At the very moment Gauteng needed humility and honesty, it received a performance. During a now-infamous visit to a reservoir in Brixton, Johannesburg, on February 11, premier Panyaza Lesufi tried to prove that politicians “also suffer” in the water crisis by explaining that he sometimes books into a hotel “so that I can bathe”.
For residents who haul 20-litre containers home from water tankers, miss school or go to work unbathed because they could never dream of affording a hotel room just to shower, the remark was a spit in the face. It sounded less like solidarity and more like flaunted privilege.
The backlash and Lesufi’s hurried apology, in which he claimed his words had been “misinterpreted”, captured the essence of Gauteng’s water moment: not only an infrastructure failure but also a leadership failure, amplified by a communications strategy that conflates visibility with accountability.
Just one day later, President Cyril Ramaphosa delivered his 2026 state of the nation address (Sona), and the contrast was instructive. Ramaphosa opened by reaching back into history ― the 1956 Women’s March, the 1976 Soweto uprising and the adoption of the constitution in 1996 ― to argue that the nation’s strength lies in what those struggles have taught it.
Ramaphosa then turned that principle into policy on the issue consuming Gauteng: “Three years ago, when we were experiencing daily power cuts, we established the National Energy Crisis Committee … We overcame what seemed like an insurmountable challenge by adopting a clear plan and delivering on it. Using the same approach, we will now elevate our response to the water crisis to a National Water Crisis Committee, which I will chair.” In other words, the president did what the Gauteng leaders failed to do: he studied what had worked before and applied it.
That instinct, to learn from history, even recent history, is the thread that should run through Gauteng’s crisis response but does not. In 2020, the Edelman Trust Barometer Spring Update recorded something extraordinary: trust in government surged by 11 percentage points to 65%, making government, for the first time in the survey’s history, the most trusted institution globally. The barometer linked this “reordering of trust” to how authorities communicated under pressure, emphasising expertise, facts and frequent explanation. Doctors and scientists, not politicians, were the most trusted voices.
Some may note that this data is now more than five years old. That objection misses the point entirely. The value of evidence is not diminished by its age, it is confirmed by it. History is precisely where we find strategies with a proven record. If Ramaphosa can invoke the women of 1956 and the youth of 1976 to chart a course in 2026, then a 2020 trust barometer that documents, in rigorous detail, how crisis communication can rebuild public confidence is not dated ― it is a blueprint.
Much of the commentary has reached for a familiar diagnosis: these politicians need better media training. It is a tempting conclusion, but it lacks insight. You cannot coach your way out of a legitimacy crisis.
South Africans experienced this blueprint first-hand during the early Covid-19 briefings and, more recently, during the electricity recovery. Minister Kgosientsho Ramokgopa, an engineer, built a Sunday ritual around the Eskom system: which units were offline, what the maintenance backlog looked like and how much capacity was returning. Eskom’s technical leaders were brought to the fore; Ramokgopa framed decisions and took political responsibility for them.
South Africans may have remained angry about load-shedding, but they received a predictable briefing from a recognisable face who spoke the language of megawatts and visibly deferred to expertise. Today, South Africa has not experienced load-shedding for more than 300 days.
Gauteng’s water response has done the opposite. Lesufi’s “hotel bath” remark reduced a structural governance failure to a story about his personal inconvenience. Water and sanitation minister Pemmy Majodina urged patience and invoked a R400bn investment gap, almost certainly accurate. Yet, offered in isolation, it becomes a technocratic cop-out for a province staring at empty taps today. Johannesburg mayor Dada Morero sat alongside them, projecting disinterest at precisely the moment his residents needed visible urgency. The combined effect was to confirm a political ritual, not a command centre.
Much of the commentary has reached for a familiar diagnosis: these politicians need better media training. It is a tempting conclusion, but it lacks insight. You cannot coach your way out of a legitimacy crisis. The problem is not that leaders are clumsy on camera; it is that they insist on being the main characters in a story that should be led by technical experts. The fix is not more polished soundbites but a different cast list. Politicians are bad at getting things done but good at fake empathy.
The core mistake has been to treat communications as image management rather than as risk governance. Residents do not need another jaw-dropping price tag; they need to see engineers at the microphone, explaining in suburb-level detail when water will return, with politicians standing beside them, visibly accountable and visibly deferring. Technical experts not only explain why things went wrong but also fix them, and the residents of Gauteng want assurances from people who can get things done.
Trust, as the Edelman Barometer reminded us in 2020 and as Ramaphosa’s own Sona implicitly acknowledged, can be built in the teeth of disaster, but only when leaders submit to the discipline of expert-led, transparent communication. In Gauteng today, trust is dangerously close to empty. Only a radical pivot, away from political performance and towards the historically proven model of honest, technically grounded explanation, can prevent the politics of dry taps from becoming the politics of open revolt and social unrest.
Lorato Tshenkeng is CEO of Decode, a Pan-African reputation management and advisory firm
Lorato Tshenkeng
www.sundaytimes.timeslive.co.za
