SA ownership debate lacks data, new paper finds

A claim that white South Africans own 80% of the economy still shapes debate over capital flows, yet no comprehensive dataset exists from which any credible figure for ownership of the economy as a whole can be calculated. That is the central finding of a new paper by Daryl Swanepoel, CEO of the Inclusive Society Institute, titled ‘Who Owns South Africa? Testing the Claim that 80% of the Economy Remains in White Hands.’

His conclusion is not that racial inequality is a myth. White households remain substantially wealthier on average than African households, and wealth is extraordinarily concentrated. The trouble lies in the certainty. Average wealth and aggregate ownership answer different questions, and treating them as interchangeable obscures both the progress the economy may have made and the inequalities still requiring redress.

For investors, operators and policymakers, the measurement gap carries practical weight. Empowerment measures capture particular transactions and company structures, but they do not collectively reveal how the country’s stock of economic wealth is ultimately distributed. Without that broader picture, it is impossible to judge whether policy is delivering its intended outcomes, or where intervention would be most productive.

The Johannesburg Stock Exchange is often deployed as shorthand for economic ownership, and even there the record is thin. Research published in 2015, examining ownership in 2013, estimated black South African ownership of the top 100 companies at 23%, white South African ownership at 22%, and foreign ownership at 39%, with a further 16% unanalysed. Black ownership in that framework grouped African, coloured and Indian South Africans under the empowerment rules. The figures are more than a decade old and cannot establish today’s pattern, yet they offer little support for an 80% claim. Nor can the exchange, by itself, stand in for the entire economy.

The deeper difficulty lies in tracing economic interest through intermediaries. A shareholder register may name a pension fund, custodian or nominee company without identifying who holds the underlying claim. The Public Investment Corporation illustrates the point. In March 2025 it managed roughly R3.05-trillion, of which the Government Employees Pension Fund accounted for 88%, and historical data for March 2017 showed about 85% of that fund’s contributing members were African, coloured or Asian. Pension entitlements differ with salaries and years of service, so the same percentage of asset value cannot simply be attributed to black ownership. But the underlying economic claims of those members do not disappear because an institution invests on their behalf.

Household wealth evidence complicates the narrative further. The Momentum-Unisa household wealth index estimated that African households held 57.3% of household wealth in 2023, against 32.2% for white households. Yet African households made up 82.1% of all households and white households 8.6%, meaning average wealth per white household was about 5.4 times that of an African household. Both findings can be true: a far larger population group can collectively hold more wealth while its average household remains considerably poorer. Caution is still required, since surveys struggle to capture the wealthiest households and complex holdings through private companies, trusts and offshore investments, and household wealth is not the same as ownership of productive businesses.

What Swanepoel proposes is institutional. The Broad-Based BEE Commission would lead a funded feasibility study into a national economic ownership and wealth index, working with Stats SA, the Reserve Bank, the South African Revenue Service, financial regulators and other institutions. The study would establish what can credibly be measured, whether records can be lawfully linked, how investments can be traced through intermediaries, and how privacy would be protected, with foreign and state ownership treated separately and assets not counted twice as they pass through structures. Costs matter too: an index imposing disproportionate reporting burdens or producing unreliable estimates would serve little purpose.

Neither business nor government should fear that test. If ownership has become more representative, acknowledging the progress could strengthen confidence without declaring transformation complete. If it remains concentrated, the evidence will sharpen the case for further action and show where it belongs. After three decades, arguing over an unsubstantiated percentage is an inadequate basis for decisions affecting investment, opportunity and social cohesion. The open question is whether anyone will fund the measurement that settles it.