Ramaphosa and the EFF's dodgy land stats

The Ratcatcher examines the racial land-ownership data on which the EWC push is being justified

The numbers in the 2017 State Land Audit report on “Private Land Ownership by Race, Gender and Nationality” released in February this year have been used extensively to motivate for Expropriation Without Compensation of white owned land. In the EFF motion to amend Section 25 of the constitution the Audit is cited as the source for the claim “that black people own less than 2% of rural land, and less than 7% of urban land”.

President Ramaphosa also used the numbers contained in the report recently in the National Assembly, saying that “While more than 3 million hectares of land was restored between 1995 and 2014, the Land Audit Report indicates that white people in our country still own around 72% of the farms owned by individuals; Coloured people in our country own 15%; and Indians 5% and Africans -- who constitute the majority of the people who live in this beautiful land -- only own 4%."

Given the centrality of these figures to the demand for either eradicating (EFF) or eroding (ANC) property rights it is important to examine them closely. It is worth noting at the start that there are numerous numerical errors in the text and tables of the report, and so the reliability of the data presented below is not beyond question.


The methodology used to conduct this land audit was to go and extract information on all property registered with the Deeds Office. This was then combined with cadastral information held by the Surveyor General to determine the extent of the properties. The nationality of origin and gender of individual property owners was acquired from the Department of Home Affairs’ population register and their race from the highly confidential census data held by Statistics South Africa.

The report states that 114m ha of 122m ha of land in South Africa is registered with the Deeds Office. The remaining 7,7m ha is unregistered trust state land in the Eastern Cape and Limpopo. The registered land falls into only three categories: “erven” (urban land - 3,2m ha), agricultural holdings (a nominal 340 000 hectares) and “farms” (110,7m ha.) The study classified land held by companies (presumably including that held by state-owned enterprises), trusts, individuals, community-based organisations as “private” and land owned by “national government, municipalities, provincial government, public entities, public schools” as state-owned. This included land held in the name of Ingonyama Trust in Kwazulu-Natal. Private land came to 94m ha (77%), leaving the remainder as all state-owned land (including the unregistered state trust land) at 28m ha (23%).

Urban land

The report states that 722 667 ha - 22,6% of the 3,2m ha total - of erven are individually owned by 6 million people, of whom 3,32m were black African (56%), 1,55m white (26%), 507 829 Coloured, 414 069 Indian (6,9%) and 173 418 (2,9%) “other.” If one looks at the land extent the situation appears to be far less balanced however.

357 507 ha are apparently owned by whites (49,5%), 219 033 ha (30,3%) by black Africans, 54 522 ha (7,5%) by Coloureds, 55 909 ha (7,7%) by Indians, 14 332 ha (2%), 14 332 ha under co-ownership (2%) and 21 365 (3%) by others. The EFF’s claim that black people own “less than 7% of urban land” is based on this figure. They have divided the 219 033 ha individually held by black Africans over the total area covered by erven (3,2m ha), which comes to 6,85%.

Incidentally, using precisely the same method of calculation one could say that “only” 11% of urban land was owned by white people. The land audit report does not provide figures for erven other than those individually owned, or falling under sectional title (another 50 000 ha). It does not provide figures for “privately-owned” land in total, or the extent of erven owned by the state and parastatal corporations. The latter must be a massive proportion of the total however.

Moreover, if one looks at white individual erven ownership by province it is clear that there is some anomaly with the Northern Cape figures. 84 041 white individuals in the Northern Cape, who make up a mere 5,4% of all white erven owners, apparently own 152 624 ha of erven. This is 42,7% of the total area of erven owned by whites, and 21,1% of the total area of individually-owned erven in the entire country. According to the land audit white individuals in the urban areas of the Northern Cape own an average of 1,8 ha each (4,9 acres). This seems implausible, and given the small population of the province, it is largely irrelevant to understanding urban land ownership patterns across the country.

Excluding the Northern Cape from these figures changes the picture significantly. Outside of this province 38.1% of individually owned erven are owned by white people, 40,2% by black Africans, 7,8% by Coloureds and 7,9% by Indians. Even here about half of the remaining extent of erven, individually owned by whites, is in the Western Cape. In the seven more eastern provinces whites own 26,5% of the area of individually owned erven, black Africans 53,3%, Coloureds 6,8% and Indians 7,8%. Accepting the figures for the other provinces are correct, according to the land audit black Africans own more individually owned “urban land” than whites in seven of nine provinces, and a majority of such land in four provinces. This is before including the massive share of urban land owned by municipalities, and other state or parastatal bodies. See table below.

Table 1: Individually-owned erven (“urban” areas) by race and province in South Africa

“Rural” land

According to the figures provided by the 2017 land audit land categorised as “erven” at the Deeds Office makes up 2,6% of the extent of South Africa, “agricultural holdings” 0,3% and “farms” 90,8%. It is important to note here that the reference to “farm” is a classification of a piece of land not a description of its use, and such land could be used for a variety of different purposes other than agriculture. For example it could form a section of a national park, game reserve, water reservoir, communal area, or cover a mining or forestry operation. The land audit report is mistaken to refer to it on occasion as “farmland”. The 6,3% remainder is the unregistered state trust land referred to earlier.

In terms of ownership 30,4% of the total extent of the country is – according to the figures in the land audit - in the hands of individuals, 24% in the hands of trusts, 22,9% in state hands, 19% in the hands of companies, 2,9% in the hands of Community-Based Organisations, and 0,7% under co-ownership. As with erven, the report provides only a detailed provincial breakdown for “farms” and agricultural holdings owned by individual owners (by race and gender) only.

It is here that one finds the basis for the claim – employed by the EFF, Ramaphosa and others – that 72% of individually owned land is in the hands of whites. See table below.

Table 2: Individual land ownership of farms and agricultural holdings by race in South Africa



% of total


26 663 144



1 314 873



5 371 383



2 031 790



1 271 562



425 537



37 078 289


The focus on this particular statistic is misleading for three reasons. Firstly, such individually-owned land makes up less than a third of the extent of land in South Africa, according to the land audit itself. This means that individually-owned land, held by whites, makes up only 21.9% of the extent of South Africa. This is less than the extent held by the state. There is substantial variation by province with under 10% of the total extent of Limpopo and KZN individually owned by white “farm” owners, 12.6% in Mpumalanga, 15,1% in Gauteng, 17,8% in the Eastern Cape. 

Secondly, as the provincial breakdown of these figures make clear, this pays no regard to the agricultural potential or value of the land. A substantial majority of this individually-owned white-owned land is located in arid or semi-arid areas in the western part of the country. 43,1% is located in the Northern Cape alone, 11,3% in the Eastern Cape (most of it in the drier western parts of the province), 10,14% of it in the Western Cape, and 14,1% in the Free State. As the land capability map below illustrates there is thus a huge overlay between such land and the arid or semi-arid areas in the western part of the country, which are not suitable for cultivation (in the absence of irrigation).

In terms of the carrying capacity for livestock of natural pasturage Ernest Pringle has pointed out previously “In the high rainfall eastern areas of the country, the average carrying capacity is 1:4, whereas in the arid western areas the average is 1:16. This means that one hectare of land in the former region can produce the same as 4 hectares in the latter, and the value of the land should therefore be 4 times higher.”

Map of land capability in South Africa (the blues and green areas are potentially arable land)

Thirdly, the ANC and EFF have seized on a metric that effectively ‘disappears’ both the land that has been transferred to black hands by the ANC government since 1994, and that was in black hands pre-1994. Ramaphosa’s suggestion that despite 3,1m hectares being “restored” between 1994 and 2014 black African people (through land restitution) “only own 4%” of individually owned land is somewhat disingenuous.

First-off the more appropriate figure for “restoration” is in fact 8,1 million hectares (6,6% of the extent of SA), as a further 5 million hectares of agricultural land has been acquired and transferred by government since 1994 through its land redistribution programme. (In an area covering another two million hectares or so financial compensation was accepted by claimants in lieu of land restitution).

Little of this land would be individually owned today as most land claims and redistribution projects had multiple beneficiaries, and furthermore since 2009 the government has held back from granting title to the beneficiaries of the land redistribution programme. In addition, government has made no effort to ensure those living on their ancestral land in former homeland areas acquire individual title to their land. It is unclear how such restituted and redistributed land was categorised by the land audit – other than as not individually owned - but it quite clearly qualifies as ‘black owned’ land, and it would have taken little to quantify it accordingly in the report.

If one measures the individually white owned land by province against former homeland areas and land restituted and redistributed post-1994 then one gets a far more balanced (though still highly incomplete) picture of land ownership patterns, notably in the wetter eastern regions of the country (the Free State being the exception here). See table below.

Table 3: Individually white-owned land vs. communally black-owned land by province 


Extent of province

Former homeland areas

% of total extent

Land owned white individuals


% of total extent

Land restituted or redistributed post -1994

% of extent

As % of ind. white owned land 2017


16 891 700

5 757 277


3 007 709


650 123




12 982 600

238 582


3 748 192


444 956




1 817 800

91 447


275 021


67 257




9 332 800

4 223 491


853 152


1 333 087




12 575 600

3 399 298


1 139 454


806 256




7 649 500

954 621


967 634


924 209




10 488 100

2 079 612


2 408 880


853 551




37 288 800

1 689 794


11 498 449


1 998 674




12 946 300



2 764 652


543 292




121 973 200

18 434 122


26 663 144


7 621 406




It is striking then that in a debate on such import for the future of the country the proponents of expropriation without compensation have employed such partial and misleading statistics. Indeed, there seems to have been a deliberate focus by the ANC and EFF on a particular metric that would exaggerate the discrepancy between white-owned and black owned-land. Most South Africans would have been unaware that Ramaphosa was talking about a minority sub-set of South African land, mostly located in the most desolate regions of the country.

To sum up then the land audit may have given the state access to highly sensitive information about the race and nationality of individual property owners via census and DHA data – essential information for any planned RET-style programme of racial dispossession – but the report itself provides little meaningful basis for discussion as to overall patterns of rural land ownership in the country, let alone of agricultural land in particular.

Politicians who continue to use these statistics are unlikely then to be awarded the Nobel Prize for Intellectual Honesty any time soon.