The Economic Freedom Fighters (EFF), a South African political party, launched its election manifesto on 2 February 2019 in the country's Gauteng province. The national election is scheduled for 8 May 2019.
The party claims that "landlessness and joblessness among black South Africans are at crisis levels, posing the biggest challenges that confront South African society today".
This report fact-checks five claims about land occupation, school dropout rates, unemployment, land redistribution and gender balance.
The EFF bases its claim on ownership data and the size of the African and coloured population.
The party’s national communications manager, Sixolise Gcilishe, explained how it was calculated:
|Total land area: 122 million hectares|
minus 67% – white-owned agricultural land
minus 14% – state land
minus 5% – land owned by Indians*
equals 14% of land “left for black African and coloured South Africans”, who make up 88% of the population
*This number is rounded off. The number the EFF provided was 4.82%.
This fact-check evaluates each of the EFF’s assumptions stated above.
Is 67% of SA land white-owned agricultural land?
In 1993, the year before South Africa’s first democratic election, a Stats SA agricultural census estimated the country had about 82 million hectares of agricultural land. The government has referred to this land as “white-owned agricultural land”.
This was 67% of South Africa’s total land.
The EFF said work by Plaas, the Institute for Poverty, Land and Agrarian Studies at the University of the Western Cape, was the source of the 67% used in its calculation.
The number is contained in a 2013 Plaas fact-check on land distribution co-authored by Prof Cherryl Walker, a professor of sociology at the University of Stellenbosch. She said the 67% was commercial farmland historically reserved for white ownership, but it shouldn't be assumed to be the share in white hands today.
Since 1994 black people have acquired some of this land through private sales and loans.
Prof Ben Cousins of Plaas described the EFF estimates as “highly misleading”. He said much of the land acquired for land reform since 1994 was “no doubt farmland, and no longer forms part of the 67% of the total land area formerly in the hands of white farmers”.
A report on land redistribution commissioned by parliament says “some land acquired or transferred may not be commercial agricultural land… Nonetheless, we may presume that most may be considered land zoned for agriculture outside of the former Bantustans”.
How much agricultural land is owned by white people? The available data is limited.
According to the 2017 state land audit, farms and other agricultural land amounted to just over 111 million hectares in 2015. The audit only provides a racial breakdown for agricultural land owned by individuals, some 37 million hectares. It estimates that white people own 72% of this land – 26.7 million hectares.
An Agri SA land audit estimated that white farmers owned 68.5 million hectares of agricultural land in 2016. This includes land owned by companies and trusts and amounts to 56% of total land.
Agricultural economist Wandile Sihlobo gave Africa Check an estimate of agricultural land not owned by white people.
It identifies 17.5 million hectares of farmland as “not owned by whites”. The calculation was done with the help of agricultural economist Prof Johann Kirsten of the University of Stellenbosch and based on data from 2015 and 2017.
It is made up of agricultural land owned by the state, redistributed and restored land and a “guesstimate” of the land bought by willing black buyers from willing white sellers. It also includes a hectare equivalent of money paid in land restitution claims.
A February 2019 Plaas working paper, co-authored by Kirsten, contains an updated calculation, which relies on 2018 data from the department of rural development and land reform.
|South African land redistribution numbers|
|Agricultural land acquired by the state since 1994 and still held by the state for land reform purposes||2,289,063|
|Private transactions (without the involvement of the department of rural development and land reform)||1,968,057|
Source: Plaas working paper (Note: This calculation does not include a hectare equivalent of compensation paid for restoration.)
Using these redistribution numbers, and an estimated 643,995 hectares of agricultural land acquired by the state for development purposes, it estimates that 64.8 million hectares of farmland (53% of total land) is owned by white people.
The available data therefore does not support the EFF’s assumption that 82 million hectares of agricultural land – 67% of South Africa’s land – is in white hands.
Is state land unoccupied?
The 2017 state land audit focused on privately owned land. The previous audit, published in 2013, identified 17.1 million hectares or 14% of the total land as state-owned land.
But Cousins said “state land” could be a misleading category. “It includes land held for nature conservation, state forests, water and defence force use. It also includes [land] in the traditional areas.”
The 2019 Plaas working paper estimates land under traditional tenure to be 18.4 million hectares – or 15% of total land.
Economist Johann Bornman, given as a source for calculations of land size in the working paper, said that in 2018 the rural development and land reform department updated its estimate of land in the former homelands to this number.
Dr Aninka Claassens, chief researcher at the Land and Accountability Research Centre, said the EFF “seems to ignore the fact that black South African have underlying rights to all the land in the former homelands, which is state land”.
(Note: In the explanation of its calculation, the EFF said it was “almost impossible to scientifically determine” how state land was occupied. We have asked Stats SA to help us determine who occupies land in the former homelands and will update this fact-check once we have the results.)
How much land do Indian South Africans own?
The 2017 state land audit found that 77% of land (94 million hectares) was privately owned.
But only the race of individual owners of 37.8 million hectares could be determined. And these numbers may not be spot-on.
The deeds registry does not record the race of land owners. In cases where the race could not be derived from Stats SA data, names and surnames were used. “The indirect extraction of race data… exposed the land audit to the risk of under- or over-reporting,” the audit report says.
The EFF calculation assumes 5% of South Africa’s total land (5.9 million hectares) is owned by Indians. But this is not supported by the land audit, which puts individual Indian ownership at 2.1 million hectares, including sectional title property.
The rural development and land reform department previously told Africa Check it did not include sectional title units when calculating land owned by individuals “as they include units in high-rise buildings such as flats, office buildings… which may distort the numbers”.
|Individual land ownership by race in South Africa|
|Race||Farms and agricultural holdings||Erven||Sectional title||Total hectares||Share of SA land|
Source: 2017 state land audit (Note: Percentages don’t add up to 100% because the categories “co-ownership” and “other” are left out.)
Indian people may own land through trusts, companies and community organisations but there is no data on this.
The same goes for African and coloured people, who own 6.9 million hectares of individually owned land or close to 6% of the total land. Coloured and African people together make up close to 90% of the population, according to StatsSA’s 2018 mid-year population estimates.
The 2017 land audit only gives data on the race of the owners of about a third of the land. A breakdown by race for other types of private ownership is not available.
It is also silent on who occupies this land – in some cases, owners do not occupy their land – as well as on who occupies land in the former homelands.
The EFF’s calculation also does not take into account erven and sectional title units owned by white people.
The EFF’s calculation makes assumptions about ownership to support a claim about the occupation of land. It incorrectly assumes white people own 82 million hectares of agricultural land and overestimates land individually owned by Indians. It is therefore incorrect.
The EFF’s claim applies to a worldwide comparison, the party’s national communications manager Sixolise Gcilishe confirmed. She referred us to three documents to support the statement.
The common thread in these documents is work done by Professor Martin Gustafsson, an education economist and member of the Research on Socioeconomic Policy group at the University of Stellenbosch. However, Gustafsson does not agree that his work supports the EFF’s claim.
“In fact, as far as secondary [school] completion is concerned – that’s essentially what this is about – we’re fairly normal for a middle-income country,” he said.
Measuring dropout rates
Gustafsson said it was “widely accepted” that the dropout rate was difficult to calculate, “especially in developing countries, where data systems are limited”.
However, he said, education planners regarded the definition of the UN Educational, Scientific and Cultural Organisation (Unesco) as useful: “The proportion of pupils from a cohort enrolled in a given grade at a given school year who are no longer enrolled in the next school year.”
Gustafsson noted that the Unesco definition related to individual grades. “So one can talk about a dropout rate for Grade 10, another one for Grade 11 and so on.”
One of the alternative methods to calculate the dropout rate is to use data from household surveys to determine what proportion of the population has obtained a matric certificate.
Gustafsson said this method was not ideal, but was an “adequate stopgap”. It was the method used by the department of basic education in response to a parliamentary question about dropout rates in June last year.
What is SA’s dropout rate?
The written reply contains an estimated dropout rate for those born between 1990 and 1992.
Information from Statistics South Africa’s (Stats SA) General Household Survey for 2014 to 2016 was used to estimate that 51.5% of these people had completed Grade 12.
We asked the department of basic education to provide more recent calculations if they were available, but it failed to respond to our query.
One of the documents the EFF referred to in support of its claim is a 2011 working paper authored by Gustafsson, but it did not include a worldwide comparison.
The working paper looked at Grade 12 completion rates in 14 countries, including Turkey, Brazil and Chile, based on data from 2003 to 2009. South Africa was ranked 12th.
Gustafsson referred Africa Check to Unesco’s upper secondary school completion rates as a source of more recent data that would allow for a global comparison.
Based on these figures, the department’s 2018 matric examination report concluded that “the upper secondary education completion [Grade 12] rate for South Africa has been equal to that of middle-income countries in general in recent years”.
We compared the average upper secondary education completion rates of 68 countries for which Unesco data were available – for the five years from 2014 to last year – and found that 38 countries had lower completion rates than South Africa.
Gustafsson said while South Africa’s secondary completion rate was “fairly normal for a middle-income country”, we should aim to increase it, as many other countries are doing.
He warned that “defining our vast educational inequalities only in terms of school completion would be incorrect. Many of our inequalities relate to the quality of education, not the years of schooling completed by young people.”
The EFF’s claim that South Africa’s school dropout rate is among the worst in the world is incorrect. Unesco data for 2014 to last year show that South Africa was in the top 50% of 68 countries based on secondary school completion.
Stats SA’s Quarterly Labour Force Survey measures unemployment in two ways – according to the narrower, official definition and according to the expanded definition.
The agency’s acting chief director for labour statistics, Malerato Mosiane, explained: “The difference between the official and expanded definition of unemployment is that the job search criterion is relaxed in the expanded definition.”
In the case of the official definition, someone has to be without work, available for work and searching for work or trying to start a business. The expanded definition does not require a person to be looking for a job – they just need to be without work and available for work.
At the time the EFF released its manifesto on February 2, the latest available Quarterly Labour Force Survey was for the third quarter of last year.
According to the report and based on the official definition of unemployment, 6.2 million people were unemployed at that time.
This means that they:
- Were between the ages of 15 and 64;
- Were not employed in the week before the survey interview; and
- Actively looked for work or tried to start a business in the four weeks before the interview, and would have been able to start work or a business in the week before the interview.
A person was also considered unemployed if they “had not actively looked for work in the past four weeks, but had a job or business to start at a definite date in the future and were available”.
A further 2.7 million were identified as discouraged work seekers. This group of people wanted to work, but did not try to find work because of a lack of jobs in their area, because they had lost hope or could not find work requiring their skills.
An additional 817,000 people without work provided reasons for failing to search for a job other than the ones contained in the definition of a discouraged work seeker. These included ill health, a disability, pregnancy, childcare duties and a lack of transport, said Mosiane.
Capability not part of either definition
The unemployed, discouraged work seekers and those with other reasons for not searching for a job make up the expanded definition of unemployment. According to this definition, close to 9.8 million people, or 37.3%, of the labour force (26.1 million) were unemployed in the third quarter of last year.
While Mosiane noted that the EFF’s claim used the term “capable”, which was not part of the official or the expanded definition of unemployment, the numbers in the claim most closely match the expanded definition.
Academics have argued that there is little to distinguish between the two definitions in South Africa. Stats SA previously told Africa Check that the expanded definition is a more stable indicator.
EFF national communications manager Sixolise Gcilishe confirmed to Africa Check that the party had applied the expanded definition in this case.
The EFF’s claim that more than 9 million – or close to 40% – of South Africans who need jobs are unemployed is correct. Almost 9.8 million people – or 37.7% of the labour force – were unemployed according to the expanded definition of unemployment in the third quarter of last year.
In 2012, then rural development and land reform minister Gugile Nkwinti used his budget vote speech to “clarify” government’s 30% target for land reform.
Nkwinti said the target applied to white-owned agricultural land – close to a third of which government wanted to distribute to the historically disadvantaged by 2014.
When the target was set in 1994, white-owned agricultural land was estimated to be 82 million hectares, he said.
The 30% came to 24.6 million hectares, which were earmarked for all three components of land reform, namely restitution, redistribution and tenure reform, according to the department of rural development and land reform.
The department pointed out that it would be incorrect to still use the 30% target as it applied until 2014. The current target is to redistribute “20% of agricultural farming land by 2030”, in line with the National Development Plan.
We were unable to confirm with the department by the time of publication whether land previously redistributed would count towards the new target.
How much land has the state bought?
EFF national communications manager Sixolise Gcilishe told us the party based its claim on a 2016 report on land reform commissioned by a high-level panel on the assessment of key legislation and the acceleration of fundamental change.
The panel, chaired by former president Kgalema Motlanthe, was tasked with assessing the effectiveness of legislation that has come into effect since 1994. Land reform was one of the focus areas of the panel’s work.
According to the report, authored by the Institute for Poverty, Land and Agrarian Studies (Plaas) at the University of the Western Cape, “about 5 million hectares” were redistributed between 1994 and 2014/15. A further 3.2 million hectares were restored through restitution by 2014/15.
This comes to 8.2 million hectares, or 6.7%, of the total land in South Africa (122 million hectares). The EFF told us this was how it reached the figure of “less than 7%”.
The latest available data, according to the department, largely correspond with the data in the Plaas report. They show that the state has bought 4.9 million hectares for land redistribution since 1994, and a further 3.3 million hectares for restitution. Land acquired for tenure reform (782,487 hectares) was included in the 4.9 million hectares, said the department.
It would therefore be correct to say that government has bought less than 7% of the total land in South Africa, but it is incorrect to say that the state has only bought 7% of the targeted 30%. The state has bought 8.2 million hectares – or 33% – of the targeted 24.6 million hectares.
Not all of the land bought or transferred to beneficiaries for the purpose of land redistribution is necessarily commercial agricultural land. According to another commissioned high-level panel report, “some land acquired may be in urban areas or in communal areas, and may have been land acquired for non-agricultural purposes – before land reform became equated with agriculture. Nevertheless, we may presume that most may be considered land zoned for agriculture outside of the former Bantustans.”
Redistribution, restitution and tenure reform
The EFF’s claim only mentions land acquired for redistribution, but land was also bought for restitution and tenure reform.
The Plaas report explains the difference: “Soon after the first election in 1994, an ambitious policy of land reform began to be implemented. This included a land redistribution programme aimed at broadening access to land among the country’s black majority; a land restitution programme to restore land or provide alternative compensation to those dispossessed as a result of racially discriminatory laws and practices since 1913; and a tenure reform programme to secure the rights of people living under insecure arrangements on land owned by others, including the state (in communal areas and the former ‘coloured’ rural reserves) and private landowners (farm workers, farm dwellers and labour tenants).”
The department confirmed that the amount of land bought for restitution excluded cases in which compensation was paid because land could not be restored or because beneficiaries chose to be compensated instead.
The claim that the government has bought less than 7% of the land initially meant for redistribution since 1994 is incorrect. Data show that it has bought a third – or 8.2 million hectares – of the land.
Women and girls make up 51% of the country’s population, according to Statistics South Africa’s 2018 mid-year population estimates. But the people most political parties choose to represent them in the National Assembly and National Council of Provinces don’t reflect this.
In its 2019 election manifesto, the EFF claims to have a “higher gender balance” in parliament.
We asked the party if this means it has a higher share of women among its representatives in parliament than other parties.
EFF national communications manager Sixolise Gcilishe didn’t respond to the question but did send us a spreadsheet. According to this, 40% – or 10 – of the EFF’s National Assembly members and 67% (4) of its National Council of Provinces members were female.
Two houses of parliament
South Africa has a bicameral parliament with two houses: the National Assembly (NA) and the National Council of Provinces (NCOP).
- National Assembly
The EFF published its election manifesto on 2 February 2019.
We used the most recent list of NA members, provided by parliament, and a list of changes in membership to work out the gender breakdown at the time the EFF released its election manifesto.
The African National Congress (ANC) had a higher share of female NA representatives, at 50%, while 44% of the EFF’s NA members were female.
|Members of the National Assembly by party and sex in South Africa (2 January to 1 February 2019)|
|Party||Female MPs||Male MPs||Share of women|
Source: Parliament of South Africa
We compared the list provided by parliament to information on the People’s Assembly website and parliament’s own website. We found one female ANC member who had been excluded from the list of female members.
- National Council of Provinces
The EFF had the highest share of women (67%) in the NCOP when the party released its manifesto, based on a list of current members and resignations provided by NCOP chairperson Thandi Modise’s office.
|Members of the National Council of Provinces by party and sex (2 January to 1 February 2019)|
|Party||Female MPs||Male MPs||Share of women|
Source: Parliament of South Africa
Total for both houses of parliament
But how do the numbers add up for both houses of parliament?
For the EFF, 48% of all its members of parliament were female when the party released its manifesto. The ANC had a higher share: 49% of its members were female.
|Members of parliament by party and sex (2 January to 1 February 2019)|
|Party||Female MPs||Male MPs||Share of women|
Source: Parliament of South Africa
The EFF claimed it had “a higher gender balance” in parliament. The data shows it had a higher share of female representatives in parliament than 11 other parties at the time it released its election manifesto. Forty-eight percent of the EFF’s members of parliament were women, compared with 49% of the ANC’s members.
|This package is part of a journalism partnership with South African newspaper City Press. The project aims to ensure that claims made by those in charge of state resources and of delivering essential services are factually correct. In the run-up to this year’s national and provincial elections, it is increasingly important that voters are able to make informed decisions. This series aims to provide voters with the tools to do that.|