The IOL article (published 5 September 2026 by Nicola Mawson) frames the long-term cost of corruption in South Africa by converting estimated losses into concrete public goods that could otherwise have been delivered: roughly 8.2 million government-subsidised homes or 1.7 million ordinary classrooms from a modelled cumulative figure of R2.1 trillion, and far smaller but still striking equivalents from the firmer Zondo Commission estimate of about R57 billion.
It stresses that these are illustrative counterfactuals, not precise audits of money stolen, and that the true damage extends well beyond the rand amounts to weakened institutions, slower growth, lost investment, and reduced service delivery for citizens.
Core estimates and how they are derived
A figure long cited in government circles (though its empirical basis has been questioned) is that as much as 20% of public procurement spending may be lost to corruption, overpayment, and related failures. With current annual government procurement around R1 trillion, that assumption implies roughly R200 billion lost per year. Extrapolating procurement trends backward to 1994 and applying the same 20% rate produces the modelled cumulative potential loss of about R2.1 trillion over the democratic period.
The article is explicit that this is not a claim of exactly how much was actually lost: consistent historical procurement data are incomplete, earlier figures are modelled, and the 20% rate itself is contested. It is presented as a scale illustration of opportunity cost.
A more concrete benchmark comes from the Zondo Commission: around R57 billion (sometimes stated as R57.3 billion) in state expenditure was tainted by state capture, with more than 97% linked to Transnet and Eskom, and the Gupta enterprise receiving at least R15 billion. President Cyril Ramaphosa has noted that more than R17 billion linked to the Commission’s work had been recovered by mid-2026, while emphasising that the true economic cost was far larger (lost investment, higher borrowing costs, collapsed institutions, forgone growth and jobs). Other estimates of the broader economic impact of state capture range up to R500 billion (Pravin Gordhan’s testimony) or as high as R1.5 trillion when including indirect effects such as weaker growth and lost tax revenue.
What the numbers translate into
Using government and project benchmarks:
- R2.1 trillion ≈ 8.2 million Breaking New Ground (BNG) houses (at the published subsidy of R255,364 each); ≈ 1.7 million ordinary classrooms (using an indicative ~R1.2 million average cost from a Northern Cape education infrastructure programme); ≈ 43,750 clinics (at an indicative R48 million per recent clinic project); ≈ 191 years of the National School Nutrition Programme (at the 2026/27 allocation of ~R11 billion feeding >9.9 million learners); or, illustratively via economic multipliers from the Institute for Economic Justice, activity supporting ~14.5 million jobs.
- R57.3 billion ≈ 224,500 BNG houses or ~47,000 classrooms (notably exceeding the Department of Basic Education’s identified national need for about 43,700 additional classrooms to ease overcrowding); ≈ 1,200 clinics; or more than five years of the full national school feeding programme. The same sum, applied illustratively to multipliers, equates to economic activity supporting almost 400,000 jobs.
These are alternatives, not additive. The underlying unit costs come from programmes at different times, so the same money could not have simultaneously produced all those outcomes. The point is the forgone tangible public goods.
Broader economic and institutional costs
The article quotes University of Pretoria economist Dr Carolyn Chisadza on how state capture weakened the state’s capacity to deliver on development goals (growth, employment, poverty and inequality reduction, investment attraction, and institutional quality). Hennie van Vuuren has cautioned that focusing only on rand values obscures the deeper costs borne by people through weakened governance. Sean Gossel of the UCT Graduate School of Business links corruption directly to wasted service-delivery capital, crowding out of the private sector, and distortion by dysfunctional state-owned enterprises; as looting intensifies, service provision declines.
On growth, a 2019 observation by South African Reserve Bank economist David Fowkes is cited: given available labour and infrastructure investment, growth in a normally functioning economy should have exceeded 4%, yet the country recorded little growth—an outcome he and later the National Anti-Corruption Advisory Council associated with the impact of state capture (equivalent to roughly 4% of GDP growth forgone). The Constitutional Court has described how corruption and organised crime stunt sustainable development and economic growth and put societal stability and security at risk.
Context and caveats
South Africa’s democracy has delivered millions of homes, electricity connections, expanded water and education access, social support, and infrastructure. The article’s question is not whether delivery occurred, but how much more might have been achieved if less money had been lost to corruption and procurement abuse. Money “was not just money. It was houses, classrooms, clinics, food and jobs.”
The piece sits within a wider pattern of reporting on ongoing procurement problems, school-level corruption cases, hospital looting scandals, and infrastructure delivery shortfalls. It does not claim that every rand of the modelled R2.1 trillion was stolen, nor that state capture alone accounts for all of South Africa’s development shortfalls; it uses the figures to make the abstract cost of corruption concrete and to underscore the opportunity cost for ordinary citizens who depend on public services.
In short, the headline numbers dramatise a long-running argument: even conservative or contested estimates of leakage from public procurement, when viewed over three decades, represent transformative volumes of housing, classrooms, clinics, nutrition, and economic activity that were never realised. The firmer Zondo figure alone would have covered the identified national classroom backlog several times over. The deeper cost, the article stresses, lies in eroded state capacity and slower growth that compound the direct losses.
