The House Oversight Committee recently decided to investigate the FCC’s decision in 2022 to deny RDOF funding to Starlink. Anybody who has been reading this blog knows that I don’t take political positions, and the timing of this announcement clearly has political overtones since it was announced as Elon Musk took the stage at a campaign event with one of the presidential candidates. However, there were a few technical and policy issues raised recently about Starlink that means the issue is worth discussing.

I should make my own position on Starlink known upfront. I happen to live in a city, but if I was in a rural area with no good broadband option, I would have been one of the first people on the Starlink waiting list years ago. I think the broadband they bring into areas with no other options is awesome. Their products aimed at folks who are largely mobile are also unique and valuable.

The House announcement praised Starlink for helping the areas destroyed by Hurricane Helene. I live in western North Carolina, and Starlink got some good press here from its willingness to bring broadband to help areas without it. Starlink offered a free month of service, and I’m sure they gained new customers through that offer. But a new subscriber still had to pay for the receiver, and the logistics of getting a receiver delivered to rural areas post-hurricane were daunting – there are still many places that can’t be reached by vehicle. Also, unfortunately, the areas that need broadband the most are still without power.

But back to the issue of RDOF. The subsidy was awarded using a reverse auction, where the ISP willing to take the lowest amount of subsidy winning the funding. RDOF had two key requirements – the ability to deliver broadband of at least 100/20 Mbps, and the ability to serve every home and business inside a Census block that got the award.

The FCC finally decided to reject Starlink on the speed issue. When Starlink applied to enter the RDOF reverse auction, it said it could meet the 100/20 Mbps speed goal. And it looked for a while like it might. In the first quarter of 2021, Ookla says Starlink had an average U.S. speed of 66/16 Mbps but improved to 91/11 Mbps by the second quarter of 2022. However, since then, the speeds have slipped, and for the whole U.S. Ookla says speeds were 67/8 Mbps at the end of 2023 and a little slower in the first quarter of 2024.

Starlink has done a phenomenal job launching satellites and now has 6,371 functional satellites in orbit. But Starlink is like any ISP might be a victim of its own success – the slower speeds over time probably reflects that the number of customers has outpaced the number of satellites. The FCC was justified in rejecting the RDOF on the speed issue – the FCC 100/20 Mbps goal was not aspirational, but a real technical requirement.

The FCC could also have rejected Starlink on the coverage issue as well. One of the benefits of winning the RDOF subsidy is that winners were protected from any other ISPs receiving subsidies to build in the same Census blocks. For all practical purposes, RDOF winners in rural places were granted a monopoly – but for that award they were expected to be able to serve everybody.

In the early years, when it didn’t have a lot of satellites, Starlink put prospective customers on a wait list. The wait lists have disappeared, and Starlink is now willing to serve almost everybody in the U.S. – although there are a few places with geographical challenges where the company doesn’t claim the ability to serve.

The issue with coverage is that not everybody is a good candidate for Starlink. I live on a steep hill and am surrounded by other hills in all directions. I would guess that 35% to 45% of my view of the sky is blocked. My situation is not unusual in Appalachia, and much of the mountainous West and Southwest. If Starlink had only applied for RDOF in relatively flat areas, this would not have been an issue. However, Starlink won a lot of geography in RDOF in places like western North Carolina and the rest of Appalachia.

Starlink reception is also hindered by heavy foliage. The suggested solution for this is to put the receiver where this is not an issue. But there were places in the RDOF award areas that are fully covered by a forest canopy.

Interestingly, Starlink is back in the conversation in the BEAD grant process. Those awards are supposed to bring a broadband solution to every unserved and underserved location in the country. In every market, there are some locations where building other technologies is infeasible. NTIA prudently decided recently that Starlink is probably the only realistic solution for such places.

Unfortunately, some states say they don’t have enough money to reach everybody with fiber and/or wireless, and Starlink might get the final laugh by winning more subsidy from BEAD grants than it was denied for RDOF.