SpaceX just scored nearly a billion dollars to improve rural internet in the U.S despite not yet launching its satellite-based Starlink ISP nationally.
Yesterday, the FCC published the results of its Rural Digital Opportunity Fund Phase 1 auction, which would provide underserved or unserved areas across the U.S. with high-speed internet: at least 25/3 Mbps as currently defined by the FCC. The first phase of this auction awards $9.2 billion to many ISPs, big and small, over a 10-year time period, assuming they meet certain requirements. While traditional wireline-based broadband ISPs as a whole received the bulk of the grant—Charter received an incredible $1.22 billion, for instance—SpaceX’s Starlink received $885 million to expand its service in 35 states out of the 49 (plus DC and Northern Mariana Islands) it applied for in its short-form application.
That’s a great deal more than other satellite ISPs received. HughesNet received just under $1.3 million to expand its services in Rhode Island only. Rival company Viasat was not included in the winning bidder summary provided by the FCC. It’s possible the company did not participate in the auction.
Either way, both HughesNet and Viasat are ISPs that have been around much longer than Starlink, operating at scale. Starlink has yet to prove without a doubt it’s not only the best satellite ISP, but that it’s also a better alternative for Americans living in rural areas who can’t get wired broadband service. Current HughesNet and Viasat customers have to contend with spotty service under certain weather conditions, small data caps, low bandwidth, and high latency—everything that makes it hard to work and attend school from home in the middle of a pandemic. Starlink promises to fix these issues with its competing service, but it’s currently only available in an invite-only form and has not yet had to meet the demands of providing internet at the scale promised in its application.
SpaceX also received more federal funds than a few major wired broadband ISPs: CenturyLink was awarded $262 million; Cox just $6.6 million; and Frontier $407 million. AT&T, Comcast, Spectrum, and Verizon were not listed on the winning bidder summary.
It also received more money in several states than some companies who have been working tirelessly for years to provide internet to rural areas. Connecting Rural America received just $10.3 million to expand broadband access across Alabama, while SpaceX received $54.6 million to expand its service in the same state.
The process in which the FCC decides how much money to give what companies is shrouded somewhat in mystery and involves some complicated math. We know that ISPs must meet certain requirements before they apply and after they receive funds, and the FCC does make its auction bidding tutorial publicly available, but it’s not the easiest thing to understand.
Qualified ISPs have to “bid” to receive any money from the FCC’s Rural Digital Opportunity Fund. A bid in the case of this specific auction is “a request for support in a given area for a given performance tier and latency at a given percentage,” according to the tutorial. In each bidding round, qualified bidders like SpaceX bid to provide service in exchange for support at percentages associated with the round—or plainly, ISPs bid for certain census blocks or areas of the U.S. where they want to provide service. The greater the bid, the more census blocks they commit to providing service to.
Bids correspond to a certain amount depends on the “reserve price of the area, the performance tier and latency waits, and the bid’s percentage.” Here comes the complicated math part:
For ISPs to figure out their implied support amount—or the amount they want to receive from the FCC—according to the equation above, they have to plug in the reserve price of an area, which can be found here, combine the performance tier and latency weight numbers from their application, and the bid percentage is a percentage of the reserve price, which is determined over the course of the bidding process.
Once a round closes, the winning ISP is required to provide service to the area that they “won.” Winning bidders’ support amounts will be at least the amount they agreed to accept by bidding. Support cannot be greater than the area’s reserve price.
Taking all that into consideration, SpaceX was not only one of the highest bidders at the auction, but it committed to providing internet service to a sizable chunk of the country that is currently underserved or unserved.
According to the auction procedures, each applicant that is awarded money from the Rural Digital Opportunity Fund “must make these services available to the required number of locations associated with the eligible census blocks covered by its winning bids.” Applicants also much provide a minimum speed of 25/3 Mbps, at least 250 GB of monthly data or U.S. median (whichever is higher), and “at least 95% or more of all peak period measurements of network roundtrip latency must be at or below 100 milliseconds (ms) (low latency) or 750 ms (high latency).”
The speed and latency requirements are straight forward—and SpaceX says it meets those requirements—but the problem with basing the amount of grant money awarded by eligible census blocks is that the FCC is working from incorrect data. An error on Form 477, which ISPs use to report how many census blocks across the U.S. are covered by their services, artificially inflated coverage numbers by millions. The FCC rolled out a new form that supposedly fixed that issue, but in choosing how it distributed Rural Digital Opportunity Funds, the FCC was working from 2018 data—data that’s based off of the old, erroneous forms.
“This systematically overstates service. It also leaves disconnected millions of people across the country who get stuck when our maps say they have broadband when they clearly do not,” said FCC commissioner Jessica Rosenworcel in a June 9, 2020 statement, dissenting in part of the planned auction. “Despite having made no efforts to improve our nation’s dubious broadband data, with the new Rural Digital Opportunity Fund we are about to hand out billions in broadband support based on maps we know are wrong.”
The other issue is that SpaceX is committing to provide service in areas where the residents can’t always afford to spend a lot on internet. The company’s new Starlink satellite service currently costs $100 a month, and the users have to shell out an additional one-time payment of $500 for equipment. Those costs are out of the question for many rural residents. A wired broadband solution, or even a wireless 4G solution, would likely be abetter option based on cost alone.
Starlink also suffers from periodic outages, according to Business Insider—not entirely surprising considering the service is still in beta, but not great either. SpaceX is hiring engineers to iron out these issues, but will they be fixed by the time SpaceX has to start rolling out its service according to its agreement with the FCC? People living in rural America already have enough trouble with spotty satellite internet. They don’t need another status-quo service.
Currently, Starlink is available by invite only, and service is only available in “the northern U.S. and Canada,” according to the website. The Starlink subreddit keeps an updated list of where users have been accepted into the beta, as well as a list of known bandwidth speeds. Unfortunately, it looks like most of the users who have reported their speeds are located in major cities.
No one type of ISP will help solve the digital divide in the U.S. Satellite alone cannot solve it. It will take a patchwork quilt of wired and wireless internet to reach the furthest corners of the States. But judging by how much money SpaceX received, it seems like the FCC is going all-in on the company’s ability to provide consistent, high-speed wireless internet even though its service has not been tested at-scale and will likely suffer the same pitfalls as other satellite ISPs as more people use the service, like constrained bandwidth speeds and higher latency. Only time will tell if that $885 million awarded to SpaceX will be money well spent or money wasted.