There are a lot of issues with the way aid is distributed. The most successful parts of both places, the wireless networks and decentralized power systems, have been totally market driven, and actually hindered more by centralized efforts and regulation than helped.
Somalia, where it is too dangerous and chaotic for aid agencies or telco regulators to operate, has cheaper Internet access and mobile calling than almost anywhere else in Africa or the developing world!
Communications infrastructure (roads, signals/fiber/wireless, electrical grid) is one of the few areas where scale can be helpful. In Afghanistan there is a lot of money coming through the military to pay for a road network (with benefits to both the foreign militaries and the local economy; dual use technology at its best). Communications is basically wireless backed by satcom, with some microwave backbone and upcoming fiber, with microwave/fiber driven by military and government customers paying top dollar, then going out to connect up the cell systems. For $20/mo you can buy 3G service.
One of the biggest issues in Afghanistan is that aside from the US Military, the other US funding agencies (primary US AID) have to operate through local contractors, and a rather inefficient process which leads to overpaying for everything, and thus not accomplishing enough concentrated change to be successful. The "Provincial Reconstruction Teams" are afraid to go out except with a huge military presence, and are pretty sheltered from everything, whereas I was riding around in an old pickup truck with a 9mm and no problems.
Irrigation systems are probably the next thing to do after roads (the military has roads under control); these can be done for $5-12k each by local labor, and 20-30 systems per district, all in parallel -- or, you can pay $500k via US AID and get 1-2 systems done after 2 years. The US AID requirement to use local contractors is "trickle down economics" -- basically the hope is that paying an Afghan with political connections $500k to do work will lead to $500k staying in the local economy, paid in wages to other Afghan laborers, etc. It turns out it turns into the connected Afghan spending the same $5-10k to get the project done, and then buying $490k in real estate in Dubai or Canada.
Government and government contracting is the opposite of the startup world in every way. Very few people seem to have done both.
There are a lot of issues with the way aid is distributed. The most successful parts of both places, the wireless networks and decentralized power systems, have been totally market driven, and actually hindered more by centralized efforts and regulation than helped.
Somalia, where it is too dangerous and chaotic for aid agencies or telco regulators to operate, has cheaper Internet access and mobile calling than almost anywhere else in Africa or the developing world!
Communications infrastructure (roads, signals/fiber/wireless, electrical grid) is one of the few areas where scale can be helpful. In Afghanistan there is a lot of money coming through the military to pay for a road network (with benefits to both the foreign militaries and the local economy; dual use technology at its best). Communications is basically wireless backed by satcom, with some microwave backbone and upcoming fiber, with microwave/fiber driven by military and government customers paying top dollar, then going out to connect up the cell systems. For $20/mo you can buy 3G service.
One of the biggest issues in Afghanistan is that aside from the US Military, the other US funding agencies (primary US AID) have to operate through local contractors, and a rather inefficient process which leads to overpaying for everything, and thus not accomplishing enough concentrated change to be successful. The "Provincial Reconstruction Teams" are afraid to go out except with a huge military presence, and are pretty sheltered from everything, whereas I was riding around in an old pickup truck with a 9mm and no problems.
Irrigation systems are probably the next thing to do after roads (the military has roads under control); these can be done for $5-12k each by local labor, and 20-30 systems per district, all in parallel -- or, you can pay $500k via US AID and get 1-2 systems done after 2 years. The US AID requirement to use local contractors is "trickle down economics" -- basically the hope is that paying an Afghan with political connections $500k to do work will lead to $500k staying in the local economy, paid in wages to other Afghan laborers, etc. It turns out it turns into the connected Afghan spending the same $5-10k to get the project done, and then buying $490k in real estate in Dubai or Canada.
Government and government contracting is the opposite of the startup world in every way. Very few people seem to have done both.