In February, finance minister Enoch Godongwana pulled the budget at the last minute when it became clear it lacked the support of the rest of the GNU. In March, he tabled a slightly adjusted version, even though most GNU parties were still not on board.
Now, as everyone scrambles to understand what happens next, a few salient truths have emerged.
The most important is that the GNU has not been ready to deal with the challenging process of passing a budget. Its bloated cabinet may have agreed on a broad medium-term development plan, but it has failed to back it with an agreement on basic fiscal policy issues, let alone a whole budget. This should have been nailed down months ago.
Given that the budget is the mechanism whereby 30% of GDP is allocated, can a government that cannot pass the national budget even call itself a government?
The second hard truth is the realisation that the ANC remains in deep denial. If left to his own devices, Godongwana would have imposed a two percentage point (pp) VAT hike on society to fund a shopping list of new priorities without introducing important expenditure cuts. This would have been a substantial shock to overall consumption spending. It would also have stoked inflation, and hurt business income and profitability.
The chief opportunity is to force the ANC to realise it cannot continue to hike taxes or borrowing each year to compensate for a failure to make hard political trade-offs over things that endlessly inflate the size of the state.
In essence, the GNU has called a halt to the ANC’s practice of kicking the can down the road; of hoping that South Africa will finally grow its way out of its fiscal problems. After more than a decade of subpar growth, that strategy has run out of road.
If this means that the cabinet finally faces the country’s fiscal predicament head-on, and realises that there are limits to tax increases, resulting in expenditure coming under better scrutiny in future, it could mark a turning point for the better.
The main risk is that the GNU will fail to reach agreement on a foundational set of fiscal policy principles that will allow parliament to not only pass the 2025 fiscal framework within the 16 days provided by law, but to pass future budgets without a repetition of the chaos of the past few weeks.
Most commentators canvassed by the FM are cautiously optimistic about how things will play out, given reports that the ANC and DA are close to reaching a deal.
But things could still unravel. For though parliament has the power to amend the budget, it doesn’t necessarily have the technical capacity to do so. That is a big risk because if parliamentarians try to amend it on their own without a proper technical framing, they could make things worse.
Wits University adjunct professor Michael Sachs, a former head of the National Treasury’s budget office, thinks the country will probably “muddle through”, but he cautions against taking a sanguine view.
“We are in uncharted territory,” he tells the FM. “Who knows what’s going to happen next? The important question is how the institutional framework adjusts and what changes emerge in this new setting, because how people behave now is going to set a precedent for the medium term ... The moment is open; anything could happen.”
Failing again on the second attempt
Budget drama is fairly common in coalition governments, but it’s still not often that a country gets a chance to redo a national budget in the space of three weeks — and then fluffs it again the second time. But that is what South Africa did between February 19 and March 12.
As widely expected, the March budget softened the 2pp VAT hike announced last month but did not bow to pressure to scrap it entirely. Instead, a 1pp VAT hike will be spread over the next two fiscal years.
Most of the other proposed tax hikes remained unchanged, but those involving personal income tax (PIT) were tightened by freezing all income tax brackets, even those of low-income earners which had been protected in the withdrawn February budget.
Last year, when PIT brackets and rebates weren’t adjusted for inflation, more than R16bn was raised. This meant that another 200,000 low-income earners were sucked into the bottom of the tax net, and everyone’s effective tax rate was raised.
This has been a long-standing trend, explains Andrew Donaldson, a senior research associate at the University of Cape Town. He estimates that over the past decade, the PIT threshold has declined from R115,000 (in today’s prices) to R95,750, bringing about 850,000 more people into the tax net.
The Treasury estimates that the March budget proposals would raise an additional R13.5bn from the 0.5pp VAT hike in 2025/2026 and a further R19.5bn from not making inflationary adjustments to PIT brackets, rebates and medical tax credits.
Even so, the March budget left the Treasury with a revenue shortfall of R52bn over the medium term relative to the aborted February budget. However, instead of cutting expenditure by the same amount, the March budget envisaged cuts worth only R31bn — and R11bn of this would be taken out of the contingency reserve.
In short, instead of taking a scythe to spending, the Treasury opted to hike taxes and run down the country’s fiscal buffers while retaining nearly all the new expenditure announced in February.
In total, the March budget will fund R232.5bn in new non-interest spending over the medium term compared with the R252.3bn proposed in February (or in net terms after offsetting items, R142bn in March vs R173bn in February).
This includes R46.6bn for infrastructure, R35bn for the one-year extension of the social relief of distress grant, and R23.4bn to cover the wage bill overrun. But the bulk of the new money — R70.7bn (previously R75.5bn) — is to bolster front-line services. This represents a conscious political decision to accommodate new and persistent spending pressures in several key areas, including teacher costs, early childhood development, and health.
The upshot is that despite the outcry over the initial proposed 2pp VAT hike, the government barely tapped the spending brakes in its second run at the 2025 budget: consolidated non-interest spending will now grow by 0.8% on average annually in real terms over the medium term, against 0.9% in the withdrawn February budget.
This is a big shift from the original plan in the 2024 February budget, which was for spending to contract in real terms by 0.5% a year on average, to stabilise the debt ratio at 75% of GDP.
This implies that the Treasury’s original plan to stabilise debt by sacrificing front-line services has turned out not to be politically feasible within the ANC. However, the new fiscal path which protects front-line services (and then some) and still achieves debt stabilisation, but only with significant tax increases, appears unacceptable to the rest of the GNU.
Rise Mzansi leader Songezo Zibi, who chairs the standing committee on public accounts, has decried the March budget as “unworkable and unsustainable”, adding that “while many have and will continue to raise the issue of VAT, our problems are far bigger than VAT, and very few want to confront them”.
Political strategist and longtime DA adviser, Ryan Coetzee, took to Politicsweb to explain the DA’s position last week.
“What is going on here isn’t a fight over a 1% VAT increase,” he wrote. “It is a battle over the future of South Africa’s economy, and the life chances of the South African people. Because if we don’t turn the economy around, a VAT increase is going to be the least of our problems.”
South Africa’s underlying fiscal problem, he says, is that growth has been too slow for the past 16 years. The country can’t borrow much more, with the debt ratio already topping 75% of GDP, and has also run out of room to raise taxes much further.
That leaves only two options: grow faster or reprioritise spending away from waste and inessentials to critical items. However, he acknowledges that even cutting expenditure will only get us so far.
“So that leaves growth,” he concludes. “But the problem with growth is that the ANC refuses to do what it takes. So many areas of public policy [remain] untouched: ports, passenger rail, tariffs, excessive regulation, localisation, race-based procurement laws, expropriation laws, labour laws. The list goes on and on.
“So confronted with a proposed VAT increase in the absence of a growth agenda that can get South Africa and the fiscus out of the hole it is in, the DA correctly said: ‘No’.”
The DA argues that neither the GNU nor South Africa will succeed unless it urgently and significantly increases growth and decreases unemployment. To achieve this, it is demanding three things in exchange for supporting a 0.5pp VAT hike: that the GNU agree to reduce the debt ratio; that it identifies a minimum of R100bn in expenditure for reprioritisation in the 2026/2027 budget; and that it accelerates pro-growth reforms.
Specifically, the DA wants the private concessioning of the Cape Town and Richards Bay ports to be brought forward, and for the City of Cape Town (and other metro authorities) to be allowed to take over running their passenger rail systems by the end of the year.
Among other things, it also wants the Expropriation Act to be amended to replace references to “nil” compensation with “just and equitable” compensation.
BER chief economist Lisette IJssel de Schepper also notes that the difficulties in passing a budget, together with unexpected tax hikes, are likely to set back business confidence, which has only recently begun to improve. This will not be good for investment.
Just look at what happened after the 2024 budget, in which nearly all the additional revenue was raised by not adjusting PIT brackets and rebates for inflation.
PIT revenue increased by more than 13% in 2024/2025 year on year, but VAT and corporate income tax collection deeply disappointed. For Donaldson, this suggests that higher taxes on personal income are partially offset by reduced consumption and declines in revenue from other sources.
In the 2026/2027 fiscal year, the DA wants the R100bn to be redirected as follows: 30% to go towards reducing South Africa’s debt burden, 20% towards reducing the PIT burden on bottom income bands, and the remaining 50% to go towards infrastructure spending.
Not sufficiently pro-growth
While it is easy to quibble over the DA’s list of demands, the real question is whether the March budget was good for growth. Unfortunately, the answer is no, according to the Bureau for Economic Research (BER).
“Some additional spending on infrastructure is eaten up by tax increases which will directly [affect] growth,” the BER said in a recent note on the budget. “Indeed, the March 2025 budget is arguably worse from a growth perspective than the February budget. It raises personal income tax which will be a drag on consumer spending and increases VAT which will both reduce consumer spending and be inflationary.”
This may be one of the reasons the Treasury shifted the burden away from PIT to VAT in the bungled February budget. However, while VAT taxes consumption rather than savings or investment and is, therefore, the least damaging of the main tax options to economic growth, a VAT hike still effectively raises the cost of living and reduces households’ spending power.
However, while the March budget reduces the severity of the VAT hit, it increases the severity of the PIT hike. (The February budget would have fully adjusted the bottom two tax brackets for inflation and provided partial relief to everyone else. In the March budget there is no inflation relief for any tax bracket.)
The net impact of the March budget’s proposed tax adjustments will be to take R28bn out of the hands of consumers in the coming year, and almost R120bn over three years — that is a huge, planned erosion of household income.
And even though the Treasury says increases in public sector wages and employment will counter this somewhat, the upshot will be to reduce demand and negatively affect business income at a time when the economy is stuttering, having grown by just 0.6% last year.
Sachs fears that the combination of raising VAT and freezing PIT brackets will hit the lower middle class especially hard. These are typically people who have jobs but who are struggling to make ends meet.
“As voters, they cut across all political parties and races, some are public servants, many are union members,” he says. “They get taxed but get very low benefits in return. If, as government, you start becoming a bit arrogant and raise VAT, what if it leads to a Kenyan moment, to protests? Then, anything could happen.”
Given these worrying implications, it seems strange that the Treasury didn’t do more to interrogate the list of new spending demands in March.
For instance, there is still R10bn set aside over the next three years for the universalisation of grade R. This would extend early childhood development coverage to 700,000 more children, but it means registering and subsidising thousands of informal crèches — a process that could easily take five years. So why not extend the budget allocation over five years and cut the current allocation accordingly?
In this way, by digging between the sofa cushions, the Treasury could easily have cut R30bn off February’s R108bn new spending list for 2025/2026. Instead, in March the list was minutely shaved down to R102bn, with home affairs bearing the brunt with a R5bn cut — the one programme that has been good for growth.
But perhaps the most extraordinary aspect of the March budget is the extra R4bn for the South African Revenue Service on top of the R3.5bn additional allocation in February. Time will tell whether this will pay off in greatly enhanced revenue collection or not.
What really matters, though, is that over the next three years, the March budget proposes taxing the economy more to accommodate over R230bn in new spending — but ignores the hundreds of spending reviews done by the Treasury in recent years and its recommendation for millions of rand worth of savings.
So, what happens now?
First, in terms of the Money Bills & Related Matters Act, parliament’s joint finance committee has 16 days (or as soon as reasonably possible thereafter) to approve/reject the broad fiscal framework which outlines government’s revenue proposals, expenditure aggregates and borrowing plans.
The Treasury has said it cannot, however, scrap the 0.5pp VAT hike, which will go ahead on May 1 regardless of what parliament decides because it is governed separately by the Value-Added Tax Act. However, should the joint finance committee not approve the revenue increase from VAT, this could precipitate a legal crisis.
If changes are proposed to the March budget by the joint finance committee, the finance minister gets an opportunity to respond, and the committee must ultimately issue a report, reflecting his views and containing a clear statement accepting or amending the fiscal framework.
However, the concern is that the committee won’t have the analytical know-how needed to challenge the minister of finance successfully. Does that mean we could end up with a technically incoherent budget, something worse than either the February or March iterations?
Donaldson’s expectation is that the budget will be accepted within 16 days as tabled in March, probably without any changes, in exchange for a strengthening of off-budget agreements to undertake spending reviews and perhaps reform the budget process.
His conviction is based on his reading of the DA’s demands as effectively insisting that broader growth and spending issues must be examined and dealt with over the period ahead — something that should not be too difficult for the president to accede to.
“It is worth bearing in mind that there is an adjustments budget process in six months’ time,” he adds, “and so an easy compromise to reach might be that [the March budget is accepted as is in exchange for there being] further work on possible spending and policy changes to be considered at the time of the adjustments.”
Indeed, this seems to also be the view of other political insiders. If so, then the risk of the committees tinkering ham-fistedly with the budget may be less than feared. In short, what happens next boils down to one simple thing: if there is political agreement the budget will be passed by parliament.
A good outcome would be one in which the whole debacle leads to more democratic deliberation and accountability over expenditure and taxation choices.
However, Zibi notes that “while it is somewhat good that the public and more political parties can engage meaningfully with the budgeting process, it happens at an unfortunate time when consensus and speed are key — and we have neither now”.
“Too much bickering and uncertainty will slow down private investment,” he warns. “It is not good that this year’s budget tabling and approval through parliament is such a mess.”
A bad outcome would be one in which a weak, ineffective executive is repeatedly held hostage to coalition attempts to govern via parliament, using the budget as a tool to gain leverage and in the process, rendering the budget process volatile and dysfunctional.
An ideal outcome would be the GNU agreeing on a comprehensive fiscal statement before the medium-term budget in October, otherwise the same uncertainty will continue to dog future budgets. But this may be too much to ask.
Says Sachs: “I imagine we won’t have ideal, explicit agreements between parties, but we will have enough agreement — they will meet each other halfway — and so the tax proposals for 2025/2026 will probably go ahead, but the medium-term ones are in doubt.”
The bottom line is that South Africa’s fiscal credibility now lies in the hands of a small team of backroom political operators. The political trade-offs they make will decide the shape of taxes, expenditure, growth and economic reform for the foreseeable future, not evidence-based policy crafted by technocrats.
Though the GNU partners are right to have rejected the two deeply flawed 2025 budgets and to push for faster growth, this is a risky development. Let’s hope it doesn’t backfire.
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.