Competition Commission’s plan – cheaper data in SA a bad idea

The Competition Commission recently released a provisional report on its inquiry into South Africa’s data services market.

Sadly, the report relies on inaccurate or cherry-picked data, leading the Commission to make fallacious conclusions. It also led the Commission to suggest potentially disastrous interventions in what is a competitive market.

It may be an unpopular opinion, but we have a tremendous amount of choice when it comes to cellular services in South Africa. Not only in service provider, but in products as well. That doesn’t mean prices can’t and shouldn’t come down, but populist nonsense has clouded reasonable discussion of how to make that happen without being self-defeating.

Vodacom and MTN are not holding a gun to anyone’s head. What stops you from choosing Cell C or Telkom Mobile as your mobile network operator? Why have you not considered Afrihost Mobile or FNB Connect as your cellular service provider?

I, for one, am a mostly-happy Telkom Mobile customer. I have also had the opportunity to travel significant parts of South Africa in the past two years to test the quality of South Africa’s mobile data networks.

Where Telkom does not have its own network, I’m generally able to connect through its roaming partner. (Telkom is busy switching from MTN to Vodacom as its roaming partner.)

Don’t get me wrong — choosing a smaller operator does lead to the occasional frustration.

That’s the trade you make, and which relatively few South Africans appear to be willing to make. You can get much cheaper prices and more consumer-friendly services, but in exchange for some reliability, stability, and good customer support.

Beware unintended consequences

It is clear from the solutions the Competition Commission suggests that, despite its inquiry, it really has not gained an understanding of mobile networking. One of the key recommendations the Commission makes to give immediate price relief for South Africa’s poorest is to reduce the per-megabyte price of smaller bundles – those below 1GB.

It suggests: “A commitment by mobile operators to reduce the price of sub 1GB bundles within an objectively justifiable and socially defensible range of the 1GB price, provisionally a maximum of 25% higher on a per MB basis.” While this may sound great, if implemented it will most likely backfire spectacularly.

South Africa’s economic climate has not been kind to our mobile network operators, with Vodacom recently posting a decline in mobile data revenue growth despite a 41.4% increase in data traffic on its South African network. It will be very difficult for Vodacom to justify data price cuts to shareholders given its current circumstances.

Cell C is not faring well financially either and is dragging down the share price of parent company Blue Label. It has a plan in place to become financially sustainable, a large part of which is a deal with the Buffet Consortium.

Given the current economic climate, it is unlikely that mobile network operators will simply reduce the price of their smaller bundles to be within 25% of the price of their 1GB bundles.

What is more likely is that networks will hike prices on larger bundles to bring them in-line with the Competition Commission’s 25% requirement. Smaller bundles may see a slight decrease in price, but not to the extent the Commission is hoping for.

The Competition Commission should consider the unintended consequences of trying to control the retail price of data in the way it is suggesting.

Network congestion control with pricing

An important issue the Competition Commission does not appear to consider in the context of pricing is network capacity. Data prices are not only a function of costs, but also capacity planning.

It’s an unpopular thing to say, which is why you don’t often hear networks talking about it, but if you make data too cheap, the spike in usage could put too much strain on your network. This will cause poor speeds across some or all of the network, with towers being completely overwhelmed by the demand.

This phenomenon has been proven many times when wireless network providers try to launch uncapped mobile data services, or give away free data as part of a promotion. Telkom and Rain are two networks that most recently tried to launch affordable uncapped LTE services on their networks.

The outcome is always the same. Shortly after launching uncapped LTE services, subscribers start complaining about poor speeds. Networks implement stricter fair use policies, leading to another wave of complaints.

Sales of the uncapped product is halted and eventually a kind of steady-state is achieved. If nothing goes too terribly wrong, the network and the service survives, but with restrictions in place.

Price, quality, or coverage: choose two

If you don’t foster an environment in which networks can rapidly and cost-effectively expand their capacity, demanding that they dramatically cut their prices will almost certainly result in a dramatic loss of network quality.

It’s a classic optimisation problem between three factors: lower prices, higher quality, and greater coverage. Unless you change the underlying parameters of the equation, like available capacity, network operators will have to sacrifice in one area to make improvements in another.

A typical counter-argument to this, which receives a mention in the Competition Commission’s report, is that Vodacom and MTN charge more for data in South Africa than they do in the other markets where they operate.

While this is generally true, the Commission fails to take into account any other data, immediately jumping to the conclusion that there has been some kind of market failure.

Yes, South Africans may pay more for data than some of our neighbours, but that’s what happens when the networks are required to provide coverage for every person in the country.

Looking at MTN’s figures, its infrastructure investment in South Africa dwarfs other African countries where it operates.

Make capacity – Release the spectrum

Though the Competition Commission’s report focuses on interventions on South Africa’s data market, at least it does not let the government off the hook completely. It highlights that the release of more spectrum is urgently needed, but also warns that the release of spectrum won’t magically cause prices to come down.

This is completely fair comment, and the Commission is justified in saying that the spectrum allocation process requires careful thought. However, when you take the Commission’s statement in its full context it grossly oversimplifies the considerations that need to be made.

“Careful thought must be given to the design of the spectrum allocation process to ensure that competition is prioritised,” it stated. Naturally, a commission focused on competition believes that the spectrum allocation process must be designed to prioritise competition.

However, I would rephrase the statement so that the use of the spectrum is prioritised and balanced against ensuring competition.

The availability of spectrum resources without effective competition won’t bring down prices, but neither will loads of competition without effective utilisation of that spectrum.

The very first thing that needs to happen before we can talk about bringing mobile data prices down is enabling South Africa’s network operators to expand their capacity quickly and cost-effectively. Whether that is by making it cheaper to build towers, releasing spectrum, or other mechanisms.

Trying to address the issue of high prices without also dealing with the cost of building network capacity is a fool’s errand.