Build for the Future. It would be a huge mistake if a rural broadband expansion builds only to meet today’s definition of broadband. Cisco recently said that the average home today needs about 24 Mbps to meet their needs, which is nearly identical to the FCC’s current definition of broadband. Historically we have seen broadband speeds for customers double about every three years. But Cisco’s latest broadband report suggests this might have slowed down to about every four years. Cisco predicts by 2020 that households will need almost 50 Mbps. Look out a decade from now and the math says that households will need over 140 Mbps.
It would be totally irresponsible to spend billions of federal dollars to build infrastructure that will be inadequate by the time it’s installed. The current CAF II program is a travesty because it is spending billions on DSL and cellular data to achieve 10/1 Mbps speeds and won’t even be completed until 2021. CAF II is not building broadband infrastructure – it’s spending gold-plated federal money to build a lead solution for rural broadband. It’s not going to be very long before all of the rural people getting CAF II networks will be screaming again for something better.
This means a federal broadband program should not be used to fund cellular wireless, point-to-point fixed wireless or DSL. Those technologies all have a place in the marketplace today, but they can’t come close to meeting tomorrow’s needs, so let’s not toss away billions of tax dollars on the wrong technologies.
Use Federal Loan Guarantees. A federal broadband program does not have to rely only on matching grants. The federal government has several loan guarantee programs that can be expanded to bring banks into the funding process. Banks love loan guarantees because they greatly reduce the risk of projects by having the federal government act as the backstop for bad loans. If the review process is done well and funding is only given to companies with a good chance of success, then there should be few loan defaults and the loan guarantee program would cost the federal government very little.
Don’t Forget the Small Towns. It’s easy when looking to fund a rural broadband solution to concentrate only on areas that are categorized as either unserved or underserved. But business plans to serve only the neediest customers are hard to make work. Rural business plans work best if they can also incorporate the small towns and county seats.
The stimulus grants ignored these towns because they are considered to have adequate broadband. That is shortsighted because small towns do not have networks that are up to snuff with urban networks. For example, they may have cable modems, but these little towns are unlikely to get upgraded to the next generation of cable electronics for a long time, if ever. If we want to have successful business plans then the funding needs to also cover the small towns in the middle of the unserved areas to help the service providers achieve an economy of scale.
Don’t Try to Serve Every Home. Any broadband program ought to have the goal of reaching the most homes as possible with the funding available. This means that there should not be rules that require that every customer within a Census block get broadband. Rural Census blocks can be large and can cover diverse topology. A census block might have most customers along a river valley with a few high up nearby mountains, or on the other side of a lake or river. In my experience when designing rural networks the hardest-to-reach 10% of the customers can easily represent 40% of the cost to build. If we want to stretch federal dollars we need flexible rules that allow for realistic business plans. There comes a point where the guy who built on the top of a mountain shouldn’t get broadband, just like it’s hard for him to get electricity or city water or other utilities. What matters more is stretching federal dollars smartly to serve as many homes as possible.