ISRO and the economics of cheap space

In 2013 India launched a spacecraft to Mars for about ₹4.5 billion, roughly $74 million at the time. The comparison that travelled around the world was with a Hollywood space film, which had cost more. The comparison is accurate and almost entirely beside the point, because the two numbers are not measuring the same kind of thing. One is a marginal mission budget. The other is a product budget with marketing attached.

What makes the Indian programme interesting is not that it is cheap in some absolute sense. It is that cheapness is produced by four specific organisational choices, each of which is copyable, and each of which has a cost that appears somewhere else in the accounts.

What the mission budget actually covers

A mission budget pays for the spacecraft, its instruments, integration and testing, the operations team for the life of the mission, and a contingency. It does not pay for the rocket. It does not pay for the launch pad, the tracking stations, the deep space antenna, or the salaries of the engineers who spent years on the design before the mission existed. Those are programme costs, funded every year through the Department of Space, whether or not anything flies.

This is not a trick, and it is not unique to India. Every space agency accounts this way, because it is the only way to decide whether a particular mission is worth doing. But it does mean that the headline number measures the marginal cost of one more mission inside a machine that already exists — and the machine is the expensive part.

The rocket had been flying since 1993

The Mars orbiter went up on a PSLV, a rocket whose first successful flight was in 1994, after a 1993 failure. That rocket did not appear for the Mars mission. It was the fourth generation of a lineage that began with SLV-3, which put a small satellite in orbit in 1980 on its second attempt, then ASLV, then PSLV, then the heavier GSLV.

Four choices that produce the cost

Reuse the platform. The Mars orbiter was built on a spacecraft bus derived from an earlier communication satellite, carrying a modest instrument complement rather than a custom-designed observatory. Reusing a bus means reusing its structural analysis, its thermal model and its test fixtures. Every one of those is a fixed cost avoided.

Write the requirement to the mission, not to the ideal. The orbiter was designed for roughly a year of operations around Mars. It was not designed for a decade, and it was not designed with the redundancy a human-rated vehicle needs. That is a legitimate engineering decision for a first attempt, and it removes most of the mass, the power and the cost that redundancy would add.

Pay engineers a government salary. This is the least fashionable factor and the largest. For decades Indian space engineers were paid on the central government scale, which is modest relative to what the same people would earn abroad. The programme got a very high ratio of qualified engineers per rupee of budget. The labour arbitrage is real, and it is temporary.

Test to the requirement, not to the handbook. A programme that intends to fly the same rocket dozens of times builds an expensive qualification regime and amortises it. A programme flying once can lean more heavily on analysis and subsystem tests. This is where the savings are largest and the risk is concentrated: a design that has skipped a full-system test has fewer chances to discover a coupling failure before launch.

What cheap does not buy

The same choices that make a first mission affordable make a launch business difficult. Commercial launch is a schedule business before it is a price business: a customer with a satellite worth several hundred million dollars needs a date, an insurance rating and a track record, not the lowest bid. India launches a few times a year out of a launch site that can only serve certain orbits; the market leaders launch dozens of times a year from sites with more favourable geometry.

The 2017 flight of 104 satellites is often quoted as a capability record. It was a rideshare — one primary satellite and a large number of small foreign ones, deployed in sequence. That is a genuine commercial achievement. It is not evidence of heavy-lift capability, and the distinction matters when the same number is used to argue that the programme is underrated.

Heavy lift, human spaceflight and long-life deep space communication are all expensive in ways that cannot be engineered around with a government salary scale. They require years of repeated testing, large facilities and a supply chain that does not exist until someone pays to build it. That is why the cheap mission and the expensive programme are not in contradiction; they are two different parts of the same organisation.

What the satellites actually did

The stated reason for the programme was never Mars. It was applications, and the applications arrived early. The INSAT series, from the 1980s, bundled telecommunications, television broadcasting and weather observation into a single satellite, because a country of that size could not run three separate systems. Weather satellites turned cyclone warning from a guess into a forecast with a track. The 1999 Odisha super cyclone killed close to ten thousand people; the storms since, watched on the same kind of imagery and followed by organised evacuation, have killed in the tens to low hundreds. That difference is the clearest return the programme has produced, and it is not a technology demonstration.

Remote sensing did the same for land. Crop acreage estimates, groundwater mapping, forestry, fisheries advisory and disaster assessment all run on data from satellites that were built for the purpose and paid for by the same annual budget. This is the part the mission-cost headlines hide: a programme spending a few hundred million dollars a year is also the reason the country does not buy its weather pictures and its crop estimates from someone else.

That is the actual reason, and it is not romantic. Data sovereignty is cheaper than data dependence when you already have the engineers. The Mars mission was the demonstration; the weather satellite is the product.

Why the model is now under pressure

The labour arbitrage is eroding. Indian engineering salaries have risen, private space companies now compete for the same people, and the engineers who built the first generation are retiring. At the same time the private sector has begun doing what the public programme could not: small launchers designed around a commercial cadence rather than a national one.

The useful conclusion is not that India does space cheaply. It is that an organisation can be extremely efficient at the marginal mission while remaining small in the global market, and that both facts come from the same set of decisions. The number that gets quoted is the symptom. The lineage, the requirement and the salary scale are the cause.

This is one of the Modern India entries in About India. The pillar's argument — that each of these systems has a stated reason and an actual one — is on the pillar page.