Currently around 80% of all of Africas voice and data traffic is carried by satellite, but this figure is likely to fall as the continent increases fibre links at all levels. The balance of traffic is almost all carried by the continents only current international fibre link, SAT3.

Based on use of its international traffic database, the report estimates that on the basis of the progress of current plans and with favourable pricing adjustments on the SAT3 fibre, just over 30% of the total market in three years time will be carried by fibre, according to the African Satellite Markets report.

Why is this transition likely to be so slow given that fibre is cheaper than satellite for high-volume traffic? There are a number of factors:

* The slow speed of competitive national backbone roll-out: It has taken Nigeria five years to get to a point where Nitel is supplying sufficient national backbone connections to SAT3 that there is now a rising flow of traffic on to the SAT3 cable. By contrast, Telkom completed this work prior to the cable opening, and now carries the majority of its traffic over the fibre link.

* The lack of inter-country links: Although both SAT3 and the proposed EASSy cable connect coastal cities there are relatively few cross-border links in place. Kenya has two sets of links being built to Nairobi by KDN and Telkom Kenya and a link is being built from Kenya to Rwanda. But other parts of the 'land-side' infrastructure are at a much earlier stage. For example, Zamtel has just announced its intention to build its connection to EASSy. And in one case Zimbabwe the transition has gone backwards: Telkom SA financed a fibre link to the country, but TelOne failed to meet the payments, so is now sending its traffic via satellite.

* The impact of high SAT3 prices on landlocked and 'no landing station' countries: SAT3 consortium member, Namibia Telecom, is a 'no landing station country', and sends 60% of its voice traffic via satellite, most of the balance being calls to SA. Why? Because the costs of transiting via SA make it more expensive than sending via satellite. Based on a pricing survey, the report looks in detail at these market distortions that have arisen from the position held by the monopoly market supplier.

* The lowering of prices on the proposed EASSy cable: Although final prices have not yet been announced, it is believed that they will fall in the $500-1000 range (the lower price probably being available after a five year period). This will give users in the largest sub-Saharan African market, SA, a much cheaper alternative, and will drive down what Telkom SA can charge. Over three to five years, this will have the effect of unlocking some of the market distortion problems identified in the previous point in the southern African region. However, it will leave similar problems in West Africa largely unaffected.

Sub-Saharan Africa has seen a fourfold increase in the level of international Internet bandwidth supplied by satellite over the last four years, from 500Mbps in 2002 to 1,86Gbps in 2006. There are now 71 satellites with full or partial coverage of Africa, and seven more are planned.