It’s been in the works for years, and the subject of numerous warnings, but it’s here. South Africa has reached its long foretold fiscal crisis. That is, the revenue expected will not cover enough of the expenditure projected. The finance minister sought to deal with this by raising VAT, but couldn’t secure the buy-in of all the parties in the governing coalition. For the first time in post-apartheid history, the sitting government couldn’t pass a budget.
This is likely to be messy. The budget is the fiscal expression of political goals, and this is particularly the case in a society like South Africa, where the imperatives of politics do not so much coexist with those of governance as overwhelm them.
South African governance has been bent into a set of large extensive patronage systems. The term patronage is used advisedly, and in a non-pejorative sense. The ANC and the government it has headed (at least until a coalition was forced into it) needed to deliver material benefits to those making up its support base. This is not a uniquely South African matter, and some of this outlay can be justified by genuine need, other parts of it rather less so. Among the former, South Africa has set up an enormous social welfare system. The keystone of this is the social grant, of which around 28 million are distributed monthly. Some 19 million people receive grants made available to cushion vulnerable categories of people – children, the elderly, people with disabilities – from the worst effects of deprivation. Another 9 million receive the Social Relief of Distress grant, a legacy of emergency measures enacted during the Covid pandemic, and which seems to be morphing into a general social support intervention.
More controversial is the large, well-remunerated public service, whose utility is of questionable value, at least in view of the very considerable sums that are poured into it (around 12% of GDP), while the global average is below 10%. This is a constituency whose largely middle-class lives are underwritten by the fiscus, and a key part of the patronage machine.
According to National Treasury and several reports from the International Monetary Fund (IMF) and World Bank, public sector compensation has hovered at around 12–13% of GDP, which is above the global average (often closer to 9%). This includes the salaries of teachers, police officers and nurses – necessary by any definition – although given the outcomes, there is something wrong with how this is being applied. On top of this, there is a vast class of office- and counter-bound civil servants doing work that makes the machinery of state function, such as it does. Recently, we’ve seen the consequences of this in high-profile action, when our failing diplomatic service – a particularly prestigious and high-cost part of our state system – proved itself ineffective in dealing with a major downturn in relations with the US.
Then there is the circle of those who draw resources from the state. These can be split into two groups. The first benefit from empowerment preferences. In other words, these are businesspeople who are leveraging premiums for providing goods and services to the state, over and above what a strict value-for-money calculation would allow. The total procurement spend is estimated at around R1.242 trillion in the last financial year.
A second group feeds off the state through corruption. This has become something of a national obsession, rhetorically, though with anaemic action to address it. And as the Zondo Commission found, it is often bound up with the fortunes of the ANC. No one really knows what the cost is – direct, indirect and opportunity together – although that it is a pathology gnawing at the innards of South African society is now common cause. As former President Kgalema Motlanthe once memorably said: ‘This rot is across the board. It’s not confined to any level or any area of the country. Almost every project is conceived because it offers opportunities for certain people to make money. A great deal of the ANC’s problems are [sic] occasioned by this.’
Corruption takes all manner of forms, often through crooked tendering, itself often justified through empowerment criteria.
This is the basic political economy of the budget. The fiscus needs to deliver to three key, legitimate constituencies, with a fourth not officially recognised, but a powerful one in shadows as a result of the operation of our politics.
Hiking taxes is very much a matter of keeping each of these groups satisfied. The problem of course is that the economy has just not grown to keep up with the calls on the revenue. As my colleague Gabriel Crouse showed in a recent paper, the South African state’s revenue to GDP ratio rose from 19.8% in 1996 to 26.8% in 2022 – reflecting a dreadful combination of increasing extraction and economic stagnation. On top of that, there is clear evidence that the growth in this tax boot print has not improved governance outcomes (far from it, as South Africa has fallen precipitously in indices that measure them), and that it is having a negative multiplier effect on the economy. This is borne out in two studies, one published by the South African Reserve Bank.
As Crouse writes: “When the fiscal multiplier is below one, that means spending R1 produces a GDP impact of below R1. The fiscal multiplier is not usually thought of as being able to enter the negative range. However, the South African state is breaking new ground, conceptually and practically. With a negative fiscal multiplier estimate, for every rand the state spends on the margin it literally destroys output. Setting the money on fire would have been less destructive than entrusting it to state spending.”
From this perspective, the proposed VAT hikes might paper over some immediate issues, but they stand to compound a larger problem.
Looking at it from that angle, a key strategic consideration must be where the trade-offs can be made, which will mean deciding which of the recipient groups should be prioritised. Here, it’s a no-brainer. The grant system is a lifeline for millions of people. I fully agree that its rollout failed to take into account the long-term funding, but there is a strong moral case for providing support to the less affluent, and now that it’s here, it’s politically impossible to remove.
That said, it’s been noted often enough that there is something perverse in taking pride in the expanding number of welfare recipients, rather than recognising the deeply concerning realities that it denotes. As the National Development Plan (NDP) put it, “not enough people work.” Grants have become the de facto substitute for employment – and this must change.
This means that the budget needs to support growth. And in this respect, ramping up the tax burden is counterproductive. And this raises the matter of the second interest group, the public service.
This is, I think, often discussed in a rather binary and undifferentiated manner. The public service wage bill has become a nightmare, though I’d suggest that this is largely a function of the lousy output of the state, rather than the number of public employees or the size of their pay packets. Besides, many of those people, teachers and nurses for example, will be needed irrespective of expense. If the economic multipliers of state spending and the overall effectiveness of governance can be improved, it could justify itself.
This is not to say that there are no cuts to be made. For example, after witnessing the debacle around the relationship with the US, it’s time to reassess the return on investment that South Africa has made in its foreign service. The NDP expressed concern about this; embassies are expensive to maintain, and too often they have been used as rewards for party hacks. It’s past time for a serious discussion about South Africa’s diplomatic presence and whether anything of the current proportions is merited.
Much the same could be said of the military. If the armed forces are not going to be properly capacitated for conflict (virtually the entire air force is grounded, while South Africa recently suffered its biggest capitulation since Tobruk), then perhaps South Africa should look at abolishing them, and absorbing some of the specialist skills into reconstituted civil institutions: border security, a SWAT-focused gendarmerie, search and rescue. (Yes, this is a more complex matter than simply rands and cents, but all things considered, the option should be on the table. Alternatively, fund it properly and make it combat-worthy.)
And there are numerous other underperforming bodies which might be dispensed with. The Sector Education and Training Authorities come to mind. They’ve had close to three decades to bring about the “skills revolution”, and for the most part, they’ve failed. Each of them should be required to justify its continued existence.
Still, as South Africa has hemmed itself in with a rigid labour regime, this isn’t going to be a short-term project.
Where real savings and (dare one say it) a transformation is possible is with the third and fourth constituencies, and the procurement systems that they are plugged into. These are businesspeople – legitimate and otherwise – using the state procurement spend to extend their profit margins. Where this is enabled by law and policy, and where they are compliant with it, one hastens to add that they are doing nothing wrong. On the contrary, using these provisions is merely rational business sense.
That does not change the fact that considerable costs are built into the preferences that the state pays, and that these inevitably take up space in the increasingly slender fiscal envelope.
One remarkable feature of the procurement regime (now codified in last year’s Public Procurement Act) is that the actual costs of these preferences have never been systematically quantified. This represents not only failure of planning and fiscal control, but of transparency. Informed public debate on this matter – and, necessarily, to calculate trade-offs that all policy demands, is only possible with accurate information.
The formal position is that state pays BEE “preference premiums” that are capped at 11% for contracts above R50 million and 25% for contracts below R50 million.
Crouse’s research on the matter estimates that the direct cost of premiums is around R17bn. This is itself a considerable sum (and is regarded by Treasury officials as “conservative”). Call this the dispersal largely to the third group, businesspeople drawing legitimate premiums.
However, this is paired with the benefits drawn by the fourth group, in other words, by those benefiting from waste, corruption or unproductive expenditure that has been justified through empowerment preferences. This is estimated at about R132bn.
Crouse’s estimate can be compared with a reference in a recent IMF procurement report to Treasury’s earlier finding that “savings from improving procurement practices could be sizable, up to 20% of the cost of goods and services procured” if “strategic sourcing” improved.
Crouse’s narrower estimate is that doing away with empowerment premiums stands to save R150 billion by combining the savings on licit and illicit expenses. Remember that the proposed 2% VAT hike that stalled the budget would have brought in an estimated R58 billion. Making these adjustments would save well over double that amount – close to three times, actually – and if only half the savings could be made, the R58 billion could be raised, with money to spare.
The Institute is not alone in pointing to the possibilities. The report of the Zondo Commission makes for interesting reading in highlighting the link between formal empowerment policy and the corruption that hides beneath it. It went on to point out that in its view, whenever tensions arose between value for money and the empowerment premiums, the former should be given preference: “There are of course many cases, one hopes the vast majority, in which the award of the tender satisfies both objectives of the Constitution, but undoubtedly there are other cases, some of which may well be high-value tenders in which one or other of these two objectives must be preferred, and it is in such cases that the legislation fails to give guidance. In the view of the Commission, the failure to identify the primary intention of the Constitution is unhelpful and it has negative repercussions when this delicate and complex choice has to be made, by default, by the procuring official. Ultimately in the view of the Commission the primary national interest is best served when the government derives the maximum value-for-money in the procurement process and procurement officials should be so advised.”
This really is the crux of the matter. South Africa needs to use its limited basket of resources prudently, which in practice demands making painful decisions. With the economy and the tax base unable to supply the patronage as before, it’s past time to prioritise how what exists will be distributed.
By any moral or pragmatic reasoning, the first call needs to be in favour of the country’s poorest people. A question mark has always hung over the sustainability of South Africa’s social welfare system, and a breaking point is approaching. Yet it has provided a livelihood for millions of people and must be protected.
Rationalising procurement – a loss to certain business communities – is the most obvious means to find the money needed for the foreseeable future. The state is no longer in a position to continue to offer premiums to a select group of the more affluent at an effective cost to society as a whole and to the least affluent (those most dependent on state support) in particular.
That all being said, we at the Institute have long maintained that the central problem confronting South Africa has been its inability to grow. Without it, every problem South Africa is currently experiencing will only get worse. It is only with growth that the employment to lower state dependency is possible, so that the state itself can switch from consumption-driven spending to investment, and that the fiscal dead-end of rising debt can be arrested.
In the South African context, one of the key constraints is the state itself, or rather, as Ricardo Hausmann’s Growth Lab at Harvard University has said, the lack of state capacity. This is yet another reason to focus on introducing a more effective and value-based procurement system. Even more, it requires that that other state constituency – the public service – up its game drastically. Value for money must apply not only to procurement, but to the services provided by the state itself through those whom it employs.
This means, in turn, changing the way its work is done and managed. What the National Planning Commission termed a “rejection of meritocracy” must itself be rejected. Those unwilling to do so must be dispensed with.
All of this means that the current budget crisis is likely only going to be resolved with some bruising conflict among powerful interests. South Africa is at the point it is because of the politics that has undergirded governance choices. These are no longer viable.
- Terence would like to thank his colleagues who provided feedback on this article, particularly Gabriel Crouse, without whose input it would not have been possible.