The growing cognitive dissonance of the European space launch debate

There is a growing cognitive dissonance in the European debate over launch. At World Space Business Week in Paris in September, launch executives increasingly spoke as if the end of SpaceX's Falcon 9 might also mark the end of cheap access to orbit. Firefly Aerospace’s Michael Creech was unusually explicit: “The customer got comfortable with a high reliability and a low price, that frankly did not support the launch market.” He later described the low-price model as a “race to the bottom”.

Others on the panel were less categorical, but the direction was similar. Among them, for instance, Blue Origin’s Jordan Charles argued that, at much higher cadence, the industry would eventually move beyond price-per-kilogram thinking towards a different orbital-logistics model — somewhat ironically, given that recent estimates, including our own work in PNAS Nexus, put New Glenn’s launch cost per kilogram below Falcon 9’s. Part of it is almost a truism: price per kilogram is not the only thing customers care about. But the repeated claim that the ‘race to the bottom’ is over and price will no longer matter is completely out of place, a perfect example of the syndrome typical of European industrial players that, when leapfrogged by global competition, build, in denial, a fictional reality around themselves.

The central point should be clear from the outset: Falcon 9 is not being retired because demand for cheap launches has fallen. SpaceX is replacing it with Starship because it expects the new vehicle to be more efficient. Treating this transition as the return of pre-reusability launch economics is the European mistake.

First of all, the claim does not exist in isolation: it is part of a broader narrative according to which SpaceX’s near monopoly is coming to an end with the retirement of Falcon 9. But it is, unfortunately, highly inconsistent. The narrative, grosso modo, goes as follows: because Falcon 9 is retiring, cheap launches are disappearing; and (supposedly) Falcon 9’s low prices were never actually sustainable (and this is why it is retiring); so customers will launch at our higher prices. This resonates well with a common proposition confidently shared by parts of the European industry: Pierre Lionnet at ASD-Eurospace wrote that “with insufficient demand even a reusable system is not profitable”, while Olivier Lemaître has argued that, despite soaring launch numbers, the “real market” is shrinking. So which is it? Did cheap launch fail because there was too little demand to sustain launch providers? Or is the problem now that there are too few launchers to meet demand? Indeed, while launch-company executives celebrated the supposed ‘end of low prices’, the same panel was full of companies planning substantial increases in cadence, while Charles said current demand signals suggested there was enough demand for multiple providers. Eric Berger has described the situation even more starkly as an industry in “panic” over launch availability; elsewhere he argues that much of the US commercial space sector is now waiting for Starship precisely because it needs more launch capacity at still lower prices. The Wall Street Journal describes a “rocket supply crunch”, with Falcon 9 availability dwindling as satellite-launch demand rises. Amazon has had hundreds of completed satellites sitting on the ground waiting for launch, while continuing to manufacture more. Reuters reports that some commercial customers have been told there is no Falcon 9 availability until 2028 or 2029. It is difficult to square all this with the proposition that cheap launch failed because there was simply too little demand for it. These are not the signs of a market that did not care about cheap, frequent space access: these are signs of a market facing a sudden negative supply shock.

Then there is the empirical question. A core claim of the narrative is that Falcon 9 was never, really, a commercial powerhouse. The Financial Times recently made precisely this argument, noting that only around a quarter of Falcon 9 missions in 2025 served external customers and taking this as evidence that the expected explosion in independent commercial demand for cheap launch never really arrived. Most Falcon 9 capacity instead went to Starlink.

This is fundamentally misleading. Falcon 9 flew 165 missions in 2025. SpaceX’s own filing records 122 Starlink launches, leaving 43 missions for external customers. Forty-three external missions in a year is hardly evidence that cheap launch found no market; it is, in fact, several times more than any other commercial launcher. But the metric is nonsensical in its own right because it relies on a very strange definition of commercial demand. Starlink (122 launches in 2025) is commercial demand. The fact that the satellite operator and launch provider belong to the same company does not make the economic activity disappear. Starlink sells connectivity to households, firms, airlines, ships and governments. SpaceX’s 2026 prospectus reports $11.387 billion in 2025 Connectivity revenue, primarily driven by Starlink, and $4.423 billion in operating income from that segment. For scale, SES and Intelsat together reported €3.512 billion in 2025 like-for-like revenue; Viasat reported $4.520 billion for FY2025; and Eutelsat €1.244 billion for FY2024–25. On a straightforward comparison, Starlink’s connectivity business is therefore already larger in revenue than these major incumbent operators combined. Calling Starlink launches “internal” and then concluding that commercial demand failed is economically misleading: it mistakes vertical integration for an absence of demand.

In fact, Starlink may be the strongest evidence we have that cheap launch did create new demand, both qualitatively and quantitatively. The architecture of Starlink — thousands of relatively short-lived satellites, continuously replenished and upgraded — makes sense precisely because launch became cheap and frequent. Lower launch costs helped make this entirely new satellite business model possible, allowing SpaceX to capture part of the downstream surplus itself. That is vertical integration doing what vertical integration often does: a firm develops an upstream technology, finds that the existing downstream market cannot fully exploit it, and builds its own downstream business around the new capability. One may dislike the resulting market concentration, but it is difficult to describe an $11 billion-plus commercial connectivity business as evidence that demand did not materialise.

There is a final nail in the coffin of the narrative that “cheap launch is over”. Falcon 9 is not being retired because SpaceX has concluded that reusable, high-cadence launch was a mistake. It is being retired because SpaceX is moving to Starship. By all metrics, Starship is designed to be cheaper to manufacture and operate than Falcon 9 ever was; this higher efficiency is why Falcon 9 is being retired. Starship is projected to achieve steep progress in reusability, payload capacity and launch cadence; for instance, where Falcon 9 was partially reusable, Starship should be fully reusable; where Falcon 9 required some refurbishment between relaunching the same stage, the first stage of Starship is designed to land back on its launch mount, undergo a quick inspection, and relaunch shortly after. As with all new technology, whether Starship ultimately achieves its promised cost savings and technological advancements remains to be seen, but most certainly the Falcon 9’s retirement plainly does not represent SpaceX abandoning the low-cost model; quite the opposite, it represents a bet that the same model can be pushed much further.

European launchers may therefore be drawing a dangerous lesson from what could simply be a temporary market window. The withdrawal of Falcon 9 capacity may indeed create scarcity, and prices may temporarily rise. Ariane 6, New Glenn, H3, Eclipse and other launchers may enjoy better margins as a result. For a while, that would be good news for these launch companies, but it would tell us very little about the long-run economics of launch, and it would be a terrible mistake to mix up a short-term bottleneck (if it ever materialises) with a long-term trend. Europe should be wary of confusing temporary pricing power with technological vindication; without reusability, there is little hope that Europe can ever escape its Space Trilemma — the tension between strategic autonomy, budgetary constraints, and privately driven technological catch-up. Falcon 9 may be out of the picture soon, but the economic model it introduced won’t go anywhere, and European space executives should do their best to make use of the little extra time they might be afforded to speed up reusability, rather than resting on their laurels.