Wireless telecoms company Comsol said last week it is putting billions of rand into a wholesale 5G network spanning most of South Africa. When it’s done in a few years, it will have 2,000 base stations and national coverage.
More than 1-million Gauteng homes are already covered, with the rest of the province to follow by next year. The Western Cape, KwaZulu-Natal and the larger regional centres will be next. Comsol will sell nothing to consumers: internet service providers, banks, retailers and new market entrants will buy the capacity and keep the customers.
Read that description again. This is, in effect, the wholesale open-access network (WOAN) that the government spent the better part of a decade trying to conjure into existence, without success.
The WOAN was the centrepiece of the 2016 national integrated ICT policy white paper. Communications regulator Icasa set aside spectrum for it. The government was at one point prepared to hand a consortium that did not yet exist, and whose feasibility was contested, a slice of the most valuable radio frequency in the country.
By early 2022, the idea was dead in all but name. The spectrum reserved for it — 20MHz at 700MHz and 30MHz at 2.6GHz — still sits unassigned, and Icasa’s 2026 international mobile telecommunications roadmap writes the WOAN out of the long-term plan altogether.
Comsol’s version needed no white paper and no cabinet memorandum. It needed spectrum, which Icasa licensed to the company in 2022; capital, which Convergence Partners, Platform Investment Partners, Wimsey Capital, Solcon Capital and RMB have now supplied; and a view about where demand was heading.
Then there’s fibre. On Monday, a group of investors put fresh equity and debt into Vox and its associated companies, Frogfoot and Hypa, at a R14.4bn valuation, or R8.4bn after debt.
The largest grouping is DNI, alongside JSE-listed Sabvest Capital, Masimong Group Holdings and Draper Gain International; two Metier Capital Growth Fund III partnerships and Simphiwe Mehlomakulu’s EM-Three are the third grouping, with no shareholder in control.
The promise attached to the money is a fourfold increase in connection velocity to 360,000 homes a year within 12 months, aimed squarely at townships and lower-income households through prepaid products.
Alan Knott-Craig Jr’s fibertime, meanwhile, has gone from 10,000 homes in early 2024 to about 500,000 across dozens of townships, selling uncapped 100Mbit/s access for R5 a day, without a cent of grant funding.
There is an irony here that’s worth noting. For two decades, Telkom’s standing argument was that fixed-line infrastructure outside the leafy suburbs did not pay. That was the justification for the copper monopoly, for years of underinvestment and for a universal service fund that collected levies and delivered close to nothing.
Very large sums of private money are now being staked on the proposition that this was never true. The economics of the townships are not marginal. They were simply never attempted by anyone with a reason to get them right.
The state is not irrelevant here. But its useful contributions have been of one particular kind only: issue the licence, publish the frequency plan, then get out of the way. Every attempt to specify the outcome in advance — to design the network, name the shareholders or mandate the wholesale model — has produced delay and nothing else.
That is a lesson worth carrying across to South Africa’s electricity sector, which is now undergoing a liberalisation of its own.
The licensing threshold for embedded generation went from 1MW to 100MW in 2021 and was then scrapped entirely. Private rooftop solar has since expanded to more than 9GW and the pipeline of advanced-stage solar, wind and battery projects runs to tens of gigawatts. South Africa has gone more than 15 months without load-shedding (thanks, in part, to the improvements in Eskom generation).
Eskom’s unbundling and the creation of a competitive trading market are the next test, and the temptation will be the one that killed the WOAN: to decide in advance who may participate, on what terms and with which shareholders, before anything is allowed to happen.
What a functioning wholesale electricity market looks like is not knowable from an office in Pretoria. It is discoverable only by people with their own money at risk — which is precisely what is happening in telecoms right now, at speed, with no policy directive to thank for it.
South Africa is closing its digital divide, and nobody in Pretoria planned it. Imagine if that approach were applied to every sector still held back by the dead hand of the state.
McLeod is editor of TechCentral
