South Africa’s slow Internet connections and high data costs have been criticised by most industry participants — including government, large companies and consumers.

The ICT policy white paper, published in 2016, emphasised the need for more competition in the ICT sector to reduce prices and stimulate economic growth. The sector is an important enabler of economic growth and bottlenecks in this sector must be addressed as a matter of urgency.

One such bottleneck is high-demand spectrum, which the industry regulator, Icasa, has not yet assigned to operators. Spectrum has become the subject of significant interest, as a new solution has been proposed in the recently released Electronic Communications Amendment Bill. The bill was published on 17 November 2017 and industry players had until 31 January 2018 to provide inputs.

The assigning of the remaining high-demand spectrum is not a new issue and Icasa has attempted to solve this problem on two previous occasions. In 2011, it published a draft invitation to apply (ITA) for the combined licencing of 800MHz and 2.6GHz bands. But this process came to a halt and was deferred until further notice.

More information on the planned assignment of prospective licences was provided by Icasa during 2015, and in 2016 the regulator issued a new ITA for a selection of spectrum bands. This process was however delayed by litigation and the spectrum remains unassigned.

The amendment bill provides a novel alternative to the problem of how the high-demand spectrum should be assigned. It proposes that the remaining spectrum should be assigned to a wireless open-access network (Woan) provider; a concept that was introduced in the white paper. There are various reasons why — in South Africa — this should be an improvement over a spectrum auction. Internationally, spectrum auctions are usually employed to assign spectrum as it has the benefit of distributing spectrum in a manner that reflects supply and demand ceteris paribus (all other things being equal).

While economists are generally in favour of relying on market forces for the distribution of resources, a spectrum auction would not necessarily support South Africa’s developmental goals. It is reasonable to assume that the outcome of a spectrum auction would be that the incumbent operators with deep pockets would obtain the most valuable spectrum. This would further enhance the structural problems in the mobile sector (high levels of concentration).

Service-based competition

The introduction of a Woan would see competition in the sector moving away from infrastructure-based to service-based competition. This would be a welcome move, as the ICT sector is characterised by high barriers to entry due to considerable network investment costs. Through the Woan, new players would be able to enter the mobile market without the need to invest in the fixed costs associated with establishing a network: they would be able to obtain spectrum via the Woan and access infrastructure on an open-access basis. The objective is more competition at the retail level to put downward pressure on the price of especially mobile broadband services.

It must be highlighted that there are not many jurisdictions where a Woan-type model has been successfully implemented. Kenya and Russia had planned to introduce a single wireless network (SWN) — a variation of a Woan where there is a single wholesale network — but most progress to date has been reported in Mexico and Rwanda. Whether the SWN in these countries will have the desired outcomes, however, remains to be seen.

The authors, Nicola Theron and Helanya Fourie, argue that more detail is needed of how the pricing and access principles would work for the proposed Woan

In principle, the Woan is a good idea for South Africa. But there are certain pitfalls which government should be aware of. Creating a SWN by requiring all high-demand spectrum to be returned to the state (an idea which is canvassed in the amendment bill) would simply establish a monopoly with all the inefficiencies associated with such a market structure, and which may translate into higher retail prices. It would also create considerable uncertainty in the market and would negatively affect investment and perceptions about property rights. The bill also plans to prohibit spectrum trading, which would further stifle market developments.

Furthermore, while the bill specifies that the Woan would provide spectrum on an open-access basis, more detail is needed of how the pricing and access principles would work. Cost-based access has proven difficult to determine in the case of mobile termination rates, for example. Finally, government would have to ensure that there was a sustainable business case for the Woan by setting reasonable network and population coverage targets. Yet, while much more work needs to be done on exactly how the Woan would be structured and implemented, the idea certainly has the potential to bring more competition to South Africa’s telecoms market.