Spectrum Policy in South Africa – Necessary But Far from Sufficient

The various reactions to ICASA’s cancellation of the Invitations to Application (ITA) for spectrum licenses in the 2.6 and 3.5 GHz bands, which were due on July 30^th, reflect an unfortunate conflation and confusion of two overlapping but nevertheless quite distinctive and fundamental issues, namely:

1. What is the best spectrum structure and procedure for allocating new (and maybe re-allocating existing) spectrum holdings; and
2. How can a healthier and more competitive, customer-friendly supply of network access and services be fostered in South Africa, given the history (and current behavior and attitudes) of incumbents that are viewed by many stakeholders as profoundly unsatisfactory?
   

Spectrum policy and management constitute one of the tools or areas for maneuvering which a regulator and Ministry can and should apply in addressing the second issue, but no means the only one. Nor is that the only goal of spectrum policy, which must also consider questions of economics and what uses of spectrum will generate the greatest overall value for society and the economy. There is room for ICASA and the DoC to exploit a richer and more diverse set of tools - including spectrum-related as well as other measures - to achieve both goals of:

1. An efficient and effective allocation and use of spectrum; and
2. A healthy competitive market.

The wise application of these tools should recognize that:

1. The techno-economic characteristics of networks mean that markets can only support a limited number of viable facilities-based operators; and
2. There are ways of ensuring that these operators do not discriminate unfairly against competitors at the services level or pursue customer-hostile paths of action and behavior, without falling into the trap of fragmenting spectrum so as to allow the entry of a large number of players who are bound not only to be inefficient themselves, but also to make it more difficult for others to be efficient.

The implication of (a) is that however much the incumbents may be disliked and mistrusted in some quarters, they and the assets they control must and will continue to play key roles in the future of telecommunications in South Africa. The goal should be to change the behavior of these incumbents with appropriate incentives and effectively enforced penalties. It would be a “shoot yourself in the foot reaction” to try to substantially fragment the market further on the supply side of transport , especially scarce spectrum resources, under the mistaken impression that this is the best way to ensure adequate competitive intensity.  Depth of spectrum is critical to techno-economic efficiency of a network yet there is only a limited amount of bandwidth to go around. Indeed the potential growth in bandwidth for mobile services is much lower than the projected growth in mobile traffic volumes, so the dangers of spectrum fragmentation will only intensify.

Any tool can be harmful and produce negative and unintended consequences depending on how it is applied. Also the relevance and efficacy of any tool depends on specific market and competitive circumstances. What works and is appropriate in London or Tokyo may be impractical or undesirable in Pretoria or Abuja, and vice versa.  So the following list should not be interpreted as specific recommendations for South Africa (both with respect to new initiatives and/or those that should be continued), but rather as an outline of the broad perspective on opportunities for steps to take that ICASA and DoC should keep in mind in deciding how best to achieve their goals and fulfill their responsibilities. Some of the measures in this list have already been implemented to varying degrees. Assessment of their value and disadvantages is outside the scope of this post. They are mentioned here to demonstrate the breadth and depth of the actions, including those not directly associated with spectrum allocations, that regulators and policy authorities can pursue to achieve the objectives of a healthy (for customers and investors) mobile broadband market and stimulate the most valuable and valued uses of finite spectrum resources. They can also mitigate otherwise undesirable consequences of spectrum holdings and allocations that may arise from a combination of legacy spectrum awards and the inevitably somewhat unpredictable outcomes of competitions for new spectrum.

Measures which either operators themselves and/or regulators can initiate, encourage, and sustain to contribute to ensuring competitiveness and/or attractive network economics in mobile markets include:

1. The application of antitrust rules rather than fixed prescriptions to spectrum holdings
2. Mobile number portability and other means to reduce customers’ switching costs between services providers,
3. Radio access network infrastructure sharing (among operators as well as with respect to authorizations of MVNOs (mobile virtual network operators),
4. Economical and ready access by mobile access network operators to wholesale facilities for backhaul and core networks,
5. Reasonable mobile termination and roaming rates,
6. Coverage obligations (“use it or lose it”) on spectrum winners to ensure that spectrum allocated to them is actually exploited for the benefit of customers in a timely manner, and to prevent anti-competitive spectrum hoarding, and
7. Rules to permit spectrum trading, i.e. creation of a secondary market in spectrum.

The effective application of antitrust rules to spectrum transactions and allocations requires close coordination between ICASA, the DoC and the Competition Commission to avoid time-consuming disputes between these different public sector entities as well as to minimize the bases for legal challenges to proposed or existing spectrum allocations that will inevitably further delay the availability of spectrum for mobile broadband network deployments.

Furthermore, in the formulation of conditions and procedures for allocating spectrum a few diminishing but possibly still lingering mistaken impressions specifically of the wireless arena should be dispelled. They include for example:

·         WiMAX is the platform for entrants and entrepreneurs, while LTE is for incumbents

·         WiMAX versus LTE is equivalent to unpaired (TDD) versus paired (FDD) spectrum

·         Competition and the principle of technology neutrality are best served by allowing bidders for spectrum to choose the amounts and locations of unpaired and/or paired spectrum they want, instead of competing for predetermined blocks of such spectrum.

However, mobile newcomers are perfectly able to choose to deploy LTE. Examples include Infotel Broadband/Reliance Industries and Qualcomm in India, Yota in Russia which is migrating from WiMAX to LTE, and the cable operator Cox Communications in the U.S. Secondly LTE will be available in the near future in both TDD and FDD modes of operation, the former initially and most notably in India and China. Finally there are potentially serious adverse implications  (less usable capacity, higher costs) due to increased risks of interference if spectrum band structure is not predetermined before it is allocated, but is left to the outcome of independent choices of competing operators. The complexity of managing FDD and TDD interference in the non pre-determined case are examined in detail in the April, 2009 white paper produced by Analysys Mason[1]: “Final Report for the WiMAX Forum, Cross Border Trigger Limits and Case Study for TDD/FDD Border Coordination in Europe”. In addition, non-predetermined band structures are likely to lead to country-specific band plans that will inevitably increase the costs of network deployment for operators and of devices, and hence the prices that customers have to pay. Operators may face more not fewer restrictions on the technologies they can deploy in practice in non pre-determined than in predetermined band structures, invalidating the claim of “technology neutrality” for the former.

Whatever path forward is eventually followed by ICASA and DoC with regard to new spectrum allocations, it should at least be founded on accurate and clear views of technological and economic realities, that cut through emotional (however understandable and driven by justified frustration) and/or self-interested biases. If these biases are acted upon they may defeat the purposes even of some of the parties who express them.

[1]Source