How could someone who’s passionate about connectivity in the developing world be opposed to a fiber-optic cable designed to make bandwidth cheaper in East Africa?
I’ve had to ask myself this question dozens of times in the past couple of years. Fortunately, recent developments on the EASSy (East African Submarine cable System) project have gone a long way to address my reservations about the project and I’m happy to celebrate the recent news that the cable will operate under an Open Access model.
Explaining why open access is important for EASSy requires some background on telecoms in Africa, a topic that’s vastly more interesting than you think, I promise. Hang in there – this stuff is really important:
Historically, sub-Saharan Africa has been connected to the rest of the global Internet via wireless connectivity – large or small satellite dishes that bounce data off geosyncronous satellites to dishes in the US or Europe. The data enters the internet from these off-continent data centers and is transmitted from there to the rest of the world. The one exception to this rule is South Africa, which has had data connections via undersea cables to Europe and the Middle East.
Satellite connectivity has upsides and downsides. On the upside – you can put a dish anywhere in the footprint of a satellite and transmit data – you don’t need to be near the coast to connect to an undersea cable. And you can build your own infrastructure, which is hugely useful in countries where telecommunications is heavily controlled by the government and inadequately developed.
But the downsides are myriad. Satellite connectivity includes a mandatory speed of light delay of roughly half a second per packet, making it difficult to offer real-time services like Voice over IP. (Not that this has stopped countless African entrepreneurs.) Because the internet access terminates on a different continent, it’s likely that two people in the same African city communicating via email are routing their packets through the US, travelling tens of thousands of kilometers out of the way to have a chat across town. (If you’re curious about this, you might check out “Internet Architecture and Institutions“, a teaching document Andrew McLaughlin and I developed years ago to teach people about internet connectivity.) And satellite connectivity, bit for bit, is more expensive than fiberoptic cable connectivity.
Map of undersea cables in use at the end of 2004 – from news.com
Africa took a big step – quite possibly a misstep – forward when the SAT3/WASC/SAFE (South Africa Telecom 3 / West Africa Submarie Cable / South Africa Far East) cable came into operation in 2002. The cable connects eight west African nations to landing points in Spain and Portugal, where it connects to terrestrial cable networks, and to South Africa, where the cable interconnects to the SAFE cable, which terminates in Malaysia.
A stated goal of the SAT-3 cable was the reduction of connectivity costs to nations connected to the cable – costs dropped, but not nearly as sharply as anyone had anticipated. The problem wasn’t a technical one – it was a political and economic one. The consortium of telephone companies that commissioned and built the cable included the dominant telephone companies – usually state-owned monopolies, or recently privatized providers – who determined that access to the cable be limited to consortium members. If I ran a telephone company in Ghana that competed with Ghana Telecom, I had to purchase access to SAT-3 via GT, my competitor… and needless to say, Ghana Telecom did not have a strong incentive to provide me with competitively priced bandwidth or swift service.
Ghana was lucky enough to have a powerful and effective ISP association, GISPA, which helped negotiate more reasonable prices from Ghana Telecom for competitive businesses. Things were more complicated in Nigeria, where upstart Globacom, realizing it couldn’t negotiate with incumbent telco Nitel, and tried to buy a share in the consortium so they could build their own fiber connection to the cable – Nitel blocked this move, and Globacom began looking into building their own cable to London… a phenomenally wasteful and stupid situation, as SAT-3 has tremendous underutilized capacity.
So when plans began to take shape for an East African cable to parallel SAT-3, people involved with African bandwidth issues (all two dozen of us) started making noise about the EASSy business model. Ronald Alden wrote the decisive paper – Just say No to EASSy – arguing that the $200 million project could be a step backwards for Africa as it reinforce the failed cartel model that SAT-3 introduced.
At least two schools of thought developed around EASSy. One suggested that it was important to ensure that universities have a seat at the table to ensure their interests were protected in a cartel situation. This group founded the Ubuntunet Alliance, which seeks high speed broadband connectivity for African universities – Ubuntunet’s plan has been to purchase a stake in the EASSy cable, ensuring affordable broadband for the universities, and increased transparency on the project, as universities would be more likely to share cost and pricing information than private companies.
The other school of thought suggested that pressure be put on international donors not to support the EASSy cable unless any organization could purchase bandwidth from the cable owners. This “open access” approach argued that connectivity was too important for African development to allow a cartel situation to evolve, and attempted to use World Bank and NEPAD (New Partnership for Africa’s Development) monies to shape the policy.
(These two schools of thought weren’t neccesarily opposed – many people supported open access but thought that Ubuntunet was a wise tactical move…)
According to a recent article by Elias Biryabarema in the Monitor (Kampala), the open access folks have won the day. “Any telecom company or institution” will be able to buy into the cable and access bandwidth, not just companies with “international gateway licenses”, the license granted by a government to companies to engage in international telephone voice traffic.
According to this arrangement, the cable would be owned and operated by the so-called Special Purpose Vehicle (SPV)-a company created specifically to manage the network and establish the pricing structure for the bandwidth.
An Intergovernmental Assembly (IGA) is to be formed to regulate the costs that the SPV will be charging user companies.
As always, the devil is in the details – will this IGA price access in a way that small telcos can afford it, or will there be large upfront fees? Will dominant telcos pull out of the agreement for fear that they won’t be able to charge enough money? I’m waiting for Russell Southwood, the dean of Africa telecoms analysts, to weigh in on his Balancing Act newsletter… but in the meantime, I’m happy to cheer this apparent victory in a very important battle that gets basically zero international media coverage.