compensation they were owed stayed stranded on the wrong side of a border.
The system had every rail it needed to carry labour in. It never laid the one to carry justice back out. That missing rail is the unfinished business of the goldfields, and for the first time the tools to build it exist.
Six years ago it looked as though that was finally being fixed. The 2019 class-action settlement with six gold mining companies created the Tshiamiso Trust and roughly R5-billion to pay silicosis and tuberculosis claims. By February 2026 the Trust had paid out about R2.5-billion across six countries. That is halfway through the money, and roughly halfway through the time it has to spend it.
The people are nowhere near halfway. Advocates count something in the order of 500,000 potential claimants. About 23,000 have been paid. In Lesotho alone, 5,464 men diagnosed since 1965 cannot be traced. Behind the Trust sits an older backlog: the Compensation Commissioner for Occupational Diseases recorded more than 107,000 unpaid claims as far back as 2017, under the century-old Occupational Diseases in Mines and Works Act (ODMWA). When the Trust’s mandate ends, whatever it has not paid does not roll over. It lapses.
Four segments of one rail
A working return rail needs four things: a law that recognises the claim, institutions that cooperate, records that mean the same thing across systems and technology that can move them. This is the anatomy of portability set out in a recent Digital Convergence Initiative report on cross-border social security. South Africa can now build three of them. Identity, data exchange and payment, the institutional and technical track, are being laid as we speak. The fourth is the legal segment, and it is the one we have left unbuilt.
The missing segment: a rule we wrote
The Trust’s own chairperson has acknowledged that strict ODMWA criteria rendered more than 70% of 83,810 certified claims ineligible. Read that again. These are men who presented themselves, submitted to a medical examination and were certified. Then they failed a legal test.
They fail it for reasons that have almost nothing to do with whether their lungs are damaged and almost everything to do with paperwork. Service records that the mines were supposed to keep and did not. Medical histories interrupted by decades and borders. Post-mortem evidence that was never collected because nobody told the family it mattered. The miner is asked to prove, 40 years later, a set of facts that his employer and the state were responsible for recording at the time.
Worse, the same claimant can get three different answers depending on which door he knocks on. The ODMWA, the Compensation for Occupational Injuries and Diseases Act and the Trust deed apply different definitions and different thresholds to the same lungs. This is a structural choice. It has stood unresolved for a quarter-century, through every commission and green paper that recommended harmonising the two regimes.
Two changes would lay this segment. The first is a most-favourable-to-the-claimant rule: where the regimes point to different outcomes for the same person, apply the outcome most favourable to him, settled by a standing dispute body. This principle already exists. It is the operating basis of the Ibero-American Convention on Social Security.
The second is to move the evidentiary burden to the party that lost the evidence. Where a claimant can show he did the qualifying work but the records are missing, the presumption should run in his favour, rebuttable by the employer. The alternative, which is the current arrangement, charges a failure of corporate and state record-keeping to a dying ex-miner in a village in Mount Frere.
Laying the other three: identity, data, payment
None of this requires new technology, and neither does the rest of the rail. South Africa is building its digital public infrastructure for domestic reasons, and it already contains the other three segments. The Roadmap for the Digital Transformation of Government sets out identity through Home Affairs, real-time data exchange through National Treasury’s MzansiXchange and modernised low-cost payment rails.
Applied here, they let the state verify a claim once against Home Affairs, the Medical Bureau for Occupational Diseases, the Commissioner and mine service records, instead of forcing an ailing ex-miner to reconstruct decades-old paperwork across five institutions that do not speak to each other.
Verifiable digital credentials matter more than they sound. An exposure history or a medical certification issued as a tamper-evident credential is evidence that survives the claimant’s move home, the closure of the mine and the loss of the file. It also prevents the next generation of unprovable claims. Computer-aided screening of chest X-rays, a field in which South African teams are doing serious work, can compress diagnostic queues without displacing clinical judgement.
The last stretch is the border. A claimant in Maseru or Maputo authenticates locally; South Africa recognises that identity and that medical certification as legally valid; payment lands with a local provider. The technology to run this exists. What is missing is a set of bilateral agreements giving those credentials legal force on both sides. The main sending states could conclude them within a year, if anyone treated them as urgent.
Finishing the rail
The Presidency’s digital programme, National Treasury and Home Affairs can and should adopt ex-miner compensation as the first cross-border use case for the build: a bounded population, a defined benefit, a hard deadline and a moral claim nobody contests. But the segments must be laid in the right order.
Parliament and the relevant departments should legislate the most-favourable-to-the-claimant rule and the evidentiary presumption first. Lay the technical track before the legal one, and better digital infrastructure will simply produce faster rejections: a modern, interoperable, biometrically verified machine for telling sick men they do not qualify.
The machine moved value out of these men’s bodies with ruthless efficiency for a century. The rail to send something back is three-quarters built, most of it in our own hands. The last segment is a rule we wrote, and can change. Let’s finish the return rail. DM
Genesis Analytics advises governments on digital transformation and social protection programmes, including work related to the initiatives discussed here.
Comments
Scroll down to load comments...