Op-eds

Land ownership in South Africa — the facts and figures, and figuring out the facts

uMkhonto Wesizwe party supporters outside the Cape Town Central Police station on February 10, 2025, in Cape Town, South Africa. The MK Party laid charges of treason against AfriForum for allegedly misleading US President Donald Trump on the issue of the Land Expropriation Act. (Photo: Gallo Images / ER Lombard)

By Cherryl Walker

10 Mar 2025

The disparity between black and white ownership of land in South Africa is not as acute as many reports suggest, and knowing this matters. What the aggregated numbers do not support are claims that land ownership has not changed since 1994. Things have changed.

The outcry unleashed by US President Donald Trump’s executive order falsely accusing the South African government of “seiz[ing] ethnic minority Afrikaners’ agricultural property without compensation” has rekindled the always smouldering debate on land reform and its discontents.

This article does not address these discontents directly. Rather, it looks at the “facts and figures” about land that pepper this debate – their use and, too often, abuse.

In what follows I review what is known about the distribution of land ownership in South Africa since 1994. But first something on the misuse of numbers.

Misuses and misreadings

A good reason for getting to grips with the land numbers is to be able to recognise when they are abused. Often this is for narrowly political ends. One example is an assertion made in August 2024 by Zwelakhe Mthethwa, an MK party member of Parliament, that “black people still, in 2024, own only 13% of land”.

This is simply not true, as a cursory engagement with numerous government reports shows. Yet Mthethwa was able to state this during a parliamentary briefing without being corrected – “13%” was simply a rhetorical flourish to his claim that “it has been over 113 years since the establishment of the 1913 Land Act and there has been no change”.

This is a blatant misuse of a familiar number (and date). However, many less-partisan reports on the distribution of land in South Africa also misrepresent the numbers. For instance, in February 2025 Reuters reported, erroneously, that “only 4% of privately held land is owned by black people, who are nearly 80% of South Africa’s population of 60 million”.

Four percent is a shockingly small figure that reinforces what most people probably regard as given – that land ownership in South Africa is deeply inequitable, that it is heavily skewed towards people who are white, and that this is a problem.

However, the disparity between black and white ownership of land in South Africa is not as acute as many news reports suggest, and knowing this matters. The actual extent of the gap – the focus of this piece – should inform any serious discussion on land expropriation and land reform.

The mistake Reuters made was to collapse several categories in a 2017 Department of Rural Development and Land Reform (DRDLR) land audit. Its figure of 4% did not apply to all privately held land in South Africa but to a subset of that land – to rural land registered in the Deeds Office, in 2015, to individual owners.

Excluded from the audit calculation were both individually owned urban land (small in terms of total area but not total value) and the significant proportion of private land in South Africa registered to trusts, companies and community-based organisations.

Bringing these categories into the equation shifts the proportion of private land in black hands upwards quite significantly, although by precisely how much is difficult to compute.

land Women protest against the government’s latest land reform strategy in Cape Town on 15 October 2020. (Photo: EPA-EFE / Nic Bothma)

Statistical literacy

Taking issue with these misrepresentations is neither pedantic nitpicking nor a strike against land reform. Knowing the extent and form of inequality matters if one wants to do something about it.

Understanding the full complexity of actual land “on the ground” is not a simple undertaking. It requires an understanding of history, ecology, economics and politics across different scales. In addition, it requires a critical approach to the aggregated data on land ownership, demography, livelihoods, etc – my concern here.

Trawling through the numbers can be tedious. Navigating frequently discordant figures requires basic statistical literacy – “the ability to interpret, critically evaluate and communicate about statistical information and messages”.

The numbers are never a perfect reflection of the reality they purport to represent (official documents do not even agree on the precise extent of South Africa) and they need to be scrutinised to ensure not only that the calculations are sound, but also that their application is fit for purpose.

A different kind of challenge is working with “race” as a self-evident category. While an area can be measured precisely once its coordinates are given, “race” is a very different unit of analysis. It is not a material substrate but a social construct – an ascribed identity, its parameters fuzzy and its meanings unstable over place and time.

In the case of the 2017 audit, populating its categories of “White”, “African”, “Coloured” and “Indian” involved not only labouring with ID numbers and census data, but also filling the gaps by using “names and surnames… to try to determine the race – conscious of the limitations that this carries” (p4). The limitations of “race” as a category need to be kept in mind in analysing the data.

Baseline data: 1994

In determining how much has changed or stayed the same over the past 30 years, one needs credible baseline data.

Given the density of land-related figures since 1910, it is possible to quibble endlessly about the precise area of land in white hands in the dying days of apartheid. However, assuming that all the land outside the scattered territories reserved by successive white governments for black people can reasonably be deemed “white”, then in 1993 about 85% of South African land was “white” and 15% “black”. This latter figure includes the frequently overlooked “coloured reserves” located mainly in Namaqualand.

This does not mean that black South Africans owned 15% of the country – reserve land was state owned. Nor does it mean that 85% of South Africa was in the exclusive ownership of white people. Some of it was state owned, for instance national parks, while here and there small numbers of black people had been able to cling onto their land through the brutal years of apartheid.

The largest share of “white” South Africa belonged to farmers, but not as much as is often assumed. In 1993, agricultural land outside the reserves covered about 82.8 million hectares – 68% of the country. A little over 500,000ha was owned by the state, while individual farmers owned about 62.4 million hectares – three-quarters of this land and 51% of the country. The balance was owned by partnerships and companies which one can also count as “white”.

It is worth noting that not all this land was prime farmland and not all farms were strictly commercial in terms of turnover and profitability. In 1993, the annual income of 6,269 “commercial” farms (11% of the total) was below R2,000.

The distribution of land in 1994 is summarised in Table 1 below:

The 2017 land audit and beyond

The Land Audit Report is not an easy read, but its aggregated numbers (summarised in Table 2) reveal clear shifts in land ownership patterns since 1994.

State-owned land

The audit found that in 2015 about 94 million hectares were in private ownership, leaving 28 million hectares (23% of South Africa) in state ownership.

The land set aside as black reserves under apartheid, now recast as traditional or communal areas, would account for nearly two-thirds of this total. The balance would have been registered to a variety of state agencies (national, provincial and municipal) and parastatals, including SANParks, the SANDF and Transnet.

Determining which state lands could be repurposed, whether for land reform or some other public use, should be a government priority. It is notable that one of the categories of land where nil compensation could be considered “just and equitable” in terms of the new Expropriation Act [section 12(3)(b)] is land held by an “organ of state” that “it is not using for its core functions”.

uMkhonto Wesizwe party supporters outside the Cape Town Central Police Station on 10 February 2025. The party laid charges of treason against AfriForum for allegedly misleading US President Donald Trump on the Expropriation Act. (Photo: Gallo Images / ER Lombard)

Privately owned rural land

Here one needs to know that the term “farm” in the 2017 audit does not refer exclusively to actual farms, but includes land owned by mining companies or used for other non-agricultural purposes, for instance solar farms. In 2017, AgriSA lamented that agricultural land had decreased by 3% in area since 1994 because of the expansion of urban areas and other non-agricultural use.

With regard to the racial profile of rural landowners, as already noted the audit found that people it classified as African (not black) owned just 4% of individually owned “farms”.

However, by 2015 this land accounted for 39% of all rural land registered as privately owned. The remaining 61% – almost 57 million hectares (47% of South Africa) was registered to trusts, companies, community-based organisations (CBOs) and “other”. Individual white owners’ share of all private rural land now amounted to 28%, substantially below the 1994 figures.

Extending the term “black” to include people classified as “coloured” and “Indian” means the percentage of individually owned rural land in black hands in 2015 rises to 24% – still deeply inequitable in relation to whites’ share, but less starkly so and, importantly, introducing different histories of dispossession into the analysis.

Also important, bringing group-owned land into the calculation pushes the proportion of private rural land in black hands still higher, although estimating by how much is difficult.

The 2017 audit did not attempt to break down non-individual ownership in terms of “race” because “these entities cannot be racially classified” (p4). But we know that the number of black beneficiaries and shareholders of land-owning legal entities has grown significantly since 1994 while the proportion of land owned by trusts and companies will have increased as well.

President Cyril Ramaphosa’s Phala Phala estate is a prominent example of private land registered to a trust that can be considered both black and African. The 3.5 million hectares of private land owned by community-based organisations in the 2017 audit includes land transferred to communal property associations (CPAs), legal entities established to hold land transferred to black communities through land reform.

By 2024, 1,735 CPAs owned a total of 2.6 million hectares, about twice the area registered to individual African landowners in 2015.

Urban land

The DRDLR audit found that urban land covered about 3.5 million hectares of South Africa (3%), the great bulk of it owned by companies, trusts, etc.

Given its tiny share of the total area, its impact on the overall distribution of land between black and white South Africans is minor. However, it is worth noting that in 2015 black people’s share of individually owned urban properties in terms of area (not value) was comparable to that of whites, with a white/black split of 49%/51%. At 30%, the share of individual property owners classified as “African” was notably higher than that for rural land.

This points to the significance of the changing geography of human settlement for land reform policy. The percentage of people living in urban areas has grown from about 54% of the total population in 1994 to fully two-thirds – 66% – today.

The country’s eight metropolitan regions cover just 2% of South Africa’s total area, but account for 40% of its population and much of its wealth. This is where the demand for land for housing (for shelter, comfort and improved life chances) is particularly acute.

Land reform

Folding the data on land reform into this overview is tricky. This is not only because the official figures have to be treated with some caution, but also because some land reform land was previously state owned and some is urban. The private land market has also been active since 2015, including among black people. Johann Kirsten and Wandile Sihlobo estimate that by the end of 2024 about 2.4 million hectares had been privately bought by black people.

Furthermore, the restitution programme provides for financial compensation in lieu of land restoration, and most claims have been settled this way. By 2024 the state had paid out a little more than R26.2-billion in financial compensation.

Kirsten and Sihlobo have also argued that taking account of the dispossessed land for which claimants have received financial compensation would boost the land reform figures by an additional 2.68 million hectares. They also argue that combining this figure with the full range of state and private transactions, the equivalent of 24.9% of formerly white agricultural land has moved from white to black ownership since 1994.

The official numbers from the state are more modest. In Parliament in August 2024 senior officials reported figures of 5.3 million hectares for redistribution, 3.9 million hectares for restitution and 30,530 hectares for labour tenant claims. The combined total of 9.2 million hectares amounts to 11% of “commercial” farmland in 1994.

Grounding the land debate

These numbers can be spun in many different directions: 72% of individually owned rural land in white hands can be touted as evidence of progress towards the deracialisation of land ownership – or confirmation of the need to speed up land reform or rethink it entirely.

The numbers on agriculture can be used to warn against its corporatisation – or decry the continued dominance of white farmers or argue that they are not representative of most white South Africans, whose landed wealth is in urban property.

The urban data confirms that African South Africans remain the most disadvantaged group in society – but also points to a growing class divide between those with property and those without.

These are all serious debating points. What the aggregated numbers do not support are claims that land ownership has not changed since 1994. Things have changed.

This is a factual statement, supported by an accumulation of data – as slippery as the category of race and imperfect as the numbers undoubtedly are. DM

Cherryl Walker is an Emeritus Professor of Sociology at Stellenbosch University. She is the author of several books, including Landmarked: Land Claims and Land Restitution in South Africa (Jacana Media and Ohio University Press). She is the co-editor with Olaf Zenker and Zsa-Zsa Boggenpoel of the open access publication, Beyond Expropriation without Compensation: Law, Land Reform and Redistributive Justice in South Africa (Cambridge University Press).

Categories:

Op-eds

Tags:

DM is much better if you are signed in...