The FCC released the first draft of its broadband map on November 18. The FCC staff deserves massive props for releasing the draft data on schedule and in such a clean format. In this post, we look at some of the high-level takeaways at the state and county level.
Before discussing the map, it is important to emphasize that because it is a draft, it is likely to differ–perhaps significantly–from the final version, at least as it applies to allocating BEAD funds. It still has to undergo the challenge process. This process is an integral part of putting together an accurate map. All data are measured with error, but this process helps reduce that error. Additionally, the map does not yet account for areas that have already been allocated broadband subsidies, for example through RDOF or ARPA. Those areas presumably now appear as unserved, but will be considered served for the purpose of allocating the remaining $32 billion of BEAD funds.1
In this first look, we use the data to show the share of fixed broadband serviceable locations (BSLs) by state and county and compare those estimates to those in the FCC’s Form 477 data. (And hat-tip to Mike Conlow for beating everyone out of the gate with the new data.). The comparison is not quite apples-to-apples for two reasons. First, the most recent Form 477 data is as of June 2021 and the new map data is as of June 2022. Second, the units of observation are slightly different. The Form 477 data shows households (and housing units) with access; the new data uses locations with access.2
The figures below show how the Form 477 estimates differ from the fabric estimates using NTIA’s definition, for the purposes of BEAD, of “reliable broadband,” which excludes all satellite and fixed wireless service using unlicensed spectrum. (Drag the slider to see the difference).
Additionally, because coverage is not a yes/no proposition, different definitions yield different estimates of the share of locations covered. The BEAD program only counts “reliable broadband,” which excludes all satellite and fixed wireless provided over unlicensed spectrum. By this measure, the new data shows more of the country unserved than the Form 477 data does. However, if we expand the definition to include anything that provides low-latency service, coverage rates are higher than shown in 477, largely due to Starlink, but also to wireless ISPs that used unlicensed spectrum.
The figure below shows estimates of coverage based on different definitions compared to Form 477. The first column shows the share of locations with access to 25/3 service under the BEAD definition of “reliable broadband.” This column includes residential and business locations. The second column includes any technology that offers low-latency service. The third and fourth columns show coverage of only residential locations, first with the BEAD definition and then with low-latency providers. The final column shows estimates from the June 2021 Form 477.
implications for bead funding by state
The new map will be used to allocate the BEAD funds. Each state will receive allocated a share of the remaining $32 billion based on its share of the nation's total unserved population. The new map changes that distribution relative to Form 477. As a result, the amount each state can expect to receive will also change. The figure below shows these estimates. The state that loses the most money from the new estimates is Mississippi, which would receive about $770 million less than it would have based on the 477 data, while Texas gains the most, receiving about $1.3 billion more than it would have under 477.
To be clear, these will not be the final allocations. Final allocations will depend on the outcome of challenge processes and the unserved areas that have already received, or are set to receive, subsidies from other programs.
<<Table below 👇 added November 29: It's a little misleading to show dollars per state without any normalization. Here we show dollars per capita per state, which allows for more meaningful comparisons.>>
100/20 coverage
BEAD defines unserved as areas that cannot access broadband at 25/3 via specified technologies, and underserved areas are those that have 25/3 but not 100/20. The map below shows how estimates from Form 477 differ from the new map.
The figure below shows these differences across states and definitions.
We'll have more analysis based on these maps in the coming days and weeks.
Endnotes
- BEAD includes $42.45 billion. 2% to program administration; $100,000,000 to each state + DC and Puerto Rico, and $100,000,000 total to the other territories; $4.245 billion marked for high cost leaves $32,056,000,000 left to be distributed.
- This difference is probably small--maybe like comparing Granny Smith to Honeycrisp apples.
