Sharing infrastructure to speed up roll-out – reaching the parts you can’t reach by yourselves
This week saw the ITU’s 8th Global Symposium for Regulators take as its theme infrastructure sharing as a means of speeding up broadband roll-out in developing countries. Sharing is one those warm, fuzzy words and as one delegate said:”My mother always told me to share.” But as those at the event discovered, some parts of sharing are fairly easy, whereas others require a greater understanding of the potential pitfalls. Somehow competition and sharing don’t seem to fit in the same sentence but competitors have been sharing for years. Russell Southwood looks at what is at issue.
With the exception of high-cost countries (like Angola, Cameroon and Gabon), soon Africa will have relatively cheap international fibre all around its coast. Not every country’s connected yet but they might easily do so in the not-too-distant future. But what good is cheap bandwidth if you have not got a high-capacity national backbone that can deliver to the main population areas within a country? Without these kinds of backbones, there will be less customers for cheap broadband Internet and the services it brings, whether individuals, the Government or the corporate sector.
But say the sceptics, why not let the mobile companies invest in a 3G backbone as there is not that much demand? Africans only really use mobiles and not PCs, so why over-build at this scale? But this assumes Africa will remain as it is now with its relatively low economic potential, not the Africa that it might become in five to ten years time. There are “fast-track” countries (for example, Kenya, Nigeria and South Africa) that are already on the cusp of significant economic changes: the middle classes in these countries have grown significantly over the last three years. Furthermore, bandwidth demand has more than tripled over a five-year period in some countries.
In these circumstances, to argue that 3G alone will be “good enough for Africa” and for keeping supply stuck well behind potential demand is not really going to give those countries that might succeed anything like the entrance ticket to the new global economy they require. Tourism, outsourcing and the logistics of exporting high-value goods all require bandwidth to be available in places that most networks either do not reach or do not yet reach with sufficient capacity and a relatively cheap price. If margins are everything, then the supplier that has to run its business using satellite is at a disadvantage over those who have broadband access over fibre.
Future services and applications will not be built if developers have to build everything around the mobile phone: you will develop repetitive strain injuries and a squint before that happens. Broadband use in Africa needs a “critical mass” if it is to be able to deliver the “people networking” effect that mobiles so obviously already do. In key countries, the growth curve is moving in that direction and it will increase with cheaper prices.
Any African country that is likely to succeed has a Government that is worrying about how it can provide its country with a broadband infrastructure that will: drive some aspect of its economic development, help Government itself communicate across large distances and deliver simple services close to where need is found. Yes, roads and schools and health clinics are all as urgent but they cannot change the pace and style of the economy in the same way as having a connected country can. But more taxes from a more successful economy can pay for a greater range of social goods of the kind that are urgently needed.
The issue is then how can this become a realistic national priority that can be achieved in a short number of years? Left to the market alone, the process will be relatively slow as operators will tend to build out behind demand rather than ahead of it in order to guarantee returns. The operators will tend to confine themselves to building metro rings and links between major urban centres.
In the old days, the responsibility for addressing this kind of national priority lay with Government and it was expressed through the incumbent telco. Nobody wants to return to that position but there are a number of “chicken-and-egg” problems where Government needs to play a facilitating role.
In a West African regulatory forum, I sat on a panel with two of the continent’s major mobile operators. They were asked: why won’t you build inter-country network links? (The question could as easily have been: why don’t you build a high-capacity, national backbone?) Each one said that there was not enough demand and that therefore this kind of project was the responsibility of Government.
Less public conversations with mobile operators (indeed most operators) make clear that where possible they would prefer to carry their own traffic as they do not trust the incumbent. In the main, the latter still treats wholesale and retail as mutually interchangeable arms of the same body and competes at the retail level with its own customers. As high-capacity infrastructure is a near monopoly except in the larger countries, there is all the usual monopoly incompetence in service delivery. In effect, the incumbent’s assets and the way they operate them have become “bottleneck facilities” and this applies as much to their near-monopoly national networks as it does to the monopoly international fibre landing stations.
There are two issues chasing each other round in a circle that needs to be broken. If there was sufficient trust in a wholesale, “carriers’ carrier”, then the operators would be able to buy this capacity from that company rather than investing significant sums of money into their own infrastructure. If the majority of them bought infrastructure in this way, the volume of traffic would be enough to finance fibre routes where none currently exist.
In effect, infrastructure investment would be shared by it being financed through a third party. Money not invested in infrastructure could be spent competing vigorously at the service level with new services and applications. At the highest level, this is the key argument for infrastructure sharing and it is clear from the description of the dilemma that it is not one that will be easily unravelled. But not to seek to do so is probably to risk slowing down the development of key markets by ten years or more.
So how to tackle this knotted cluster of issues and provide incentives for investment? Since the responsibility of Government has been invoked, let’s be clear what it should be doing. It is not a strategic role of Government to run infrastructure “forever and day”. However, it may need to make a tactical intervention in order to create the circumstances for solving some of these problems. In brief, it need to “get in, get it done and get out”.
This is the Achilles’ heel of structures like South Africa’s Infraco where the Government has sought to solve a problem without announcing when it is going to “get out”. Worse still, if ANC MPs and the Minister have their way, then the Government will have investments in three entities, all of which are supposedly involved in providing some part of the services to solve the same problem. So what might be done to remove blockages to the development of the market?:
Rights of way: If sharing is a tactic to “get it done” then the safest ground on which to start is the simplest. Government can help simplify the process of obtaining rights of way from all the different parts of Government that have control over them. It can either give them away at a very low cost as an investment incentive or it can accept fibre capacity for its own operations in exchange for the value of the rights of way.
Duct sharing: Governments and the regulators can insist that all ducts built must be shared and that “dark fibre” will be made available to other operators on the same terms as it is used by the company investing in the infrastructure. Alongside duct sharing comes rights to connect to the network through co-location points where other operators have the physical space and access to be able to deploy their own servers.
In much the same way, mast sharing can be imposed, particularly in areas where there is currently no coverage. This means sharing physical space on the tower and in the cabinet and sharing generator costs. Wi-MAX backbones may also share these same masts. A number of countries have mast operating companies that mediate this process in ways that allow competitors to trust each other in doing this.
National infracos and joint venture companies: African Goverments often have more fibre than they know what to do with. The power utilities, railways, gas and oil pipeline companies and water utilities all have fibre to manage their operations and build in surplus capacity. These assets can be drawn together in a National Infraco to speed up roll-out. Furthermore the costs of stringing fibre on power towers or in sewers is almost marginal compared to trenching.
The key issue as ever is trust. The Ghanaian Government is about to create such a vehicle using the fibre assets of power utility Voltacom. The National Communications Company will remain a wholly Government-owned entity with a Government-appointed Chief Executive and it will inherit the existing staff. Does this create trust in the market? Probably not.
The Kenyan Government will either appoint a private company to run the “infraco” or give the contract to the former incumbent (now the newly privatised) Telkom Kenya. With the former, the right appointment might well create the kind of market trust needed. With the latter, there is the risk of simply recreating the old incumbent vs competition dynamic of old.
The Nigerian and Ugandan backbone projects seem to have retreated from the broader national priority and to be focusing on Government traffic alone with any other sales as marginal activity. Extracting Government traffic from the market in this way is hardly likely to contribute to market development, even if there is sound economic logic to Government buying capacity in bulk.
Joint ventures: One way to overcome these issues is to look at how operators might be encouraged to become investors in jointly-owned vehicles involving the private sector. At an international level, this is the basis of the TEAMS fibre project. The logic of a national joint venture is therefore broadly speaking the same. Government’s involvement can be to get everyone to work together and to take the project forward at a cracking pace.
It can also invest with the operators and the public money can be used to reach more marginal rather than market-ready areas. The difficulty here is getting the “lion and the lamb” to lie down together and Government having the subtlety and skills to provide both trust and speed of action.
If the market for infrastructure investment is buoyant and many routes are in the process of being built out, the Government might focus on un-serviced or under-serviced areas by creating smaller scale “greenfield operators” that receive a capital kick-start (using universal access funds) for providing wholesale capacity. This new network will attract existing operators to extend their coverage at the service level and/or drive the creation of local service operators.
Functional separation: In a small but growing number of developed countries, regulators have insisted on a considerable degree of functional separation between the wholesale and retail elements of the incumbents. Vincent Affleck of the UK’s Ofcom told the conference that after BT Open Reach was created the number of complaints about BT has gone down dramatically but that there was an increase in complaints from BT’s retail services side.
The dream ticket of hardline sharing would be to bring the incumbent’s fibre assets together with all the Government’s utility fibre assets. The challenge is that functional separation is by no means a simple task but as experience builds elsewhere, then maybe so will the relevant expertise to achieve it.
The dilemma for the vertically-integrated mobile operators who have become the new incumbents is whether they wish to replace the old incumbent in the infrastructure business. At least two of the major mobile operators on the continent have convinced themselves that they will. But it is not clear if the position is tactical or strategic. As one of them told me, “We’ll keep building out national fibre routes until the incumbent starts bringing its prices down.” But whatever attitude the vertical integrators take, they will not be embarking any time soon on connecting the more marginal areas of a country. So Governments in a hurry to see their economies develop need to consider which of the “levers” suggested above that that they might pull to “get things done.”