During every budget cycle stakeholders argue that the government spends too much on operations or too little on infrastructure. But this binary view obscures a deeper structural problem that no amount of budget reallocation can solve.
Since 1994 the government has built thousands of schools, clinics, hospitals and police stations. Infrastructure spending in 2018-24 exceeded R1.3-trillion. Yet visit any public clinic and you’ll find two-hour waiting times posted like operating hours. Police stations stand understaffed while crime climbs and so on. We’ve built the hardware but cannot afford to run the operating system.
Government budgets oscillate between capital expenditure and operational spending as political priorities shift. When infrastructure needs dominate, capex swells and operations suffer. When service delivery crises erupt, operational budgets expand and infrastructure decays.
The 2024 budget allocated roughly 58% to payments while infrastructure transfers declined from 4.1% of GDP in 2018 to 2.8% by 2024. We’re trapped between building things we cannot maintain and maintaining things that we cannot afford to operate.
The revenue blindspot
Budget debates rarely address this: the government treats itself primarily as a redistributor of tax revenue, not a generator of it. This wasn’t always the case. Transnet once contributed billions annually to the fiscus. Eskom was profitable. Telkom’s privatisation unlocked R5.6bn in 1997. The Industrial Development Corporation returned dividends to the National Revenue Fund.
Today, the equation is reversed. Eskom required R254bn in bailouts in 2018-24. Transnet’s losses exceeded R5bn in 2023. SAA’s business rescue cost more than R16bn. State-owned enterprises shifted from assets to liabilities, bleeding the Treasury. While attempting to fix these entities remains critical, the government has surrendered the principle that the state can be entrepreneurial in generating revenue.
The South African Reserve Bank transferred R4.6bn to the Treasury in 2023. Yet conversations about optimising state capital management within existing legal structures remain nascent. This isn’t about compromising the Bank’s constitutional mandate or operational independence. Rather, it is recognising that within its current framework the Bank manages significant capital: gold and foreign exchange reserves exceeding R1-trillion, investment portfolios and operational assets.
International precedent suggests central banks can optimise returns through sophisticated risk management and diversified investment strategies while maintaining their primary mandates. The question isn’t whether the Bank should change what it does, but whether the technical sophistication applied to these holdings — asset allocation, risk-adjusted returns, currency management — fully reflects contemporary global practice.
As the government addresses state-owned enterprise dysfunction, a parallel conversation about how state capital held by constitutionally independent institutions might be optimised within their existing mandates seems overdue.
The procurement markup problem
Not all government spending creates equal value. When the government contracts a firm to build a bridge or construct housing, that procurement employs engineers, creates jobs and advances transformation. The value chain is productive.
But consider a different transaction: a department needs stationery. A tender goes to a connected intermediary who purchases from wholesalers, adds a 40% markup and invoices the state. No jobs are created. No skills transferred. The economic activity is pure rent extraction.
This isn’t hypothetical. Municipal procurement scandals routinely expose this pattern. In 2023 the auditor-general flagged irregular expenditure of R88.5bn, much involving supply chain failures. A significant portion involves markup-based procurement adding no value.
Recent procurement reforms provide a pathway. The National Treasury’s digital platform enables direct competitive bidding. If departments access wholesale markets directly, the markup layer disappears. Suppliers still win tenders. Jobs at wholesalers remain. But the intermediary margin — conservatively 25%-40% — returns to the fiscus.
Conservative estimates suggest 15%-20% of goods procurement involves pure markup. On a national procurement spend of R900bn annually, goods represent roughly 30%, or R270bn. If 15% involves pure markup averaging 30%, eliminating it could save about R12bn annually.
This isn’t total state internalisation. Engineering and specialised services should remain tendered. The distinction: where procurement creates jobs and competition, retain it. Where it merely adds margin without economic benefit, streamline it.
The immediate temptation is hiring more teachers, police and nurses. But one-off savings create permanent liabilities. Hire a nurse today and you’ve committed to salary and inflation adjustments for 30 years.
A workable model exists in plain sight. The Government Employees Pension Fund manages more than R2.1-trillion through the Public Investment Corporation, generating returns that the fund pensions for 1.3-million active members and 450,000 pensioners. The governance framework — parliamentary oversight, Treasury control, professional management — already functions.
An analogous structure for operational shortfalls could work similarly. Procurement savings and one-off revenues capitalise a service delivery sustainability fund with strict governance: Treasury oversight, Reserve Bank custody, competitive professional management. The capital stays intact. Only dividends and interest fund designated operational gaps — additional police, healthcare staff, teachers — with direct budgetary tracking.
International precedents exist. Chile’s Economic & Social Stabilisation Fund, capitalised from copper revenues, smooths countercyclical spending while preserving capital. Botswana’s Pula Fund, seeded by diamond revenues, has sustained government operations for decades through professional market management.
The framework is conceptual: operational sustainability requires revenue generation, not just budget reallocation. Not all procurement creates equal value. One-off savings shouldn’t fund permanent liabilities. Professional capital management, properly governed, can generate returns that expand fiscal space without expanding taxes.
The question isn’t how much to spend on operations versus infrastructure. It’s how the government positions itself to sustainably fund both. That requires treating the state not only as a spender but as a capital allocator generating returns that serve public purpose.
We’ve spent 30 years arguing about how to slice the pie. Perhaps it’s time to discuss how to bake a bigger one.
• Mafinyani is risk advisory & financial modelling partner at DiSeFu, a specialised financial technology and risk advisory firm operating in the Sub-Saharan region.
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.