Property rights, complementarity, and capitalization
The complaints against SpaceX point to fundamental issues in the economics of orbit use
You may have heard SpaceX is launching a lot of satellites. For their second generation of satellites, they filed two mutually-exclusive constellation configurations with the FCC.
Amazon isn’t happy. Amazon’s central claim is that SpaceX is doing a bad thing in filing for two mutually-exclusive constellation configurations. They can only use one of the two, but their choice has implications for other (current and future) orbit users. Other orbit users will have to plan around both configurations until SpaceX or the FCC narrow it down for them, at which point their work on the unused configuration is wasted. This isn’t the first time another prospective constellation operator has taken issue with SpaceX’s plans. The Amazon-SpaceX exchange is ongoing. It’s also not the only complaint against SpaceX — Viasat recently filed (and lost) a lawsuit.
I’ve seen discussions on the Amazon-SpaceX exchange focus on two angles:
The personal Bezos/Musk dimension. Between the personal beefs, the growing societal focus on inequality, and general interest in the new developments in space, there are fun Bezos/Musk angles to write about.
The market competition between Starlink/Kuiper, specifically that Amazon is trying to block SpaceX because Amazon can’t compete. The comment sections of the Ars articles linked above have a lot of talk about market competition. I’ve seen this angle discussed in the Viasat case as well.
I have nothing to say about the personal angles, and I’m sure at least a part of this is about market competition and using the regulatory process to raise a rivals’ costs. (Most of the Ars comments focus on Amazon trying to raise SpaceX’s costs by complaining, but SpaceX’s dual-configuration proposal also uses the regulatory process to raise its rivals’ costs.)
But I think there’s another, underdiscussed, economics angle to this: property rights, complementarity, and capitalization. I think this angle explains part of why other constellation operators are going at SpaceX right now. It points to a fundamental incompatibility between the way space law treats orbital property rights and the incentives orbit users face.
Definitions: “Property rights” here means the right to exclude someone else from using a thing. In my use here, you have a property right to a parcel of land if you can legally prevent me from using it. “Complementarity” between two things means they’re more useful together. Left and right shoes are complements. “Capitalization” refers to the value of one asset (typically non-traded) being reflected in the price of another (typically traded). In the US, the value of living in a good school district is capitalized into the price of houses in that district.
Raising rivals’ costs
Let’s start with the market competition angle, pretty widely discussed in the comment sections of the linked articles. Filing two plans makes perfect sense from SpaceX’s perspective. As far as I know, there’s no actual law saying they can’t file two plans. Even if they’re not thinking about their rivals, they get option value from filing two plans. They might even get feedback from the FCC and others about their constellation configuration, which could help make their design better.
The problem from a social perspective is that it’s basically costless to file a plan, but it costs something to plan around someone else’s plans. SpaceX is the first-mover when it comes to megaconstellations. Being vertically integrated (owning the rockets and the satellites) likely gives them a natural cost advantage over other would-be megaconstellation operators. Adding some uncertainty to their rivals’ plans further cements SpaceX’s advantage. Intentionally or not, SpaceX’s “novel approach” is raising their rivals’ costs. Naturally, said rivals are likely to complain.
It’s also totally rational for Amazon to complain and use the regulatory process this way even if SpaceX wasn’t filing two plans. Lawyers are probably cheaper than rockets and satellites. Resource value capitalization only adds to this market competition incentive.
Property rights and complementarity: staking claims to orbital real estate
Orbit use is kind of like real estate development with more dimensions: you want to put a structure in a place, and location matters.1 But it’s different in one key respect: (in most places) you can explicitly buy and sell land. I think the lack of orbital location trading rights is a key factor behind Amazon and Viasat’s complaints against SpaceX.
Say a developer wanted to plan to build two mutually-exclusive structures on land they owned. As long as they comply with the relevant laws and regulations, it’s their land, they can do what they want with it. Planning two buildings would likely impose some costs on others — folks will likely want to know about parking and traffic impacts and so on — but the developer bears most of the costs. Since they own the land, nobody else is able to use it anyway. The developer is free to lay out mutually-exclusive plans on as many parcels of land as they like; it doesn’t matter to anyone else until they own or try to own the land.
The Outer Space Treaty limits property rights assignment over orbits. At least, that’s how Article II’s “not subject to national appropriation by claim of sovereignty, by means of use or occupation, or by any other means” language is often interpreted. Notionally, this ought to help keep space available for everyone’s use. But practically, it’s come to mean the first entity to use an orbit gets the orbit. If SpaceX starts using the orbits described in configuration 1 (which they say is their primary configuration), the only ways a new entrant could repurpose those orbits are to
coordinate a handover with SpaceX, subject to FCC approvals if they’re serving US markets;
buy those satellites from SpaceX, potentially replacing them with their own satellites (or buy Starlink/SpaceX outright), again getting regulatory approvals as necessary; or
launch to those orbits from outside the US with the intent of serving non-US markets (to avoid the need for FCC approval), accept the risk of collisions, and let international courts sort out who gets to use what.
1 and 2 likely involve the entrant paying SpaceX, so are sort of like buying orbits in a market. 3 is likely the most hostile, and likely to cost the entrant more than 1 and 2 or the “market price” of an orbit.
OST article II says orbits can’t be appropriated. This means the only way to repurpose an orbit is through market transactions that start to look like we’re buying and selling orbits. But because the orbits themselves can’t be traded separately, their value gets capitalized into the satellites that occupy them and the inputs necessary to use the orbits. If satellites were traded, I’d expect to see shoddy satellites in useful orbits selling for high prices. Satellites are only useful for telecom applications with radio spectrum, so if spectrum use rights in particular orbits were traded I’d expect the spectrum prices to reflect the value of the associated orbital slots too.
To me, this looks a bit like issues with land and water in the American West under prior-appropriations water rights. Here water rights are like orbital use rights, while land is like satellites or spectrum use rights. When you can’t sever the water from the land it flows through, the value of the land ends up incorporating the value of the water as well.
This isn’t novel. I can’t own a nice view of a lake, but I can own a house that has that view. When I sell the house its price will incorporate the value of the view. As a result, I might feel incentivized to lobby hard to limit any construction that would block the view. Property rights emerge and get valued in interesting ways.
In a world with explicit orbital property rights Amazon would have less incentive to complain about how many configurations SpaceX is filing for. If SpaceX was really keen on having the option to use either configuration, they could just buy the orbits outright until they decided. Until they did, their plans would be irrelevant to Amazon. Amazon might even want to encourage SpaceX to spend money buying backup orbits “just in case” to tie up their resources.
The FCC awards orbital property rights
Not explicitly of course. Any space lawyer worth their salt will tell you such a thing is ludicrous. “You can’t own orbits, that’s just basic space law”. Nevertheless, complementarity and capitalization mean the FCC or any other spectrum-allocating authority is in or will end up in this position.
You need radio spectrum to use an orbit for telecom: in econspeak, they’re perfect complements in telecom satellite service production. If orbits and spectrum were both were tradable, folks would probably buy and sell them together like left and right shoes. If only spectrum were traded, the value of the orbit will be capitalized into the price of spectrum, and you’d see people selling spectrum use rights with descriptions of the orbits they’re cleared for attached: “2 Mhz Ku band slot for sale, can be used at the following sets of orbital parameters.”
Right now neither is traded. There’s no body you go to for permission to use LEO “slots”. You get permissions to do other stuff necessary to use the slots, like launch rockets or use spectrum. Spectrum is the bottleneck here. You can launch from the US, Kazakhstan, India, French Guinea, wherever, and mostly reach whatever orbit you want. You need specifically the FCC’s permission before your satellite can start broadcasting to customers in the US. If your business model involves reaching US customers, the FCC is important. Part of SpaceX’s filing is about buying a call option on valuable orbital real estate.
Part of Amazon’s complaint against SpaceX is about outbidding a rival for valuable orbital real estate. It doesn’t matter if Amazon never intends to use the orbits SpaceX is describing. There may be other aspects of SpaceX’s use of these orbits Amazon (rationally) objects to, independent of competitive concerns. Think of residential zoning — people get feisty about views or “neighborhood character”. The FCC controls a key input to producing satellite services, one that’s a perfect complement to orbital real estate. As a result, the value of orbital real estate has been capitalized into the value of complaining to the FCC. Lawyers are cheaper than satellite constellations, too.
Orbital property rights are inevitable
None of this is new to folks who’ve been thinking about this, but I think it’s worth highlighting. OST’s prohibition on explicit orbital property rights turns spectrum regulatory agencies into orbital zoning boards. OST’s prohibition on orbital property rights will always turn someone else into an orbital zoning board.
It’s important to note that this is not about treaty interpretation, norms around slot filings, or legal authorities. It’s an implication of production functions and asset pricing. As long as the FCC controls a perfect complement in the production of satellite services and there are no other entities explicitly allocating orbital slots, the FCC will also be an orbital zoning board. In the current status quo, complaining to the FCC is the closest thing to trying to outbid a rival land developer.
This is effectively what’s happened in GEO. To the extent that the ITU controls the relevant spectrum, it’s also become a GEO zoning board.2 The ITU is relatively toothless — it can’t levy fines or seize property, for example — but the ITU is the one to issue GEO satellite spacing and disposal guidelines. GEO is also different from LEO for two reasons. First, the high fixed costs of entry limit the degree of market competition and thus the demand for GEO slots. Second, the risks of collisions and excess debris growth are much lower in GEO due to its unique physics, constraining the planning costs imposed on others by multiple filings. Still, GEO slots are valuable and folks have engaged in plenty of legal shenanigans to secure them. Lawyers are cheaper than launches to GEO.
This isn’t a case against the OST. Prohibiting explicit orbital property rights isn’t necessarily a bad thing. Having severable, tradable rights maximizes allocative efficiency — perhaps not investment efficiency, oh well — but we might (rightly) be concerned about things other than allocative efficiency. Maybe we care about allocative equity and are willing to give up some allocative efficiency in exchange. I have a hard time believing that the FCC is currently balancing allocative efficiency and equity, but we can agree to disagree if we must.
As long as orbits are congestible resources, orbital property rights will always exist or have their value capitalized into some other property right. Rational orbit users will recognize this and contend for the property right they can secure with more vigor than if the two rights were severable, not complements, or both. The fundamental policy questions here are about how rights to use orbits will function, where their value will be capitalized, and whether these are the outcomes we want.
I’m using “orbit” here to mean “a set of Keplerian elements”. It’s harder to think in higher dimensions, but for our purposes an orbit is just a high-dimensional parcel of land.
ITU is also involved in allocating GEO slots. So the analogy is a bit imperfect, because the two assets (spectrum and slot use rights) can be allocated separately. But their complementarity in producing telecom services, combined with GEO being mainly useful for telecom, makes doing separate allocations a bit silly. Anyway the point is that ITU’s real bite comes from its ability to deny/revoke spectrum access.