Click in the black circle in the image below and move the slider to compare aerial pictures of Nkandla taken in 2009 and 2013.
President Jacob Zuma’s rural homestead in Nkandla, KwaZulu-Natal, first developed into a saga in 2009. Over the years it has seen multiple court cases, numerous reports, reams and reams of reporting, shifting allegiances, changes of heart, and enough drama to confuse veteran writers of daytime soap operas.
The story is far from over, but many broad trends have emerged, and many developments have played out. The Mail & Guardian believes that Nkandla is of enormous public interest, and we have spent large amounts of time and money on trying to unravel it. This is an attempt to capture the essence of the Nkandla saga.
1. The story breaks
‘President Jacob Zuma is expanding his remote family homestead at Nkandla in rural KwaZulu-Natal for a whopping price of R65-million." This is how the world learned, on the first Friday in December 2009, the details of extensive work that had recently started at what would later come to be known simply as Nkandla.
The discovery was an accidental one. It came not through a tip-off from a concerned civil servant or a telephone call from a political opponent who smelled a potential advantage – those would come only later, as information on the project dribbled out. Later there would be tens of thousands of words written, thousands of pages of documents published, hundreds of pictures taken, and more press conferences and statements and parliamentary questions than you could shake a stick at.
But the article that launched it all came about because a nosy journalist talked her way into a prefab office and looked at the wall.
Former Mail & Guardian political reporter Mandy Rossouw, who died tragically young in March 2013, had set off to Nkandla at the end of November on what government officials would term an inspection, but which journalists simply call good journalism.
Jacob Gedleyihlekisa Zuma had taken office in May 2009 and South Africa was still trying to figure out what manner of man it had as president.
Zuma, a determined traditionalist, had deep roots in what was then an inaccessible part of KwaZulu-Natal. As the end of the year rolled around, with its slower pace and room for deeper rumination, the M&G wanted to see from where Zuma hailed. The idea was to speak to his neighbours, describe the environs, perhaps provide some deeper insight into his nature, and what it would mean for South Africa during his term in office.
That idea, of course, was soon overtaken.
Rossouw and Chris Roper, who later became the M&G’s editor-in-chief, set off prepared to do battle with poor roads; their expectations were fully met.
“Nkandla boasts potholes that look like the work of stray meteors. Even for those wanting to visit, getting into this area requires time, commitment and, if not a helicopter, at least four-wheel drive,” Rossouw wrote in one subsequent article.
“Two roads lead to Nkandla and neither is an easy drive,” she started another piece.
When they arrived, by dint of slow going and directions from locals along the way, they did indeed speak to neighbours, and Roper took what would become the first published pictures and video of the swelling homestead.
What they found was not exactly explosive: Zuma was a little better off than other residents of the area, perhaps, but was hardly living ostentatiously. Those living nearby spoke of a man who still showed deference to local traditional leadership despite holding the highest office in the land. The area had been connected to the electricity grid, but so had many other rural settlements, and there was no sign of undue preference for Nkandla in terms of state development spending.
It was pretty much exactly what could have been expected, and would have made for an interesting article, but hardly one that would launch multiple formal investigations, bouts of angry accusations and counter-accusations, and become a major pain point for Zuma throughout the rest of his political career.
Then the always convincing and always determined Rossouw wheedled her way into a cramped and hot site office, and suddenly everything changed.
Rossouw had noticed some building work going on around the residence. It was rather desultory stuff, featuring about a dozen men, some earth-moving equipment, two cement mixers and two water tanks. It was clear the renovations or expansion would entail more than just a fresh coat of paint, but nothing suggested that, eventually, more than R200-million in tax money would be spent.
Not until Rossouw saw the architectural drawings taped to the walls of the project office. The plans showed a first phase comprising a guest house and two new residences, but also provided a hint of what was to come: a helicopter pad, a large parking lot, a clinic, a visitors’ centre and a police command post.
Denials, deceit and dirty spin tactics
From there things got very weird very quickly. For starters, the story nearly immediately came to feature anonymous sources, the first of many the M&G would come to protect in return for information that could not be obtained in any other way.
In addition, the next few days would see the first examples of what would become a four-year pattern of behaviour for the state: denials, deceit, misleading statements, accusations of invasion of privacy and dubious motives, ill feelings on all sides and dirty spin tactics.
By the end of the week Nkandla had seen its first public battle in what would become a long and grinding war between the media and the government.
M&G’s then editor-in-chief, Nic Dawes, wrote a story that accompanied the initial article (See “Softening the blow” in shaded box).
“The presidency has damaged the relationship of trust that we had developed with officials there,” he wrote.
“If government communicators make it impossible for us to trust them with basic courtesy, we will struggle to share information with them, and two things will suffer: their ability to try and shape our opinions, and the willingness of journalists to seek all sides of every story. Ultimately, that is bad for democracy.”
Dawes was spitting mad because, although details of the planned work had been published in that Friday’s edition of the M&G, the paper had been scooped the previous day.
It was not because another news organisation happened to stumble into the project office in some sort of punchline for a cosmic joke on coincidence. The country learnt that something was going on at Nkandla on Thursday because the presidency said so.
Earlier that week the M&G’s story was brewing. It had confirmed that the Nkandla project was expected to cost R65-million, but had early indications that the cost would be considerably more. It knew that such spending on the private residence of a president, who not only had official residences elsewhere in the country, but who was also still under a cloud from a judge’s finding that his former financial adviser had been guilty of corruption, would be controversial. So, as is standard practice, the paper turned to the presidency and the department of public works for comment, so that those organisations would have several days to respond.
That turned out to be a mistake.
The presidency, through its communications arm, simply refused to speak on the matter. The department of public works did respond, but with what would be exposed within days as a flat-out lie.
“Please note that there is no work or extension project taking place at President Jacob Zuma’s homestead at Nkandla,” the department responded by way of a spokesperson.
Behind the scenes there was a scramble to figure out what had happened. One spokesperson suspiciously asked Rossouw how the story had originated, and where the initial information had come from. “We drove there,” Rossouw later told friends she had responded, complete with a laconic drawl.
The presidency later also said it had no knowledge of a project, and insisted that the state had no business in the private home of the president.
Then-presidential spokesperson Vusi Mona – who would go on to a fairly controversial career at the South African National Roads Agency (Sanral) in defence of urban freeway tolls, or e-tolls – promised to raise the matter with Zuma during the course of Thursday, but was not available later.
And that was what was reflected in the article finalised on Thursday morning ahead of going to print later that evening. Despite the denials, the M&G had a story it knew would stand up to scrutiny, and the paper was not in the habit of holding back on matters of public interest because it would raise the ire of the powers that be.
But, late on Thursday afternoon, things took a dramatic turn.
At a time that seemed carefully calculated to be too late for the M&G to make alterations to the newspaper, but early enough to catch the evening news cycle, the presidency issued its very first statement on Nkandla.
Instead of the earlier denials to the M&G, it admitted to the construction work, but said that there was nothing to concern either the public or the media about a government project that would be entirely separate from upgrades being done by the Zuma family.
“We urge the media to leave the family alone to conduct its business, and reject any insinuation that there could be any untoward abuse of state resources by the president or his family,” the statement read.
“The demarcation at Nkandla is very clear, and there can be no reason to confuse the private construction work in the Zuma household and the state facilities that will be constructed outside the perimeter.”
In hindsight, one phrase in particular stands out. “No government funding will be used for construction work.” Today the statement is not to be found on the websites of the presidency, which reflects speeches and statements right from the start of Zuma’s term, or of the government communications hub, which documents such releases back to 1999. At the time, however, the government’s communicators wanted it seen, and sent it to every major news organisation in the country.
The M&G saw the statement as an obvious attempt to limit the impact of a major breaking story, the kind of exclusive that is vital for investigative newspapers. The result was a sequence of tense discus- sions between various editors and the government, with warnings that newspapers would switch to providing the minimum amount of time for comment rather than the maximum if they faced being ambushed in such a fashion.
An agreement of sorts was subsequently reached, but it was already too late. The story of Nkandla was born amid lies and suspicion, and over the following years little would change.
2. Tough time for the president
Jacob Zuma was not secure in the universal love and admiration of the country before Nkandla broke, and his life did not become any easier after it.
The ANC’s overwhelming support for Zuma as its leader at the party’s 2012 elective conference in Mangaung changed the equation, but that was later. In late 2009 and early 2010, Zuma had not yet consolidated his support, and it is hard to imagine that the slow-motion train wreck that was Nkandla could have come at a worst time.
Almost exactly two years before the publication of the first Nkandla story, Zuma wrested control of the ANC from Thabo Mbeki at the historic Polokwane conference. The fight was nasty, and the outcome was not unity: 2 329 delegates representing ANC members across the country – the constituency that in effect chooses the president – voted for Zuma, giving him 60% of the support. But 1 505 voted for Mbeki. A full 40% of the ANC did not want to see Zuma in the top job. Some of them voted with their feet, forming the ill-fated Congress of the People as an act of protest.
But the distrust he faced within the party was nothing compared with that of the nation more broadly.
Zuma left the deputy presidency under a cloud in mid-2005. A court had found his financial adviser Schabir Shaik guilty on two counts of corruption and one of fraud — with Judge Hilary Squires saying in so many words that Shaik had had a corrupt relationship with Zuma. Mbeki noted that Zuma himself had not been found guilty, had in fact not even been tried, and that Shaik had not exhausted all his appeals, but fired Zuma nonetheless.
Or, as Mbeki put it so delicately in Parliament: “... as president of the republic I have come to the conclusion that the circumstances dictate that in the interest of the honourable deputy president, the government, our young democratic system, and our country, it would be best to release the honourable Jacob Zuma from his responsibilities as deputy president of the republic and member of the cabinet.”
In December 2005 Zuma had been charged with rape. By May 2006 he had been found not guilty, but only after a slew of intensely embar- rassing revelations that would follow him for the rest of his life: he had had sex with the daughter of a friend, he had had sex with an HIV-positive woman without using a condom, he had taken a shower as a precaution against contracting HIV.
Over the next four years, South Africa came to know Zuma through legal action against him – and more often brought on his behalf. He was charged with corruption, the case was struck off the roll, he was indicted again, one court ruled evidence against him to be admissible, another declared the charges against him unlawful. By the time the National Prosecuting Authority controversially dropped the charges, in April 2009, the legal situation had become incomprehensible to the layman.
The nonlegal situation was not much clearer. Was there a conspiracy to prosecute Zuma, or at least time his prosecution for maximum damage? Did that make him innocent? If he really wanted his day in court, why did his legal team seem to use every conceivable opportunity to stall?
It was into that context of questions and questionable conduct that Nkandla was first exposed. But it would take years for the scandal to mature fully. In the meantime, there were other incidents that cast doubt on the character of the man who, by then, had become a president with approval ratings approaching respectable numbers: urban approval was 58% by the end of 2009, according to one poll.
Some commentators worried about the cost to the state of a first family many times larger than any the country had ever seen; others mused about the implications of having a staunch traditionalist as a leader.
In January 2010, as if to invite scrutiny of Nkandla, Zuma married Thobeka Madiba, his fifth wife, at the homestead. Front and centre was not the bride’s dress but the president’s polygamy and his 19 acknowledged children.
In February 2010, Zuma admitted he had fathered a child out of wedlock with the daughter of yet another friend – not voluntarily, but after unsuccessfully trying to weather an exposé by the Sunday Times. “Has President Jacob Zuma finally overstepped the morality mark?” asked the M&G in an article after the acknowledgment. Judging by public outrage,the answer appeared to be a clear “yes”, and Zuma’s urban approval fell to 43%.
By March 2010 the story had moved on. Zuma had failed to declare his assets and interests after taking office, an ominous breach of what should have been a formality.
During April 2010, there was a choice of stories to follow.
Zuma’s son Duduzane had been linked to Imperial Crown Trading and its enormously suspicious acquisition of prospecting rights on an existing iron ore mine; Zuma’s nephew Khulubuse Zuma had been linked to the increasingly controversial Aurora mine; Zuma himself had been linked to the still relatively unknown Gupta family’s efforts to create a newspaper friendly to the ANC; and Zuma had revealed the results of an HIV test (negative), and so almost forced the country to hark back to his rape trial.
By the middle of 2010 the portents were ominous. Leaders within the ANC and its alliance partners the South African Communist Party and trade union federation Cosatu were anonymously pondering whether Zuma had become a liability.
The Juju factor
One leader with an unusually loud voice was willing to put his name to the criticism – although at the time Julius Malema, then leader of the ANC Youth League, was rather more circumspect in his criticism of his nominal boss than he would later become, once he had finally parted ways with the ANC. At the time, he was forced into an apology that nicely summarised what his approach to Zuma had been.
“I accept that these statements had the effect of undermining the stature of the President of the ANC and of the Republic,” read the apology Malema was forced to issue after a disciplinary hearing.
“It further may have had the effect of undermining the confidence of our people in the leadership of the ANC and of creating serious divisions and breakdown of unity in the organisation.”
Jacob Zuma was not having an easy ride. But Nkandla, even then being expanded at a breakneck pace, got barely a mention as these various dramas unfolded. It would not be until deep into 2012 that the houses on the hill in a remote part of KwaZulu-Natal would storm the headlines – and stay there for more than a year.
3. A town called Zumaville
The call came from a professional who closely watched tenders in KwaZulu-Natal. Had the M&G heard of the enormous development being planned just about in the backyard of the presidential homestead, he wanted to know. It was big, it was state-funded, and it was being pushed ahead with nearly unseemly haste. Might it be worth investigating?
The M&G had not, in fact, heard of the development, but, yes, it did indeed seem as though it could be of interest. Specifically, it seemed to be both in the public interest, and something the public might be interested in, that all-too-rare confluence. The basic details were confirmed in short order and it was not long before a dedicated team, stretching across the political, investigative and news desks, were worrying at the issue.
That team hit pay dirt within a week, and one of the most memorable – and on occasion confusing – news monikers of the year was born. “Welcome to Zumaville”, read the headline of the story published on August 3 2012, before putting together the eye-popping numbers: R1-billion in estimated government spending with another R1-billion in private money required, just 3.2km from the Zuma residence, the first new town South Africa would build since the advent of democracy. At a time when the president was already being accused of looking out first for his family, and second for his neighbourhood, it raised many an eyebrow.
Unlike the original Nkandla article, the M&G did not break the news of what was officially known in planning documents as the “Umlalazi-Nkandla Smart Growth Centre”. There had been meetings about it, sketches had been presented, and there had even been tender documents. Cabinet ministers had visited the site, and the idea had been mentioned in Parliament. But thanks in large part to the innocuous name, and the lack of hard spending numbers and context, it had gone almost unnoticed.
Giving it a relatable name and costing the exercise changed everything, although it was the images that clinched it in the public mind. The concept drawings for Zumaville, secured shortly before the August 3 story was published, showed gleaming buildings, lush green areas, and infrastructure that would be the envy of any number of struggling towns.
Yet it was the name that caused the most trouble.
Dubbing the project “Zumaville” made it memorable, but also emotive and confusing. The close proximity of the new town development and the presidential homestead led to the two being conflated, and “Zumaville” has since been used many times to refer to the Zuma residence.
Although not entirely unfair – the greater compound does, after all, include a tiny commercial district and enough housing and “public” amenities to qualify as a village – there will be a considerable difference between the two projects.
The emotive nature of the name was a somewhat bigger problem. Government departments involved pointed to the choice of “Zumaville” as proof that the M&G, and the media in general, was trying to link the president to matters that he had no direct hand in.
Rural development was a priority of the government, the presidency said. Several similar projects were under consideration, the department of rural development said. The fact that Zumaville would be built so close to the president’s private residence was, well, just a coincidence, came the justification from far and wide.
However, the strange distortion of reality that permeates so much in the vicinity of Nkandla was in operation again. Investigations showed, time and again, just how closely linked Zuma and Zumaville actually were.
First, of course, there was the inescapable proximity. Many bedroom windows in the Nkandla complex would look out on to what concept drawings showed would be a sprawling town of modern structures inter- spersed with trees.
Second was the origin. Zumaville was conceived by Masibambisane, a rural development organisation based in the area. Masibambisane, which under one proposed structure would in effect control land use in Zumaville and receive rent from commercial tenants, was chaired by Zuma.
Third was the excessive haste with which several portions of the project were handled. Decisions were made at breakneck speed, and deadlines for certain steps, such as the town planning, bordered on the ludicrous.
“You’ll have the president watching you,” town planners were told during a tender briefing attended by the M&G.
If it is described as a presidential project, originates from a group run by the president, happens in the backyard of the president, and is seen as being watched over by the president, is it unfair to associate it with the president? Some in government thought so, especially once the M&G reported that families would have to be relocated in order for the project to go ahead. Information was scarce to begin with, then the tap was closed altogether.
Zumaville would not be a national key point. There would be no security concerns to be cited in its development, no claims to privacy in refusing to provide information on public spending. So, instead, the government simply told the media to wait.
“Government has noted that some members of the media have been asking departments to provide information concerning their involvement or financial contribution to the Nkandla development project,” said a statement issued just weeks after the initial Zumaville article appeared.
“The 2011-2012 financial year has been completed and government departments are in the process of finalising their annual reports.
“Information concerning any project undertaken by departments and budget expenditure will be outlined in these reports. The reports will be presented to Parliament towards the end of September 2012 and all interested parties, including members of the media, are encouraged to monitor the presentations and interact with information contained in the reports accordingly.”
In what came as no surprise at all, the reports by the departments of rural development, agriculture, education and various others did not detail what commitments had or would be made towards Zumaville. However, the departments did promptly stop any pretence of answering questions, and functionaries at various levels became peculiarly averse to taking telephone calls on any related matter.
With reality setting in, it became clear that construction on Zumaville would not be able to start in 2013, and much of the focus shifted back to individual projects in the area, run and funded by individual departments, rather than a single grand scheme. Nkandla would not have a shining town pop into existence in its starkly underdeveloped valley.
At the time of publication the higher education ministry was trying to get construction of several further education and training colleges under way, two in the greater Nkandla area. The department of agriculture was tinkering with infrastructure improvements and training in the area, and roads were still being upgraded. But the area bore very little resemblance to the concept sketches of what should have been an entire new town in the making.
4. The vanishing report
Nkandla has been probed and investigated, questioned in Parliament (including in the context of an attempt to launch a vote of not confidence against President Jacob Zuma) and documented in all kinds of ways.
At the same time there have been turnarounds and flip-flops, lies that became truth and disputed facts that became just plain facts – and secrets that turned out not to be all that secret.
Amid all this, however, a single document has come to embody both the multiplicity of reports and the dizzying changes the Nkandla story has come to encompass.
Under different names and in different guises, it has come to be referred to simply as the government’s Nkandla report. And it is this document that has been cited in three different ways during its three different incarnations to show that Zuma is definitely, absolutely, and in no way whatsoever culpable for everything that went wrong at Nkandla.
The three different incarnations and the drama that went with each seemed, at the time, simply farcical, a symptom of bungling and floundering of the first order. In hindsight, however, things look a little different.
Seen as a whole, 2013 was the year in which the official Nkandla narrative gradually changed. Gone were the attempts to ignore the issue until it went away. The attacks on the media and opposition parties grew gradually more muted and the defence that the spending had been by the book (through the imaginative reinterpretation of various laws and regulations) was all but abandoned.
Instead, the government departments involved regrouped around a new message: things had gone wrong, badly wrong, and there may even have been criminal acts behind the massive Nkandla price tag, as well as criminal negligence by the officials who had allowed it – but Zuma himself could positively, absolutely, definitely, under no circumstances whatsoever be blamed or held responsible for any of it.
South Africans had been ripped off, yes, by government officials and rapacious tenderpreneurs, but not by its president.
There was just one problem. Thanks to the veil of secrecy drawn over everything Nkandla it was next to impossible to really bring that message home. Something had to give and, eventually, it was the secrecy. That, in turn, came with its own set of complications and embarrassments.
Around about October 2012, Public Works Minister Thulas Nxesi commissioned the officiously titled Inter-ministerial Task Team on the Security Installation at President Jacob Zuma’s Nkandla Private Residence report. The investigation was initiated two months shy of a full three years after the first questions had been raised, but also just months before Nxesi and Zuma would be forced to answer questions on Nkandla in Parliament. Coincidence, perhaps?
Nxesi said only that the team had been formed “in response to concerns about the cost of the security upgrade at the Nkandla residence”. Why those concerns had not merited an investigation in the years before was never addressed.
So, late in 2012, all questions led to the task team report.
- Could the department of public works provide a breakdown of the costs of Nkandla? — Doing so would pre-empt the task team investigation.
- Was any action being taken against bureaucrats involved in decisions around Nkandla? — That would depend on the outcome of the task team report.
- Could the department answer this or that specific question about Zuma’s involvement in the Nkandla project? — Wait for the task team report.
That approach allowed the government to see out the remainder of 2012 – the year in which the R200-million price tag was attached to Nkandla by a City Press exposé – without new trouble.
Declarations of innocence
Early the next year, Nxesi made his first attempt at declaring Zuma innocent; 2013 was going to be a fresh start, it seemed.
“There is no evidence that public money was spent to build the private residence of the president or that any house belonging to the president was built with public money,” Nxesi said in a statement on January 27 2013.
The first incarnation of the government’s Nkandla report contained its own bombshell: confirmation that the project had cost more than R200-million. That number had been officially disclaimed, derided and scoffed at previously. Now it was stated as simple fact. Media outlets felt vindicated.
But, according to Nxesi, the top finding of the inter-ministerial task team was Zuma’s innocence, and he expounded on it extensively during a press conference.
“President Zuma is not involved in this process whatsoever,” Nxesi said in one of several formulations of the same theme.
It seemed a peculiar finding for two reasons. For one, the task team was not actually tasked with considering what Zuma knew, or what his involvement was. Its carefully framed terms of reference said it had to determine when Nkandla was declared a national key point, where supply-chain rules may have been broken, what recommendations had been made on upgrading security, and details of “recommendation(s) made by statutory entities”. Entities of the nonstatutory persuasion, such as, say, an architect in the employ of the president, or even the president acting in his personal capacity, were not included.
The second peculiarity was that in late November 2012, after the establishment of the task team but well before it concluded its work, the M&G had already reported that Zuma had indeed been involved in the upgrade process.
In an article under the quarrelsome headline “Nkandla: Documents call Zuma’s bluff“ the M&G reported that:
“President Jacob Zuma was provided with exhaustive details about progress on the security project at his Nkandla complex in November 2010. They cast doubt on his vehement denial in Parliament last week that he was unaware of the scale of construction.”
In one of the documents, a letter from Gwen Mahlangu-Nkabinde, at the time the minister of public works, Zuma is addressed directly. Another showed the close involvement of ministers in Zuma’s cabinet in the practical details of the project, yet another showed that a security director in the presidency had helped to scope the work.
What the documents did not provide was absolute proof that Zuma ever read or even received the progress reports sent to him, or that he was given rand amounts for the work. They did, however, show that several people had the opportunity, and perhaps even the duty, to inform him, and had tried to do so.
Nxesi and his department were promptly confronted by the letter, which the M&G published in full on the same day as Nxesi’s presentation of his team’s findings, on January 27 2013. His department first questioned its authenticity, then said it would ignore the letter as it had bigger fish to fry. The report stood unquestioned, as far as the government was concerned, and though there was lots more to do – such as investigating and prosecuting those who had taken advantage of the public purse – the matter of Zuma’s involvement had been settled.
Calls to release the full report
Naturally, not everyone was mollified. The department of public works soon started to receive letters from organisations not content with receiving a summary of a report. After all, the credibility of the exercise was questionable. The report was commissioned by Nxesi, the terms of reference set by him and his cabinet colleagues, the investigators were picked by those same ministers, and their identities were never revealed.
Nonetheless, a number of organisations badly wanted to see the actual report, as opposed to a summary presented by Nxesi. Among those who started making legal demands for the whole thing were amaBhungane (the M&G Centre for Investigative Journalism), Corruption Watch (the body established with the help of the ANC’s alliance partner Cosatu) and, of course, the Democratic Alliance.
Though phrased differently, the demands all came down to the same thing: Nkandla is a matter of public interest, and the public has the right to see the inter-ministerial task team report. Give it up.
At which point Nxesi became the minister who said “no”, repeatedly and in an imaginative number of ways. amaBhungane simply received no response, despite a legal obligation to respond. The DA was told that Parliament would be dealing with the matter, so demanding a copy of the report was “premature”, a response not made provision for anywhere in law.
But questionable legality aside, it appeared that only the courts could force the release of the report, and that would take many months. Meanwhile, the report did go to Parliament, after a fashion, and there it had its second coming.
It took four months for the Nkandla report to make it into parliamentary proceedings, four months of vicious fights and furious negotiations. The report was, apparently, secret, but was vital for the legislature in its oversight of the executive government. Whatever was to be done?
A report on the report
The solution painfully arrived at was this: the report would be put before the Joint Standing Committee on Intelligence (JSCI), a joint committee between represented parties that deals in strict secrecy with matters that require it. Members of the opposition would get to see the report, but speaking out of turn about the work of the committee could come with a jail term.
So it was that, in June 2013, the government report was finally tabled in Parliament, and whisked off to be dealt with in secret. It would take another five months, but in November 2013 the JSCI released a report on the report, and it had its second incarnation.
This time the finding was buried in 40 pages of analysis rather than trumpeted again and again by a minister, but it was still unequivocal.
“The committee noted that the task team found that there was no evidence that the department of public works paid for the construction of the private houses of the president,” the JSCI said in its report on the report.
The second incarnation, though, was a lot more detailed. The JSCI dwelled on the qualifications, integrity, trustworthiness, reliability, expertise, honesty and professionalism of the anonymous task team. It described the terrain in Nkandla and the fraught political history of the area. It also landed a few new bombshells.
“Residents within rural communities are generally considered soft targets of crime and the family of the president may be reviewed as an even easier target for retribution on different political viewpoints,” the JSCI said in a section dealing with the reasons why a massive security upgrade was called for at Nkandla.
“Rape is reported as being particularly on the rise, among other crimes. This is but one of the factors that necessarily place an obligation on government to tighten security even more around the residence of President Zuma.”
Other threats the president had to be protected from included earthquakes, floods and political violence.
More intriguing, but more subtle, were other new details, the full import of which were not immediately evident. The JSCI spoke of the hazard of fire and a resulting “water reservoir”, and a piece of ground stabilisation engineering known as “the amphitheatre”.
There was much that was tantalising, but in the spirit of the striptease that Nkandla revelations had become, all would not be revealed.
“Unfortunately declassification of the report does not lie within the authority of the JSCI nor does it lie within the authority of the task team,” the committee said. Parliament had spoken: the government’s Nkandla report was to remain secret.
And secret it remained, until the ANC decided otherwise. On December 3 2013, the ANC said that the task team report – not the JSCI report about the report, but the actual document – should be released. Urgently.
Or, as ANC secretary general Gwede Mantashe put it, the ANC was asking nicely.
“We call on the full report and not findings of [the] inter-ministerial task team should be made public,” Mantashe told a press conference, speaking on behalf of the national executive committee of the ANC.
It was an extraordinary call, leaving ministers tasked with security – and maintaining secrecy – with an invidious choice. They could either reverse their previous position utterly, and be seen to be doing so because the ANC demanded it, or they could refuse a “request” from the top of the ruling party that had put them in office.
All is revealed
Mantashe and the ANC were not acting on a whim, of course. Just days before that strange “request”, the M&G had published details of a draft report the public protector was crafting on Nkandla, and from the point of view of a party preparing for elections, it was a shocker.
The public protector, the M&G said, would find that Zuma had derived undue benefit from the Nkandla upgrade, and that he should repay some of the money to the state. That article had also revealed the existence of the now-infamous “fire pool”. The country was buzzing about Nkandla. The ANC needed the government report, the one that exonerated Zuma, in the public domain, fast.
And fast it was. Just two days later, the cabinet announced that it had perused the report, “endorsed the recommendations and directed that the report be released to the public”.
Why could the recommendations not have been endorsed nearly a year before? Why so many refusals to release the document, such insistence that it contained sensitive security details, such determination to handle it behind closed doors – only to finally release it?
The suspected motive, of course, was to gainsay the public protector’s report, but officially the time had simply come.
And so the government’s Nkandla report had its third incarnation, on December 19 2013, and Jacob Zuma was once again declared innocent. This time it came at the end of a list of conclusions, but the words were identical: “No public funds were used to build the private houses of the president.”
Zuma had not ordered the improvements, the country was told for the third time, had not been involved in the details, and had not even accidentally unduly benefited from the work.
“President Zuma did not ask for security installations,” Nxesi said, in an eerily familiar summary of the report.
“As per normal procedure, SAPS [the South African Police Service] and department of defence conducted a security assessment, as per their mandate. As it will be shown in the findings of the task team report, no state funds were used to build the president’s private residence.”
That was not the end of it, of course. The task team report, first summarised in January, then described in November, was supposedly released in full in December. There were no redactions, no pages missing, no gaps in the numbers, no bold black lines through details of just how Zuma was being protected. If this was the report that had to be kept secret, then a cabinet minister and a parliamentary committee had lost themselves in paranoia.
The only other plausible explanation was that the report released in December 2013 was not, in fact, the unadulterated item.
“The DA is not satisfied that the document released on December 19 2013 is the actual report submitted by the task team in January 2013,” said the opposition party, on filing yet another demand through the courts for access to what it considered the genuine report. “It would ... appear that the report has been rewritten in a way that hides the nature and extent of any exclusions.”
Not so, said the department of public works in a court filing, in February 2014. The January 2013 report and the December 2013 report were, in fact, the same report. There was only one report. There had only ever been one report.
Opposition groups, most notably the DA and AfriForum, reacted with what could only have been mock horror, or at the very least gave an overly dramatic portrayal of confusion. If the full report had already been released, and it clearly contained no security-specific information, why had it been classified and handled in such strict secrecy and generally been so determinedly withheld from public view? Was it possible – and say it wasn’t so – but was it possible that the ministries involved had merely used the excuse of secrecy to shield the president from possible embarrassment?
At the time of publication, Nxesi was still facing the consequences of these questions, phrased as a criminal complaint in one case, and as charges that Parliament had been misled in another version. So the government’s Nkandla report may not have a fourth coming – but the consequences of its pronouncement of Jacob Zuma’s innocence linger on.
5. Key point
Nkandla has seen many an awkward press conference filled with convoluted logic and self-defeating arguments, and cabinet ministers doing hairpin turns while claiming to be on the straight and narrow. But State Security Minister Siyabonga Cwele seemed hardly to be trying at all when, on November 21 2013, he utterly outclassed the competition and took the prize for the most stupid thing ever done in defence of the presidential homestead.
“No one, including those in the media, are [sic] allowed to take images and publicise images even pointing where the possible security features are,” Cwele told a regular post-cabinet media briefing, with reference to pictures of Nkandla.
Had the minister misspoken? Would he later amend and expand his remarks? Indeed not. Under extended questioning he insisted that publishing pictures of Nkandla was a criminal trespass against the National Key Points Act, and that the media were being requested – not yet threatened with prosecution, but definitely asked – to refrain from doing so.
The media were not in a giving mood. “So, arrest us” blared the front page of the The Times the next day. “The picture the state does not want you to see,” said the Cape Times. “Look Away: what ministers don’t want you to see,” read The Star.
Social media was well ahead of the outrage curve. Almost as soon as Cwele had made his call, South Africans started publishing photos of Nkandla, with everything from satellite images to children’s drawings thrown in for good measure. The scent of insurrection was in the air.
The call was never repeated, and no action was ever taken. It turned out the media had indeed somehow, mysteriously, misconstrued the instruction, a government statement the following day said.
“Government has no problem with the media publishing pictures of national key points, including President Jacob Zuma’s Nkandla residence, as it is part of their daily line of duty,” read the statement, in contrast to what had been said less than a day before.
“However, zooming into safety and security features of national key points is a challenge as it compromises national security.”
It was the National Key Point Act (NKPA) that gave rise to the great picture insurrection of 2013.
The NKPA is a piece of apartheid legislation that was introduced in 1980 in response to struggle sabotage; it gave vast and sweeping powers to, at the time, the defence minister. Once a building or installation is designated a key point, the law says, a minister (initially of defence, later of police as structures shifted, but always somebody in the security cluster of ministries) may “take any necessary steps which, in his or her opinion, may be necessary for the security of these areas in relation to third parties”.
Though blatantly unconstitutional in a number of respects, such as its criminalisation of free expression and delegation by Parliament of law-making powers to a minister, the law survived a 2007 attempt at amendment.
Law and liability for security upgrades
On Nkandla, the NKPA has been cited as a reason for massive spending, on why that spending could not be confirmed, why details of how that spending was allocated cannot be discussed, and why pictures of Nkandla are illegal. Through a slow process of attrition, those claims have mostly been rolled back.
Yet the invocation of the NKPA as the Nkandla temperature rose during the course of 2012 brought new questions.
The law calls for the owners of designated properties to be held liable for the cost of security upgrades, with the state only to dig into its own pockets where absolutely necessary.
If tax money is spent on a key point, the law says, money must be drawn from a special account. That account may, or may not, actually exist, but it is well documented that money for the Nkandla project was drawn from a range of other accounts, primarily “prestige” projects (dealing with housing for cabinet ministers and above), conceivably making the project outside of the law.
But the NKPA wasn’t the only officious defence brought into play.
The now infamous Ministerial Handbook quietly became an official document in February 2007, when it was approved by cabinet. In the following years it would be used primarily as a way to excuse government spending on flashy cars and expensive hotel stays in the formulation “it is allowed by the Ministerial Handbook”. That is exactly how Public Works Minister Thulas Nxesi dealt with questions around Nkandla, before, during and after the NKPA defence was invoked.
“I would like to state categorically that everything that has been approved and carried out at the private residence of the current president is in line with the Ministerial Handbook as far as it relates to security arrangements for private residences of the president,” Nxesi told the City Press in October 2012.
The beauty of the handbook defence was that it was considered a classified document, so, though cited, it could not be examined.
What Nxesi failed to appreciate at the time was that the M&G had obtained a copy of the secret handbook, and published it, in April 2011 (to much castigation from the government). And the handbook was clear: state spending on the security of ministers at their private residences was capped at R100 000.
Nxesi and others would subsequently explain that the R100 000 cap did not, in fact, apply to presidents, and that the handbook was never, in fact, the basis for spending at Nkandla.
So all that was gained by reaching for the handbook defence was to highlight that spending on Nkandla was 2 000 times more than what was considered adequate for the tier of national leaders just one step below Zuma.
Other attempts at obfuscation have been equally ludicrous. In a mem- orable, televised press conference in October 2012, triggered by City Press first pegging the Nkandla costs at over R200-million, Nxesi said he had launched an internal investigation to discover who had leaked that information. Once found, those culprits would be jailed, added his acting director general Mandisa Fatyela-Lindie. She could not disclose the cost because then she herself would go to jail, in terms of legislation such as the Defence Act.
At the last check neither official had been jailed, or even questioned by police.
5. Opposition on the march
Opposition parties love Nkandla with a burning yet enduring passion. They love it for its popular appeal in a country where scandals come and go: Nkandla fades only to come back even stronger. They love it for its close personal link to Jacob Zuma, playing the man rather than the issue, or the party. They love it for the whiff of opulence amid squalor it carries, for the hypocrisy it evidences, for its intertwining with hot-button issues such as state secrecy, rural development and wasteful government spending.
And on their part, Nkandla’s many suitors have been true in their affections. The DA, the Congress of the People, the Freedom Front Plus, the Economic Freedom Fighters (EFF), all have assiduously wooed Nkandla, contributing to the unravelling of the story or just adding their own flavour to the proceedings with marches, events and oh-so-many speeches and statements.
Yet none have managed to make Nkandla their own. For the EFF in particular, its dalliance with Nkandla has had, at best, mixed results.
Thanks to its position atop a gentle hill, the Nkandla complex looks out over a scattering of other, far more humble homes. Some are pleasantly welcoming. Others are less so: crumbling mud huts with leaking, patchy roofs, cold in winter, baking in summer, and altogether in sharp contrast to the luxury that lies just beyond the two security fences that separate the presidential compound from its neighbours.
To one such mud hut, as the story was later told, a woman named S’thandiwe Hlongwane took leader of the EFF Julius Malema.
Malema’s party, it would subsequently emerge, had noticed her very humble dwelling in the foreground of pictures of Nkandla, and on closer inspection had confirmed it represented an opportunity.
“The mud houses did not have furniture, no proper doors and did not have beds, and children were sleeping on the mud floor,” the party said in a statement, after the fact, about the conditions it had found.
For all its failings, the house had enviable geopolitical positioning; it stood just a few hundred metres from the Nkandla fence and, thanks to a fortuitous lay of the land, it could be photographed with the Nkandla complex in the background, unobscured by the acacia trees that dot the area. In the lingo of the EFF, suddenly turned political estate agents, it was “next to the museum of corruption”.
As a place to make a statement it was, in fact, ideal. So the EFF built a house, with red berets passing bricks hand to hand and wielding paint rollers in what could have been a heart-warming sort of way were it not for the faintly menacing political overtones, the threat of imminent violence hanging in the air.
The house the EFF built was not anything special, certainly nothing on the level of the high-walled home Malema once had in Polokwane (until seized by the South African Revenue Service) or the mansion he had partly built in Sandton. It had simple lines, a neat tile roof, and startlingly white walls – all of which showed up nicely on photographs.
What the EFF house lacked in size or noteworthy features, though, it made up for with exquisite timing and symbolism. The final handover to the presumed-worthy recipient just happened to be on the same weekend as the ANC launched its 2014 election manifesto, drawing much attention away from what should have been the ruling party’s news cycle.
It also came shortly after Zuma had handed over the Mala Mala game reserve to land claimants, a little exercise that cost the government R1-billion; the contrast between a land reform policy in deep trouble and a humble home to a single family could not have been greater.
The handover also created an opportunity for the EFF to show its mettle. The ANC may not approve, the newcomer party said, it may try to defend its turf, as it is wont to do – but the EFF would not abide no-go areas.
With the stage so set, and the house nearly complete, Malema, a group of supporters, a great many well-armed police and a sizeable media contingent descended on Nkandla on January 11 2014 to put “final touches” on the donated home and officially hand it over. They found an ANC welcoming party just begging to make headlines for all the wrong reasons.
The tacit understanding among ANC supporters around Nkandla, if not direct party orders, was that Zuma and his reputation required protection. So the faithful turned out to deny Malema his publicity stunt, through the simple expedient of denying him access.
But, of course, Malema is no stranger to street politics.
The small but vocal ANC crowd blocked the convoy on the road.
Malema got out and proceeded on foot. Police on the scene, choosing between the best of bad options, quickly formed up around him, pro- viding a perfect image: a leader once lauded by the ANC, including by Zuma himself, requiring a heavy police presence to keep him safe from ANC members while trying to do a good deed.
Malema, meanwhile, got in the soundbites. “Any African who stays in a beautiful house like that and his neighbours stay in shacks must be ashamed,” he said, looking stern but unconcerned for his own safety, though somewhat straining the buttons of his stylish shirt.
It would only get better, from the point of view of the EFF, in fact if not in imagery. With their area denial plan foiled, the ANC group resorted to hurling objects at the EFF group and its leader, bringing out the rubber bullets, tear gas and a water cannon – and, by the police count, the arrest of 30 ANC supporters, but not a single one of the EFF group. For the EFF, trying to build a reputation for fearless militancy but with utter discipline, this was an unearned bonus.
The good news continued, as Zuma and his deputy Cyril Ramaphosa both condemned the ANC-linked violence in the following days. And then it all went badly pear-shaped for the Fighters.
“Malema Nkandla stunt backfires – family had two others”, declared the January 26 edition of the Sunday Times, explaining that the recipient of the house, S’thandiwe Hlongwane, “is married to a senior state official who earns about R250 000 a year and owns at least two other homes”, plus two cars, and other accoutrements unusual for those living in destitution.
Malema was left to explain that he had been convinced Hlongwane was unmarried, and he was ridiculed for empowering those who did not seem to need the help all that badly.
The EFF tried. It disputed some facts, confirmed others, decried media sensationalism and solemnly shook its collective head at those it said were trying to distract it from a duty to improve the lives of people living in mud huts. The damage, though, was most thoroughly done. The house the EFF built had gone from a clear political victory to a muddled, confused, arguable draw.
But the EFF had at least learnt a thing or two from a similar headline-grabbing attempt by the DA two years earlier – one that saw the DA likewise achieve, at best, mixed results.
DA leader Helen Zille had found a renewed sense of purpose, or at least vigour, about Nkandla in late October 2012, issuing demands for information, making strong statements, and scheduling an “inspection” of Nkandla. She was not there to build a house, but she was intent on finding the truth, as she put it.
What she found instead was a determined thin blue line, and a large, organised mob.
For the Zille visit, on a sunny Sunday in early November, the ANC was well organised. The crowd of several hundred alighted from buses with printed posters declaring “President Jacob Zuma has a right to privacy” and shouting “Stop attacking Zuma!!”. Zille, on the other hand, had a hat.
It is that hat, mostly obscuring her face, and a waggling finger that made one part of the most enduring image of the day. A line of unimpressed police, with which Zille was remonstrating, made up the other part of the image.
Zille’s anticipated, scheduled convoy had been met by police along the way, and escorted towards Nkandla. But not far from the homestead was another group of police officers, beyond which the ANC crowd did their thing. The second contingent of police were resolute: allowing the DA group any further progress would create a dangerous situation.
In this they were probably understating the fact. In the preceding week various ANC structures and branches had used words such as “provocation” in threats of violent retaliation for the “inspection” that were not so much veiled as paraded. Again there was never any evidence of formal orders, but much evidence that members of the crowd knew what was expected of them – as, apparently, did police intelligence.
Zille, and her hat, lost the altercation; her attempt to approach Nkandla on foot failed, where Malema would later succeed. Instead of triumphant soundbites, she was reduced to the equivalent of verbal foot-stomping in front of the cameras. She promptly took the fight to the local police station, where she insisted on laying criminal charges.
That, as it turned out, was another mistake. The ANC crowd also relocated to the police station, with the result that Zille and her small entourage were rushed away under police guard, in what seemed like nothing so much as an ignoble retreat under fire. Or, to put it differently, the direct opposite of portraying the official opposition as feisty, crusading searchers for truth.
6. Who owns Nkandla?
Who owns Nkandla? It is a deceptively simple question considering the implications of the answer. If the land belongs to the Zuma family, then every cent poured into it will ultimately benefit that family in one way or another.
If it belongs to the state, then Jacob Zuma and his family may face expulsion at some future point, at the whim of whatever administration is in charge at the time.
If it does not belong to the state, the entire Nkandla project is at least highly irregular, and heads should certainly roll. If part of the land belongs to the state, then everything may be above board, in a sense, and with a sufficiently apologetic interpretation of events.
All of which assumes that ownership itself is a simple, binary state: you either own something or you do not. When it comes to Nkandla there are very many shades of grey involved, enough to keep entire teams of lawyers in caviar and cigars for years.
Because, ultimately, Nkandla is best described as belonging to nobody in particular, and to lots of people in principle.
And that is largely the fault of apartheid.
Ingonyama Trust land
As the Constitutional Court summarised the situation in a March 2013 ruling which was unrelated to Nkandla, successive apartheid governments first set aside portions of land for the use of black Africans and in effect confined them to those areas, then declared those areas to be tribal trusts, and later turned the trusts into nominally self-ruled homelands.
In what is now KwaZulu-Natal, the Zulu homeland was sprawling, making up around a third of the total of the province. In the fullness of time, that homeland again became a trust area, run by the Ingonyama Trust, and it is on Ingonyama Trust land that Nkandla stands.
Not that Ingonyama owns Nkandla, not entirely, and not outright, and, in a way, perhaps, not at all. Ingonyama is a complex, sometimes confused entity, one with many eccentricities all of its own.
And that is largely the fault of the democratic transition.
April 1994 was a turbulent time, with the painful transition seemingly in peril on multiple fronts. One of those fronts was represented by the extraordinary brinksmanship of the Inkatha Freedom Party (IFP), a powerful force in KwaZulu-Natal at the time, as well as a powerful enemy (sometimes on the level of bloody confrontations) of the ANC.
Facing a certain and almost certainly massive electoral defeat, the IFP refused to take part in the elections of April 27 1994, reversing its positions only days before ballots were actually cast. The reversal was finessed through concessions and guarantees about the protection of traditional leadership in KwaZulu-Natal, with a modicum of self-rule for the Zulu people.
Land, clearly, needed to be part of the equation; the Zulu homeland would in effect be under government control after the elections, and the government would be under ANC control. So the Ingonyama Trust was born, a structure that would have outright ownership of all Zulu tribal land, and which would in its turn be solely controlled by Zulu King Goodwill Zwelithini on behalf of his people, with a sort of hazy, opaque oversight by Parliament.
There was no time to bother with the details, and insufficient consensus to go beyond broad strokes anyway. All concerned knew it was a sketch solution, but at the time the political horizon was only days, and sometimes only hours, into the future, and the fix was to be temporary, transitional.
Ingonyama helped to get the IFP on the ballot, making the elections far more legitimate than they would otherwise have been, and arguably staved off continuing bloody violence in a province that had seen far too much already.
The underlying legislation was amended in 1997, and the political reality changed, and new processes and interpretations arose, and by early 2013 the Ingonyama Trust itself argued that it was, basically, an organ of state, the equivalent of the English king in the days of “crown land”, and that its vast holdings in KwaZulu-Natal – including the land on which stands Nkandla – are to all effects and purposes state land.
The reality is nowhere near that simple. Thanks to a tangle of laws and constitutional duties, messy definitions and tenure obligations, the state would find itself in very deep waters should it try to exercise its ownership by selling any portion of the Ingonyama Trust land.
The same goes for the trust itself, and anybody who lives on such land would be laughed right out of the Deeds Office should they try to transfer ownership of houses they, in many cases, built with their own hands or have lived in all their lives.
Instead of ownership, Ingonyama provides long-term leases on property it owns, typically for 40 years, with the option to renew for another 40 years. Obtaining such a lease is by no means a simple task, and requires the consent of local traditional leaders plus, in theory, the local community, and transferring such a lease to another person (whether in exchange for money or not) borders on the impossible. Transferring the lease to, say, a bank, as security for a loan, is not so much impossible as just laughably legally complex.
Contemplating such mechanics became something of an unexpected priority, though, when President Jacob Zuma stood up in Parliament on November 15 2012 and declared: “I engaged the banks and I am still paying a bond on the first phase of my home ... Those rondavels were paraded accompanied by lots of allegations. Yet I am still paying a bond to this day.”
It was a curious statement, for a number of reasons. For one, it was an entirely new defence to the swirl of allegations and questions that had been around for nearly three years. It would explain, well, just about everything, in a fashion documented and cross-referenced to the satisfaction of even the most nit-picky. Yet it had never been brought up.
The bond explanation also suffered a further credibility problem, considering that it was not in keeping with the other facts on the table. Some things did not quite fit, others did, but in curious ways.
At the time, Zuma’s former financial adviser and funder Schabir Shaik had already been convicted of corruption. His trial had brought to light a considerable stash of documents, some of which related to Zuma’s financial affairs. Trace those through court papers to an original indictment and a related report, and you’ll discover references to a home loan for Jacob Zuma, which became germane because not all the payments on it were made by Zuma.
Look a bit further and there is evidence that the granting bank, First National Bank (FNB), had been well aware of the political implications of the transaction it was contemplating, as it evaluated the Zuma loan application.
On the other hand, just two weeks before Zuma made his statement in Parliament, amaBhungane had received a response to a set of questions submitted to the Ingonyama Trust, part of a tenuous thread of evidence that held the promise of explaining the relationship between the state and the Zuma family and their various land holdings.
Here is what Ingonyama said about control of land:
“In the case of the Nkandla Presidential Complex there are two leases for two portions of land and appropriate land use, there is a portion that is exclusive to the Zuma family and the other that of Public Works.”
So far, so good. Right from the start of the Nkandla saga the government has claimed that it was building what is in effect a little village for members of the security forces that abuts, but is not of, the Zuma homestead.
The government had, from the start, entirely neglected to mention just how close the security village and the family complex were, or how much work had been done by the state within the family complex as well, but two separate leases was just about right.
Then, however, things got weird. The Ingonyama Trust also provided the dates it had on record for the two lease applications. The Zuma family application was dated September 2010, the state lease application January 2011. But the Nkandla project was well under way by the end of 2009.
So two remarkably curious things appear to have happened. First, Jacob Zuma had funded an upgrade of his family residences using a loan that was guaranteed by a property that he not only did not own, but at the time also did not even have a formal lease for.
Secondly, the state budgeted, planned, and actually started building what it continues to maintain is state infrastructure, on land it likewise did not own – well, not directly – but definitely did not have the formal right to use.
The use of Ingonyama Trust land without a formal lease is common; informal, sometimes simply handshake land-use agreements between residents and local chiefs are the historic norm. That has been changing over the years, however, because of a concerted push by the trust itself.
One of the reasons for that push? Because a handshake agreement is not the kind of surety banks base loans on, and commercial investors (and, hopefully, the state) are more than a little loath to pour capital into projects on land to which they hold no formal rights. To allow the wheels of finance to turn, formal leases are essential.
After Zuma’s statement in Parliament, things got all the more complicated. Several careful searches of deed records showed no bond registered over the Zuma property. FNB, the bank in question, rejected the notion that it would grant home loans for trust-land property.
The presidency, on seeing reports to this effect, confirmed that Zuma had not misspoken, insisted that there was a bond, lambasted the media for intrusive questions, and refused to provide any further information.
The country’s banks, meanwhile, including FNB, cited customer confidentiality when asked whether they held any presidential loans.
And there, for the time being at least, the road came to an abrupt end. There was no more information to be had, no further clarity forthcoming and, at that point, no investigative agency with both the power to demand documents and answers and the ambition to get to the bottom of Nkandla.
So, who owns Nkandla?
If you define ownership as the ability to sell, then nobody in particular. But if you take a more liberal view of ownership, considering control and future usage, lots of people in principle.
In one sense Nkandla is owned by Ingonyama, which technically acts on behalf of all subjects of the Zulu king. In another sense, part of Nkandla is owned by the Zuma family, and part of it by the state.
There is a related sense in which the state owns all of Nkandla, including the Zuma complex. In yet another sense – according to the president – Nkandla is at least part owned by a bank, or will be if mortgage payments fall far enough behind.
But just what that tangle means in terms of broken rules and regulations, undue benefit and irregular spending, that is a harder question altogether, one that the government investigation into Nkandla answered not at all, and one that the findings of the public protector, who was asked to investigate the matter at the end of 2011, may come to define.
7. The racist compound
It must have seemed a good idea at the time, in the heat of battle, as it were. “She’s called the president’s house a ‘compound’, a word used for hostels and migrant workers,” said Mac Maharaj, the former presidential spokesperson and firebrand, during what amounted to a live debate with Democratic Alliance leader Helen Zille.
“She’d never use that for a white person’s home.”
It was an absurdity he would come to regret.
Why Maharaj decided to go on air on Talk Radio 702 alongside Zille in early November 2012 is in and of itself an interesting question. In the years before, and in the time since, Maharaj, the presidency and the government as a whole had been largely unavailable for questions, never mind debate. Government departments refused to provide information, politicians dodged questions, and Jacob Zuma ignored the issue altogether.
But by the last quarter of 2012 the strategy of ignoring the issue was wearing thin. It was time to fight back – not with facts and figures, but with emotion.
On November 5 2012, Maharaj took to the airwaves to decry Zille and her party as racist. Ten days later, Zuma expressed much the same sentiment (though far more obliquely) during a memorable question session in Parliament when the president spoke, in a tone alternating between outrage and disappointment, about how aggrieved he felt, and how he took exception to some of the questions being asked about Nkandla.
“People are speaking without knowing, saying I have spent so much of the government’s money. I have never done so,” railed Zuma. “It is unfair, but I do not want to use harsher words, because you believe that people like me cannot build a home.”
It was a new spin on an old approach. Over the years various government departments, the ANC, and the presidency itself, had some success – mixed though it was – with the shoot-the-messenger approach: Nkandla was a creation of a tjatjarage (yappy) media.
Coverage of Nkandla was born into antagonism three years before racism came into it directly, and the approach rarely wavered, although the details changed from instance to instance.
The media were out to get Zuma, embarrass him, weaken the ANC, support the political opposition, were disrespectful of traditional values, urbancentric, lying, misrepresenting, misunderstanding, breaking various laws, and generally acting in a fashion unbecoming.
As evidence mounted, some of the specific accusations fell away, but the underlying theme continued in statements to Parliament, in media statements, and in speeches: on Nkandla, the media were being unfair.
Now, however, the media and the opposition were being racist, or – should one strain to interpret Zuma’s words as not dealing with race – dismissive of a man from a rural backwater without much in the way of formal education.
In hindsight, it was an approach doomed to failure.
Three years into the story, Zuma’s primary defence was still ignorance, and that was starting to look mighty disingenuous. In his parliamentary answer, Zuma again said he did not know how much various aspects of the Nkandla project had cost, that he simply did not know where the money had gone, just shy of three years after the first questions had been raised.
As an academic paper on the coverage of Nkandla would put it in early 2014, to maintain such ignorance “over the time span of the coverage is scarcely becoming but also unconvincing. Why, it must be asked, did [Zuma] not address his ignorance in order to make a public statement consistent with the responsibilities of his office?”
Bringing race into the mix would, in and of itself, also be neither becoming nor convincing. Adding race on top of ignorance just compounded the mistrust.
Timing and context aside, choosing “compound” as the exemplar of racist motivation was also ill-advised, as Maharaj would soon discover.
Had it remained just a passing comment, made off the cuff in a medium that is more transient than most, the “compound’ debacle would not have been all that embarrassing to Maharaj and Zuma. Then the South African Broadcasting Corporation (SABC) came to the party.
A memo issued to staff by SABC head of news Jimi Matthews, shortly after Maharaj decried the term, and leaked almost immediately, stated:
“You are hereby notified that, with immediate effect, President Zuma’s Nkandla home should be referred to as the President’s, or Mr Zuma’s, ‘Nkandla residence’, and not a ‘compound’ or ‘homestead’ or any other such term.
“Please also refrain from using imported terminology in reporting on the controversy surrounding the infrastructural developments around the residence, such as ‘Nkandlagate’, ‘Zumaville’ and such like.”
On “Nkandlagate” Matthews had a point. In analysis and interpretive journalism it is great, and entirely justified shorthand, but for a news organisation that has old-school ideas about objectivity it could be troublesome.
But on “compound” and “homestead”, it looked as though the SABC was taking marching orders from the presidency, letting Maharaj police its language.
With political influence invariably suspected, and over the years occasionally plainly observed at the SABC, the combination was unfortunate.
The SABC, like just about every news outfit that ever touched the issue, had used “compound” in many instances for many years before November 2012 to refer to Nkandla.
So, a quick review of government documents showed, had the government itself. The department of rural development and land reform had not shied away from using the term when speaking about clusters of dwellings, the department of public works had happily referred to the “security compound” that forms part of Nkandla, the term had been used in Parliament to refer to Nkandla, as well as to high-end gated communities, and so on and so forth.
If Maharaj had expected sympathy from the public, or self-censorship from the media, he must have been greatly disappointed.
In comments online, letters to editors, in calls to radio stations and idle conversation far and wide, there was either outrage at what was seen as a cynical ploy to divert the Nkandla debate, at the SABC’s pliant attitude, or just general mirth at such a ham-fisted bit of spin.
And a few months later, even the ANC started quietly reintroducing the phrase “Nkandla compound” in its daily roundup of news, as the communications strategy shifted from ignorance and victimhood to innocence – an approach that would not be without its bumps and false starts either.
8. Nkandla files
A project such as Nkandla generates vast quantities of paperwork, veritable mountains of memos, letters, contracts, invoices, updates, lists, statements, plans. From these documents a skilled investigator can extract an entire narrative, and find at least the starting points for further questions that can unearth skullduggery of all kinds.
In the case of Nkandla the most important of those documents, or at least a representative sample, passed through the official channels of the department of public works, which was in overall charge of state spending on the project. And as deficient as government bureaucracy in general and public works in particular (by its own admission before Parliament on charges of inefficiency and worse) may be, hanging on to paper and email should not be all that hard.
Getting hold of it, or any subset of it, is a different matter.
Enter the Promotion of Access to Information Act (Paia). Passed in 2000, and thoroughly tested in the courts since, it set out to reverse apartheid secrecy and “actively promote a society in which the people of South Africa have effective access to information to enable them to more fully exercise and protect all of their rights”.
Speaking inexactly, Paia allows any individual to ask anyone else – private company, state organ, government department – for information needed to exercise a constitutional right. Need to know whether the spooks are spying on you, to exercise your right of privacy? You can ask.
And though the recipient of a Paia request can demur, it must by law respond to say why it will not provide the information. Much of the law deals with mechanics or logistics, and so pre-emptively closes loopholes.
It is, by any standard, as progressive and effective a piece of law as anyone can boast.
Paia contains protections for information relating to third parties, limits on what can be requested and, from the point of view of news organisations, maddeningly long time periods for responses to requests.
However, it also puts the onus on the government to provide everything it can to anyone who has a legitimate reason to ask, and to explain itself in detail if it fails to do so.
In theory, it works well. In practice, the system is not so smooth. The M&G spent six years in various courts trying to gain access to a 14-year-old report on elections in Zimbabwe, for example, which then went missing under strange circumstances. Not everything that should be made public, even though a string of courts say it must, actually makes it into the public domain.
Nkandla, with its long-suspected corruption and over-billing, its apparent easy spending of public money, and its implications about the level of threat against the president that security agencies perceive, does not lack for reasons to demand information in the name of public accountability.
So in July 2012, amaBhungane did just that. In a carefully crafted Paia request, they formally asked the department of public works for documents relating to public expenditure on upgrades to President Jacob Zuma’s private residence in Nkandla – but only information that did not contain sensitive security details.
Security had long been used as a shield against questions on Nkandla, sometimes to a ludicrous extreme. At times various government officials had claimed that the total spending on Nkandla was in itself a security matter, as were the identities of contractors who did the work and the extent of work done.
Paia requests to government departments are sometimes entirely ignored; in some cases that is the default option, as the M&G and others have learnt through dozens of examples over many years. This time the department of public works did reply, however, but only to say that it would not be complying with the request.
It cited several pieces of legislation and regulation that pertain almost exclusively to matters of national security, implying that its hands were tied.
In terms of Paia a requester can appeal a decision internally, meaning the appeal must be lodged with the party that refused the request in the first place. amaBhungane appealed, and this time was in effect ignored, receiving only a letter to confirm receipt of the appeal.
That finally placed the matter before the courts, where it was always destined to end up.
amaBhungane said the rejection of the request had been illegal, because much can be revealed about a security project without compromising its security. Not so, said public works; the documents relating to Nkandla are so saturated with security-sensitive information that the mundane cannot be separated from the dangerous, and therefore it was simply impossible to provide anything whatsoever.
Nonsense, said amaBhungane, pointing to previously leaked documents marked “top secret” for no apparent good reason. One of the most egregious examples was perhaps a document that contained nothing but an apportionment of costs between the state and Zuma containing little but rand amounts and clearly in the public interest without a hint of security detail.
amaBhungane lined up a civil engineer who had worked on public projects to testify that an invoice is hardly a threat to life and limb.
And so it could have gone for several more rounds through several different courts over several more years. But in June 2013, nearly a full year after the initial request, the government department had a change of heart, or had suddenly been swayed by the legal argument. On reflection, public works said, it turned out that some nonsensitive documents could indeed be gathered and presented for public scrutiny.
Not long after, it provided 42 files filled with 12 253 pages of largely mundane but occasionally explosive documents. The photocopying bill for the documents came in at just a hair under R28 000, which, given the contents, amaBhungane considered cheap at the price.
Although not at all obvious at the time, mid-2013 was also when the government approach to Nkandla had become, or was starting to become, one of breaking down the walls of secrecy. That was the time that the public works department’s report was finally made public, making it all the more untenable to argue that no documents could be released.
At the same time, not releasing supporting documents was undermining the credibility of the public works report; if the findings were so clear, the tone of questions went, why is it that the evidence could not be presented too?
Whatever the reason, the results of the Paia request were spectacular. For the first time reporting on Nkandla would not be based on leaked documents whose provenance could not be proved or on the painstaking connecting of dots, but on the paper that oiled the wheels of the machine.
Budgets thrown out the window
The initial set of articles showed how government officials had fallen all over themselves to meet completion deadlines set by Zuma himself. They showed that money had been shifted from other government programmes within the department of public works to accommodate the rising price tag, while the usual budgets and cost controls for government projects were thrown out of the window.
They showed that secrecy had been such a big motivator that money not due was paid out at least in part because of a threat that legal action could expose information damaging to Zuma, netting one contractor a tidy amount.
They showed that, despite Zuma’s continued insistence that he had not been aware of the cost of the project, various officials had claimed to have received instructions from him and provided him directly with updates.
And they showed that initially R10.5-million of the project cost had been seen as being for Zuma’s account, but that account dwindled to what was, in effect, zero, without any clear reasons for this dramatic change of organisational heart.
Yet, as with all things Nkandla, the story did not end there. An examination of the Nkandla Files, much of it by way of checking cross-references, showed that many documents were missing. Most notably absent were minutes of meetings about Nkandla that featured top officials, at least two of them government ministers, and any document bearing the signature of the president.
One document that was released, amaBhungane said in subsequent legal papers demanding more documents, “makes it clear that the Principal [Zuma] issued ‘deadlines’ [and presumably other directions] to the ‘top management’ of the department [of public works] and that they, in turn, kept him informed of the progress of the Nkandla upgrade.”
What, exactly, Zuma was told, and what he instructed to be done, would be of extraordinary significance. Sadly, though, the department of public works told the same court that such documents either did not exist or could not be found.
“The [department has] done everything reasonably possible to locate all the documents related to the Nkandla security upgrade and to make those available ...” reads a response from a government official, before explaining the sheer volume of paperwork public works deals with.
“I submit that it would be unreasonable to expect the respondents to examine thousands of files because the possibility exists that some of the missing documents may be in one or other of these files.”
Will the missing documents, perhaps containing a smoking gun or two, ever be found? At the time of publication legal action demanding more documents in fulfilment of the original Paia request was still at a stage at which it was impossible to say.
It bears keeping in mind, though, that there was a time when every document in the Nkandla Files was considered secret – and today they are available on the internet for all to see.
In the world of high art and entertainment it is a pejorative term, though in newspaper parlance a little less so. A potboiler keeps food on the table, often by playing to what the public wants rather than sticking to a stricter standard of worth.
The story of Nkandla has not been all big investigative leads and major breakthroughs. Along the way there have been a fair number of potboilers, sometimes because other angles were not entirely ready, sometimes simply because the stories were amusing. If they served the additional purpose of keeping the public eye on Nkandla and the pressure on dissembling officials, well, that was an added benefit. But sometimes sharing something interesting or amusing is in itself enough.
There was, for instance, the day the M&G counted the president’s cattle.
Though that is how it worked out, the original intent was not actually to count the president’s cattle. The president’s cattle are clearly the private property of the president, and have not to anyone’s knowledge been enhanced, upgraded or refurbished using state money.
The same cannot be said for the kraal in which the president’s cattle reside overnight.
It was July 2013, and amaBhungane had just published the first set of stories that drew from the Nkandla Files and South Africa’s understanding of Nkandla was considerably enhanced, if very far from complete.
So a visit to the homestead seemed called for, to see whether the new information would provide new avenues of investigation.
But the thing about Nkandla is that one does not simply drop by for tea with the president. It is possible to approach the security fence around the compound on all sides, and the M&G and others have often done so. It is also possible to catch glimpses of work inside the compound, to watch contractors come and go, to speak to neighbours, and otherwise to ferret out information. Yet all of that amounts to nothing so much as sniffing around the periphery.
Among the documents in the Nkandla Files were references to the Nkandla kraal, a feature not visible from the outside, and previously little reported on. The kraal, it seemed, benefited from a state-funded upgrade, and rough calculations put the value of that upgrade somewhere in the region of R1-million. In other words, the presidential cattle were apparently sleeping in state-funded quarters of some opulence.
And it turned out that the presidential cattle are available for public viewing.
Every morning the cattle are led from their refurbished kraal down the slope in front of the compound through a specially constructed culvert and up the hill, skirting the edge of the compound. Before they disappear into their communal grazing area they pass several easily accessible and public choke points, in single file. Counting them and even estimating their health is simply a matter of being in the right place at the right time.
There are, of course, those who specialise in the trading of cattle, and are familiar with market prices. Provide them with enough information and they can provide an estimate of the worth of the cattle. In the case of the Zuma herd, that estimate came in at around R500 000 – or half the likely upgrade cost of the new kraal.
The presidential cattle, it would seem, were boarding in a kraal worth twice as much as they were, and a potboiler was born.
Other potboilers have been the result of impertinent questions or astute observations by members of the public. Perhaps chief among these was the revelation that the alleged chimney of Nkandla could also be its biggest fire hazard, if it was indeed a chimney, which it possibly is not.
The origin of the resulting article is somewhat baroque. An acquaintance of a man by the name of John Clarke, who subsequently outed himself in public, made an offhand comment. "The chimney on the main house of President Jacob Zuma’s Nkandla home is too short. It is a dangerous fire hazard,” he said, as Clarke recounts the story.
Clarke, in a fit of civic-mindedness, wrote to public protector Thuli Madonsela about the chimney, warning that it could present a danger to the president, and might require intervention.
The M&G would never have known about the letter had Clarke not sent it on to the newsroom. The chimney may have gone entirely unnoticed, lost in the detail that has come to swamp the Nkandla story. But a quick check of the facts confirmed that the chimney appeared to be outside the safety regulations that apply to thatched roofs, and that a more thorough investigation was required. That investigation turned up what came to be the most peculiar aspect of the chimney story: the chimney may not be a chimney at all.
Buried in the documents of the Nkandla Files, which were supposedly scrubbed of security-sensitive information, was a proposal submitted in 2009, dealing with the filtered air supply to the system of bunkers buried underneath the homestead. The fortified bunker requires ventilation, and the possibility of attack with chemical or biological agents means the air must be filtered. But where must it be drawn from? An air intake, the proposal said, could be “concealed as a high chimney at the main house”.
Perhaps the chimney is just a chimney and smoke will billow from it for many winters to come, hopefully without setting the surrounding roof on fire. Or perhaps the chimney is part of an elaborate defence system. On matters Nkandla, organs of government are loath to comment; on matters directly relating to security arrangements they nearly never do, so the truth remains elusive. Either way, the army of engineers and experts, security officials and health officials had missed a trick, and a potboiler was born.
Anyone of a betting disposition would have put their money on Moneymine, nominative determinism aside. Moneymine, more correctly Moneymine Investment 310 CC, was a solid contender as the most interesting of the many suppliers, service providers and contractors involved in Nkandla – at least some of which the state firmly believes took it for a ride.
Moneymine, after all, was there first. It was, according to the government narrative of events, appointed as a contractor sometime in 2008, right at the beginning of the Zuma family’s upgrades to the homestead. It stuck around throughout, as the project ballooned out of all proportion, and took in just shy of R53-million in revenue, about a quarter of the total spending, when an official tally was done in November 2012.
Moneymine worked in what was designated the high-security area of Nkandla, working on the buildings meant for the sole use of the Zuma family, and it built dwellings for the relocation of what the government termed “neighbours” over the years.
Just a bit of digging revealed why Moneymine may have originally come to the attention of the president, or, alternatively, what the associated benefits could be for landing a presidential gig: its sole director, Pamela Mfeka, made at least one foreign visit as part of an ANC business delegation to China, along with her husband, towards the end of 2009. How close she is to Zuma and the ANC is unclear, but documents show that Zuma was comfortable having Mfeka do work on the inner sanctum of his compound.
And, of course, Moneymine, like every other contractor involved in Nkandla, got the job by way of a “nominated process”, which is government-speak for ignoring tender procedures entirely and just appointing a chosen supplier to get the job done, with little or no (in the case of Nkandla, no) shopping around for cheaper alternatives.
Moneymine, in short, looked as though it would be the centre of attention if that attention ever shifted from government bungling to the companies that took advantage of the opportunity.
But in the final analysis Moneymine was easily trumped by Bonelena, formally Bonelena Constructions and Projects, when it came to most notable supplier.
Bonelena was the single biggest recipient of Nkandla money, paid just over R78-million (or 38% of the total spending) for the work it completed, even though all the work it was responsible for was not completed to general satisfaction, and things got nasty towards the end of its tenure.
Bonelena did a lot of the more boring work that went into Nkandla, including putting up a fence, with its most notable publicly acknowledged contribution being the relocation of the family tuck shop.
Thandeka Nene, Bonelena’s sole director, would, during the course of events, emerge as the most colourful character in the Nkandla cast.
Nene started her professional career as a teacher, only going into construction with any seriousness in 2005. By November 2008 she was successful enough to be boasting of heading for a listing on the JSE, the stock exchange in Johannesburg, in an interview with the Sunday Tribune, before Nkandla trouble made her considerably more shy of the media.
By the end of 2011 her company had won tenders from the KwaZulu-Natal government worth more than R450-million, much of it involving the construction of schools, and her penchant for luxury cars and luxury homes was being well funded.
At Nkandla, though, things were going wrong. Documents on the management of the project show many complaints about deadlines Bonelena had missed during the course of 2011. The lack of performance from Bonelena – which would later emerge to have been because of cash flow troubles at the company – was risking delay to the whole caboodle, some of the officials involved warned.
By early 2012, the company was on the verge of being thrown off the site and being denied a final tranche of payments.
Between some surprising political pull and a threat of, sadly, being forced to make the whole thing public by going to the courts, Bonelena still came out ahead, and was paid almost everything it said it was due.
Public Works Minister Thulas Nxesi got personally involved, the Industrial Development Corporation came to the party with a R10-million loan that it insists was above board but seemed awfully risky, and somehow Bonelena stayed afloat. Its methods, however, resulted in it being accused of something very closely resembling fraud.
Referring to Nene’s approach during negotiations for final settlement – and her failure to disclose that the company was in final liquidation, which meant she did not have the authority to negotiate on its behalf – a senior public works department official asked: “Can it be labelled as fraudulent?” in an internal email in September 2012.
The department ultimately did not label the negotiations as fraud, and did not press charges.
In December 2013, however, a court in the Seychelles found Nene guilty of attempting a R7-billion fraud in that country, making her a clear winner as the most interesting of the Nkandla contractors.
The Seychelles authorities found that Nene and two others had tried to pass fake papers worth €500-million on the island, an attempt that was quickly thwarted. Nene pleaded guilty, was given the choice between eight months in prison and a fine equivalent to just under R50 000, paid the fine, and was removed from the country.
Once safely out of reach of the Seychelles, Nene’s lawyers recanted her guilty plea, saying there had been no merit to the case against her, and that the entire episode had been a “political manoeuvre”, without explaining why she said otherwise under oath.
Her next brush with the law may not provide such an easy exit. Nene, Moneymine’s Mfeka, and a host of other suppliers could yet face a raft of charges related to Nkandla before the final act of the saga.
In searching for a reason behind the enormous price tag for the Nkandla upgrade, a government task team made four broad findings: the president needed protection, and that can be costly; things cost more in Nkandla because it is remote and relatively poorly serviced; the department of public works tore up, then set fire to the rules meant to contain costs; and suppliers ripped off the state.
Or, as the task team put it, “large variation orders and the high percentage spent on consultancy fees point to the possibility of over pricing and collusion” and “there is a possibility of inflated prices and overcharging”.
A giant mark-up is not in itself a crime, but faking numbers to arrive at a price could constitute fraud. Likewise, not all forms of collusion are necessarily criminal, but engaging in a conspiracy, for instance, can be.
If any Nkandla suppliers are definitively found to have played fast and loose with state money, they may find it hard to continue doing business with the government; in 2013 the ANC made a clear call for any such companies to be blacklisted, and regulations require nothing less.
Whether or not criminal charges will follow was unclear at the time of publication. What was clear was the burning ambition of those in government deeply embarrassed by Nkandla to see heads roll.
We know it was close, we just don’t know – officially – how close. Only that the attack lawyers were on standby right until the end.
“The investigation has had an unprecedented number of threats to litigate right up to the eve of the release of the report,” public protector Thuli Madonsela wrote, in usual dry fashion, in her report titled Secure in Comfort when it was made public on March 19 2014.
According to decidedly off-the-record sources, one thing that would have seen the ANC apply for an interdict against the release of the report was a finding that President Jacob Zuma had misled Parliament. That, a subset of the party had apparently decided, would require a quick trip to the courts for yet another attempt at delaying the report.
As things turned out, Madonsela did find Zuma may have misinformed Parliament – but did so unintentionally – and that seemed to be enough. After all the threats, all the drama, all the attempts to pre-emptively discredit the public protector’s report on Nkandla, it was officially published.
And it was damning.
Zuma, Madonsela found, should be held responsible for a “reasonable percentage” of that section of the Nkandla project that could serve no real security purpose, including the swimming pool (no coy references to a “fire pool” for the protector) and the cattle kraal. She did not seek to quantify that, but previous calculations by amaBhungane put the total sum he could be on the hook for at R20-million; small change in the context of the R250-million price tag, but a bundle for even the highest-paid government employee.
That would have been enough of a blow, perhaps. For four years Zuma, through his security-cluster ministries and others, had managed to dodge any such responsibility for Nkandla. Now, finally, a highly respected public office with a mandate directly from the Constitution had held him culpable.
But the Madonsela report went even further. It found Zuma in violation of the formal ethical code for the executive in his failure to protect state resources, turning what gave every appearance of being a wilful blind eye to the project, its scale and its problems. It was a breach, Madonsela ruled, the kind that Parliament is all but obliged to act on.
In terms of the findings and the action required, that was as bad as it got for Zuma, and anyone scanning the final version of the report quickly (a high-ranking representative on the legal decision-making side of the ANC, for instance) could be forgiven for thinking it safe to send into the world.
Buried in the 443 pages of the report, however, was enough dirt to bury even a president with a previously flawless reputation.
Zuma’s office, Madonsela recounted, had dilly-dallied for nine months in providing some of the information she had requested during the investigation, even though every organ of state is enjoined, in no uncertain terms, to aid the public protector in any way possible.
Oh, and Zuma had actually failed to provide all of the information required from him, Madonsela added. In particular, a line of inquiry that, with a little bit of luck, could conceivably have uncovered orders on Nkandla from Zuma himself.
There was more. In one of the more absurd strategies ever employed on Nkandla, which is really saying something, at one point Zuma’s lawyers argued that the entire investigation was null and void on the basis of a convoluted technical point.
An investigation into a violation of the Executive Ethics Code by a member of the Cabinet (such as that the president lied to Parliament or failed to look after state cash) must, by law, be concluded within 30 days, or the protector must at least acknowledge that the deadline cannot be met. What with the 2012 festive season and the complications of the Nkandla investigation, it turned out that Madonsela had advised Zuma only 48 days after receiving a complaint on his conduct that she had yet to complete the investigation into it.
Zuma’s lawyers pounced on that failure, contending, as Madonsela put it in her final report, that “non-compliance would negate the validity of the investigation”.
Put more plainly, Zuma’s position was that an 18-day delay cancelled out the entire two-year investigation and all its findings.
Madonsela’s response, which ran to 12 paragraphs, bordered on the snotty, and definitely strayed into snooty territory, and lashed Zuma both directly and in general.
“An executive members’ public accountability, in my view, cannot lapse if the 30-day period is not met,” she wrote. “I am particularly saddened by the fact that such a submission was made by the ultimate custodian of executive accountability ... The need for a speedy resolution of these matters cannot be contested. However, an approach that supports public accountability and, accordingly, the rule of law, would not seek to aid members of the executive to evade accountability.”
Note that Madonsela stops short, if only by millimetres, of saying that Zuma sought to avoid accountability. In much the same way, her report stopped short of actually calling Zuma a nest-feathering, unethical, self-centred obstructionist generally unfit to hold his office – but the facts and their presentation painted exactly that picture.
By any measure, Zuma had been eviscerated.
Yet the focus of the Nkandla investigation had not been on Zuma and his actions (or, mostly, inaction) regarding it. He was collateral damage of a sort.
Madonsela likes to have one foot in a perfect universe, where everything goes as it should.
Sure, part of her job, as she puts it, is to figure out what happened when she investigates failures of the state, the overextension of government authority, the underdelivery of what the government should provide, unlawful behaviour and maladministration.
But Madonsela considers it her primary role to match those excesses and failures against a sort of Platonic ideal, the better to figure out what should change to prevent more instances of failure. “Our job is to fix the foundation, not just the cracks,” she once said about her office.
So, woven into Secure in Comfort there is an alternate history of the project, one that shows how things could have been at various stages over the years. And it is there, rather than in her actual findings and recommendations and narrative of events, that Madonsela truly damns Zuma and his cabinet on Nkandla.
With just a little extrapolation, by filling in around the edges and bringing colour to the outline Madonsela provides, one can compare the ideal to which the president and his subordinates should have been striving with the grim reality.
Nobody involved in Nkandla at a high political level comes out of that exercise looking fit to hold office.
In an ideal world someone in the presidency, maybe Zuma himself, would have picked up the December 4 2009 issue of the M&G, taken one look at the front page and called an immediate halt to the project until some important details could be pinned down – such as who would pay for what, how much the whole thing would cost, where the money would come from and whether it really needed to be spent.
In reality, Zuma spent the next four years claiming a highly implausible level of ignorance that, if real, amounted to intentional blindness.
In an ideal world the various bits and pieces of the Nkandla project would have been distributed in a way that would best suit the long-term needs of the community rather than a single family. Security demands a formal helicopter landing site close to the homestead? Build it next to a medical facility where it could help to save lives.
The South African Police Service needs housing for VIP protectors in the area? Build those next to the police station, where they would find purpose long after Zuma (and the requirement for his protection) passes away.
Instead, tens of millions of rands in infrastructure was built far away from where it could do broader good, and ring-fenced for perpetual use by a single family.
When Madonsela asked why, she was variously told that no better sites were available, or that she did not understand the intricacies of providing security or medical care for a head of state.
In an ideal world, security for Nkandla would have been provided in the most functional, and cheapest way, not by choosing the most grand and opulent option that could be dreamt of. In this world, a last-resort bunker for the president would have been a fancy hole in the ground costing R500 000, not a network of tunnels that cost R19-million.
In an ideal world a minister of public works would not present information “riddled with inaccuracies and inconsistencies”, various officials would jump at the chance of aiding a real investigation into what went wrong rather than playing silly secrecy games and government employees would not seek to undermine the office of the public protector.
And, in an ideal world, a number of people would have spotted rather obvious problems with, for instance, appointing the president’s personal architect as the project leader for a government project, and linking his remuneration to the cost of the project, creating a perverse incentive to drive up costs rather than control them.
The public protector’s report on Nkandla was, in a word, bad. Bad enough that the M&G called for the ANC to remove Zuma from his post in an editorial published the day after the report was released.
“Given the massive challenges facing South Africa, we cannot afford another five years of a bloated and self-serving administration,” the paper held, after considering the many ways in which the ANC had sought to protect Zuma over previous years.
“The protection and enrichment of one man, his family and cronies cannot be a state priority. It is time to protect all of South Africa’s people, not only Number One.”
Others, too, thought the Madonsela report was grounds for the president to go — even to jail. Both the Democratic Alliance and the Economic Freedom Fighters laid criminal charges against Zuma on the basis of the report, in a blaze of pre-election publicity, naturally. ANC leaders branded them “charge-office activists” and rubbished the allegation that Zuma had breached the law.
In that the ANC was probably right, the public protector’s report hardly held evidence of theft or corruption that could meet the test for criminal action, and where it found breaches by Zuma himself it was in a code of conduct, not a law.
But for all the DA and EFF attempts to make political hay, the misinterpretation or misrepresentation of the report by those parties paled into insignificance next to those of the government and the ANC.
The ruling party and the executive immediately took positions so illogical, so unsound, so unfounded that the only possible charitable interpretation was that they had been provided with a fake report that held no resemblance to the real thing. (The alternative, that the government and the ANC were again reaching for absurdity in order to spin Nkandla is, of course, completely implausible.)
Both these sadly misled organisations held that, in the words of the government statement, even as Madonsela was still answering questions about her report, that the public protector had confirmed “the essential findings” of the thrice-trumpeted interministerial task team report into Nkandla.
Had this been true, it would have been a remarkable feat: an independent body would have come to the same conclusions as a government team appointed by and reporting to the very cabinet ministers most seriously implicated in the scandal; an institution with a wide mandate to investigate government excesses and failures would have made the same findings as a group that had a narrow set of questions to answer.
Sadly, it was simply not true. While both reports did indeed find a wide range of failures of process and a wild disregard for taxpayer money, the “essential finding” highlighted time and again in the government report was that Zuma himself was blameless – and beyond that lay many stark differences.
Both the ANC and the government said the Madonsela report found Zuma had not misled Parliament, apparently missing the bit where she said he had, actually, misled Parliament, but was excused for doing so because he had done so by mistake.
And both the government and the ANC held that Madonsela had confirmed there had been no political interference in the Nkandla project. Again, Madonsela had actually reported “an atmosphere that was perceived as political interference or pressure”, if one perhaps created unintentionally.
The contortions were magnificent to behold. One minute the ANC would declare the public protector was nothing like a court of law, the next it would assert the public protector should act as a court of law when dealing with delays from “respondents”.
Having more time to study the admittedly hefty report in detail seemed only to increase the gulf between reality and political interpretation. In one instance, perhaps intended as self-flagellating satire, the ANC said the Madonsela report had not found impropriety on Zuma’s part, conveying the impression the party considered an ethical failure to be proper behaviour.
All the distortion in the world could not, however, change the facts.
From the first discovery in December 2009 it had seemed that Zuma was directly, personally and unduly benefiting from the Nkandla project. Evidence to that effect continued to mount even as attempts to prove the contrary had fallen short. The report by the public protector laid out the evidence, and provided the detail – then said what had so long been suspected:
“The excessive and improper manner in which the Nkandla project was implemented resulted in substantial value being unduly added to the president’s private property.”
It was, evidently, rather hot under the glare of public scrutiny. In the first hour of his presentation police minister Nathi Nhleko wiped his face 31 times. Sometimes it was just his brow, sometimes small wipes to the right of his neck or his left cheek.
Occasionally he would go for heroic mops of the entire face, pausing ever so slightly in his delivery, but mostly it was quick, twitchy movements, complete with sundry fidgeting and scratching and complimented during convenient lulls by the tiny sips of water familiar to anyone who has ever been struck by the peculiar dry mouth produced by anxiety.
Minister Nhleko was not having a good time.
Perhaps he was struck by prescience, aware that his “Report by the minister of police to Parliament on security upgrades at the Nkandla private residence of the president” would make him a laughing stock before that Thursday in late May 2015 was out.
Or perhaps he realised that misleading Parliament was not something to be done lightly, even if his boss had already got away with it on the basis that he had made a genuine mistake.
A little more than a year before, in March 2014, public protector Thuli Madonsela’s voluminous report into Nkandla had found Zuma could have been “legitimately construed as misleading Parliament” when he said state spending on his homestead had not benefited him or his family. But in the same breath she also found that Zuma appeared to have made a bona fide mistake, which does not an ethical violation make.
Nhleko was upping the stakes. After years of lies, untruths, misstatements and misapprehensions by government officials on Nkandla, he was trying something new: he just made stuff up.
To be sure Nhleko played the old favourites, ignoring that which was inconvenient to his thesis — that everything government money paid for at Nkandla was necessary for security — and conflating into obscure complexity that which gainsaid his recommendation — that Zuma should repay the state the sum of zero for the benefit he thus had not derived.
He also, however, invented things, creating a quotation here and a reference there until his 50-page report seemed more the product of wishful thinking than the “inescapable” logical conclusion of a professional process Nhleko insisted it had been.
“This is one report that I do not have a problem with it,” he told dubious journalists. “I mean, if it has got to be scrutinised, let it be scrutinised.”
Scrutiny was hardly required.
Nhleko had been tasked by Zuma to determine how much money, he, Zuma, owed the public purse for the Nkandla improvements. In doing so, Nhleko and Zuma agreed, they would satisfy the requirements of both public protector Madonsela that such a costing exercise be done (although she had directed Zuma to work with the national treasury to get to a number) and the order of an ad-hoc committee of Parliament to much the same effect.
Four pages into the formal document he had presented first to Cabinet, then to Parliament, and finally to the public, Nhleko stumbled.
The Nkandla story, Nhleko said, had “dominated public discourse since November 2011 when the matter first appeared in the Mail and Guardian”. The publication was right but the time off by; the first article had appeared in December 2009.
Five pages later Nhleko presents a table that baldly declares Madonsela had found “no public funds was used to build the President’s house(s).
In a statement the very next day, Madonsela disagreed. “This,” her office said, “could not be further from the truth.”
Perhaps the misunderstanding came about because Nhleko was working off an illegible version of Madonsela’s report.
In between those two errors he provides page and paragraph references to the public protector report he says defines what Madonsela considered “non-security” items at Nkandla.
In actual fact the paragraphs he references deal with allegations that the President’s brother improperly benefitted from state spending (for which she found no evidence) and with the relocation of Nkandla’s tuck shop (with which she did not have a problem.)
A particularly poor reproduction, the kind streaked by too many generations of photocopies made of photocopies made of photocopies, may also explain how Nhleko missed other elements of the Madonsela findings.
His report to Parliament, the police minister repeated over and over again, dealt only with four “non-security comforts” Madonsela had found the state had built for Zuma: an amphitheatre, a cattle kraal, a visitors’ centre, and the famous firepool.
That missed the fact that Madonsela had also expressed reservations about the private medical clinic built for the Zuma family, and rather expensive paving and landscaping done at the compound.
A similarly poor reproduction of the August 2014 Special Investigating Unit (SIU) report into Nkandla could explain how Nhleko could have come to think the SIU had not dealt at all with the matter of public money building the president a home.
The SIU — which does its investigations based on a proclamation by the president — had not recommended Zuma repay any money spent on Nkandla. Instead it went after the contractors, and specifically Zuma’s architect Minenhle Makanya.
The claims against Makanya for extra work he tagged onto security improvements at Nkandla, the SIU said, had only one logical conclusion.
“It is... implicit from the claims based on the increase in the scope of works that the value of the President’s, or the Zuma family’s, residential complex was enhanced. Clearly, to the extent that these claims are well founded, the President or his family were enriched.”
On the other hand, poor multi-generation photocopies would not explain how the police minister came to make up a quote, as well as an entire new hierarchy of world wide web addresses.
When is a retaining wall an amphitheatre?
“According to Wikipedia, and I quote, ‘the walls of the amphitheatre are normally constructed in stepping dwarf walls...’” Nhleko started his explanation of why an amphitheatre in an entertainment area of Nkandla is a must, security wise.
Dubious as Wikipedia can be as an authority, the end of the quote may have bolstered his case, pointing out that “the theatre would normally have electrical points to provide for entertainment or sound system in an appropriate stage.”
Setting aside the little matter of Wikipedia being a crowd-edited encyclopaedia of sometimes highly variable accuracy, an authority that states an amphitheatre must have electricity plus an Nkandla construction that lacks such electricity would make the Nkandla construction a non-amphitheatre.
Nhleko’s quote, however, does not exist. It does not exist at the web address cited in his written report —www.wikipedia/amphitheatredefinition.com — because that is not a valid web address and does not conform to the structure of domain addresses.
Nor does such an assertion appear in the Wikipedia entry on amphitheatres current at the time of the compilation of the report or, in fact, in any of the 885 different versions of that Wikipedia entry created since the first one in May 2002. The quotation now appears online elsewhere — but only in reports on Nhleko’s report.
The same weirdness crops up elsewhere in Nhleko’s analysis of the amphitheatre, as he explains that it is actually a wall to retain soil.
“Retaining walls prevent soil and substrate from moving because of erosion of gravitational pull,” his reports quote another entirely invalid web address, http:www.soil retentionmethods.za, as saying.
At the time of the report that exact phrasing could be found on only one website: cinderblock.com, an online product catalogue of sorts for a company based in Baltimore, Maryland.
The company does not hold itself out as experts on the difference between security features and entertainment areas. “We don’t do the design, we just manufacture the blocks,” its salesperson for landscape products Ken Sperber told the Mail & Guardian.
The problem Nhleko sought to solve by elevating Sperber and his colleagues to landscaping experts is this: the amphitheatre at Nkandla was built as an amphitheatre. To turn it into a security feature, as he eventually did in his report, is difficult, apparently difficult enough to warrant heroic measures
“I noticed during my site visit that the swimming pool forms part of an area close to the Visitors’ Centre that also includes an amphitheatre, which can easily accommodate about 100 people, and a lawn area that was flattened during the landscaping process to accommodate a marquee tent,” Madonsela wrote in her report, which found the amphitheatre to be an amphitheatre.
“My general impression was one of an enormous entertainment area that has very little connection with issues of security.”
The enormous stack of documents Nkandla has generated over the years backed up her impression, speaking of a “social node”, created to host functions. And officials were ready to talk about how aesthetic considerations called for an amphitheatre.
On this seemingly solid bastion of fact the minister of police launched a multi-pronged attack.
Because the architect and many of the contractors on the Nkandla project did not have security clearance (a major blunder in itself) they could not be told what all the features they were building were to be used for, Nhleko contended.
So a contractor may think he is building an amphitheatre when, in actual fact, he was building an emergency assembly area. Security agents would then pop by later and install the secret security equipment the contractor is not allowed to know about, such as a sign that reads “emergency assembly point”.
Then there are the dimensions. If the designers thought they were creating an amphitheatre they failed, because the steps in the non-amphitheatre are 65cm high, Nhleko held, too high for comfortable seating. This presumably came as a shock for the many owners and operators of amphitheatres around the world that feature even taller seats.
Finally, the trump card. “It would be unimaginable that people would be seated in an aloe garden with irrigation system; this could not be regarded as an amphitheatre under the circumstances,
”Nhleko declared — to an accompanying picture of suspiciously juvenile aloes planted where a traditional outdoor amphitheatre would have welcoming grass.
Mercifully Nhleko did not pursue the matter of electrical points for the stage being a prerequisite for an amphitheatre to be a space for entertainment, or many an ancient Roman would have been spinning in their graves.
Cattle in a high-security zone
In contrast to the convoluted amphitheatre explanation, Nhleko’s finding on the state-built cattle kraal at Nkandla is surprisingly logical. The security system at Nkandla, be it ever so incomplete, is supposed to rely heavily on motion-detection sensors. Animals moving about would play havoc with such sensors, so the animals had to be moved from the “inner high-security zone” to the “periphery”.
“... it is clear that the President’s family, in agreeing to shift the animals to the new enclosure, made a compromise in the interest of security,” Nhleko said.
The fatal flaw with that argument is not contained in Nhleko’s report itself, but in what Zuma told Madonsela: the Zuma family’s concession on the kraal was more on the order of a demand.
“Worth noting is that you did not take the opportunity to dispute some of the evidence outlined in [the draft public protector report provided to Zuma before its publication for his comments], including the fact that by your own admission, the larger kraal was built at your request,” Madonsela wrote to Zuma in September 2014.
Madonsela’s final report referenced a conversation between herself and Zuma in which, by her account, Zuma “indicated that he would be amenable to refunding the state for the cost incurred” on the kraal.
On the visitors’ centre the police minister found the most solid ground. Who is to say the people of Nkandla should travel to a major city if they want to talk to the president, like all the other citizens of South Africa?
And why should the president not receive another head of state at his private residence, instead of making them go to one of his official residences, all of them conveniently close to major airports? And if Madonsela visited Nkandla to find the state-built visitors’ centre had just been used for a private function, “which confirmed the impression that it has little, if anything to do with the security of the President,” so what?
Dwelling on the visitors’ centre may have been a good idea, that being the only one of the four features where doubt could be introduced. Nhleko’s report dealt with it in just six short paragraphs and two small pictures. The visitors’ centre did not get a video to explain its importance, or an exercise to show how it functions.
That kind of attention was instead lavished on the firepool.
The firepool, or the “Fire Pool and/or Swimming Pool”, as the Nhleko report has it, is uniquely difficult to defend among the many lavish upgrades the state made to the Nkandla compound. It is visible from space, and more importantly, on satellite photos.
Its central location in what seems awfully like an entertainment areas, decorative blue tiling, and standard safety netting (aimed at preventing drownings while being quick to remove for use, as opposed to preventing evaporation from a long-term storage container) all conspire to make it look like a swimming pool.
And documentation throughout the course of the project to build it shows that it was intended, at least in part, as a swimming pool.
“The firepool and/or swimming pool is a strategic asset useful in fire fighting and therefore is a security feature,” he told a nation that had already been prone to giggle at any mention of a firepool, a word unknown to South African English before Nkandla, and used since its introduction more often as shorthand for tax-funded nest-feathering than as a serious designation.
The firepool section of Nhleko’s report provided perhaps the most sublime illustration of the power of self-delusion the Nkandla saga has had to offer. As the anthemic orchestral opening of the beloved Neapolitan song “O Sole Mio” rose above a video clip titled “FIRE POOL DEMONSTRATION”, journalists burst into laughter. (Watch the videos on YouTube)
A confused Nhleko was quickly assured that the laughter was due to the music, and never seemed to grasp that the joke was, in fact, him.
It cost how much?
The perfect illustration of wilful ignorance was left to public works minister Thulas Nxesi, who stoically sat through Nhleko’s three-hour reading of the report to be available to answer questions at the end. His contribution did not exactly ameliorate things.
“You see, the problem is that what was speculated by the media of R246 million, we don’t know where it came from,” Nxesi lectured a journalist during one exchange, holding firm to a total government spend on Nkandla of R216 million instead.
Nxesi was right that the R246 million price tag was an incorrect number, the actual number being R246,631,303.04, as Madonsela had explained in a handy colour-coded table in her Secure in Comfort report. The incorrect rounding down to R246 million instead of up to R247 million was Madonsela’s mistake, not a media invention.
The simplest explanation for the farce would be that a minister of police, who serves at the absolute and untrammelled pleasure of the president, may find it difficult to adjudge how much money that president owes the public pursue, and may unconsciously stretch the truth to fit a career-supporting hypothesis. But this, Nhleko said, was definitely not the case.
“Am I conflicted? ... I don’t feel conflicted at all. My approach to issues of this nature and issues of governance in general is that issues of governance do not work on the basis of whether I like you or I don’t like you, you know. There are policies, there is law and that’s what we’ve got to stick to.
“But we should also at all material times be governed by principle. In this instance for me the major preoccupation was what is the actual truth about these things. How did these things come about, what informed them, and so forth. So the question of I serve at the pleasure of the president and so forth for me doesn’t arise, it doesn’t apply as far as I’m concerned.”
Writer: Phillip de Wet
Project editor: Tanya Pampalone
Project production and layout: Laura Grant
Photography: Madelene Cronjé, Delwyn Verasamy, Chris Roper, Rogan Ward, AFP, Gallo Images, Reuters, Jackie Clausen, Beeld, Department of communications
Graphics: John McCann and Laura Grant
Timeline: Sarah Evans
Project production and layout: Laura Grant
This longform project is possible only because of the fine work by many individuals at the Mail & Guardian, and by our colleagues at other media organisations, most notably City Press. The back of the Nkandla story was broken by the relentless legal legwork of amaBhungane, the M&G Centre for Investigative Journalism. That work, in turn, is made possible by amaBhungane’s external funders. For that, our thanks to: the Bertha Foundation; the Claude Leon Foundation; the Millennium Trust; the Open Society Foundation for South Africa; the Open Society Initiative for Southern Africa; the RAITH Foundation
A longform journalism project published by the Mail & Guardian ©2015
This is an updated version of an ebook first published in 2014.