Rocket launch costs are falling faster than steamship freight did in the 1800s — and a new study of every orbital launch since 1960 finds that every time total cargo volume doubled, the price per kilogram fell by 21.2 per cent.

The strange thing about rockets is that they are starting to behave like an industrial product.

For most of the space age, that was not obvious. Rockets looked like national projects, military assets or scientific one-offs: expensive, bespoke machines built to solve a mission rather than to grind down cost through repetition. A launch was an event. A payload was precious. The price of orbit sat in a category of its own.

A new economics paper argues that this old picture is breaking.

In PNAS Nexus, Alessio Terzi of the University of Cambridge and Francesco Nicoli of the Politecnico di Torino assembled what they describe as the largest standardized dataset of rocket launches to date: 4,405 launches from 1960 to 2025, spanning more than 330 rocket configurations and 16 spacefaring geographical entities.

The headline result is blunt. The average cost of sending a kilogram to orbit fell from $87,023 in 1960 to $3,868 in 2025, in 2024 dollars. That is a reduction of about 96% across six and a half decades.

But the more important result is the curve underneath it. The researchers found that each time cumulative payload sent to orbit doubled, the average cost per kilogram fell by about 21.2%.

That is not just cheaper rocketry. It is a learning curve.

The Steamship Comparison

To understand why the 21.2% number matters, the authors compare space launch with one of the great transport revolutions on Earth: steamship freight in the 1800s.

Steamships did not merely move goods faster. They helped rewire world trade by making long-distance freight cheaper and more reliable. Grain, cotton and manufactured goods could move across oceans at costs that reshaped markets, food systems and empires. The 19th-century steamship was not just a machine. It was an economic accelerator.

The Cambridge summary of the new study says launch costs are now falling faster than steamship freight did during the 19th-century transport revolution. In the paper’s historical comparison, steamship freight costs for goods such as wheat and cotton fell by about 15.5% each time cumulative cargo volume doubled. Space launch, at 21.2% per doubling of cumulative payload, is steeper.

That does not mean rockets are already as economically important as steamships were. The total volume is still tiny by terrestrial freight standards. Space remains difficult, risky and infrastructure-poor. But the comparison is meant to show the slope of change. By that measure, launch technology is getting cheaper faster than one of the transport systems that helped globalize the modern economy.

Terzi put the point sharply in Cambridge’s release: space technology has achieved steeper declines at a far smaller scale than steamships did. That suggests the industry may still have a long way to run before the learning curve flattens.

Wright’s Law Goes To Orbit

The model behind the study is Wright’s Law, an old but powerful idea from manufacturing. In 1936, aerospace engineer Theodore Wright observed that aircraft production costs tended to fall by a predictable fraction each time cumulative production doubled. The more a technology is produced and used, the more experience accumulates, and the cheaper each unit can become.

Wright’s Law is not magic. It does not say costs must fall forever. It captures a pattern: repetition can reveal efficiencies, standardize processes, improve supply chains, reward better designs and shift demand toward cheaper systems.

Terzi and Nicoli adapted that logic to launch. Instead of counting factory units, they counted cumulative payload to orbit. Their result was that a 1% increase in cumulative payload was associated, on average, with a 0.34% fall in cost per kilogram. Mathematically, that implies the 21.2% cost drop for every doubling of total payload.

The study also found that the curve fits the data especially well since 2000. That matters because the modern launch market is not simply repeating the Cold War. It includes commercial satellite demand, standardized launch vehicles, reusable boosters, constellation deployment and a market increasingly shaped by private operators.

In other words, the falling price of launch is not only about better engines. It is about volume.

A Payload Economy

Rockets get cheaper when more payload flies, but more payload flies because rockets get cheaper. That feedback loop is the heart of the story.

The PNAS Nexus paper reports that costs fell quickly during the 1960s and early 1970s, then flattened after Apollo and the phase-out of Saturn V. The decline picked up again from the early 1990s onward, as commercial activity and international cooperation expanded. The authors split the data at the end of the Cold War and found that costs fell much faster after 1989: roughly 44% per payload doubling after the Cold War, compared with 17% before it.

The interpretation is provocative. State competition produced spectacular capability, but not necessarily efficient access to orbit. Commercial demand and private-sector involvement appear to have pushed cost down more aggressively.

That does not make governments irrelevant. NASA programmes, military launch demand, public research and regulatory decisions all helped shape the market. SpaceX’s Falcon 9, for example, emerged from an environment in which NASA deliberately supported commercial cargo capability. But the economic pressure changed. Launch vehicles were no longer only prestige machines. They became tools in a growing payload economy.

Once that happens, volume starts to matter in a different way. A rocket that flies often can spread fixed costs. A booster that returns can carry experience into its next mission. A factory that builds for a steady manifest can improve faster than one building rare custom vehicles. A market that rewards lower price can shift payload toward cheaper providers, pulling the average down.

Why One Launch Price Is Misleading

People often talk about launch cost as if it were a single number: the price of a Falcon 9, the price of an Ariane, the price of Starship if it works. The study avoids that trap by asking what it costs, on average, to send one kilogram to orbit across the entire launch system.

That distinction matters. A rocket can be cheap per launch but expensive per kilogram if it carries little payload. A heavy-lift vehicle can look expensive as a mission but cheap per kilogram if it carries enormous mass. A fleet can also become cheaper not only because each rocket improves, but because more payload shifts toward the rockets that already have lower cost per kilogram.

The authors tried to standardize costs to low Earth orbit in constant 2024 dollars. That does not remove every uncertainty. Launch prices are often opaque, subsidized, bundled or politically shaped. Payload masses and mission profiles vary. Military and scientific launches do not always behave like commercial ones. The paper acknowledges those limits.

Still, the dataset gives economists something space has often lacked: a long, comparable launch-cost curve rather than scattered anecdotes.

That is why the steamship comparison is powerful. It does not require rockets to be like ships in every way. It asks whether access to orbit is now following a recognizable transport-technology pattern: more volume, lower cost, more volume again.

The Future Is Not Automatic

The study’s projections are dramatic. In its central estimate, average launch cost falls to about $1,600 per kilogram by 2030 and about $300 per kilogram by 2040. Cambridge’s release notes an even more specific central 2040 figure of $273 per kilogram, a roughly 93% drop from the 2025 average.

Those numbers depend on the curve continuing. The authors are careful about that. They warn that several forces could slow or distort the fall: orbital debris, geopolitical fragmentation, national launch duplication and market concentration. A cheaper launch market can become less efficient if access to orbit becomes a bottleneck controlled by too few providers or divided into protected national systems.

The risk is not only technical. It is economic and political. A reusable rocket can cut cost, but a monopoly can keep prices higher than costs. A crowded orbital environment can add insurance, tracking and collision-avoidance burdens. A world of rival blocs can choose expensive independence over cheaper shared launch capacity.

So the study does not say space access will inevitably become cheap. It says the experience curve so far is unusually steep, and if the forces that produced it continue, orbit could become far less expensive than it is today.

A Transport Revolution Above The Atmosphere

Steamships changed the world because they changed the cost of distance. Railways did the same across continents. Container shipping did it again in the 20th century. Solar panels changed energy economics because cost fell with deployment faster than many institutions expected.

Rocket launch may now be entering that family of technologies: not because it is cheap in an everyday sense, but because it is getting cheaper as use expands.

That is the quiet significance of the 21.2% learning rate. It turns launch from a heroic event into a scaling problem. Every doubling of orbital cargo does not simply represent more satellites. It represents accumulated experience, larger markets, more standardized operations and pressure on the cost per kilogram.

If that pattern holds, the next space economy will not be unlocked by one launch. It will be unlocked by many launches, repeated often enough that the extraordinary begins to behave like freight.

That is what makes the steamship comparison so striking. The 1800s made oceans cheaper to cross. The 2020s may be doing something similar to the atmosphere.