Broadband Mapping NOFA: "NOFA" Enough, But Not "BADFA" A First Try

Anyway, let's have a look at the good, the ugly, and the indifferent of the information that prospective mapping fund-seekers are being asked to provide by the NTIA.  First, the indifferent--the static/backward-looking inventory of broadband availability by census tract, speed advertised, "typical" speed available during a peak period, and technology by which the service is offered.  This is my potentially minor source of criticism, if you could even call it that.  The BDIA requires the FCC to conduct periodic consumer surveys seeking the very same information (Section 103(C) of BDIA, 47 USC 1303(c)).  Hopefully, the Commission will do the consumer surveys on the "census tract" level, so that the NTIA doesn't end up paying "mappers" for data that the FCC is already required to gather. If anything, the NTIA should offer to use some of the $350 million to get the FCC to modify its survey to conform to the NOFA requirements.

Now, for the "good" part of the NOFA--I was pleasantly surprised to see broadband demand, and not just service availability, make its way into the data being collected for the map.  Appendix A, (3) (a) and (b) include aggregation points--both "last mile" and "middle mile/backbone transport", respectively.  This is good and deserves credit, because this is a "value add" from the NTIA staff--neither the BDIA nor the Recovery Act really recognizes the importance of wholesale demand, and aggregation points as an incentive for the private deployment of non-last-mile, broadband transmission facilities.  Additionally, the data collection also includes demand from "community anchor institutions" (Appendix, A, 4.)), which is also a good inclusion. The "market" was already moving in this direction, prompted by the promise of broadband grant assistance.  Leading wholesale carriers, like Level 3 Communications and Zayo Bandwidth were actively seeking out aggregators of retail bandwidth demand in underserved areas as project partners, and the NTIA's decision to focus on providing demand information serves to facilitate more private deployment.

My only two suggestions here would have been for the NOFA to: 1) have asked for reasonable short-to-near term (like one year out) projections of future demand by all of the aggregation point consumers, and 2) possibly the inclusion of demand, and projected demand, from the top 5, non-residential consumers, not already covered on the carrier side (the aggregation points in App A, 3) (a) and (b) and the community anchor institutions covered in section 4 of Appendix A.
 
Finally, . . . the just plain "ugly."  While both the Recovery Act and the BDIA authorize/require the NTIA to create a broadband map, neither law (nor the requirements for the map outlined by Congress) charges the NTIA with collecting pricing information.  In fact, while the Recovery Act, makes no reference to pricing information, in the BDIA, Commerce is the only agency not charged with collecting pricing information.  The FCC is required to collect information about the amounts consumers pay per month for broadband service (Section 103), the Comptroller General is charged with determining the average price/megabit/second of broadband service (Section 104), and the Small Business Administration Office of Advocacy is required to analyze the impact of speed and price of broadband services on small businesses (Section 105).
 
However, despite all of the various formats in which pricing data is already being collected by other agencies, the NTIA--quite inexplicably--chose a new, different, undoubtedly more expensive (because it's not already being collected), and completely irrelevant, format to task the "broadband mappers" with gathering . . .  "average revenue per user" ("ARPU"). 

There are many variables that make this information useless to any would-be entrant, or consumer, but difficult, costly, and competitively-sensitive for broadband service providers to report.  Moreover, the NTIA asks the broadband mapping applicants to collect ARPU information from service providers only with respect to residential users--though it is total demand on a network (from all sources, not just residential) that determines the average cost per user.  Differences in the network provider's choice of technology, too, can influence the average cost per user . . . and entry by a prospective broadband service provider depends on both costs and revenue.  But, what if the "revenue" attributable to broadband services (assuming it were easy to allocate, and that residential consumers did not typically buy broadband services in conjunction with other services at a "bundled" price) wasn't even "real"?

As we noted in an earlier post, last month, the FCC issued a report to the House Energy and Commerce Committee, listing the highest "per line" recipients of Universal Service Support from the High Cost Fund. One company reported receiving almost $17,000 per line, per year; many others reported well over $1,000 per line, per year in high cost support.  Moreover, USF support (revenue?) is unique to the carrier receiving it; e.g., a "competitor"--regardless of technology--would not be eligible for any new high cost support, much less the amount the existing recipient is recovering.  Despite its significance as an economic matter, the "per user" amount of a carrier's USF support is not addressed at all in the Broadband Mapping NOFA "ARPU" calculation instructions (Appendix A(2)(a)(6)).  You might note that USF doesn't support broadband services, but it doesn't really matter, because broadband services are provided over the same network that gets USF support for voice services.

But, I've already digressed too much.  The bottom line is that, of all the broadband pricing data already being collected by other government agencies, the NTIA decides to pick its own version of pricing data for its broadband mappers to try to collect . . . which, in itself adds expense, but the particular pricing data the NTIA chose is the most resistant to comparison among geographic areas, technologies, or network service providers.  In other words, the "juice" just isn't worth the "squeeze."

Still, to stay positive, the NTIA did a pretty good job of trying to create a map that will continue to be valuable for network planners, and not just policy bickerers.  A couple of "tweaks" would probably stretch the ability of the map to produce value to service providers and consumers, and cut expense.  So, to quote my old Brooklyn buddies, watching a James Bond movie . . . if 'dis wuz me heeeya, 'dis is wut I wood do: 1) use your original information requests to expand your existing, demand side information requests by: a) expanding the consumer base for which demand is sought (per census tract) to include the highest, non-carrier, non-community-anchor-institution, commercial users, and b) ask for 1 year out projected demand; and 2) use the standard "off the shelf" pricing info the FCC is already required to collect under the BDIA.