Posts
The Auction
Machine
What Auctions Replaced
The spectrum assignment system that auctions replaced was not working. Before 1993, the FCC assigned spectrum licenses through two mechanisms: comparative hearings, in which competing applicants submitted detailed proposals and the FCC chose among them based on public interest criteria; and, beginning in the 1980s, lotteries, in which licenses were assigned randomly among qualified applicants. Both systems were slow, politically manipulable, and produced no price signal that connected spectrum value to spectrum use.
Comparative hearings could take years. The administrative record in a contested licensing proceeding could run to thousands of pages. Applicants had strong incentives to challenge competitors' proposals on technical grounds, to litigate FCC decisions, and to engage in the political process that influenced how the public interest standard was applied. The hearings were not corrupt in any simple sense. They were captured — shaped by the interests of participants with the resources to engage in years-long regulatory proceedings, at the expense of the public interest they were nominally designed to serve.
Lotteries were faster but produced their own pathology. Because the license had significant value and the lottery was random, applicants had no incentive to demonstrate that they would use the spectrum productively. The lottery winner's optimal strategy was often to sell the license to a carrier that could actually use it — creating a secondary market in lottery winnings rather than a primary market in spectrum use. The winners were not necessarily the users who valued the spectrum most. They were the users who got lucky.
The auction mechanism was a genuine improvement over what it replaced. Comparative hearings produced political capture disguised as public interest review. Lotteries produced random assignment and secondary market arbitrage. Auctions at least produced a price signal and assigned spectrum to users who were willing to pay for it. The problem is that "better than lotteries" is not the same as "adequate for governing a public resource."
The economic case for auctions was made most influentially by Thomas Hazlett, a communications economist who had argued since the 1980s that spectrum should be treated more like property — assigned to users who valued it most, with flexible use rights that allowed reallocation over time, and with interference rights that could be adjudicated rather than administratively managed. The Coasean property rights argument: if spectrum rights are clearly defined and freely tradable, they will flow to their highest-valued uses through market transactions without requiring the FCC to determine who deserves what allocation. The auction mechanism was a partial implementation of this vision — capturing the price signal benefit without fully implementing the property rights regime that the theoretical argument required.
$200 Billion in Public Resource Revenue
The auction record is the revenue record of the American wireless industry's construction — the sequence of spectrum transfers from public ownership to private operation that built the cellular infrastructure Americans now depend on. Each auction is also a concentration event: spectrum flowing from public allocation to the carriers with the capital to bid for it, deepening the oligopoly with each cycle.
Three Carriers — Most of the Spectrum
The auction machine's most consistent output, across thirty years and dozens of auctions, is concentration. Each major auction cycle has produced a spectrum landscape in which the three national carriers — Verizon, AT&T, and T-Mobile — hold the majority of licensed spectrum in the bands that matter for commercial wireless services, while smaller regional carriers, rural operators, and new entrants hold progressively less.
This outcome was not inevitable. The auction mechanism does not inherently produce concentration — it assigns spectrum to the users who value it most, and in principle, that could mean a wide variety of users across different bands and markets. In practice, the users who value licensed spectrum most in competitive bidding are the users with the largest balance sheets, the most established customer bases, and the most to gain from denying competitors access to the spectrum they do not win. The three national carriers meet all of those criteria. Small regional operators and new entrants meet none of them.
The concentration table above understates the competitive consequence because it does not capture the geographic dimension. National carriers hold licenses in virtually every market. Regional carriers hold licenses in specific geographies. The spectrum a regional carrier holds in rural Pennsylvania does not help a customer in rural Wyoming. The spectrum the national carriers hold is a national asset; the spectrum smaller operators hold is a local one. In competitive terms, this means that any customer who travels — any trucker driving I-81 from Pennsylvania to Tennessee, any traveler crossing state lines, any logistics operation spanning multiple regions — depends on the national carrier network. The regional operators serve their home markets. The nationals own the road.
Market structure: The U.S. wireless market is effectively a three-carrier oligopoly in national coverage. Verizon, AT&T, and T-Mobile collectively serve the substantial majority of U.S. wireless subscribers. The Sprint-T-Mobile merger approved in 2020 — over the objection of state attorneys general who argued it would harm competition — reduced the national carrier count from four to three, consolidating spectrum holdings that Sprint had accumulated over decades.
New entrant barriers: The capital requirements to compete as a national carrier are prohibitive for new entrants. A competitive national spectrum position requires billions in auction expenditure, followed by billions more in infrastructure deployment, in a market where the incumbents have already deployed networks and locked in subscribers. No new national carrier has entered the U.S. market since the auction system was established. The auction mechanism has been effective at assigning spectrum. It has been structurally hostile to the competitive entry that the public interest rationale for spectrum licensing was meant to promote.
Secondary market consolidation: The secondary market for spectrum licenses — where carriers trade, lease, and swap licenses after the initial auction — has accelerated concentration rather than dispersed it. Large carriers acquire secondary market spectrum to fill coverage gaps, expand capacity, or deny competitors access. Smaller carriers sell secondary market spectrum when capital pressure requires it. The secondary market is not a competitive correction to auction concentration. It is a mechanism for extending it.
DISH/EchoStar spectrum warehousing: DISH Network accumulated approximately 100 MHz of spectrum across multiple auctions over more than a decade — a position worth tens of billions of dollars — without deploying it at the scale its buildout obligations nominally required. The FCC's buildout requirements, the public interest condition that was supposed to prevent spectrum warehousing, proved enforceable only at the margins. DISH's spectrum position was eventually sold to T-Mobile and SpaceX in transactions that, whatever their competitive merits, demonstrated that spectrum acquired from the public via auction can be held as a speculative asset rather than deployed for public benefit without effective regulatory consequence.
What the Market Cannot Justify Covering
The dead zones on I-81 are not accidents. They are the predictable output of a spectrum governance system designed to maximize auction revenue in the markets where spectrum is most valuable — which are, by definition, the markets with the most people, the highest population density, and the most revenue per square mile. Rural America has fewer people per square mile than urban America. The market logic of the auction system produces less coverage in rural America. The physics of wireless propagation requires more infrastructure per subscriber in rural America. The combination produces the coverage map that every American who has driven a rural highway has experienced.
The auction system was designed to assign spectrum efficiently. It was not designed to ensure universal coverage. Those are different objectives, and the mechanism optimized for the first has systematically underdelivered on the second. The buildout requirements attached to spectrum licenses — the public interest conditions that nominally require carriers to deploy coverage in their licensed territories — have been the primary tool for addressing the gap. They have not closed it.
The trucker on I-81 running Channel 19 is not in a dead zone because the physics is wrong. The physics in rural Pennsylvania is exactly as it is everywhere else. The trucker is in a dead zone because the auction system assigned the spectrum that could cover rural Pennsylvania to carriers who calculated that the revenue from rural Pennsylvania did not justify the infrastructure cost of covering it. That is not a market failure. It is the market working exactly as designed.
The structural consequence of the rural failure extends beyond individual coverage gaps. The logistics industry — trucking, rail, agriculture, energy — depends on wireless connectivity for safety, efficiency, and coordination. The dead zones that a line haul driver experiences on a rural interstate are the same dead zones that impair precision agriculture, remote infrastructure monitoring, and the IoT sensor networks that modern supply chains depend on. The rural coverage failure is not a consumer inconvenience. It is an economic infrastructure failure whose costs are distributed across industries that do not appear in the FCC's auction revenue calculations.
The policy responses to the rural failure — the Universal Service Fund, the Connect America Fund, FirstNet for public safety communications, and the BEAD program for broadband — are subsidies layered on top of a market system that the subsidy programs acknowledge cannot, on its own, produce the coverage the public interest requires. Each subsidy program is an implicit admission that the auction mechanism's public interest claim was incomplete. Each represents additional public expenditure to correct the coverage failures that the auction system's revenue maximization logic produced.
How Temporary Permissions Became Permanent Property
The Radio Act of 1927 was explicit: licenses are not property rights. They are temporary permissions, revocable by the government for failure to serve the public interest. The Communications Act of 1934 preserved this framework. The FCC has repeated it in licensing orders for ninety years. No court has found that spectrum licenses create compensable property rights that the government must pay to reclaim.
And yet. A Verizon spectrum license in the C-band — acquired at the 2021 auction for approximately $45 billion — is, in every practical sense, a permanent private franchise on a public resource. It will be renewed. It has been renewed in every band for every carrier in every renewal cycle since the auction system was established. The FCC has revoked licenses for technical violations and regulatory non-compliance, but has never revoked a major commercial carrier's spectrum license for failure to serve the public interest in any substantive sense. The legal framework says revocable. The operational history says permanent.
License terms: Most commercial spectrum licenses are issued for initial terms of ten to fifteen years. Renewal is the norm — not a fresh competitive review, not a reassessment of public interest performance, but an administrative renewal with a presumption of approval. A carrier that has deployed infrastructure under a license has, in practice, a near-absolute right to renewal. The investment creates the permanence that the law does not formally provide.
Renewal presumption: The FCC's license renewal framework creates a presumption in favor of renewal for licensees who have substantially complied with buildout requirements and have not been subject to major enforcement actions. Competing applications for a spectrum license at renewal are rarely filed and rarely succeed. The renewal process is not a competitive reassignment. It is an administrative confirmation of the incumbent's continued franchise.
Secondary market valuations: Spectrum licenses trade in secondary markets at valuations that reflect their permanent franchise value, not their temporary permission value. A spectrum license that could legally be revoked at any time without compensation would not trade for tens of billions of dollars. The market prices these licenses as permanent property because the operational history of the FCC has made them so, regardless of what the statute says.
The Ligado precedent: Ligado Networks' decade-long effort to repurpose L-band satellite spectrum for terrestrial 5G use — ultimately granted by the FCC over DoD objections in 2020, then contested in courts and Congress — illustrates both the difficulty of repurposing licensed spectrum and the property-like expectations that license holders develop. Ligado's spectrum position, which it acquired through the secondary market and inheritance from prior satellite operators, was treated throughout the regulatory proceedings as a property interest that the government was obligated to respect, regardless of the statute's formal position on license revocability.
What the Revenue Architecture Establishes
The auction machine solved the assignment problem and created three new ones. Comparative hearings and lotteries were genuinely worse at assigning spectrum to productive users. The auction mechanism replaced them with a system that produces a price signal, assigns spectrum to willing buyers, and generates federal revenue. It also produces concentration in three national carriers, systematic underinvestment in rural coverage, and a permanent franchise dynamic that has converted the 1927 framework's revocable temporary permissions into permanent private property on public airwaves. The problem it solved was real. The problems it created are also real.
The $200 billion in auction revenue is the price the public paid to give its own spectrum to private carriers. The spectrum was always public. The government did not create value by auctioning it — it transferred value from public ownership to private franchise in exchange for payments to the federal treasury. Whether that transfer was worth the price depends entirely on whether the carriers' use of the spectrum has served the public interest better than alternative governance arrangements would have. The rural coverage record — the dead zones, the subsidy programs layered on top of market failures, the logistics and safety costs of inadequate rural connectivity — is the evidence that the answer to that question is not obviously yes.
The concentration outcome was predictable and has been reinforced by every subsequent auction cycle. The auction mechanism favors large, capitalized bidders. Each auction cycle deepens the advantage of incumbents who already hold spectrum and infrastructure. The secondary market extends concentration rather than correcting it. The result is a wireless industry in which three carriers hold the public spectrum that determines whether Americans can communicate — and in which the public interest standard that is the legal foundation of every license those carriers hold has been effectively redefined as whatever the highest bidder chose to do.
The permanent franchise reality is the auction machine's most consequential output. The legal framework says temporary and revocable. The operational history says permanent and untouchable. That gap — between the statute's public interest framework and the industry's permanent franchise reality — is the governance failure that Post 3 examines in its operational form: the capture of the FCC itself, the protection of incumbent allocations against more productive uses, and the systematic defense of the status quo by the carriers whose $200 billion in auction payments have purchased the political relationships to sustain it.
The Auction Record — What Post 2 Establishes
| Finding | Source | Status |
|---|---|---|
| FCC auction authority granted 1993; first auction 1994 — mechanism replaced comparative hearings and lotteries with competitive bidding for public spectrum licenses | Omnibus Budget Reconciliation Act 1993; FCC auction history | Documented |
| $200 billion+ in gross auction bids since 1994 — including ~$81B C-band auction 2021, ~$45B AWS-3 2015, ~$20B 700 MHz 2008 | FCC auction records; CTIA industry data | Documented |
| Three national carriers — Verizon, AT&T, T-Mobile — hold dominant positions in low-band, mid-band, and the majority of commercially deployed licensed spectrum | FCC license database; carrier financial filings | Documented |
| No new national carrier has entered the U.S. market since the auction system was established — capital barriers to competitive entry are prohibitive | FCC market structure analysis; industry record | Documented |
| Rural coverage failures are systematic — auction mechanism assigns spectrum to highest-revenue markets; subsidy programs (USF, CAF, BEAD) required to address coverage gaps market logic produces | FCC coverage maps; USF/CAF program records; NTIA BEAD documentation | Documented |
| Spectrum licenses trade in secondary markets at valuations reflecting permanent franchise value — legal revocability has not been exercised against any major commercial carrier for substantive public interest failure | FCC renewal records; secondary market transaction data | Structural Finding · Supported |
| The auction machine converted the 1927 public interest licensing framework into a permanent private franchise system — revenue maximization replacing public interest as the operative governance standard | Structural inference from auction record and regulatory history | Structural Finding · Supported |
No comments:
Post a Comment