| Getting Mapping Right |
| Monday, 14 September 2009 20:42 |
 The primary purpose of the U.S. broadband map is to help bring broadband to the “unserved,” as Obama promised during the campaign. The early state maps lacks some crucial information about the “unserved” areas that I hope the $100M project quickly resolves. The current maps have almost no information about the facts in the “unserved” territories. Since these are the focus, we need to know what facilities are available.
Which telcos offer voice in the broadband “unserved” territories? Which cable companies are in the area? Are there wireless towers that can be used? This is basic stuff needed to target the spending. Nearly half the homes can get cable TV but not Internet and can be upgraded for less than $500/home. For about a quarter of the $7.2B stimulus dollars, 3M and possibly half of the homes can be offered 50 megabits. That's just a no-brainer. If towers are available, wireless should also be inexpensive to offer, albeit at 5-10 meg rather than 50 meg. Between those two groups and a patchwork of cable/fiber extensions, DSL remotes and extenders, etc. the $7.2B can bring decent speeds to 70-85% of the “unserved.” Doing less is wasting public money. NTIA deserves credit for bringing the cost of broadband mapping down from a budgeted $240M to what is now expected to be $100M. That's still 3 or 4 times as much as it should cost according to actual costs in several states and the experts I and the AP consulted, probably because the states performing the work have little relevant experience. In a compromise with the telcos, NTIA effectively prohibited competitive bids for the work. They allowed only “non-profits” to bid, knowing that only one “non-profit,” the massively AT&T supported Connected Nation had significant experience. AT&T among others had threatened to sabotage the mapping effort, and Larry Strickling decided not to fight the carriers on this because he needed the maps so badly. It would have been trivial for any major carrier to tie things up in court until after the stimulus money was gone.My reading of the 2200 summaries confirms some of the worst fears of those who don't want to waste public money. Nearly never are there two or more bids. The system of oversight needs immediate change to deal with negotiating the single bidder contracts. Mark Seifert provided two additional explanations for the higher spending. NTIA is sensibly getting data like locations of schools and fiber connections, which does increase the cost. Seifert also told AP costs will be high because “NTIA wants the information to be independently verified.” http://bit.ly/TYsc7 The carriers generally have excellent data on where they do and do not offer service, because otherwise they would be massively taking orders they couldn't fulfill or turning away valuable customers. That's easy to confirm by checking a few addresses with one of the reseller websites or the companies directly. Verizon's records in 2000 were atrocious, part of the DSL Hell problem, but nearly all problems like that were corrected years ago. A handful of incompetent carriers may have bad records, but the states can identify those in a few days checking addresses. Essentially, Seifert is saying the government is spending $millions because the carriers are lying. It shouldn't cost millions to verify, unless NTIA has reason to believe the carriers are submitting fraudulent data. I hope he turns out to be wrong and early checks of the data show such extensive auditing isn't necessary. That would allowbringing the cost even lower. Creating a broadband map of the U.S. should cost less than $30M if done frugally, as I reported in May http://bit.ly/bOkCZ and August http://bit.ly/YwSX7. Joelle Tessler at the AP advanced the story with several expert opinions confirming that estimate http://bit.ly/TYsc7. |