The Orbital Commons

The problem

Since 1994 the United States has auctioned rights to use the radio spectrum over its own territory, raising well over $230 billion for the Treasury. Proposal № 015 argued that this money should build a permanent fund rather than plug a year's deficit, and the case there rested on a simple premise: the spectrum is a public asset, and the public should be paid when a firm is granted exclusive use of it.

Now look up.

Orbital slots and the satellite spectrum that goes with them are allocated by the International Telecommunication Union on a first come, first served basis. A firm files a request, meets the coordination requirements, launches within the deadline, and holds the position. The payment to the public, anywhere in the world, is approximately zero.

This mattered very little when there were a few hundred satellites. It matters enormously now. Active satellites in orbit have passed 10,000, the large majority of them belonging to a single constellation launched in the last seven years. Filings lodged with the ITU for planned constellations, taken together, propose well over a million satellites. Some of those filings come from countries with no space industry whatsoever, lodged on behalf of operators elsewhere, for the same reason ships fly Liberian flags.

Low Earth orbit is finite in exactly the way spectrum is finite: not because you run out of room, but because interference and collision risk rise faster than occupancy. And there are only about 1,800 usable slots in the geostationary belt, a genuinely fixed quantity, currently allocated by queue position.

Then there is the debris. Roughly 40,000 objects larger than ten centimetres are tracked, and the untracked population in the lethal-but-invisible range runs to the millions. Every one of them is a bullet, and the Kessler scenario, in which collisions generate debris that generates further collisions, does not require any new launches to proceed. Nobody is funding remediation because no revenue is attached to the activity that creates the risk.

We are allocating the most valuable new real estate of the century by the rules of a nineteenth-century land rush, and paying for the cleanup out of nothing.

The proposal

Auction American orbital-spectrum licences instead of granting them, and require a posted deorbit bond for every satellite. Bond forfeitures fund debris removal. Auction proceeds fund the endowment.

How it would work

1. Auction the licence, not the orbit. The United States cannot sell outer space and does not claim to. What it grants, and what has real market value, is the right to operate a satellite system serving the American market and using spectrum over American territory. That is an FCC licence, it is already required, and it is currently given away. Auction it on the same statutory authority that governs terrestrial spectrum.

1. A deorbit bond. Every licensed satellite posts a bond at launch, refunded on verified disposal within the FCC's five-year post-mission window. A satellite that is abandoned in orbit forfeits its bond. This converts an externality into a line item on the operator's balance sheet, which is the only place externalities ever get managed.

1. Forfeitures fund removal. Bond forfeitures and a small per-satellite congestion charge capitalise an active debris removal programme, procured competitively. Today there is no money for this at all, which is why there is no industry for it.

1. Auction proceeds to the Fund. Licence receipts go to the American Permanent Fund as principal under the locks of № 023, joining terrestrial spectrum from № 015.

1. Priced congestion, not banned congestion. The per-satellite charge scales with orbital shell density and with the object's collision cross-section, so an operator choosing a crowded shell pays for the crowding it adds. The goal is not fewer satellites. It is that the cost of the last satellite in a shell is borne by the firm that put it there.

1. A small-operator floor. Licences below a size threshold, university cubesats, research missions, and early-stage constellations, pay nominal fees. This is a charge on commercial scale, not on access to space.

The numbers

Terrestrial spectrum auctions have averaged something on the order of $8 billion a year over three decades, extremely lumpy.

Satellite spectrum is worth less per megahertz than prime terrestrial mid-band, because it serves fewer users per unit of bandwidth. But the constellations now being licensed are businesses with subscriber revenues in the billions of dollars a year, and they hold their positions in perpetuity for filing fees.

Assume American orbital-spectrum auctions and congestion charges together settle at $1.5 billion a year once the regime matures. That is a deliberately conservative figure against constellation revenues of that order or larger.

To the Fund: $1.5 billion a year, compounding at 5 percent real to roughly $20 billion of corpus in a decade, paying out about $1 billion a year, permanently. Per citizen, about $3 a year.

Three dollars, and the same concession as the rest of the catalog applies.

The debris half is where the immediate value sits. A deorbit bond of $100,000 per satellite across a licensed population heading toward tens of thousands puts billions of dollars of posted collateral behind disposal compliance, at a cost to a compliant operator of approximately nothing, since compliant operators get it back. That is the elegant part: a bond punishes only the behaviour you are trying to stop, and the firms already deorbiting responsibly pay nothing in the end.

The honest objections

"The Outer Space Treaty forbids national appropriation of outer space. You cannot auction what nobody may own." Legally the strongest objection and the design answers it directly rather than dodging. Article II of the 1967 treaty bars claims of sovereignty over celestial bodies and space itself. It does not bar a state from regulating and charging its own nationals and market participants for licences, which every spacefaring state already does. The United States is not auctioning an orbit; it is auctioning permission to use American spectrum and serve American customers, which is unambiguously within its jurisdiction. The distinction is real, but it is also load-bearing, and a sloppily drafted statute that reads as a property claim in orbit would invite a genuine treaty problem.

"Operators will simply file through another country and serve America anyway." The flag-of-convenience problem, and it is already happening: ITU filings have been lodged by states with no domestic space sector on behalf of foreign operators. The answer is that market access is the choke point rather than the filing. A constellation that wants to sell service to American subscribers needs American market access regardless of whose flag its filing carries, and that access is what gets auctioned. This works precisely as well as America's willingness to enforce it, and no better.

"Auctions favour incumbents. The firm with the most capital wins the sky permanently." A serious concern, and terrestrial spectrum auctions have a mixed record here. Item 6 exempts small and research operators, and licences should be term-limited with renewal rather than perpetual, so a position won in 2028 is not owned in 2078. But it is true that auctions allocate to whoever values a resource most in dollars, and that this correlates with existing size. The alternative currently in force allocates to whoever filed first, which correlates with existing size even more strongly and collects nothing for it.

"This raises the cost of satellite broadband for rural users, who have no alternative." Some of the cost passes through, and that is a real conflict with № 036, which promises rural connectivity. Two mitigations: the congestion charge is small relative to constellation revenue per subscriber, and licence conditions can carry rural service obligations in the way broadcast licences historically carried public-interest ones. It remains a genuine tension inside this catalog and we would rather name it than let a reader find it.

"Debris removal technology does not really exist yet. You are funding a fantasy." Close to fair. Active debris removal has flown demonstration missions and nothing at scale, and the economics have never been tested because no customer exists. That is the actual argument for item 3: there is no industry because there is no revenue, and there is no revenue because nobody has ever been charged for the mess. Creating a funded, competitively procured demand signal is how you find out whether it can be done, and the money comes from forfeited bonds rather than from taxpayers.

Sources

• FCC spectrum auction proceeds since 1994, cumulatively exceeding $230 billion (fcc.gov)
• ITU Radio Regulations, first-come first-served coordination for satellite network filings; geostationary arc capacity (itu.int)
• Active satellite population and tracked debris objects; ESA Space Environment Report and US Space Surveillance Network catalogue (esa.int)
• FCC 2022 order requiring post-mission disposal of low-Earth-orbit satellites within five years
• Donald Kessler and Burton Cour-Palais (1978), on collisional cascading in low Earth orbit
• Treaty on Principles Governing the Activities of States in the Exploration and Use of Outer Space (1967), Article II
• Proposals № 015 (The Spectrum Dividend); № 023 (The Debt Covenant); № 036 (The Rural Compact)