Illustration: KAREN MOOLMAN
Illustration: KAREN MOOLMAN

SA is at a crossroads on how to make the broadband market more competitive and ultimately affordable for all South Africans. Broadband affordability rather than availability is a primary barrier to such access. After all, companies such as Vodacom lay claim to 99% 3G population coverage and 75% LTE population coverage, yet about half the population are still not part of the global information society.

If affordability is the problem, and campaigns such as #SocialMediaBlackout are not the solution, what can be done to reduce data cost? Four steps will make a large and immediate difference: releasing more spectrum to mobile networks, regulating aspects of wholesale mobile network access and pricing, actively promoting Wi-Fi services in public spaces, and lowering the device-entry barrier.

Regarding the release of more spectrum, the only alternative to spectrum for increasing data capacity is to build more sites, which is costly. Lowering costs creates conditions for lowering prices. The recent Constitutional Court ruling on encryption in the digital migration policy hopefully paves the way for broadcaster spectrum to finally be released for mobile services. The next step is resolving the telecoms industry dispute over how best to use that spectrum.

The original white paper idea of giving all spectrum to a single wholesale open-access network (Woan) provider is a poor one. Not only is infrastructure competition critically important to driving innovation, service improvement and roll-out, but establishing any new Woan is many years away, if it is even commercially feasible at all.

Any cost savings from consolidating networks envisaged through a Woan is achievable through infrastructure sharing without compromising network competition. The reported compromise of reserving some spectrum for a Woan but releasing the rest to the industry is sensible but must not be the cause of further delays.

Spectrum should also be released in a way that enhances network competition and avoids untargeted but expensive conditions — both criticisms of the initial Independent Communications Agency of SA invitation to apply for this spectrum. The condition that each network achieves 30 megabits per second (Mbps) speeds across 100% of the population involved needless and costly duplication that is unprecedented globally. It also misconstrues the problem as availability rather than affordability. Universal service obligations are important, but will be more effective if they are better targeted and don’t negatively affect the ongoing cost of service provision.

Releasing spectrum to reduce costs but not considering the competitive implications does not guarantee these savings are passed on to consumers in lower prices.

Regarding regulating aspects of wholesale mobile network access and pricing, the conventional wisdom in telecoms regulatory circles globally is that if wholesale markets are sorted out, retail markets will take care of themselves. The regulation of termination rates and the subsequent reduction in voice call prices bears this out.

Similarly, the Telkom settlement agreement with the Competition Commission and the move to separate out Openserve has been relatively successful in enhancing competition in the fixed-line market.

There have long been calls for more wholesale regulation of mobile markets, and the minister’s suggestion of a Woan seems to have been born out of frustration with the lack of progress on this front. There are two levels at which this needs to occur.

Infrastructure access and sharing between mobile operators have to be regulated. As the cost of network provision escalates with each new generation of technology, there is a move towards active infrastructure sharing to reduce costs. This means sharing the base station tower as well as radio transmission equipment. We need to do the same.

However, the skewed mobile market structure in SA means the bargaining dynamics between large and small networks do not necessarily yield a fair or socially optimal outcome. Regulating the commercial terms of such infrastructure sharing as well as network roaming is therefore also probably necessary.

The second level concerns deep wholesale access for other licensed operators. Virtual mobile operators have not been a strong feature of the South African market, unlike countries such as the UK. As the UK experience demonstrates, virtual operators can be effective competitors if they have much deeper radio access-network wholesale access. The market structure is not conducive to such models emerging in the ordinary course and ways of opening up a more effective virtual mobile industry must be considered.

Regarding actively promoting Wi-Fi services in public spaces, the solution to lower data prices cannot be a mobile strategy alone. Providing greater opportunities for consumers to offload data usage from more expensive mobile to cheaper or free Wi-Fi networks reduces their cost of data use. If pervasive, it may even start to place constraints on mobile data pricing.

While wealthy consumers often have such options through home and office networks, as well as malls, this is not generally the case for poorer consumers. If home access is constrained, there needs to be a concerted effort to roll out public Wi-Fi extensively in social places where poorer consumers can be reached. This is precisely the type of targeted intervention where universal service obligations and independent government action can be effective. Internet access at schools, zero-rating data access to educational sites and city-level public Wi-Fi initiatives are already making a difference. However, to have a broader effect, these need to be more ambitious. Leveraging the spectrum sale to impose more targeted universal service obligations of this sort is one means. Local city-level targets are another.

To be truly successful, the potential for greater commercial provision must be unlocked too. Already some operators provide access to a large number of geographically dispersed Wi-Fi sites through a single subscription, often leveraging the existing Wi-Fi networks of restaurants, malls and even households. Providing scope for commercial operators to leverage off any city, government and universal-service Wi-Fi initiatives is a way to reduce the costs and accelerate coverage.

Regarding the device-entry barrier, at an individual level, affording a data-enabled device (smartphone, PC or tablet) often represents the first barrier to internet access. The price of these devices is falling, making them gradually more affordable. This is aided by targeted initiatives such as the programme to provide tablets to all matrics. The question is whether there is a means to accelerate the adoption process universally.

Mobile and fixed-line operators benefit from consumers shifting to data-enabled devices. Apart from presenting data-revenue opportunities, it also frees up spectrum and network costs supporting current 2G devices. Despite these benefits, there is little incentive for any network to subsidise device replacement as prepaid customers churn so frequently, they are unlikely to capture the benefit. An industrywide initiative may be able to overcome these free-rider problems.

The launch of the #Internet4All initiative at the recent World Economic Forum meeting seeks to find solutions to providing universal internet access in SA by 2020. If that goal is to be realised, some bold and decisive action will be needed.

Hodge is a partner in the Competition and Regulatory Economics practice at Genesis Analytics. He writes in his personal capacity.

Please login or register to comment.