Fast Net News, DSL Prime, DOCSIS Report

The House stimulus bill included $B for wireless, a major surprise because most of the public advocates where thinking the money would go for fiber and landline subsidies. One key reason is that a CTIA study on "ubiquitous 3G wireless" was circulated around the transition team, along with a comment that spending perhaps a quarter of the stimulus money on wireless towers would be by far the most effective use of the money. Erecting towers and providing backhaul and in two years "The President will be able to announce 98-99% of American homes can connect at megabit speeds." That was very attractive to the politicians. The several dozen decisionmakers had many motivations, but the notion that wireless towers were the best way to reach the unserved was powerful. In the rush to get a final bill, the conference committee stripped nearly all the details from the final bill, but the NTIA is very aware of the issue.

The Costquest study estimated "ubiquitous" wireless broadband would require about 16,000 towers. At $500K/tower including backhaul, that's perhaps $8B. There are other costs, including equipment, but also additional income from the additional customers served.  The country can reach "near-ubiquity" with significantly fewer towers, bringing the cost closer to $5B. When I wrote in January, I estimated that a 25% subsidy would be sufficient, and less than $2B from the stimulus would get us to 98-99%. Working on a statewide wireless proposal, I discovered the revenues from those towers would likely be less, and a somewhat larger incentive would be required.

The world is going mobile, so ensuring wireless everywhere is good public policy. A wireless build is much cheaper and faster than rural fiber/coax, and is a natural complement. I believe 50+ meg to every home is worthwhile while LTE and WiMAX speeds will typically be 2-10 meg. We need both.

Stimulus money is one way to make this happen, but the FCC's Adelstein has a proposal that doesn't require public money:"Use It Or Lose It" wireless licensing. By far the least expensive and most sensible way to get near universal service with voice and megabit wireless and a pretty obvious step forward. When wireless licenses come up for renewal, require that the operator actually is servicing the territory. Perhaps require 92% population coverage for the first renewal, about the current U.S. average. Raise that to 98% coverage for the second renewal, with very limited exceptions. Since U.S. wireless licenses are typically 10 years and most territories have several, in a few years 98+% would be covered without a penny of government subsidy.