The new FCC order takes 583,000 locations off the board for BEAD

Yesterday the FCC released a Report and Order creating the Enhanced Alternative Connect America Cost Model (EA-CAM, I guess) program which extends subsidies for rural ISPs for 10 additional years (beyond the remaining 5 years) at a cost of $1.27 to $1.33 billion annually to the FCC’s Universal Service Fund. In exchange, any ISP that elects this subsidy would be required to deploy 100/20 broadband to everyone in their service area. According to my estimates, if all these ISPs accept the EA-CAM offer and build fiber, cable, or licensed fixed wireless (more on that later), that takes 582,675 locations off the board for the BEAD program.

The FCC picking up the tab for some locations will significantly help some states. In Nebraska, the ACAM areas include 24,776 Unserved and Underserved locations, 23% of their total Unserved and Underserved locations. That’s a lot!

Let’s look at what the FCC is paying for these locations. At $1.27 billion new spend for 10 years, that’s $12.7 billion. In net present value terms that’s $10.8 billion to serve 583,000 locations, or $18,589 per location. Using my RDOF reserve price-based cost model, my estimate of the cost to serve these locations is $14,195 per location, or $8.3 billion total. (There are certain limitations to my RDOF reserve-price cost model. In particular, it likely underestimates cost for the most expensive locations (the highest cost location I have is $32,400 per location) and overestimates cost for partially served blocks and Unserved locations in dense areas.)

It’s possible that not all 583,000 locations will get a cable or fiber broadband service. In “declin[ing] to depart from the Commission’s long-standing policy of technological neutrality at this time”, the FCC will accept licensed or unlicensed fixed wireless to fulfill the 100/20 EA-CAM obligation. An ACAM ISP that deploys unlicensed fixed wireless would still find those locations in BEAD; by contrast fiber, cable or licensed fixed wireless would make the location ineligible for BEAD. I think most of the EA-CAM locations will get a qualifying BEAD technology (after all, it keeps a competitor out of the area), so I include the full 583,000 locations.

(I did a post on this proposal in March. Big picture, without a plan to fund the Universal Service Fund long-term, I remain skeptical of anything that adds long-term multi-billion dollar USF obligations. But, as I pointed out then, and am again here, taking locations off the board is helpful to high cost states that may struggle in BEAD.)

As an example, here’s the area around Fremont, Nebraska. The green shaded areas are ACAM blocks, and dots are shaded green for Served, yellow for Underserved, and red for Unserved. There’s an ACAM area to the southwest of Fremont that is Served by a fiber coop; an area to the northwest that is Unserved; and an area to the east which is Underserved with 25/3 DSL and unlicensed fixed wireless. If the ACAM ISP accepts the EA-CAM offer, which I suspect the vast majority will, and deploys cable, fiber, or licensed fixed wireless (including CBRS), all of these locations are no longer eligible for BEAD because they’ll have an enforceable commitment from another federal program to deliver BEAD-qualifying 100/20 broadband service.

This significantly improves Nebraska’s ability to serve residents with fiber. Nebraska was allocated $405 million in BEAD. In the previous estimates, pre-EA-CAM, they needed $840 million to reach everyone with fiber. When you exclude locations in ACAM areas as already served, Nebraska’s shortfall is basically cut in half: they only need $680 million to reach all the Unserved and Underserved with fiber. And we still haven’t factored in any ARPA spending.

Methodology and other notes: There’s a unit of geography called “study areas”. The links on this page to download Study Area data don’t work. Though I found a shapefile here. The files that are available of funded ACAM areas — and they are linked in the FCC’s R&O — use 2010 Census blocks as a unit of geography. So I use those Census blocks. (There’s a map below of the overlap between Study Areas and funded 2010 ACAM Census blocks. They mostly overlap). To match between the 2010 Census blocks and the current status of the location according to the National Broadband Map, I apply 2010 Census blocks to every location in the country.

The FCC order is 118 pages and represents months (years?) of back and forth between the FCC and ACAM ISPs. There are a lot of weeds. For example, paragraph 74 states:

Furthermore, because a primary purpose of this ongoing support is to ensure the maintenance (or improvement) of service to locations that would otherwise be unserved, we further extend the support for ILEC-only served locations to locations that were ineligible for prior A-CAM offers but which are not served with 100/20 Mbps or faster service by a competitor.

It’s possible this is saying there may be more locations that are eligible beyond the existing ACAM blocks.