In late May, the New York Times ran a story by Eric Lipton titled “Elon Musk Dominates Space Launch. Rivals Are Calling Foul.” In response, the antitrust community largely shrugged its shoulders. I went back and give it a read, along with related stories in the Wall Street Journal (“Elon Musk’s SpaceX Now Has a ‘De Facto’ Monopoly on Rocket Launches”), the Washington Post (“SpaceX could finally face competition. It may be too late.”), and CNBC (“SpaceX’s near monopoly on rocket launches is a ‘huge concern,’ Lazard banker warns”). Having reviewed the theories of competitive harm and the publicly available evidence, I conclude that there is a monopolization case worth pursuing here.
Lipton’s piece in the Times contained two noteworthy allegations (emphasis added):
Jim Cantrell worked with Mr. Musk at the founding of SpaceX in 2002. When he started to build his own launch company, Phantom Space, two potential customers told his sales team they could not sign deals because SpaceX inserts provisions in its contracts to discourage customers from using rivals.
Peter Beck, an aerospace engineer from New Zealand, met in 2019 with Mr. Musk to talk about Mr. Beck’s own launch company, called Rocket Lab. Several months later, SpaceX moved to start carrying small payloads at a discounted price that Mr. Beck and other industry executives said was intended to undercut their chances of success.
The first allegation refers to what economists consider an exclusionary contract: You can buy from me only if you commit to not buying from my rival. Other exclusionary provisions include demanding that buyers fulfill a large portion of their needs with the seller or that buyers give the seller a right to match. The second allegation sounds like predation, which requires pricing below a firm’s incremental costs and a likely chance of recoupment. Both are well-recognized restraints of trade that can generate anticompetitive effects under certain conditions, the first of which is when the restraint is employed by a dominant firm.
SpaceX is dominant in space transportation
Firms that are not dominant in a market can engage in exclusionary tactics without fear of exposing themselves to antitrust scrutiny. It is the combination of market power plus an exclusionary restraint that generates anticompetitive effects. Obtaining market shares on a privately held company like SpaceX, is admittedly difficult. But the New York Times story tell us that in 2023, “SpaceX secured $3.1 billion in federal prime contracts, according to the data, nearly as much as the combined amount the federal government committed for space transportation and related services from its nine competitors, from giants like Boeing and Northrop Grumman to startups like Blue Origin.” This statistic implies that, at least as a share of government spending for space transportation, SpaceX commands nearly a 50 percent share. The article also tell us that “SpaceX’s 96 successful orbital launches during 2023 contrast with seven launches to orbit from the U.S. in total last year by all of SpaceX’s competitors,” indicating a share of over 93 percent when measured in terms of launches. In the same story, Musk himself reckons that as of 2023, SpaceX delivered 80 percent of the world’s cargo to space. According to BryceTech, in the fourth quarter of 2023, SpaceX lifted nearly 90 percent of all pounds sent into orbit. Any share in this range (50 to 93 percent) would be consistent with dominance, particularly when combined with evidence of entry barriers.
SpaceX’s market share is protected by entry barriers
By the time SpaceX launched its 63rd mission of 2023, ULA, the next largest U.S. rocket competitor, had completed just two launches. Each rocket launch leads to new data, the same way that each drive by a Tesla owner gave Tesla new information over its electric vehicle rivals. (A similar incumbency advantage owing to learning economies prompted policymakers to endorse subsidizing charging stations and even forcing Tesla to open its stations to EV rivals.) The Washington Post story has a line from the CEO of Firefly Aerospace that supports this effect: “You could see a scenario where one provider has such a lead … that it is literally impossible to catch up on the order where there will be true competition.” Moreover, SpaceX has “deep ties to NASA and the Pentagon, which have awarded it billions of dollars in contracts and elevated it to prime contractor status.”
There are myriad other natural barriers to entry:
- High fixed costs: Lars Hoffman, then senior vice president at Rocket Lab, estimated in 2020 that it cost $100 million to get a rocket to its first launch;
- Long development periods: Development periods of at least three to five years for rockets are common. Blue Origin’s first orbital launch is more than three years behind schedule and now planned for later this year; and
- Strategic launch schedules: According to the Wall Street Journal, “SpaceX’s grip on the launch business means many government agencies and satellite operators must tether their ambitions to the company’s timetables and capabilities.” Per the Washington Post, “SpaceX’s perch atop the industry has allowed it to dictate timelines and prices for satellite launches that favor its launch cadence and schedule, industry officials said.”
In an attempt at journalistic balance, Lipton suggests that competitive entry is picking up despite these natural impediments:
Jeff Bezos’ Blue Origin is close to its first launch for its New Glenn rocket. RocketLab is building what it calls Neutron, and Relativity Space is working on its TerranR, among other new entrants. After years of delays, Boeing is soon expected to start launching NASA astronauts into space on its new Starliner spacecraft.
Lipton ultimately concludes, however, that the ability of the United States to reach orbit in the near term “remains largely dependent on Mr. Musk and his Falcon 9 rocket.” The aforementioned high fixed costs, long development periods, and strategic launch schedules can counter any evidence of initial entry. Even if these natural barriers could be overcome, entrants would still have to hurdle the artificial barriers erected by SpaceX’s two forms of exclusionary conduct.
SpaceX’s ride-sharing program might be predatory
Recall that Mr. Beck of Rocket Lab alleged that SpaceX started carrying small payloads at a discounted price that Rocket Lab could not match. Here’s more on the predation allegation from the New York Times:
[Beck] and other industry executives said they were convinced that SpaceX had set the price for its Transporter service — where small satellite companies can book slots on a Falcon 9 launch — with the explicit goal of undermining the financial plans of emerging competitors. Transporter’s low price — initially $5,000 per kilogram — was below what some industry executives calculated was SpaceX’s basic cost. They concluded that SpaceX could only offer such a low price by subsiding those flights with some of its government contracting revenue.
Beck also asserted that SpaceX was selling flights on its new Bandwagon service, which offers satellite makers launches to orbits that provide them better coverage over key sections of the world, “far below its own costs to undermine its competition.”
To know whether such pricing is in fact predatory, one must estimate the incremental (that is, avoidable) cost for SpaceX’s ride-sharing missions. Adding one payload to a rocket likely imposes no incremental costs for SpaceX. Thus, the test should be performed on a per launch basis.
The best estimate of SpaceX’s marginal costs per launch comes from Musk himself at $15 million under a “best-case” scenario. But that number excludes other avoidable costs, including “the costs to refurbish the first stage rocket booster, and the cost to recover and refurbish fairings.” Musk also claims that, with regard to manufacturing costs, SpaceX incurs “$10 million to manufacture a new upper stage [rocket] and that this stage represents about 20 percent of the cost of developing the rocket.” If SpaceX replaces this upper-stage rocket every mission, then the incremental costs are $25 million.
Turning to the revenue side of the equation, SpaceX’s average incremental revenue per launch has declined to roughly $22.5 million (equal to $300k per payload times the average of 75 payloads per launch). This would not cover the incremental costs estimated above, and to the extent these numbers are accurate, would be predatory. Of course, these estimates are based on publicly available information. An antitrust agency pursuing an investigation would be able to obtain more precise estimates.
I also find the evidence on the likelihood of recoupment to be highly persuasive. The Washington Post story offers this line on ride-sharing: “One example of how SpaceX made it tough on competitors was its move a few years ago to launch smaller satellites in bunches at very low prices in a ‘rideshare program’ that was seen in the industry as a tactic to target smaller launch companies such as Rocket Lab by taking away customers.” The aforementioned evidence of the high fixed costs and long development periods also make recoupment more likely. Finally, the rocket industry is subject to considerable scale economies, so any practice that denies rivals the ability to achieve scale could be seen as exclusionary and consistent with the classic raising-rivals’-cost framework.
SpaceX’s contracts with customers seem to be exclusionary
The second potentially anticompetitive restraint employed by SpaceX is exclusionary provisions in contracts with its customers, comprised largely of government agencies and satellite companies (many of whom compete against Starlink). Here is a little more detail from the New York Times on SpaceX’s contracting:
Mr. Cantrell, whose company Phantom Space has received funding from NASA to help build its new launch vehicle, said his sales team had been told by Sidus Space and a second company that SpaceX had demanded contract provisions intended to limit their ability to hire other launch providers.
Carol Craig, the chief executive of Sidus Space, confirmed in an interview that SpaceX had a “right of first refusal” provision in a deal she had signed for five launches, allowing SpaceX to counter any offers from its competitors.
A right of first refusal, sometimes called a right to match, can foreclose competition to the extent it discourages rivals from making competitive offers to the customer. Why would a rival launch provider bother formulating a costly bid if the incumbent (SpaceX) can end the competition by simply matching the rival’s offer? Economists recognize that such provisions can generate anticompetitive effects when employed by a dominant firm and when the associated “foreclosure share” is economically significant (typically over 30 percent).
The foreclosure share, as the name suggests, is the share of the market that is foreclosed by an exclusionary contract. Consider a market in which a dominant firm supplies 80 percent of the market and half of its customers buy pursuant to a contract that contains the exclusionary provision. In that case, the foreclosure share would be 40 percent (equal to the product of 80 percent market share and 50 percent of customers with the provision). To the extent that most (or all) of SpaceX’s customers have such a provision in their contracts, the foreclosure share should easily clear the 30 percent threshold.
SpaceX could be favoring its own satellite broadband company
Predation and exclusionary contracting fit squarely within antitrust’s orbit (pun intended). Self-preferencing, on the other hand, is harder to police. A classic example is Amazon favoring its own merchandise over that of a rival merchant. SpaceX might be distorting competition in satellite broadband, a vertically related service to rocket launches. That satellite broadband rivals like OneWeb, Kacific, and Echostar rely on SpaceX for launching into space raises natural concerns about preferencing SpaceX’s affiliated satellite broadband company (Starlink). Per the Wall Street Journal story: “’It’s of course a very uncomfortable situation, where you have a supplier that wanted to go down the value chain and start competing with its own customers,’ said Christian Patouraux, chief executive at Kacific, a satellite internet company focused on Asia and the Pacific region. SpaceX launched a satellite for Kacific in 2019.”
Musk insists that SpaceX charges unaffiliated satellite broadband rivals the same as others, but query what SpaceX is charging Starlink (if anything) for launches. Ownership of Starlink also creates a conflict for SpaceX when it comes to scheduling launches for customers: “If Starship doesn’t ramp up as expected, there will likely be a shortage unless SpaceX allocates more of its Falcon fleet for customers instead of Starlink.”
Will the agencies launch a case?
SpaceX’s exclusionary contracts with customers have all the markings of an anticompetitive restraint. While predation cases are rare, SpaceX’s pricing seems oddly low relative to its incremental costs, and the chance of recoupment is high. If an antitrust agency were considering filing a Section 2 complaint against SpaceX, it should push the boundaries by challenging SpaceX’s self-preferencing as well.
Rocket launches are considered a must-have input in the process of transporting satellites, spacecraft, and astronauts in orbit. The launch industry is important to U.S. national security, and the defense agencies should aim to avoid making the government overly dependent on a monopolist, especially a predator. For the foregoing reasons, the antitrust case against SpaceX might soon have liftoff.