Farmers’ interest group AgriSA last week released its own land audit. This filled in a major blank in the land reform and policy field: How many black emerging farmers have bought farms outside government’s land reform programme?
We now have a part answer: they bought 4.3 million hectares. For this major research effort, which has long been needed, we should all be grateful.
The audit is based on research into the Deeds Registries records of land transfers for 1994 to 2017. The methodology is carefully laid out making sensible conservative assumptions over names to derive racial categories of purchasers. The problems of post-1994 boundary changes are acknowledged and dealt with; different kinds of land transactions are explained and decisions on inclusion or exclusion of data are laid out. In total 221610 transactions were individually analysed and incorporated into the database. The data is then set in a context of data from other sources, with introductory maps.
Now the bad news
They have a good story to tell and good intentions: “To ensure effective and sustainable policy formulation, land policy should be based on a factual foundation. This land audit provides such a foundation ….The aim of this report is to provide a more detailed indication of the racial make-up of land ownership in SA” and “this report aims to clarify many of the facts pertaining to landownership”.
But they tell the story badly. By unhappy coincidence, the audit is now seen through the prism of the Black Monday protests.
We don’t know how many black purchasers bought the 4.3 million hectares because the report does not say; likewise it does not give a figure for white farm purchasers, though both figures must be in their data to derive percentages and are very relevant to the debate. In fact, reading the report you would not know there are white farmers in SA as they receive no mention at all.
Percentages rather than numbers
The audit predominantly uses percentages to draw conclusions not actual numbers. More seriously, the executive summary differs from the audit itself. It compares apples (land value) with oranges (land potential) and even pears (ownership share) – farmers should know better.
“On a national level: in terms of value, the share is 29.1% and in terms of land potential, the share is 46.5%. These statistics compare favourably against an ownership share of only 14.9% in 1994 (based on land size).”
The term “government and PDI” (previously disadvantaged individuals) fudges together two categories of landholders, which should be clearly differentiated. Emerging black farmers are not the same as government. This smacks of spindoctoring.
Unfortunately, a critical piece of information is omitted: Were the PDI purchases from government – state land – or from white farmers (directly or via recent land reform government purchase)?
In a land reform context there is a major difference, especially as the 30% target aims only at white-owned farmland. Government land disposal is a separate issue, controversial but with no target.
The most important statement of the audit is: private purchases by PDIs amounted to 4.3 million ha, while government only purchased 2.2 million ha.
This totals 6.5 million ha, but the audit conclusion says: “From the 1994 to 2016 transactional data, a total of 8.9 million ha was bought by PDIs and government”.
Well, which is it?
Other terminology carelessly used
The term “ownership” is used throughout and is misleading because the focus of the audit is “government and PDI-owned” land, most of which is communal. Communal land is legally unregistered state land, but is also land held in trust for the communities who occupy it. It is not “owned” by anyone. It does not appear in the Deeds Registers which are the fundamental data source for this audit.
The significance here is that indigenous concepts of land holding and land rights differ radically from the western concept of “ownership” – this is the core of the land tenure/reform problem of communal land.
The consequence is that large-scale commercial agriculture cannot happen on communal land because it is not “owned” in the Deeds Registry sense and cannot therefore, be used as security for loans from banks.
The 93.3-million hectares of “agricultural land” in SA is the base of AgriSA’s audit. It includes crop land, irrigated land and grazing land (by far the biggest proportion). “Potential agricultural land” is a big part of AgriSA’s argument, but this is perhaps a confusing term. To be more accurate it means high potential arable or croppable land. Most is in the eastern half of SA, with higher rainfall and deeper soils.
AgriSA wants to emphasise that much of this land is black “owned” saying 46.9% is owned by government/PDIs. Most is government state land or communal land; a small portion – 4.3-million ha, has been purchased by black people since 1994; an unknown small portion was already owned by blacks before 1994.
Careless use of the audit’s terminology has already happened, putting a false spin on what ought to be a factual product. In Farmers Weekly (surely a vital audience) Dan Kriek, AgriSA president no less, is reported stating “the eastern part of SA held the most agricultural potential, and PDIs owned 46.5% of this land”.
No they don’t – in spite of the executive summary saying so, the audit actually says government and PDIs do.
News24 repeats the same mistake. The consequence is that readers are given the false impression that black emerging farmers own 26.9% of farms or 46.5% of agricultural potential – they don’t.
What would have been useful ….
While understanding the difficulties of putting companies and trust owners in a race category, it would have been useful to state at least the size of that group as compared with government and PDIs and private white farmers. A major portion of agricultural production in SA comes from such farms.
In dealing with the Eastern Cape the audit reports that the provincial level of ownership by PDIs and government in 2016 was 48.3% of the province, an increase from 28% in 1994. An increase of 20% of the province’s 16892 000ha is 3 378 400ha. Frankly this is not believable; it is approaching the size of Transkei. If true, where are all these farms?
Major blocks of state land transferred from government to black farmers (PDIs) such as Pitseng, Beestekraal, Port St Johns and Seymour don’t count as they remain in the same fudged category of government/
PDIs. Major restitution settlements such as Dwesa-Cwebe and Mkambati have also never been transferred to black beneficiaries so don’t appear in the Deeds Registry data. Government land has decreased through disposal since 1994 so the 20% must be in former white commercial farm areas.
Part of the reason for this confusion is that the audit base figure of 28% in 1994 is wrong. (Government’s own land audit in 2013 said it is 9% – even less believable.)
State land in 1994 – Transkei, Ciskei, SADT: 4741010 ha
Provincial state land: 506609ha
Department of Public Works: 606463ha
Total: 6026312 = 36%
Consequently the increase is 12%, (not 20%) or 2027040ha or an average of 120000ha per year since 2000 when land reform started to move.
(In 2004 the EC Department of Land Affairs transferred 90 farms totalling 20000ha. Did DLA really pick up the pace of transfers by over six times for more than ten years? It seems unlikely since the Department of Rural Development and Land Reform 2016-17 annual report for land reform shows 80000ha acquired for the whole country and no transfers.)
Indeed, it is difficult to locate 2000000ha of white farms transferred to black farmers in the province. That’s roughly the whole area between the N6 and N10 or the Karoo portion of the Eastern Cape If the audit mapped its findings the situation would be much clearer.
Government land in Eastern Cape
What constitutes state land is not straightforward. It’s complex and difficult to analyse usefully and may be the source of some confusion. Technically it includes communal land (nearly a third of the province). State agricultural land expropriated by the apartheid government in the 1970s for incorporation into Transkei and Ciskei has only partly been transferred to black farmers. It includes domestic state use such as for police stations, clinics, but also forestry land which may be under long lease to companies by communities under restitution settlements. A major part of provincial state land is nature reserves.
It is still administered by two separate national departments – Public Works for former RSA land and RDLR for former homelands and acquired farms – and the provincial DPW.
More questions than answers
So although the intention of the audit was to clarify the land situation, we have only one major fact, some percentages and a number of unanswered questions. Hopefully a second edition of the report will make more use of the data which now exists but has not been made available.
Mike Coleman is a specialist in land, rural development and agricultural planning in the Eastern Cape