My comments to U.S. State Department ITAC group. One of the issues at the ITU is whether the cost of peering/transit should be part of the discussion on affordability of access. Africans tell me this is perhaps the largest single impediment to bringing down the cost of the Internet in Africa. Backhaul is 10x the cost in Europe or the U.S., only a small part of which is due to the cost of undersea cables. I infer a (weak) cartel and strongly urge the ITU to produce the data to convince anyone objective something should be done.

Some people are trying to exclude this from the development agenda. They come from countries with companies that benefit from high transit prices. (Surprise.) By accident, I’m on the U.S. State Department International Telecommunications Advisory Committee and wrote this. By my choice, I do not report what anyone else says at that committee. This is my opinion.  

On Peering

Carlos Slim at the Broadband Commission told me $50 mobile phone prices will allow two billion more people to connect, a goal i know we all share. Since then, I’ve looked closely at the actual cost factors delivering Internet to the poor in emerging nations. I also discussed it with numerous people at ITU Busan. (Article below) Peering/transit costs appear the largest single international issue in affordability in Africa especially, as I report below. According to an African at the Busan Plenipot, the saving from reducing peering/transit costs is 5-10x anything likely from IXPs. Several Africans saw peering/transit costs as a key issue and welcome ITU involvement.  

It is extremely disrespectful for the U.S. to oppose the Africans getting the actual data on peering/transit, especially since U.S. companies are profiting. 

That peering/transit costs are a crucial factor in access costs is confirmed by my research, which found costs of backhaul in Africa 10x higher than Europe or the U.S.. Only a small part of that is related to the actual costs of the necessary undersea fiber. Pricing has gone down as Africa has been connected by fiber, but remains so high I infer a (weak) cartel. 

I’ve removed all comments except my own from this reply so I can cc this to Michael Kende and Kathy Brown, Jane Coffin’s colleagues at the Internet Society. Michael, a good economist, is one of the world’s experts on IXCs and is close to the data sources here. He can probably run approximate numbers on an envelope. Kathy – it would be enormously useful if Michael did a thorough study here. The U.S./EU poloicy people seem not to have the data. 

Until we’ve reviewed the data, I strongly urge the U.S. not to object to including peering/transit costs in the work.

Dave Burstein

Fairness for Africa: Cheaper Transit, Reasonable Royalties, Fair Taxes (10/30/2014)

“Affordability is the single most important lever to bring the Internet to everyone,” Hamadoun Touré tells us in Busan. Here are some things the richer countries can solve to help bring affordable broadband to the poor. I’m focused on correcting flaws in the system that suck 

1) High backhaul/transit costs double or triple the cost of providing broadband in most of Africa. A megabit costs $0.50-$3 in most of the U.S. and much of Europe. The same megabit in Lagos costs $170, 100x as much. 50-90% of the difference is cartel pricing, based on undersea costs where there is more competition.

2) High royalties may soon double or triple the cost of a low end smartphone.  Hundreds of millions fewer people will connect. Carlos Slim of Telmex told me at the Broadband Commission the $50 smartphone will connect two billion more people. On a mass product like a smartphone, a “reasonable” royalty would be something like 5-10%. Intel calculates royalties may soon be $140 on a mid-priced phone. Every international agreement calls for “reasonable” royalties and it’s time to reduce the Qualcomm tax.

3) Multinational giants should pay reasonable taxes. France and England can’t get Google or Apple to pay taxes. What chance does Cameroon or Thailand have without strong support from International groups like ITU and the giant’s home countries. Collecting through a “sender pays” is a terrible system which clearly is anti-consumer in the developed world. But changes in telecom pricing have sucked hundreds of millions in taxes/fees and probably more from some of the the poorest countries. Most of the benefit goes to carriers and consumers in richer countries, starting with carriers like AT&T that now aren’t pay the old termination charges. 

i’ve been asking those from the emerging world what countries like the U.S. can and should we do to support affordable broadband development. I’ve spoken in the last few days to people from Sudan, South Africa, Malawi, Niger, Zimbabwe, Guyana, Bangladesh, Saudi Arabia, Costa Rica, Egypt, Nigeria, China, Mali and others.

    One question I asked is “What should richer countries and International organizations do to help those in need?” The answers were surprisingly consistent, focusing on a few situations where richer countries/their companies are extracting capital from poorer countries. 

    Most of the work climbing out of poverty and building networks will need to be done in-country. Fortunately, the domestic private companies are finding the finance they need to build mobile networks for the cities of the world. Econet Wireless in Zimbabwe is completely privately funded and has reached nine million subscribers in one of the poorest countries in the world. Zimbabwe has more mobile subscriptions than people, and Econet is expanding across Africa.