The National Treasury has suffered the indignity of having its 2025 budget rejected by the cabinet on the very day of the budget, but it is possible the fallout could lead to a strengthening of the country’s fiscal architecture and more decisive policymaking. It could even accelerate the pace of economic reform.
Consider, for starters, that the cabinet is now being forced to confront the disparity between its expansive policy agenda and the country’s limited means. It has no choice but to make the harsh policy trade-offs that have long been put off.
Before last Wednesday, few cabinet ministers could have told you what the country could buy for a two percentage point (ppt) VAT hike (R60bn). The aborted budget sets out the shopping cart in detail (see table).
It shows, for instance, that had the budget opted for a more politically palatable 1ppt VAT hike (R30bn), the cabinet would still have had to choose between a R350 a month social relief of distress (SRD) grant for 10-million people (about R35bn annually) or covering the public sector wage bill overrun (about R7.5bn on average annually) together with the extension of frontline services in health and education (R25bn annually).
Making hard public policy choices has proved to be impossible for the ANC government — hence the decision to opt for the 2ppt VAT hike. But, in fact, the Treasury went way beyond what was necessary, laying out a “maximalist” budget in which it “caved in to everything”, says Wits University adjunct professor Michael Sachs.
As a former head of the Treasury’s budget office, Sachs puts the crisis down to the fact that “of the two largest parties in the GNU, one (the ANC) cannot contain expenditure and the other (the DA) cannot countenance tax hikes”.
To put it in economic terms: growth is too low and spending too high, given the size of the country’s tax base. Last Wednesday, this simmering contradiction was finally pushed out into the open.
On the positive side, this presents a unique opportunity for the GNU to come together to design more democratic budget processes and develop lasting fiscal solutions. It doesn’t have to fix everything by March 12 when a new budget will be tabled, but it does have to make a credible start.
In the DA’s leaked plan for how to resolve the impasse, it argues the country’s options are close to exhausted.
“We cannot continue driving up our debt-to-GDP ratio,” it says, “since we are already spending 21% [of revenue] servicing debt.” Further borrowing will just push the country’s borrowing costs up even higher, further crowding out funding for critical economic and social priorities.
It argues that the option of cutting expenditure further is also limited since “there is only so much cutting a government can do before the economic and social consequences become extremely painful”. Nor does it believe there is much more room to hike taxes.
“And so, we find ourselves now in an economic Bermuda Triangle,” writes the DA, “unable to borrow, tax or cut our way to success. There is only one way out: economic growth.”
Now that coalition politics has forced tax hikes off the table, politicians are scrambling to work out how to cut R60bn from the budget. But that is starting from the wrong premise. The VAT hike added new money to the budget baseline, money that was going to allow the Treasury to relax fiscal consolidation and pay for a whole lot of new things.
As the DA says in its proposed plan: “We simply cannot fund every spending demand in this coming budget. The R60bn additional spending contemplated is largely a choice, not an obligation. We can and must be prepared to just say no.”
Unfortunately, it’s not as simple as that. Several of the new expenditure items in the botched 2025 budget cannot be wished away. These include: the wage bill overrun (R7.7bn a year on average for three years); the additional year extension of the SRD grant (R35bn); and the extension of early childhood development coverage at about R5bn in year one, since the recently passed Basic Education Laws Amendment (Bela) Act makes grade R compulsory.
(It is, however, possible this Bela provision could be deferred for a year or two. In fact, it would help enormously if the entire Bela Act could be put on ice, since the Financial & Fiscal Commission has put the total cost of implementing it at a fiscally disastrous R17.7bn.)
In total, the withdrawn budget would have added a net R173.3bn in new money to the 2024 budget baseline over the next three years. Most of it (R75.6bn over three years) would have gone towards bolstering frontline services in education and health care, including hiring more teachers and young graduate doctors, rebuilding the commuter rail system and strengthening capacity at home affairs.
In short, it would have reversed some of the fiscal austerity of recent years of which the labour-intensive departments of health and education have borne the brunt.
The Bureau for Economic Research notes in its budget commentary that, had the withdrawn budget been passed, it would have allowed consolidated non-interest spending to grow by 0.9% on average annually in real terms over the medium term, whereas previously the plan was for it to contract in real terms by 0.5% a year on average.
In other words, it would have represented a loosening of the fiscal stance. While this would have been achieved without a meaningful deterioration in the country’s debt metrics, the burden of adjustment would have shifted from the users of public education and health services (the poor) onto all consumers through a higher VAT rate. Higher-income earners would have been additionally burdened by the proposal not to adjust personal income tax (PIT) brackets and medical tax credits for inflation for the second year in a row.
The clamour against hiking VAT stems not just from the disproportionate effect this would have had on the poor (since they spend a larger proportion of their income on essential goods and services subject to VAT), or the short-term impact on the cost of living and inflation, but also from the fact that a higher tax burden would reduce demand and inhibit growth — something that would be counterproductive in a country that is battling to get growth going.
When it comes to PIT, it is clear that this tax base has been maxed out, certainly at the top end. As the Treasury admits in the 2025 Budget Review, increasing PIT rates further is likely to be inefficient, would reduce the incentive to work and save, and probably fail to deliver the anticipated revenue. (See box: Getting blood from a stone.)
Corporate income tax (CIT) is similarly dismissed by the Treasury as it is a smaller and more volatile source of revenue than VAT. It is already high by international standards, so further hikes could impede the country’s competitiveness.
At a more fundamental level, the crisis has highlighted that government spending is structurally too high. It has fluctuated around 32% of GDP for several years. Despite tax rate increases, tax revenue has struggled to broach 28% of GDP. The upshot is that budget deficits have become stuck at around 4% of GDP, causing borrowing and debt service costs to mount relentlessly.
The obvious conclusion is that spending must adjust to the reality of the tax base. If South Africa is to reduce its budget deficit to the more acceptable international benchmark of 3%, then spending cuts of at least 1% of GDP are needed. (See box: What the IMF thinks)
“If the tax increase is to be avoided, then the spotlight will have to fall on the expenditure side of the budget,” says University of Cape Town senior research associate Andrew Donaldson, writing in The Conversation Africa. “This is a far harder discussion than tax policy — there are a thousand options to consider, and there are vested interests wherever you look.”
A wholesale review of spending with a view to closing nonperforming projects and programmes has become urgent and unavoidable. But while there is undoubtedly fat in the system, the notion that a bloated backroom bureaucracy exists that can be easily culled without having an impact on frontline services has been debunked by research by the Public Economy Project at Wits University.
It found in 2022 that within core public services, the balance between professional and administrative staff appeared “stable and sensible”. Any “bloating” was concentrated in economic regulation, infrastructure services and public administration. Even so, total employment in all public administration departments was less than 40,000, against more than 1-million in health care, education and criminal justice.
Quite simply, because two-thirds of government consumption spending is concentrated on basic education, health care and criminal justice, and these core public services are inherently labour-intensive, accounting for more than 85% of government employees, these sectors are going to bear the weight of fiscal consolidation as there is limited scope to reduce employment elsewhere.
This implies, according to the report, that further “large and damaging” cuts will be required to the number of nurses, teachers and police officers on top of the significant downsizing these departments have endured in recent years.
Already government education spending had fallen to about R16,500 per learner by 2021 in real terms from about R20,000 per learner in 2009, the Wits study found.
The basic education department now estimates, based on its medium-term budget allocations, that by 2027/2028, the sector will have been underfunded by between R79bn and R118bn since 2021/2022.
So, with further significant tax hikes and higher borrowing seemingly off the table, what is the government to do about the large spending requirements in health, education and other critical frontline public services?
In the short term, it means further routine tax hikes remain inevitable — so brace for an increase in the fuel levy, above-inflation increases to sin taxes, and probably the continued freezing of PIT brackets on March 12.
It is also likely the new budget will defer some new spending, such as the extension of early childhood development or the continuation of public employment programmes, until the Treasury has figured out a sustainable way to fund them.
The debacle has exposed a very deep, fundamental flaw at the centre of the budget process: ,” says Sachs, “the fact that all the good aspects of the budget — the order, the rationality, the democratic debate — apply only to the expenditure side
— Michael Sachs
Over the medium term, the government would be wise to sell valuable, underutilised land and buildings. The department of public works & infrastructure has compiled a list of R155bn worth of government assets that could be sold. In addition, the state owns 2.6-million hectares of agricultural land. At an average price of R10,000 per hectare, that could bring in another R25bn.
But, most importantly, economic reforms must be urgently accelerated to raise the growth rate. If events of the past week allow that penny to finally drop at the top of government, then the entire budget bungle will have been worth it.
It is still hard to understand why the Treasury went for the big-bazooka approach of hiking VAT, the least politically palatable tax. It is almost as if it wanted to provoke the cabinet to confront the contradiction between the ever-increasing expenditure pressure being heaped onto the budget and the country’s limited resources.
So, though the debacle is a setback for finance minister Enoch Godongwana, it need not be a wasted crisis if the opportunity is seized to reform the way budgets are put together.
Sachs says: “The debacle has exposed a very deep, fundamental flaw at the centre of the budget process: the fact that all the good aspects of the budget — the order, the rationality, the democratic debate — apply only to the expenditure side. They do not apply to taxation, because this is regarded as too important to subject to the democratic process or too technical to involve politicians.”
Historically, the detail of tax policy announcements has always been kept to a very tight circle made up of the finance minister, his director-general, the president and a few top tax officials. For 30 years, the dominance of the ANC allowed this system to work even through the introduction of the new 45% PIT bracket in 2017 and the 1ppt VAT hike in 2018. As long as the finance minister had the president’s support, the rest of the ANC fell in line.
The same thing would have happened last week had that approach not run aground on the rocky shores of the GNU.
The solution now, Sachs feels, is for all coalition party leaders to do something that should have been agreed to at the very start of the GNU: sign a fiscal charter or compact that sets out a list of realistic fiscal goals or parameters for the next five years. It will also have to be aligned with the GNU’s new medium-term development plan because if the plan is not backed by fiscal resources, or aligned to fiscal realities, then it will be worthless.
The bottom line is that when Godongwana stands up on March 12 to take a second stab at tabling the 2025 budget, some routine tax increases are inevitable. But above that, he will have to announce a convincing raft of measures to both slash spending and grow the economy. It is also important that he gets the ball rolling on ways to democratise the budget process. Anything less and South Africa may never recover its fiscal credibility.
Getting blood from a stone
South Africa is a country of vast inequality with huge unmet developmental needs, but squeezing the affluent for more tax is easier said than done and could even be costly and counterproductive.
When this was attempted in 2017 by abruptly raising the marginal personal income tax rate from 41% to 45% for those with incomes of more than R1.5m, revenue collection dropped by R6.48bn instead of generating the extra anticipated R5.46bn.
This happened because taxpayers responded to the tax hike far more vociferously than expected, reducing their declared income by about 10%. This meant the tax revenue collected on personal incomes above the R1.5m threshold was less than it would have been without the reform — a disastrous outcome.
The lesson from that exercise was that tax authorities need to tread carefully when they tax the incomes of the rich, as the efficiency losses can be considerable. Or to put it more simply: as a tool, personal income tax appears to be maxed out at the top end.
Many economists agree that instead of hiking taxes, the focus should be on raising the country’s growth rate, broadening the tax base and increasing compliance measures to bring illicit activity into the tax net.
What the IMF thinks
The International Monetary Fund (IMF) has, for the past decade, been begging South Africa to get a handle on its rising debt levels. It also believes that cutting inefficient spending is essential to restore fiscal sustainability. This should be accompanied by the GNU adopting a prudent fiscal anchor supported by a fiscal rule, while also increasing the pace and ambition of pro-growth reforms, it says in its 2024 Article IV report on South Africa released last month.
It takes a tepid view of South Africa’s growth prospects: it expects real GDP growth to average just 1.6% over the medium term against the National Treasury’s latest average forecast of 1.8%. The IMF’s forecast assumes unemployment will remain high, that fiscal deficits will remain sticky and that gross government debt will rise from 75% now to 80% by 2026 — and keep climbing.
The problem, it explains, is that public debt will automatically continue to rise given the combination of weak growth and the high average interest rate on public debt. While the slightly faster growth anticipated in the next few years should offset some of these unfavourable debt dynamics, it will be insufficient to lower the debt ratio.
This view suggests that far more emphasis should be placed on bolstering growth through faster structural reform and by shifting expenditure towards infrastructure investment in parallel with “more ambitious” fiscal consolidation.
* Bisseker is an economics writer and researcher at the Bureau for Economic Research
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.