In March of 2021, the Federal Communications Commission concluded what is now known as the C-band auction. Three carriers — Verizon, AT&T, and T-Mobile — collectively spent $81 billion to acquire licenses to transmit on a slice of mid-band spectrum between 3.7 and 4.0 gigahertz. The auction took twelve weeks. It generated more revenue than any prior spectrum auction in American history.
I have been thinking about this auction lately, because the C-band spectrum is what makes the 5G coverage on my phone actually feel like 5G. The marketing got ahead of the network for several years. Towers were upgraded; phones were sold; the speed indicator on the screen changed from "LTE" to "5G." The actual speeds did not improve much. The C-band auction is the reason they finally did.
Why spectrum is auctioned
Radio spectrum is, in a literal sense, a public commons. The frequencies through which all wireless communication moves — radio, television, satellite, mobile, Wi-Fi, garage door openers, baby monitors, military radar, weather radar — are limited. There is only so much electromagnetic real estate available. And different parts of the spectrum have different physical properties. Low frequencies travel far and penetrate walls but carry less data. High frequencies carry enormous amounts of data but stop at the first inconvenient leaf or window.
The federal government holds the spectrum in trust for the public, then licenses it out for commercial use. The licenses are auctioned because auctions, in theory, allocate the resource to whoever will use it most productively. The companies willing to pay the most are presumed to be the ones with the best plans for using the spectrum.
This is mostly true, in the way that all theories about markets are mostly true.
The political layer
Spectrum auctions are deeply political in ways that are easy to miss if you read only the trade press. The decision of which spectrum to auction — and on what terms, with what coverage requirements, with what build-out timelines — is a federal policy decision that shapes the structure of an entire industry.
The C-band auction is a useful case. The 3.7-4.0 GHz band was previously used by satellite operators for distributing television and radio content to local stations. The decision to clear that band for mobile use required relocating those satellite operators, compensating them, and coordinating a multi-year transition. The total cost of that relocation, paid by the winning bidders on top of their license fees, exceeded $13 billion.
Why mid-band? Because mid-band spectrum is the goldilocks zone for mobile. Low-band spectrum (the kind T-Mobile got from its 600 MHz auction in 2017) covers vast geographic areas but offers modest speeds. High-band spectrum (the millimeter-wave auctions of 2018 and 2019) offers extraordinary speeds but barely covers a city block. Mid-band is the practical sweet spot — fast enough to feel different from 4G, far enough to actually deploy at scale.
Who won and what it meant
Verizon spent $45.5 billion in the C-band auction. AT&T spent $23.4 billion. T-Mobile spent $9.3 billion. The asymmetry tells a story. T-Mobile had already accumulated substantial mid-band spectrum through its acquisition of Sprint, which closed in 2020. Sprint had owned a large 2.5 GHz block — also mid-band, also goldilocks-zone — and that holding became T-Mobile's 5G foundation.
This is why T-Mobile was, for several years, ahead of Verizon and AT&T on actual 5G performance. Verizon and AT&T spent the C-band auction making up for not having mid-band spectrum. T-Mobile already had mid-band spectrum and could focus on deploying it.
The competitive landscape of American mobile in 2024 and 2025 is in large part a story about spectrum holdings.
The auctions to come
Federal spectrum policy is now in a strange interregnum. Several auctions are pending. Some bands previously held by federal agencies are being considered for commercial relicensing. The political question of how aggressively to free up spectrum for commercial use — versus protecting incumbent users like government, science, and military — is, as of 2026, genuinely contested.
What gets auctioned next, and on what terms, will shape the next decade of American mobile. The decisions are being made now, mostly out of public view, in regulatory proceedings whose outcomes most people will never read about.
The literary point
I find spectrum policy interesting partly because it is almost the opposite of how technology is usually written about. The story is not about brilliant engineers or visionary entrepreneurs. The story is about regulatory proceedings, multi-year auction designs, satellite operator relocations, and the slow grinding work of public policy. The technology is downstream of the policy.
And yet the policy almost never makes the news. Most people experience the C-band auction only as the moment their phone got faster.
The air around us has a politics. We mostly do not see it.
