Paul Colmer argues in TechCentral that the Competition Commission should refuse the application ACT has made on behalf of its members for a five-year exemption covering parts of rural network deployment. He makes his case carefully, and it deserves a direct answer rather than a restatement of our position.
Start with his strongest point. He accepts that co-location, facilities leasing, roaming and open-access infrastructure all have a place. His objection is to competitors aligning their future plans. Sharing what has already been built, he argues, reduces waste while leaving each operator to decide independently where to compete.
That distinction is real. It is also the reason we applied.
In the most marginal parts of the country, the realistic alternatives are not several independent networks or one shared network. They are shared infrastructure capable of carrying competing services, deployment that arrives years late, or deployment that never becomes commercially viable at all.
The underlying economics are not in dispute. A site must be acquired, built, powered, secured and connected to backhaul before it serves a single customer. In dense urban markets, several operators can carry those fixed costs and still compete. In a low-density area, where every operator must cross the same threshold to reach the same small revenue base, infrastructure ownership can become a barrier to entry rather than a guarantee of competition. Requiring each operator to reproduce the same costly assets before it is allowed to compete is itself a constraint on entry and expansion.
Competition, not duplication
This is where I part company with Colmer on principle rather than on detail. The object of competition policy is competition, not duplication. Operators compete on price, coverage, service quality, network performance, innovation, products and customer experience. Owning a tower or a fibre route does not confer control of the retail market.
Sharing infrastructure is a different thing from sharing customers, allocating markets or coordinating prices: the first concerns how connectivity is delivered, the second concerns how operators compete for the people who buy it. Duplicated infrastructure adds value where it increases capacity, resilience or genuine choice. It is not, on its own, evidence that a market is competitive.
Colmer’s warning is that in a thin market the first operator to build may become the only one with a rational reason to be there, and that exclusivity then arises without anyone signing an exclusivity agreement. But that argument does not hold, because it leaves out three things:
- The first is that infrastructure sharing is already a requirement rather than a favour: access on open, fair and non-discriminatory terms turns a first build into an entry point rather than a closed position.
- The second is that retail competition does not depend on owning towers. Mobile virtual network operators compete for customers on networks they do not own, and a single physical network can carry several competing providers.
- The third is that Wapa’s own members would be able to use this infrastructure – the smaller operators who have spent years making marginal areas work would gain access to sites they could not have built alone.
Arrangements that foreclose access, entrench incumbents or raise rivals’ costs should still attract competition scrutiny. Neither benefit nor harm should be assumed from sharing alone. Both have to be shown.
South African competition law already draws this line. Section 4(1)(a) of the Competition Act prohibits agreements between competitors that substantially prevent or lessen competition, unless a party can show technological, efficiency or other pro-competitive gains that outweigh that effect. Section 10 provides a separate route, allowing defined agreements to be exempted where they contribute to statutory objectives — among them the effective entry into, participation in and expansion within a market by small and medium businesses and firms owned or controlled by historically disadvantaged persons. The two provisions do different legal work. Both proceed from the same premise: that cooperation between competitors is not automatically the suppression of competition, and that the difference is established on evidence.
Sharing is not competition-neutral, and I would not argue that it is. Competitors remain competitors. Commercially sensitive information unrelated to the infrastructure purpose must stay protected. Retail pricing, customer strategy, product decisions and market allocation must remain matters of independent competition.
Clear limits on scope, information barriers, aggregation and anonymisation where needed, independent oversight, non-discriminatory access and continuing competition-law compliance are how those risks are managed. A theoretical risk is not a demonstrated harm. Competition analysis requires evidence of likely effects, of market conditions and of the link between the conduct and the alleged harm. Treating a possible risk as an inevitable outcome is not a substitute for that analysis.
On affordability I will be direct, because Colmer is entitled to a straight answer.
Improving the economics of deployment does not guarantee lower retail prices, and I am not going to claim that it does. Retail pricing should remain an area of vigorous competition, and coordinating it is precisely what this application excludes.
Open, evidence-based process
What efficiency buys is something different: capital that is not spent repeatedly on the same duplicated assets can go into expansion, capacity, resilience, upgrades and coverage in places where investment is otherwise difficult to justify. Coverage that people cannot afford does not deliver inclusion. Neither does an affordable service in a place with no network. Both problems have to be solved, and neither is solved by insisting on duplication for its own sake.
That is the case as I would put it. It is not a case that should be settled by op-ed, mine or Colmer’s. The Competition Commission’s process exists to test exactly these claims against evidence and submissions. ACT supports an open, evidence-based process, and I would encourage Wapa, the smaller operators it represents and every other interested party to put their arguments there, on the record, where they can be weighed.
