How Kenya’s bold open-access telecom plan came unstuck
South Africa presses on regardless
Kenya’s 2011 national broadband policy contained details of a government project for a “fast track” rollout of a wireless 4G wholesale open-access network (WOAN).
The network would be built, financed and operated by a “consortium” that would include the main telecommunication operators.
The consortium would be a public and private partnership (PPP), in which the government and the telecoms operators would own stakes that reflected the capital they had invested in this proposed joint venture.
It was proposed that the mobile network operators would provide the financial investment while the government would provide frequencies. It was suggested that the network might use spectrum in the 700 and 800 MHz and 2.6 GHz bands.
The government planned to issue an invitation for a new consortium to tender to build and operate the wholesale network, as opposed to just auctioning the spectrum and awarding mobile licences to existing individual operators.
The arguments in support of the government proposal suggested that it would avoid duplication of infrastructure and would use spectrum efficiently.
The consortium network would offer open-access wholesale capacity to new and existing service providers.
The stated objectives were countrywide connectivity, affordable high-quality broadband services, increased penetration, economic growth, and improved “general social well-being”. The planned network would cover 98% of the country.
There were, however, difficulties in securing the cooperation of the mobile providers.
In 2012, Kenya’s largest operator, Safaricom (40% owned by Britain’s Vodafone, which has management responsibility), reportedly objected to the decision to roll out the network in the 2.6 GHz band. Safaricom felt that starting with the 700 MHz band would have been more efficient.
In 2013, Safaricom, complaining about the “glacial pace” of the proposed public project, became impatient and withdrew from the project.
Mortimer Hope, the sub-Saharan Africa spectrum and public policy director of the London-based global club of telecoms operators, the GSMA, said that Kenyan IT policy was unusually consultative and that getting feedback from all relevant stakeholders took time.
By 2014, it was noted that Uganda and Tanzania had newly-established 4G networks, whereas Kenya still had only 3G networks.
Safaricom announced it would build its own LTE network. Safaricom’s CEO Bob Collymore told shareholders that the company had successfully tested LTE after its acquisition of yuMobile and the spectrum rights that came with it, and that Safaricom planned to launch 4G services in the next financial year.
The government, in January 2014, insisted that its planned wholesale network was still on track, despite Safaricom’s departure.
The result would be cheaper internet for operators and end-users, declared the telecoms authority. “Some companies want to have their own infrastructure, but they stand to reap more from shared access in terms of their capex,” the authority’s project manager Andrew Lewela told reporters. “We prefer a market approach in having them invest in shared infrastructure.”
The GSMA’s Mr Hope said that single networks’ main problem was that they operate as monopolies and were thus less motivated to stay on the cutting edge. “There isn’t the incentive for them to invest and upgrade technology as quickly as a private enterprise would. So as with all monopolies they would keep their prices high and their service levels would usually be quite bad.”
This lack of technological agility, combined with political squabbles about who would benefit first from network rollout, was bound to make such a project cumbersome and slow.
That would not inspire confidence in investors, so raising new capital will be difficult when the government eventually finds its WOAN too costly to maintain.
That wasn’t to say that governments shouldn’t be addressing poor network coverage in hard-to-reach areas, said Hope. But public funds could be used to encourage private operators to expand their services into less profitable parts of the country. Existing operators could bid for subsidies on condition that the network built would need to be open-access.
But instead, Kenya’s planned WOAN would be competing with Safaricom’s network. Hope warned that if public contracts are channelled toward the government-run network as seemed possible, the telecoms authority could end up defending itself in court on charges of anti-competitive practices, which was not a battle the government would necessarily win.
“The world had moved away from monopolies,” Hope said, and the long-term viability of a state-sponsored WOAN was questionable.
In 2015, the Kenyan WOAN was being described as “dead in the water”. In the absence of Safaricom’s participation, it was hard to see how the WOAN could attract enough traffic to be economically viable. The possibility of a WOAN had thus been overtaken by events, it was said.
In 2016, observers concluded with confidence, although no official announcement had been made, that the Kenyan government had abandoned its WOAN plan.
This was evidenced by the lack of mention of the network in recently published draft ICT policy and framework documents and by the recent assignment of 800MHz spectrum to existing mobile operators who had since commenced the deployment of broadband services using the spectrum.
In 2017, the GSMA stated that the fate of the Kenyan WOAN project highlighted how complicated the single wholesale network model was. The network had been proposed through a public-private partnership in order to “fast track” rollout of LTE services. That never happened.
The 2016 telecoms regulations contained another government plan, also now discontinued. The regulations sought to restrict the building of new masts and infrastructure where there was an option of sharing with an existing infrastructure provider. The Daily Nation newspaper observed, in August 2017, that this appeared to have fallen through as well.
The telecoms authority had also published guidelines in 2016 that would compel telcos to share up to 30 per cent of new infrastructure. The authority argued that sharing would eliminate “wasteful duplication” by competing networks.
It has not yet been determined whether the authority will continue pursuing those guidelines. The authority would do well to heed an empirical assessment by British economists which found that telecom coverage by population and area was significantly higher with network competition than without.
The South African government seems not to have heeded these conspicuous precedents. It has been pressing ahead with its own project for a wholesale network that would be granted exclusive access to high value spectrum. This would remove the benefits of infrastructure competition, which would be achievable if multiple operators were licensed to compete using these frequencies. As the Australasian telecoms advisors Webb Henderson reiterate, a monopoly has poor incentives to operate efficiently and provide a high quality of service at a competitive price.
The draft South African communications bill of 2017 intends to oblige network-service licensees to provide wholesale open access to their networks and facilities on request at cost-based pricing.
This will remove all incentive to improve and upgrade networks and facilities. Why work to introduce improvements if others can insist on partaking at cost price?
Gary Moore is a South African lawyer and senior Free Market Foundation researcher
For case studies on WOANs in Kenya, Mexico, Russia and Rwanda: http://www.freemarketfoundation.com/Article-View/case-studies-on-wireless-open-access-networks-