Incremental Infrastructure

The PCR Project highlights an interesting article from the Boston Globe on using incremental infrastructure to improve conditions in Africa.  It starts with a great success story about an [entrepreneur] named Alieu Conteh who started up a mobile-phone network in the Democratic Republic of the Congo in the midst of a civil war.   "Now his phone network is the most important piece of infrastructure in the country, the only way most Congolese can communicate with their neighbors and with the wider world. Conteh's company, CWN, grew into Vodacom Congo, with 3 million mobile phone users and a market valuation of $1.6 billion."

Infrastructure is typically thought to require some level of public investment.  Why?  Because it tends to require coordination and tends to be available for all, whether they contribute or note.  However, as a friend of mine who's spent a lot of time in Africa has pointed out, many governments do a bad job with infrastructure.  The good ones will take on manageable projects then maintain them.  The bad ones ones will build super-highways then let them fall to pieces.

Incremental infrastructure provides a different economic model for infrastructure:

Successful Internet and phone projects suggest that there are at least three common characteristics of successful incremental infrastructure projects. These projects are atomic: A small part of the infrastructure is useful by itself, like a single mobile phone tower that allows people in a single city to make calls to one another. The projects are financed in part by users, lowering the costs for the operator: Mobile phone users buy their handsets and Internet users purchase their own computers. Finally, these projects are providing capabilities that weren't available before: they're new services, not an upgrade of existing systems.

Beyond mobile phones the article gives a few more examples of the potential of incremental infrastructure:

• "[H]igh-quality toll roads between a small village and a city."  (Although railroads don't work under this model).
• Mobile phone already have to invest in generators to run their services.  They could take the next step, upgrade their generators, and start selling power to locals.  (Ideally they'd use some renewable sources).
• "[Congo] has more than 200 airports, most of which are unpaved landing strips at the edge of dense forest... The investment to run one of these incremental airlines is modest -- a refurbished Antonov aircraft, two pilots, and a mechanic -- and their safety record is abysmal. But this air infrastructure makes trips that would take weeks by water possible in a few hours, and opens up trade between regions of the country that would otherwise be impossible to reach.
• These projects still will require millions.  "Governments that encourage foreign direct investment -- especially investment from their diasporas -- are more likely to see incremental infrastructure develop."

I buy this argument.  I think it's the equivalent of a scaled up version micro-finance.  That said, I'm no libertarian.  As the article points out, this technique can only work in certain cases.  As the overall complexity rises, there's going to be some need for government support of standards if nothing else. 

So why don't we see more of this already?

1. Foreign dollars might not have funded these projects in the past because in some ways they're inherently inefficient.  Infrastructure has economics of scale and can be freely made available to all citizens.  So part of the problem might be theoretically efficient government projects were competing with more achievable private projects. This actually comes up some in Iraq, I've heard that we continue trying to work with large power plants rather than a wide-spread network of generators because we don't want to admit defeat.
   
2. Technology is changing too quickly.  Technological advances and mass production of items like mobile phones and small airplanes take time to disseminate throughout the world.  Could be not everyone's figured out this trick yet.  Note that this wouldn't explain why toll roads aren't more common unless the trick is EZ-Passes and the like.
3. Security.  Incremental infrastructure still needs to be protected from those that would destroy or prey on it.  The incremental nature helps some, since there's a constituency that might be willing to defend it.  However, security limitations may kick in long before other forms of economic limitations do.  This I think is the most worrisome problem and one that any major pushes for incremental infrastructure needs to address.

[Update: Title added and typo fixed]