Here is a controversial suggestion: shut down the Government Employees Pension Fund (GEPF) and use its assets to pay off government debt.
Why? Start with the numbers. The GEPF portfolio has earned an average return of 7.2% over the past 10 years. Over the same period, the government has paid around 8.5%-10% to borrow in the market. This is like saving into an account while borrowing on a credit card at a higher rate. It is not rational.
Closing the GEPF and paying down debt would instantly improve public finances, with the government taking on the liability of paying civil servants’ benefits directly thereafter (while of course still collecting pension deductions from them).
To see why this is possible, consider what the GEPF actually is. It is the biggest investor on the JSE and one of the biggest public pension funds in the world, with R2.69-trillion in assets managed by the Public Investment Corporation (PIC). It is a defined benefit fund, meaning members get set benefits irrespective of how its assets perform. And it is, in theory, fully funded, holding assets to cover its future liabilities to members.
That structure has two implications. First, underperformance of GEPF assets is not a problem for members. Their benefits are set by a formula based on years of service and final salary; if the assets fall short, the government (that is, taxpayers) must make up the difference.
Second, viewed on a consolidated basis, the portfolio is effectively government savings backing government liabilities. Segregated portfolios are common in defined benefit corporate schemes too, existing primarily to protect members if the company goes under. But companies are still on the hook to make good any shortfall and the government is meant to be practically a risk-free debtor.
A fully funded scheme like the GEPF is unusual globally. Brazil, India and most of Southeast Asia run civil servant pensions as unfunded defined benefit promises, with civil servants relying on the government to stay solvent and be willing to pay. Advocates of funded schemes argue that they force governments to recognise pension obligations explicitly rather than pushing costs onto future taxpayers. Perhaps, but that discipline comes at the price identified in South Africa’s case: paying more on debt than the savings earn.
So why does South Africa have one? History. The GEPF was put together in 1996 from several predecessor schemes. In the run-up to 1994, the apartheid government switched many civil servants from pay-as-you-go arrangements, where pensions were paid out of current contributions (or by government directly), to funded ones.
One consequence was to provide civil servants with a buffer against any perceived credit risk associated with the incoming democratic government. It was a considerable drain on an apartheid state already deeply underwater, a measure of how important protecting these benefits was felt to be.
The GEPF consolidated the funded approach, reassuring apartheid-era civil servants that their benefits wouldn’t be diluted or defaulted on, though at first it was considerably underfunded. The awkward reality is that for years afterwards, the GEPF was effectively a scheme for apartheid-era retirees that was being funded by the democratic government. That has changed: it now has 550,000 beneficiaries and 1.27-million active members, increasingly reflecting today’s diverse civil service.
The credit risk that motivated all this never materialised. The democratic government has proved a far better fiscal risk than the apartheid state, and clear regulation plus powerful public sector unions give employees strong protection over their benefits. But the mistrust survives. Civil servants would still rather have a portfolio backstopping the promise than rely on the state’s word. That, more than anything, is why closing the fund would be practically impossible.
It also leaves the GEPF in an awkward position: a R2.69-trillion pool of public money with no clear financial rationale. In the absence of one, the portfolio has been treated as a mechanism for the state to pursue all manner of policy objectives: transformation, economic growth, occasionally saving banks.
Which brings us to the PIC, which must manage the fund under its formal mandate to members while also meeting the expectations of its shareholder, the government. Serving two masters, where the underlying financial logic is already tenuous, has proved difficult. The PIC has repeatedly struggled with governance, with investment decisions across multiple administrations shaped by political expediency.
The Zuma era was the nadir, with the PIC funding dubious deals for dubious reasons. But as the latest chaos at the institution shows, even after the well-considered recommendations of the Mpati commission, it cannot find an even keel. Nor will it, while it remains a hybrid — serving fund members and beneficiaries on one hand, the government’s policy objectives on the other.
My opening suggestion was tongue in cheek. No government is going to close the GEPF. But the thought experiment exposes an uncomfortable question: if the fund exists, what public purpose does it serve? Until that question is answered, the PIC will continue to be asked to pursue conflicting objectives … and remain ungovernable.
- This is my first column for the FM, the place I had my first real job 26 years ago. Much has changed since then. The contested role of the GEPF, sadly, has not.
Dr Theobald is founder and chair of research-led consulting firm Krutham