The scourge of slow growth and rising unemployment that has plagued SA over the past decade has had two big consequences: rapidly increasing government debt and worsening poverty. Now, as demands grow for increased income support for unemployed adults, doubts are being raised about the sustainability of the country’s tax base.
Many South Africans are worried that, in the absence of faster growth, the small, ageing tax base will buckle under the burden of higher taxes — especially if a budget-busting basic income grant (BIG) is instituted by a ruling party desperate to shore up waning support.
Over time, the government has come to rely increasingly on personal income tax (PIT) as the dominant source of tax revenue. It now accounts for more than 35% of total tax revenue, against the Organisation for Economic Co-operation & Development (OECD) average of just 23.5%.
Despite that, taxpayers are seeing little service for the taxes they’re paying. And they’re becoming more and more irate about the burden they’re shouldering. The FM estimated last month (Cover Story, August 11-17) that the average middle-class family of four is spending an additional R15,000-R40,000 on private health care, education and security to fill the governance vacuum. Now, with ailing power utility Eskom requesting a 32% tariff increase for 2023/2024, the burden of indirect taxes is set to grow — as is the public backlash.
As it is, there’s a growing popular perception that SA is relying on a shrinking pool of high-income earners — an increasing number of whom are emigrating. The concern is that the country, as a result, will become less fiscally sustainable unless the state reins in expenditure or resorts to continuous — even draconian — tax hikes levied on those who remain.

While these fears do contain strong elements of truth, the facts paint a less dire and more nuanced picture, according to new tax data from the SA Revenue Service (Sars), shared exclusively with the FM, and released here for the first time.
The first message from the data is that despite the economic devastation caused by Covid, rising unemployment and emigration, SA’s PIT base isn’t shrinking. It’s growing — but it’s doing so very slowly.
The number of unique individual taxpayers from whom Sars received IRP5 and IT3(a) tax certificates has stayed constant on a five-year view, rising from 14.6-million individuals in 2016/2017 to a pre-Covid peak of 15.15-million in 2019/2020 before falling back to 14.6-million in 2020/2021 and 2021/2022.
Similarly, the number of individuals with a pay-as-you-earn (PAYE) liability (excluding those below the tax threshold who qualified for full refunds) grew from 6.63-million in 2016/2017 to peak at 7.22-million in 2019/2020, just before Covid struck. It fell to 6.72-million in 2020/2021, before staging a mild recovery to 6.85-million in the tax year ending in February 2022 (see “A Bumpy Ride” graph).
So, while the PAYE tax base declined sharply during the pandemic, shedding almost 500,000 individuals year on year, it has recovered some of this lost ground. While the number of personal income taxpayers has, therefore, not grown substantially since 2017, it has grown.

More interesting have been the shifts between income groups.
Contrary to the popular belief that most of the growth has been in the top income bracket, making Sars increasingly dependent on the top 5% of income earners, it has in fact been the middle-income segment that has grown faster and is shouldering a rising share of the PIT burden (see “Sharing the Burden” table).
The Sars data shows that between 2017 and 2022, the share of PAYE contributed by the wealthiest 5% (group F&G on the table) fell from 45% to 40.3%. At the same time, this segment shed 36,000 taxpayers, to number just under 400,000 by 2021/2022.
By contrast, it was taxpayers earning R20,000-R30,000 a month (group B on the table) that grew the most over these years, adding about 118,000 individuals, to number 1.56-million by 2021/2022. This segment also increased its share of PAYE the most — almost 2 percentage points (pp) — from 12.8% to 14.7%.
Taxpayers earning about R30,000-R56,000 a month (groups C and D) together added 60,000 taxpayers. Group C’s share of the PAYE burden rose by 1.5pp to 14%; group D contributed 0.8pp more, raising its share to 12.7%. The R56,000-R72,000 a month bracket (group E) remained much the same size, but nudged its share of PAYE up by 0.6pp to 10.4%.
What this all means is that, over time, there has been a gradual shift in taxpayers to higher income groups with a concomitant rise in their tax burden. But the top end has clearly not fully recovered from the Covid setback.
It is also clear that the tax base rests on the shoulders of comparatively few individuals: 40% of PAYE comes from just the top 5% of taxpayers — the 400,000 people earning more than R72,000 a month.
Though the data understates the size of the top end because it excludes provisional tax payments, it is nevertheless flashing warning lights about the slow erosion of this group.

‘R100bn undercollected’
One would expect Sars commissioner Edward Kieswetter to be having sleepless nights, given the pressure on the state to ratchet up tax-financed welfare spending to support the masses of unemployed. Especially at a time when there is growing evidence that SA’s best and brightest are emigrating.
But, in an interview with the FM, Kieswetter cuts a sanguine figure in his large Sars corner office on Cape Town’s foreshore. He is calm and forthright, and has compelling reasons for why he maintains a positive outlook on the health of SA’s tax system.
First, he argues, PIT collection has been “resilient”, outpacing GDP growth in recent years.
Given how difficult the past decade has been both for Sars as an institution and in terms of slowing growth, PIT revenue has been astonishingly buoyant, rising from 7.3% of GDP in 2011 to 9% in 2018. For the past two years it has been slightly below 9%, which was nonetheless a strong performance in a lockdown economy.
Between 2017 and 2022, total PIT revenue grew by R129bn or 30%. This translates into an average annual growth rate of 6% a year, ahead of both nominal GDP growth and inflation. This buoyancy is partly because wage growth has exceeded GDP growth in recent years.
As Kieswetter sees it, the tax system is benefiting from two apparent trends: “upward mobility”, as people shift up from lower- to higher-income brackets, and “a compliance dividend”, as Sars recovers from its nadir during the state-capture years.
While there is certainly some truth in this, careful analysis of the data shows that the upward mobility and PIT buoyancy is partly an illusion caused by the state not fully compensating taxpayers for fiscal drag. In other words, slightly higher numbers of people are moving into slightly higher income brackets because they’re being dragged up by inflation — not because the economy is growing rapidly and creating thousands of new jobs.
At the same time, the state has tightened up tax-reduction regimes involving pensions, medical scheme credits and travel allowances. However, it is not clear that there is much scope for further gains from policy tightening without faster growth.
Kieswetter also takes comfort in the fact that provisional PIT compliance is very low, at 56%, as this suggests there may be scope to bring more income tax into the net. PAYE payroll compliance, on the other hand, is much higher at 74%, signalling that employers are mostly voluntarily compliant.
He estimates that SA has a PIT gap “well in excess of R100bn” and he’s aiming to reduce it by R10bn-R20bn a year through increased compliance efforts and administrative efficiencies.
“If the level of compliance was already 90%, I would be worried,” says Kieswetter. “But because it’s so low, it means there’s still a lot of headroom.”
He adds: “I believe, given all the data, that the least-cost opportunity for us to improve SA’s fiscal integrity is by continuing to drive compliance efficiencies. We believe that just from PIT, coming mainly from provisional and wealthy taxpayers, there’s at least R100bn being undercollected.”

Compliance is slowly improving
Since Kieswetter took the reins in May 2019, Sars estimates it has raked in R515bn in additional revenue purely from better compliance. In the 2021 tax year alone, R215bn, or almost 14% of the R1.55-trillion collected, is credited to administrative improvements that raised payment compliance by 4.2% for corporate income tax (CIT), 1.4% for VAT and 1.8% for PAYE.
“The compliance revenue generated over the past years provides us with the confidence that we should be able to extract overall compliance revenues in excess of R250bn per annum, with PIT contributing significantly to this,” says Kieswetter.
“It’s too early to declare victory. We have a long way to go, and still drop the ball far too often, but it’s important to appreciate the progress already achieved,” he adds.
For instance, most of the Nugent Commission’s recommendations to address governance failures at Sars have been implemented. In addition, the tax authority’s IT infrastructure and customs division are being modernised, additional skilled staff employed, and work on illicit activity intensified.
The decision to relaunch the large business and international unit closed by former Sars commissioner Tom Moyane has certainly been vindicated. In the 2021 tax year it dealt with 345 cases of transfer pricing, base erosion and profit shifting, yielding almost R12bn. This helped it raise R111bn more in revenue year on year and increase compliance by 3% to 92% in this segment.
Our biggest crooks aren’t those who are leaving; our biggest crooks are those who are staying and not paying
— Edward Kieswetter
Similarly, the new high-wealth individuals unit raised an additional R265m and improved compliance for this segment by 4% to 92%. It has developed an integrated view of high-wealth individuals in terms of their assets and related entities, which provides a fuller picture of their tax obligations than simply relying on their income from regular employment.
Last week, Sars chief revenue officer Johnstone Makhubu got the green light to create a new team of 10 people to bring into the tax net wealthy individuals who remain unregistered. It will be examining third-party data, including invoices from luxury boutiques, and working with the banks to check the sources of large inflows into these individuals’ accounts.
By leveraging big data, machine learning and artificial intelligence, the tax authority is both improving its engagement with honest taxpayers and its ability to crack down on negligent taxpayers and criminals, according to Kieswetter.
However, it will need specialised skills and resources to triangulate the data to raise compliance.
Confidence in the tax authority is also on the rise. According to PwC’s 2022 Taxing Times Survey, 45% of respondents say their trust in Sars has improved in the past 12 months, due mainly to media coverage of large investigations and fraud that has been uncovered.
The latest example is what Sars refers to as the “transnational plunder network” of Gold Leaf Tobacco, first reported by Daily Maverick. It has taken nine Sars officials three years to plough through 100 terabytes of data to crack the allegedly illegal cigarette syndicate, something Sars estimates could have robbed the fiscus of as much as R3bn. To have outsourced the work to a forensic services company would have cost R100m, according to one quote Sars received.
Kieswetter wants to hire 3,000 more people, taking Sars’s headcount to about 15,500, and has appealed to the National Treasury to increase his budget by R15bn over the next three years. Sars’s annual allocation is now R11.8bn. In 2014 it was R10bn. So, in real terms, it receives a smaller budget now than it did eight years ago.
“It’s ridiculous,” says Kieswetter. “The tragedy is that if we could double the workload, we could double the compliance. Don’t tell me we need a wealth tax or to increase the VAT rate! The lowest-hanging fruit in SA is the revenue we could get from increasing tax compliance. It’s the least-cost option for improving tax revenue.”
Judge Dennis Davis, a consultant to Sars’s high net worth unit, shares Kieswetter’s view.
“He’s dead right,” he tells the FM. “We don’t have enough staff and expertise. Sars is being underfunded by R4bn-R5bn a year. I can see [Kieswetter] is unbelievably frustrated. I share it. Give him R4bn and he’ll turn it into R30bn.
“If I held the purse strings, I’d spend on data specialists because just a small number can raise a huge amount more tax. And if you plug the gap, you could reduce taxes, not raise them.”
Emigration — it’s not as bad as you think
Kieswetter is not overly concerned about the rate of emigration, given that the amount of tax leakage due to noncompliance significantly outweighs what he believes SA loses to emigration each year.
“Our biggest crooks aren’t those who are leaving; our biggest crooks are those who are staying and not paying,” he says with a wry chuckle.
And he has the emigration certificates Sars issues each year to prove it.

Officially, 6,801 people financially emigrated in 2021/2022. The reason Kieswetter still refers to emigrants jokingly as “crooks” is because more than 70% of this group implausibly claimed to be among society’s poorest, earning less than R70,000 a year. Only 63 admitted to having incomes exceeding R1m.
This suggests that most of those leaving have either not fully disclosed their taxable income, or their wealth is being disclosed as coming from sources other than income. For example, they may be pensioners living off interest and dividend income, or the children of wealthy parents.
Either way, the total number of emigrants is small relative to the size of the tax base, whether taken as the 14.6-million people from whom tax certificates were received or just the 6.85-million taxpayers with a PAYE liability.
The rate of financial emigration is also vastly outpaced by the rate at which SA is growing the tax base. For every taxpayer who emigrated in 2021/2022, SA gained six new taxpayers.
This is based on Sars’s estimate that a net 44,575 new taxpayers contributed to the total tax take in that year — including individuals and newly registered VAT vendors. This group generated an additional R4.7bn net in tax revenue.
Even so, it’s hard not to be unnerved by the upward trend in financial emigration — from 4,097 individuals in 2018/2019 to 6,801 in 2021/2022. This represents a 66% increase during a period that encompasses the Covid pandemic, when travel was hugely restricted.
“Of course, every taxpayer that we lose because of emigration has to be a concern because it’s not only a loss of tax revenue, it’s an erosion of confidence in SA,” says Kieswetter. But he doesn’t view emigration as an alarming or growing problem.
However, Sars’s data captures only those people for whom formal emigration certificates have been issued. The actual total could be more than double this number, judging by unofficial research reports based on statistics on SA immigrants issued by destination countries.
David Kaplan, professor emeritus at the University of Cape Town’s school of economics, and Australian emigration researcher Thomas Höppli used destination-country data to estimate that SA lost 1,500 people a month (18,000 a year) between 2000 and 2017 to other English-speaking countries, mainly Australia.
According to their 2019 research paper, this exodus was particularly strong between 2015 and 2017, when state capture was at its height, with more than 60,000 South Africans emigrating on a net basis during this period.
Using Australian Bureau of Statistics data on migrants’ skills levels, they revealed that in 2016, 54% of SA emigrants to Australia over the age of 15 had tertiary qualifications. More than half (51%) were either professionals or managers, 13.8% were administrators, 11.5% were technicians or tradespeople, and 8.4% were in community or personal services.
The top five occupations of SA nationals entering Australia were (in descending order) accountants, IT and software-related professionals, engineers, teachers, and metal fitters and machinists.
In November 2019, HSBC Securities economists David Faulkner and Thato Mosadi used UN data to estimate that the total stock of SA migrants rose by 14,000 a year (or 70,000 in total) over the 2015-2019 period, with large increases in emigration to the UK and Australia.
They also found that SA emigrants living in OECD countries in 2015/2016 were relatively well skilled, with 55% possessing a university degree compared with less than 6% of South Africans back home. This 49pp gap was higher than for any other emerging country, bar India and Malaysia.
The skills makeup of SA’s emigrant population suggests that further emigration is likely to be concentrated among better-educated and higher-skilled workers — the same group on whose taxes SA depends. Just last week, Bloomberg reported on a survey by the Joburg-based Social Research Foundation, which found that 53% of graduates in a survey pool of about 3,200 respondents are thinking of leaving SA. And 43% of those earning more than R20,000 a month are considering emigration.
SA’s personal income tax base is growing slowly, but the middle-income segment is shouldering a rising burden as the top end erodes
— What it means:
The latest FNB Estate Agents Survey shows that emigration-related residential property sales peaked in the second half of 2019 at about 18% of national sales. The number slumped to about 10% during the pandemic and over the past four quarters has hovered around the 8%-9% mark.
FNB’s estimates align roughly with Sars’s emigration figures in numbering just a few thousand individuals a year. But FNB senior economist Siphamandla Mkhwanazi stresses that the data is drawn solely from housing sales, and that not everyone who emigrates sells their house or even owns a house.
More importantly, FNB’s estimates show that SA is experiencing net emigration — those selling to emigrate outnumber incoming property buyers by two to one. This exodus has created an excess supply in the housing market, exerting downward pressure on property values, according to Mkhwanazi.
“Values in higher-priced segments came under enormous pressure at the height of this selling-to-emigrate trend in 2019,” he says, while acknowledging that prices have since recovered from these lows.
Also interesting is that most of these sales took place in higher-priced housing brackets above R3.5m, with about 70% of those selling due to emigration aged between 35 and 44 years — people often at the height of their careers.
Faulkner and Mosadi noted in their 2019 paper that SA’s rising emigration trend risked slowing productivity gains, denting competitiveness and ultimately undermining SA’s growth prospects by worsening the country’s skills constraint. But perhaps the biggest risk, they warned, was to the fiscal outlook, as any reduction in PIT collections would weigh heavily on overall tax revenue.
The Sars data has shown that though this is not yet of grave concern to the taxman, the upward trend in emigration is undeniable — and not something SA can afford to be complacent about.
Tax hikes ahead?
But even if Kieswetter is not concerned, that still doesn’t answer the question posed at the outset — whether SA’s tax base is growing fast enough to ensure the country’s long-term fiscal sustainability without the government having to resort to continuous or draconian tax hikes.
The short answer is that while the tax system is holding up better than many expected, we cannot possibly know how fiscal policy will unfold in the future.
Andrew Donaldson, a former National Treasury deputy director-general and now a senior research associate at the University of Cape Town, explains that, in SA right now, “GDP growth and government spending decisions are more important than anything that can be done on the tax policy side”.
The tax system is doing fine; the problem is growth
— Andrew Donaldson
When it comes to government spending, the big-ticket items include the public sector wage bill and fee-free higher education, neither of which the Treasury has succeeded in whittling down by much.
Political decisions have yet to be made on whether to fully pursue National Health Insurance (NHI) and/or institute a BIG, and these may go against the prudent advice of the Treasury, which is committed to stabilising SA’s debt ratio at about 75% of GDP through continuous fiscal consolidation.
If either NHI or a BIG are pursued to scale, it will require a significant step-up in taxation, most plausibly the hiking of VAT and PIT rates by several percentage points each. These increases would have to be broad-based, affecting not just the wealthy, but also the middle-income earners who are the mainstay of the PIT base.
Without such a spending shock, or any sizeable departure from the Treasury’s medium-term spending plans, Donaldson believes there will be no need for tax increases other than inflation-related adjustments.
“Tax collections are again going to exceed budget estimates this year, thanks largely to the commodity boom, which will ease the Treasury’s debt problem,” he says. “Growth and investment are the central policy imperatives now — this argues for tax relief, not further increases.”
In fact, he believes the major tax increases that SA needs have already occurred. These include the moderate rise in the PIT burden from 2014 to 2019 caused by government not compensating taxpayers in full for fiscal drag, a 1pp increase in PIT rates for nearly all taxpayers in 2015 and the addition in 2017 of a new tax bracket at 45% for taxable income above R1.5m a year. In 2018 the VAT rate was also increased by 1pp.
His chief conclusion is that while the tax base has recovered from the impact of Covid, there is unlikely to be further upside potential without stronger economic growth. In fact, he believes the Treasury should be looking for opportunities to ease the tax burden to stimulate growth.
In short, he says, “the tax system is doing fine; the problem is growth”.















Would you like to comment on this article?
Sign up (it's quick and free) or sign in now.
Please read our Comment Policy before commenting.