SpaceX Is Making Space Cheaper and More Fragile

The strangest thing about SpaceX’s dominance is that the best argument against it begins with admiration. I do not think Elon Musk’s rocket company has captured the space market by accident, subsidy alone, or some lazy monopoly trick. SpaceX has executed. It has flown, landed, reused, repeated, and lowered the psychological barrier to orbit in a way the old aerospace world plainly did not.

That is exactly why I think the United States has a problem.

SpaceX is not just another successful contractor. It is becoming the load-bearing wall of American space policy. On May 15, NASA’s CRS-34 mission launched on a Falcon 9 with a Dragon cargo spacecraft, part of SpaceX’s 34th commercial resupply mission to the International Space Station, according to NASA’s mission page¹. Dragon is SpaceX’s reusable spacecraft family for crew and cargo. Falcon 9 is the reusable rocket that has become the company’s workhorse. Low Earth orbit, or LEO, means the near-Earth region where the space station, many military satellites, and SpaceX’s Starlink internet network operate.

The old question was whether private companies could make spaceflight cheaper and more routine. SpaceX has answered that. The better question now is whether the commercial-space market is becoming too dependent on the one company that answered it first.

Start with the scale. SpaceX flew 165 Falcon 9 orbital missions in 2025, according to Space.com’s year-end launch tally⁴. The same tally said those launches represented about 85 percent of all U.S. orbital launches that year. That is a stunning industrial achievement. It is also the definition of concentration. When one company flies at a cadence that leaves not merely other companies but whole nations behind, the market has gained capacity while losing slack.

This distinction matters because too much of the public discussion treats dominance as either a moral failing or a deserved trophy. I think both frames miss the engineering reality. A bridge can be brilliantly designed and still be dangerous if every truck in the region has to cross it.

NASA’s own watchdog has been unusually clear about this. In its 2025 report on agency management challenges, the NASA Office of Inspector General said the International Space Station faces “lack of redundancy and limited capabilities” for transporting supplies, science, and crew to and from the station. It also said some commercial partners at times rely solely on SpaceX’s Falcon 9, only SpaceX can return cargo from the station, and NASA relies on Crew Dragon for U.S.-based crew transportation while Boeing’s Starliner works toward human-rating certification, according to the NASA OIG report³.

That is not a blogger’s hot take. That is the institutional risk statement. Commercial crew, in NASA-speak, means NASA buys astronaut transportation from private providers rather than owning and operating a shuttle-like system itself. NASA awarded Commercial Crew Transportation Capability contracts to Boeing and SpaceX in 2014 to carry astronauts to the station from the United States, and NASA describes Dragon as an autonomous spacecraft designed to deliver crew and critical cargo to orbiting destinations, according to NASA’s Commercial Crew spacecraft explainer². The model worked. But at the moment, it has worked much better through SpaceX than through Boeing.

The Starliner episode made that visible to everyone outside the space world. NASA’s watchdog said Boeing’s first crewed Starliner test in June 2024 experienced propulsion-system problems that led NASA to bring the astronauts home on SpaceX’s Dragon instead, turning an expected 10-day mission into a roughly nine-and-a-half-month stay aboard the station, according to the same OIG report³. Dragon was the rescue path. That is a compliment to Dragon. It is also a warning about the rest of the system.

The Pentagon has the same tension in a different costume. In April 2025, Space Systems Command awarded National Security Space Launch Phase 3 Lane 2 contracts to SpaceX, United Launch Services, and Blue Origin, according to the Space Force announcement⁵. Lane 2 is the more demanding bucket of national-security launches, the missions where failure tolerance is low. SpaceX was expected to receive 28 missions, about 60 percent of the Phase 3 Lane 2 missions from fiscal years 2025 through 2029; ULA was expected to receive 19; Blue Origin was projected for seven starting in the second order year, according to the same Space Force release⁵.

That is not a monopoly on paper. It is not even bad procurement. The government is buying from the providers that can credibly serve the mission. But resilience is not measured by how many logos appear on a contract vehicle. It is measured by what happens if the lead supplier is grounded, delayed, attacked, distracted, or pulled into a governance crisis.

Blue Origin’s New Glenn is the current stress test. New Glenn is Blue Origin’s heavy-lift reusable rocket, built to compete for commercial, civil, and national-security payloads. On April 19, its third mission achieved a real milestone by reusing and relanding a booster, but the upper stage placed AST SpaceMobile’s BlueBird 7 satellite into an off-nominal orbit that was too low to be useful, according to TechCrunch⁶. The FAA then required a mishap investigation, meaning New Glenn could not return to flight until the company completed the probe and regulators accepted the result, according to The Register’s account of the FAA action⁷.

An FAA launch license is the regulator’s legal authorization for a commercial launch or reentry after safety, environmental, insurance, and other reviews. It is not a ceremonial stamp. After Starship mishaps, the FAA has said SpaceX could not launch again until the agency accepted the final mishap investigation report or made a return-to-flight determination and all other licensing requirements were met, according to the FAA’s general statements on commercial space mishaps⁸. The same regulatory logic now matters for Blue Origin. The irony is sharp: the competitor needed to reduce concentration risk is itself showing how hard it is to build redundancy in rockets.

Starship makes the concentration risk bigger because it moves the dependency from today’s space station to tomorrow’s moon program. Starship is SpaceX’s fully reusable heavy-lift system under test, meant to carry far more mass than Falcon 9 and, in NASA’s Artemis architecture, to serve as the basis for a lunar landing system. SpaceX is targeting May 19, 2026, for Flight 12, the first test flight of Starship V3, according to Space.com⁹. That test is supposed to try dummy Starlink deployment, an in-space engine relight, and splashdowns rather than a booster catch, according to Space.com’s description of the mission⁹.

The point is not that Starship might fail. Test programs fail. The point is that NASA has already built major plans around capabilities Starship has not yet proven at operational scale. The NASA OIG reported in March 2026 that SpaceX’s Starship lander would not be ready for a lunar surface mission by June 2027, that Starship development for Artemis III had been delayed at least two years, and that a large-scale vehicle-to-vehicle cryogenic propellant transfer test had slipped to March 2026, according to the OIG audit of Human Landing System contracts¹⁰. Cryogenic propellant transfer means moving ultra-cold fuel in space from one vehicle to another, a central feature of SpaceX’s lunar architecture. The OIG called it one of the most significant technical challenges facing the provider and said the processes SpaceX plans to use have never been done vehicle-to-vehicle, according to the same audit¹⁰.

NASA is not asleep at the wheel. The OIG also found that NASA’s insight into lander development covers more than 1,100 focus areas across SpaceX and Blue Origin, with deeper insight into high-risk areas such as engine development, cryogenic fluid management, and crew training, according to the same report¹⁰. That weakens the most alarmist version of the critique. SpaceX is not being handed the moon on a handshake.

But oversight is not redundancy. NASA can watch risks closely without having another vehicle ready to take over.

The coming IPO, if it happens, would widen the circle of exposure. An IPO, or initial public offering, is when a private company sells shares to public investors. Reuters reported in December that SpaceX was looking to raise more than $25 billion in a 2026 IPO that could value the company above $1 trillion, driven largely by Starlink expansion and progress in Starship for moon and Mars missions, according to Reuters via Investing.com¹¹. Those details remain reported plans, not final public terms. Still, the logic is clear: public investors would not be buying a simple rocket company. They would be buying an interlocked platform of launch, broadband, government services, crew transport, lunar ambitions, and strategic dependence.

The strongest counterargument is serious: constraining SpaceX would be stupid. I agree. If the government tried to cut SpaceX down before competitors could replace its capacity, NASA would have fewer flights, the Pentagon would have less launch certainty, satellite operators would face longer waits, and American space power would be weaker. SpaceX’s scale has disciplined the old market by showing that reusable rockets can fly often and that orbital access does not need to be treated as a once-a-quarter sacrament.

But “do not punish SpaceX” is not the same as “do not worry.” My position is narrower and, I think, harder to dismiss: SpaceX has strengthened the commercial-space sector while also becoming too essential to fail. The solution is not to kneecap the company. It is to make sure NASA and the Pentagon stop treating future redundancy as if it were present redundancy.

That means paying for boring resilience: certified alternate crew transport, cargo-return capability beyond Dragon, national-security launch manifests that test real surge capacity, range infrastructure that can handle multiple high-cadence providers, and milestone-based support for New Glenn, Vulcan, Rocket Lab’s Neutron, Sierra Space, and others only when they prove they can fly. It also means that if SpaceX goes public, regulators and index funds should treat concentration risk as a disclosure issue, not merely a space-policy footnote.

My threshold is simple. I will stop calling SpaceX too essential to fail when NASA can rotate crew and return cargo from low Earth orbit without Dragon, when the Space Force can absorb a six-month Falcon grounding without major mission reshuffles, and when at least one non-SpaceX heavy-lift provider flies often enough that a delay is an inconvenience rather than a national planning problem. Until then, the indicator to watch is not whether Starship Flight 12 explodes or succeeds on May 19. It is whether anyone else can turn one good launch into a repeatable schedule.