Institutional Networks and Cherry Picking

My friend, Geoff Daily at App-Rising.com, has questioned the wisdom of running fiber to all anchor institutions.
There's been a lot of buzz around the benefits and relative viability of wiring all community anchor institutions (schools, libraries, hospitals, etc.) with fiber as the way to get the best bang for the broadband buck. But recent conversations with my fiber-deploying friends have led me to worry that doing this could be a big mistake. ... The reason is simple: if you build a network to serve community anchors, then those institutions won't be available to serve as anchor customers for a community-wide deployment. Without those community anchors as customers, the economics of deployment, especially in rural areas, becomes much harder and may actually make robust, sustainable broadband impossible in some areas.
This is a question I have wrestled with also, in trying to help communities understand the real impacts of decisions they make on whether to build their own broadband network. My first reaction is on philosophical grounds - public institutions like schools, police departments, etc., do not exist to prop-up the business models of cable or telephone companies. Large entities like municipal and county governments should own their own network because it will save them money and expand their capabilities. When will the tea-party protesters start protesting government paying exorbitant fees to telephone companies for slow T-1 lines and the like? After all, these are our tax dollars and they should be spent wisely. My second reaction is that I seriously doubt removing these institutional networks will impact the business model significantly. Maybe it would have last decade, but now we know that Comcast and probably many more have ">massive margins in their broadband operations. Losing the libraries and schools will do little to their bottom lines. Even if it takes a bit out of their profits, they won't go missing meals. But really, the answer is more complicated. Many municipalities already get "free" services from their cable company as a part of the video franchise. To gain access to the right-of-way, cable companies have often given "free" (meaning paid for by the subscriber base) services via an I-Net. Though this has been helpful for communities it was never a particularly fair, efficient, or rational means of solving connectivity issues for local governments. It wasn't fair because cable subscribers paid for the costs of local government that should be paid by all citizens. It wasn't efficient because cable companies often did not live up their responsibilities or franchises did not require modernization of networks over the many years of the agreement. And it wasn't rational because neither entity had an incentive to build the kind of network local governments need to do their jobs effectively. But the right-of-way is a valuable asset and communities should have the freedom to negotiate access to it as they choose. Those choices are also constrained by what state and federal laws allow (I said this was complicated, right?) So - getting back to the question of whether building fiber to these public buildings is a good idea or not, I say it absolutely is ... if it is locally owned and the local community is responsible for it. In the unlikely event that such a network causes private companies to cease investing in the community (though continue refusing to invest in the community is likely a more accurate description), the community should take initiative to build the last-mile networks necessary for future vitality. Either this is an essential infrastructure or it isn't. If it is, local governments must take a stronger role in ensuring everyone has access. If it isn't essential, then we can continue watching private companies deploy networks wherever they decide it is profitable. Update: In an attempt to be more clear, I will say that I think federal policy should make it a priority to make funding available (loans where possible, mixing in grants where absolutely necessary) so that local communities can connect their anchors. Local ownership is paramount. Statewide networks are a poor approach in that it would de facto prevent communities from building their own networks. I don't think these networks will interfere with business plans of those private companies who have already made investments - but I also don't think this should be a major concern because local government's mission is to serve the needs of the community, not those of absentee-owned cable or telephone companies. To the extent that people in the community need better networks, local government must be ready to step in -- just as they do with roads, water treatment plants, and other elements of infrastructure.