In mid-December 2021, National Treasure released discussion documents concerning planned and possible changes to South African retirement funds. There has been a lot of subsequent discussion but not much has changed yet.
I write this article two months after the February 2023 budget speech. Therein it was confirmed that the two-pot retirement system would indeed be going ahead with effect from 1 March 2023. We still await final legislative clarity on the details of how the two-pot system will work, although National Treasury has recently provided verbal feedback on this.
The auto-enrolment proposal that was also shared in the December 2021 discussion papers is yet to proceed, even in discussion phase, because of the focus on the two-pot system and we still await a new version of the national social security fund idea from the Department of Social Department, who still appear to passionately believe in this idea.
Although the plan is well known, the details of the two-pot system need to be clarified. One area is how this will be applied to defined benefit retirement funds: how would benefits be adjusted if a member accesses part of their retirement benefits in advance? Most defined benefit retirement funds were closed to new entrants many years ago and their membership is dwindling. The one notable exception is the Government Employees Pension Fund (GEPF) with its 1.2 million active members. The GEPF will soon (?) also be required to comply with the two-pot system and the impact is likely to be significant particularly because the GEPF’s annuity/gratuity benefit structure does not fit into the idea of flexible access being based on early access to the one-third commutation option that members have at retirement. The GEPF is not the only retirement fund with such a benefit structure, some other retirement funds in the public sector and in local authorities have similar concerns.
‘Legacy’ retirement annuity funds are another area of concern where clarification is required, as the structure of many of these entities also does not fit well into the twopot model. There is an expectation within the pensions industry that these funds will be allowed to apply for exemption.
Even the idea of seeding capital (that is, moving some of the existing benefits into the accessible savings pot), which sparked intense debate, needs finality. National Treasury has said that the intention is to allow for 10% of existing accrued benefits to be used for seeding capital, subject to a maximum of R25 000. The regulations, which are expected to be released in June/July 2023, should provide further guidance on this.
Another key consideration that should not be overlooked is how pension backed housing loans will work in the two-pot world? Would members choose to take a loan against their savings pot if they could simply access the money instead? can the preservation pot with its preservation requirement be able to be used to back housing loans? Will the two-pot system lead to the demise of pension backed housing loans? Is this a good thing or a bad outcome? The discussions around reform ideas have generated many different views on this.
The South African government’s different departments seem to all agree on the need for some sort of mandatory retirement savings system and also the need for a national pension fund. However, the devil will be in the details.
Mandatory retirement savings can work well for those in the formal sector, who can make contributions that are regular and fixed in nature. National Treasury’s autoenrolment idea is therefore expected to work well for formal sector workers who have an employer.
The examples of the British auto-enrolment system (which requires all employees to be placed into a workplace pensions scheme with a minimum total contribution rate of at least 8% of earnings) and the Australian superannuation system (which is increasing the minimum total required contribution rate from 9% to 12% of earnings over time) are well known. It will be interesting to see what minimum contribution rate is required under a South African auto-enrolment system and how this impacts on existing retirement funds which often have relatively low contribution rates. Will the South African government dictate that contributions be based on employees’ full cost to company remuneration? Defined contribution schemes fit better with the concept of cost to company remuneration, how would the minimum contributions and prescribed remuneration definition be applied to DB funds?
Both Nigeria (in 2004) and Ghana (in 2008) introduced compulsory occupational retirement savings for their formal sector workers. These countries have relatively small formal sector employment and, even within their formal sectors, compliance seems to be at only about 50%. By comparison, South Africa and its neighbours Namibia and Botswana have roughly 60%
contributory pension coverage in the formal sector despite the absence of a mandatory occupational retirement savings contribution scheme requirement.
The Kenyan government and its national social security fund have been fighting court battles for years to introduce a new social security system. In February 2023, this matter was finally concluded with the relevant legislation being allowed to reach implementation stage. Part of Kenya’s new pensions model makes occupational retirement savings compulsory for formal sector workers, with the alternative being higher contributions to the national social security fund as a default.
Zambia has had its current national pension fund since 2000, but outside of the national fund contributory pension membership is relatively low in the formal sector. The Zambian government has therefore been debating for years (prior to the COVID pandemic) the need to introduce mandatory occupational retirement savings. South Africa is therefore behind some other African countries in this debate, but implementation has been a challenge for these other countries.
A big part of the pensions debate in South Africa and the rest of our continent remains how to include informaljatypical workers. As compulsion will not work, any system must be voluntary in nature. Here too, other African countries have been trying for years to include their relatively larger informal sectors in contributory pensions. The South African government hopes to follow suit by using a combination of incentives, FinTech and a government default fund. However, the experience from other African countries shows that it is unlikely that we will be able to successfully include even 25% of these workers in a contributory pensions system.
Such workers’ incomes are typically volatile, and they are not able to contribute regular, fixed amounts. They also lack often trust in both the public sector and private sector. A key difference between South Africa and other African countries is the social grant system. The South African state old age grant is designed to support those who were unable to save for their own retirement.
Compared to the rest of the continent, our noncontributory pension system has wide coverage and pays relatively generous benefits. These benefits may seem to be very low when compared to average salaries in the formal sector, but it is above the food poverty line quoted by StatsSA and research shows that it makes a major difference in the lives of many people in our country.
With the ongoing infrastructure collapse in South Africa, many of us have concerns about the idea of a national pension fund. However, in the private sector we have also struggled to find a suitable solution for informal sector and low income workers. Therefore, any viable solution for such workers will likely need to be based on some sort of private public partnership while at the same time realising that it is never going to be the main pillar on which pension provision for these workers is built.