James Myburgh on why the French economist's belief that 'more ambitious land reform' is key to addressing inequality in SA is misplaced
There is a broad and widely shared narrative that very little of the land settled or conquered and occupied by the ancestors of modern day white South Africans in the 19th century has been transferred to black Africans under the African National Congress government since 1994. The ‘slow pace of land reform’ has been by the ANC government to justify the scrapping of the ‘willing buyer willing seller model’ and the salami slicing away of existing property rights through various pieces of legislation (passed or pending).
The EFF meanwhile are pressing for all caution to be thrown to the wind in this regard and for the land to be forcibly expropriated. In his recent address to the American Chamber of Commerce the EFF leader Julius Malema stating that “this main issue of ownership, which is mainly the land” needed to be fixed. “You’ll never be confident if you do not have ownership” he told the audience. ”Whites have ownership, black people don’t. Your survival depends on the selling of your labour to white people.”
The French economist Thomas Piketty too has blamed the absence of large-scale forced land redistribution in South Africa, post-1994, for enduring inequality in the country. He told aBloomberg interviewer, during his recent visit to South Africa, that: “You never had this kind of big phase of redistribution of property and probably that explains why the legacy of apartheid is still very much there in terms of inequality.”
In his Nelson Mandela Annual Lecture Piketty elaborated on this theme:
“Let me just say that if we take a broad international historical perspective, we see in many countries, in history, much more ambitious land reforms than what we have seen in South Africa since the end of apartheid; I think it’s fair to say that black economic empowerment strategies, which were mostly based on voluntary market transactions, [and or not] market values were not that successful in spreading the wealth and limiting the extreme concentration of wealth from which we start in South Africa. So I think we need to think again about more ambitious land reform.”
Piketty and Malema’s remarks beg the question of how much land is actually in black hands, and whether the government’s redistribution efforts have been as limited as they suggest.
The lie of the land
On the face of it, the assumption that most land across the country remains in the hands of white farmers appears to be a matter of simple arithmetic. On the one hand the 1913 Land Act, and successor legislation, restricted black African land ownership to ‘Native reserves’ (later ‘homelands’) covering some 13% of the total area of South Africa. On the other, the ANC has fallen far short of its 1994 target of redistributing through restitution or land redistribution 30% of commercial agricultural land.
In 1993 an estimated 14% of the area of South Africa fell in the soon to be abolished homeland areas. This was mostly state trust land. A further 12% was state-owned land, in the former ‘white South Africa’, which fell under the control of the democratic government in April 1994. At the beginning of the democratic order then some 26% of the area of South Africa was thus state or former homeland land.
Two years later, in 1996, a government survey found that there were 82m hectares (ha) of commercial agricultural land in South Africa (67,4% of the total area), and this is the figure against which the ANC has subsequently measured progress in meeting its redistribution and restitution targets.
By January 2015 4,5m ha had been purchased by government as part of its land redistribution programme, and a further 3,2m ha for land restitution. If one adds in the additional 2m ha where claimants accepted financial compensation in lieu of the return of land (as of March 2013) the total area redistributed or restituted since 1994 rises to some 9,6m Ha. This falls well short (39%) of the ANC’s 1994 restitution and redistribution target (of 24,6m ha).
9,6m ha is not an insignificant amount of land however. Historically, it is an area almost equal to (94%) the amount of formally ‘Native-owned’ land recorded around the time of the 1913 Land Act, and 60% of the area of all the homelands in 1982. In European terms this area is three times the size of Belgium, twice the size of Switzerland or the Netherlands, larger than Ireland, Austria and the Czech Republic; about the same size as Portugal; and almost three quarters the size of England, a third the size of Italy, and a quarter the size of Germany.
In overall South African terms, however, it amounts to only 7,9% of the total extent of the country. Why then has the purchase by the state of such a lot of land, achieved relatively little in terms of meeting overall percentage targets?
Generally speaking when people think of land reform the goal that comes to mind is the transfer of well-tended, productive farmland from white to black hands, not tracts of wilderness or semi-desert. However, the land in South Africa is generally of very poor quality. Rainfall is often poor or unreliable, the soils shallow and slopes steep. Only 12,6% of the country’s land is classed as potentially suitable for any kind of cultivation, and another 11% appropriate for intensive grazing. The amount of productive land in South Africa is thus extremely limited.
Except for the winter rainfall areas in the far west of the Western Cape most of such land lies in the Eastern half of the country. Mpumalanga is the area with the most and best arable land, in absolute terms, followed by KwaZulu-Natal and then the Free State and Limpopo. By contrast most of the western half of the country is semi-desert. This is an area which was not settled, historically, by black African agropastoralists, with the original inhabitants being Khoi Khoi pastoralists and San hunter gatherers, followed by trekboers. It also happens to be where most of the commercial farmland was located in 1996. In provincial terms 60% of commercial farmland lay in the Western, Eastern and Northern Cape; a further 13,8% in the Free State; 0,9% in Gauteng; and the remaining 24,6% in the North West, Limpopo, Mpumalanga, and KwaZulu Natal combined.
Map of South African land capability (red and orange low capability vs blue and green higher capability) and graph of extent of commercial farmland in 1996:
For understandable reasons the government’s land redistribution and particularly its restitution efforts have been disproportionately focused on the higher rainfall areas in the Eastern side of the country. These were also the areas of original black African settlement; enduring black tenant farming on white-owned farmland post-1913; and of apartheid-era forced removals after 1948. Just over half of the extent of land redistributed and restituted since 1994 has been located in the North West, Limpopo, Mpumalanga and KwaZulu-Natal.
The land in ‘black hands’
If one looks at land ownership and redistribution figures by province one gets a far better sense of the land in black hands, without the distortion caused by the presence of huge tracts of semi-desert in the west of the country. Earlier this year Freedom Front Plus leader Pieter Mulder asked Minister of Rural Development and Land Reform Gugile Nkwinti “how much of the land surface of each of the nine separate provinces covers land that previously resorted under homeland areas?” The figures provided by Nkwinti – which are estimates based on mapping of old homeland boundaries against today’s provincial borders –are as follows:
Table 1: Area of each province made up of former homeland land
Former homelands as % of total
5 757 277
16 931 000
13 001 100
1 818 200
4 223 491
9 445 100
3 399 298
12 580 600
7 654 400
2 079 612
10 523 800
1 689 794
37 827 700
13 152 200
18 434 122
122 934 100
These figures are probably slightly overstated (by a single percentage point overall), however they do provide a reasonable indication of the areas of each province made up of former homeland territory. As can be seen former homeland areas make up 44,7% of today’s KwaZulu-Natal, 34% of the Eastern Cape, 27% of the Limpopo and 19,8% of the North West and 12.5% of Mpumalanga. If one includes the 1996 figures for commercial farmland, and then uses that to estimate the remaining mostly state and urban land, the distribution of the land in each province was approximately as follows in the mid-1990s:
Table 2: Former homeland areas, commercial farmland (1996) and other land as a proportion of each province
Former homeland areas
Commercial farmland (1996)
Other land (mostly state-owned or urban)
As can be seen from the table commercial farmland (commonly defined as the ‘land in white hands’) made up less than half of the extent of Gauteng, KZN and Limpopo; and about 60% of the Eastern Cape, Mpumalanga and the North West. In Gauteng much of the land not in former homeland areas or classed as commercial farmland (in 1996) would be urban land, including land owned by municipalities. In the other provinces it would have been more weighted far more towards nationally and provincially owned state land (game reserves, forests, military bases etc.)
In order to get a rough sense of the land now in ‘black hands’, by province, one would need to combine former homeland areas, land redistributed and restituted post-1994, and state land. (This is not a perfect measure however as it may include some state land that was restituted post-1994.)
Table 3: Area redistributed and restituted (R&R) in each province
Area R&R / commercial farmland in 1996
Area R&R / total area of province
Former homeland + R&R
Former homeland + R&R + state and urban land
As can be seen from the table above the ANC has redistributed or restituted an area equivalent to 35,3% of the commercial farmland in KZN in 1996, 31,9% in Mpumalanga, 23,9% in Limpopo and 14,5% in the North West. This means that close to 60% of the total extent of KZN today is composed of former homeland areas and land redistributed or restituted since 1994. The equivalent figure for the Eastern Cape is 39,2%, 37,4% for Limpopo, 31,4% for Mpumalanga and 28,3% for the North West.
If one includes state-owned land the ‘land in black hands’ rises to over 40% in the Eastern Cape, close to 50% of the North West, some 60% of Mpumalanga, about two-thirds of Limpopo, and around 70% of KwaZulu-Natal. These figures do not include the land purchased on the open market by individual black South Africans since 1991 or land owned by companies in which black South Africans are today shareholders.
It should be also borne in mind that the figures for commercial farmland are now almost twenty years old. In that period the number of commercial farmers in South Africa has halved from over 60 000 to around 30 000. Even established medium-sized farms have often struggled to make a successful go of it. Rapid urbanisation across the country would have reduced the extent of commercial farmland as well, and the expansion of coal mining activities in a province like Mpumalanga has cut into South Africa’s high capability commercial farmland.
As the discussion above has attempted to illustrate the amount of ‘land in black hands’ – outside of the arid western half of the country and with the exception of the Free State – is far greater than many people seem to assume. In Malema’s home province of Limpopo and Jacob Zuma’s province of KwaZulu-Natal it is safe to say that well over half the area is already in ‘the hands of the people’, to use ANC terminology.
The area of land that has been restituted or redistributed over the past twenty years (some 9,6m hectares) is also substantial in absolute terms. In terms of some of those countries commonly cited as successful examples of land reform, post-1945, it is an area just under the size of the whole of South Korea, two-and-half times the size of Taiwan, and three-and-a-half times the extent of land redistributed in Japan between 1947 and 1949. Close to R80bn rand (in 2013 Rand values) has now been spent on these programmes.
It is not quite true either that this was all done on a ‘willing seller, willing buyer’ basis. There have been increasingly questionable land claims made and granted, often against farms owned and occupied by the same (white) family from well before the 1913 Land Act was passed. The owners concerned have often chosen to take the pay-out rather than fight the matter in court. The high levels of farm attacks and farm murders have not exactly acted as an encouragement to white farm owners, or their children, to stay on the land if they don’t have to.
Although it would have served little real purpose, far more land could certainly have been distributed had government gone about buying huge tracts of cheap and arid land in the old Cape Province. Or, if government had chosen not to prioritise (over land purchases) other more important matters such as the R70bn arms deal post-1998, or the increasing of public servant salaries post-2006. Or, indeed, if government had administrated the process efficiently and without corruption.
Piketty’s remarks on the need for more aggressive land reform are misguided for three basic reasons. Firstly, he seems to underestimate quite how extensive land restitution and redistribution has already been in the areas where it actually matters. This is unless he is suggesting that one of the answers to inequality in South Africa is to forcibly expropriate and then subdivide the massive areas of commercial farmland located in the Karoo and Namaqualand for the purposes of small holder agriculture.
Secondly, the substantial experience of land redistribution thus far strongly suggests that it is not an answer - at least when administered by ANC cadres and while other govt policies render much smaller scale agriculture uneconomic - to ‘inequality’. In 2010 Minister Gugile Nkwinti acknowledged that of the 5.9m hectares' worth of farmland - which had been active and generating revenue – redistributed by government “90% of those are not functional, they are not productive.” Government actually pulled back from purchasing more land precisely because of this, in order to ‘recapitalise’ these failed farms.
Thirdly, cases of successful land reform - such as South Korea, Taiwan and Japan – usually involved the purchase and transfer of title to the land from absent semi-feudal or foreign landlords to the tenant farmers who actually lived on and worked the land (often having done so for generations). South Africa no longer has a similar tenant farming or peasant class, able to step into the gap. This is illustrated by a recent survey which found that while 18% of respondents regarded the land question as an important issue for government to address, 0% said it was an important issue to them personally.
The demand for ‘land reform’ in South Africa is driven by primarily by ideological objections to an ‘alien’ racial minority owning, and successfully farming, black ‘African’ land. The African nationalist model of ‘land reform’ – which Malema is agitating for and the ANC pursuing in crab-like fashion – involves forcible racial land redistribution whereby highly skilled and productive commercial farmers are dispossessed of the land they own, live-on and work, and their workers displaced, inevitably in favour of the politically powerful actors (including chiefs) and/or their clients. As the recent Zimbabwe example vividly illustrates the result is the evisceration of property rights, the destruction of capital intensive agriculture and the collapse of food production - with hunger and starvation only averted through external food aid and remittances from abroad.
Only a Western economist of Piketty’s extraordinary brilliance would think that promoting this model is advisable in the febrile political climate that currently prevails in South Africa. As George Orwell once remarked: “One has to belong to the intelligentsia to believe things like that: no ordinary man could be such a fool.”