As South Africa moves into its next phase of retirement reform, the global experience offers hard-won lessons – and some uncomfortable truths – about who really gets covered.
South Africa has a savings problem decades in the making. World Bank Open Data places gross domestic savings at 15.3% of GDP in 2025 – below Brazil at 17.4%, the United States at 22.7%, the European Union average at 21.8%, and India at 32.6%. The comparison is telling: too little of what the country produces is being converted into savings for future resilience, investment and retirement security. The issue is not only preservation within the retirement system, but broader participation in long-term saving.
The two-pot system, introduced on 1 September 2024, was an important first step. By preserving two-thirds of future
contributions until retirement, while allowing one-third into an accessible savings pot, it tackled the historic leakage caused by
members resigning simply to access retirement savings. For many members, it has already changed the conversation about what retirement money is for.
But preservation alone was never going to be enough. The bigger reform is auto-enrolment: a system that would require employers, through legislation, to enrol all employees into an approved retirement fund. National Treasury’s 2021 discussion paper set the groundwork. The question is increasingly less about whether auto-enrolment will happen, and more about how it will be designed, who it will reach, and whether it will expand coverage rather than simply formalise what already exists.
The coverage problem
Auto-enrolment works best where there is already a clear and enforceable employer-employee relationship. The UK’s programme, launched in 2012, remains the most-cited success story. Participation among eligible workers rose from the mid-50% range in 2012 to about 88% by 2021/2022. A phased roll-out, starting with the largest employers and gradually drawing in smaller businesses, helped build momentum rather than resistance. When minimum contribution rates increased in 2019, the anticipated wave of opt-outs did not materialise.
But the UK has something South Africa does not: a formal sector that covers the vast majority of working people. South Africa’s challenge is structurally different. According to the Statistics SA 2022 Labour Force Survey, roughly four in ten formal sector workers are still not covered by any retirement fund. And that is before we consider the millions who work informally, intermittently, or across multiple platforms. The domestic worker, the Uber driver, the seasonal farm labourer, the part-time caregiver: auto-enrolment may still leave them behind unless it is deliberately designed to reach them.
Our African neighbours provide a cautionary note. Nigeria introduced compulsory retirement savings for formal sector workers in 2004; Ghana followed in 2010. Both reforms were important, but both countries continue to face compliance and coverage challenges, particularly outside larger employers and stable payroll environments. The lesson is not that compulsion is wrong, but that compulsion on paper does not guarantee coverage in practice.
Kenya and Rwanda have taken a different approach, using voluntary, digitally enabled micro-pension or long-term savings platforms – Mbao and Ejo Heza – to extend coverage into the informal economy.
South Africa’s approximately 854,000 domestic workers, and a fast-growing gig economy workforce, deserve at least as much thought.
Who should run it?
The governance question is one South Africa cannot afford to get wrong. There is a tension between two visions: a centralised, government-administered National Social Security Fund, and a model that routes contributions through existing approved private funds with a government default option for those whose employers do not choose a specific provider.
In my view, the more durable model is an independent, transparent framework built on public-private collaboration. This is not ideological. Government’s fiscal position is constrained, with debt-to-GDP in the mid-70% range, while the Old Age Grant, currently R2,400 per month, places pressure on the fiscus. Adding a new pension obligation without careful design, governance and accountability could stretch an already strained system and weaken trust before the reform has had a chance to work.
South Africa’s private retirement sector manages close to R6 trillion in assets. It has deep institutional capacity. Routing auto-enrolment through existing, well-regulated approved funds – with a properly governed government default – could harness that infrastructure without duplicating it. It is, in fact, the direction National Treasury’s own 2021 paper pointed toward.
A phased approach is a design principle, not a compromise
Every successful auto-enrolment implementation globally has one thing in common: sequencing over speed. Start where the infrastructure already exists – with large employers, established payroll systems and regulated providers. Set initial contribution rates low enough not to trigger mass resistance. Increase them gradually as evidence shows workers and employers can absorb them, then watch whether participation holds. Make re-enrolment mandatory on a rolling basis, so early opt-outs do not become permanent exits from the system.
South Africa will face real friction: employers worried about payroll costs, workers already stretched thin, and data and regulatory systems that will need significant upgrade to handle millions of new members. These constraints are not reasons to delay indefinitely, but reasons to design carefully. Financial literacy campaigns should be load-bearing parts of the design, helping workers understand not only that money is being deducted, but why long-term saving matters and how the system protects them.
None of that makes the project less urgent. Retirement poverty is a slow emergency – invisible while it builds, devastating when it lands. The two-pot system was a necessary correction to a broken preservation model. Auto-enrolment, if implemented with the right sequencing and genuine coverage ambition, could be the next – and far larger – task.
The global playbook exists. The question is whether South Africa learns from it, and applies those lessons thoughtfully.
