Rural America still does not have reliable internet. That is a real problem, not a talking point, and it deserves a real infrastructure answer. What it got instead, in a rule change to a $42.45 billion federal program, was a shift toward satellite dishes built by the same billionaire who spent the last two years sitting inside the government writing the rules that fund them. I do not think that is a coincidence, and I do not think Starlink is the right tool for this job even setting the conflict of interest aside.

Starlink has grown fast. It went from 4.5 million subscribers in early 2025 to more than 10 million by February 2026, and it is easy to see why in places that had nothing else. But the price keeps climbing, and it loses badly to fiber over any real stretch of time.

Starlink Fiber
Monthly cost 55 to 130 dollars, up 5 to 10 dollars in June 2026 50 to 100 dollars
Upfront hardware 349 dollars Typically included
Two-year total Roughly 3,379 dollars 1,200 to 1,920 dollars
Speed Baseline Two to ten times faster

Run the math over two years and Starlink costs nearly double what a comparable fiber plan does, before you even count the speed gap.

Capacity is the other problem nobody selling Starlink likes to mention. Each satellite serves a fixed geographic cell, and once that cell fills up, new customers in the area go on a waitlist regardless of how much they are willing to pay. A technology that runs out of room is not a serious answer to universal coverage. Buried fiber does not have that ceiling. Every additional household gets its own line.

Infrastructure you have to keep repurchasing

Fiber laid in the ground today will likely still be carrying traffic in forty years. Starlink satellites do not last that long. They sit in low orbit, degrade, and deorbit on a cycle of roughly five years, at which point they burn up in the atmosphere by design, and SpaceX has to build and launch a fresh one just to hold the network at its current size. Each satellite runs somewhere between 2.5 and 3 million dollars once you count manufacturing and launch. With close to 11,000 of them in orbit as of mid-2026, that is nearly 30 billion dollars already spent just to stand the current constellation up, and every satellite in it is on a five-year countdown to reentry. That is not a network you build once. It is a network you rebuild in full roughly every five years, forever, and every rebuild is a fresh round of revenue for the company that owns the only supplier.

Every dollar of public money that goes toward satellite service is a recurring payment to a private company for infrastructure that never becomes a public asset, and that SpaceX has to replace on a timer no matter how well it performs. Every dollar that goes toward fiber buys something a community will own and use for decades after the check clears.

The rule change and who wrote it

The program at the center of this is BEAD, the Broadband Equity, Access, and Deployment fund, 42.45 billion dollars passed in 2021 specifically to close the rural broadband gap. It originally favored fiber, for exactly the durability reasons above. In June 2025 the Trump administration rewrote the rules under a new “technology neutral” framework it called Benefit of the Bargain, opening the door for satellite and fixed wireless providers to compete for the same money on equal footing with fiber. Fiber still took the largest share of planned locations at 63 percent, but low earth orbit satellite, meaning Starlink, the only company operating at that scale, picked up around 23 percent, with fixed wireless taking most of the rest. NTIA framed the change as saving 21 billion dollars. Critics framed it as diverting billions away from permanent infrastructure and into a subscription service.

A conflict of interest nobody bothered to hide

Elon Musk ran the Department of Government Efficiency during the exact window this rule change took shape, with real influence over federal agency decisions, while owning the one company best positioned to benefit from redefining “technology neutral” broadband policy. Congressional Democrats demanded an investigation. Watchdog groups called it out by name. None of that stopped the rule from taking effect. This is not a subtle case of regulatory capture happening somewhere behind closed doors. The person shaping government spending rules and the person who owns the company collecting the money were the same person, in public, while it happened.

What actual rural infrastructure looks like

The United States has done this before and gotten it right. The Rural Electrification Act of 1936 did not subsidize kerosene lamp companies to hold rural America over. It built power lines, publicly, deliberately, into places private utilities had decided were not profitable enough to bother with, and those lines are still standing. That is the model for broadband too: direct federal investment in fiber trunk lines and 5G tower density in the places the market has decided are not worth serving, built to last, owned by the public or by public-adjacent cooperatives, not rented indefinitely from a satellite company because a rule change made it easier to qualify for the money.

This is not an isolated broadband story either, and it is what got me thinking about this piece in the first place after watching Technology Connections lay out the same pattern for renewable energy. The same administration spent 2025 and 2026 withdrawing offshore wind lease areas, halting construction on projects already underway like Empire Wind and Revolution Wind, and signing a bill that guts the tax credits solar and wind projects have relied on since the Inflation Reduction Act. Courts have blocked some of it, ruling the wind shutdown illegal, but the intent is consistent. Durable public infrastructure that would outlast a single administration, whether it carries electricity from a turbine or broadband from a fiber line, keeps losing to arrangements that funnel public money toward private companies on a recurring basis instead.

Where that leaves rural broadband

Starlink can fill a real gap for the household that truly has no other option, but it should be the exception the country reaches for after it has tried to build something permanent, not the default a $42 billion federal program gets steered toward because the person writing the rules also owns the satellites. Rural America does not need a subscription. It needs wires in the ground and towers on the hills, paid for once and owned for good. What’s your take on how rural broadband should actually get built? Find me on Bluesky.