THE sun always shines in space, so it was no surprise when Sir Paul McCartney called the crew of Atlantis, the last space shuttle, on Friday and sang Good Day Sunshine to them. Later in the day President Barack Obama called and told the astronauts that their mission “ushers in an exciting new era to push the frontiers of space exploration and human spaceflight”.

The last shuttle mission actually ushers in an era when the only hope of getting into space for the few remaining American astronauts will be to hitch a ride on a Russian or Chinese rocket. Most of them will have to find jobs elsewhere. And however brightly the sun shines, the day when the United States finally gives up on manned space flight is not a good day.

US rockets will still put satellites into orbit. The older ones were built by the military or the National Aeronautics and Space Administration (Nasa); the newer models will be built by private companies that claim they can boost cargo into space at a much cheaper price. But they won’t be able to put a human being in orbit for a very long time, if ever.

This is not to say that the US should have kept the shuttles going indefinitely. They weren’t safe: two of the four original shuttles were lost, with 14 crew, in a total of only 135 trips. They were not cost-effective either: they each flew on average only once a year during their 30 years of service.

Nasa had perfectly sensible plans to replace the shuttles. In 2004, former president George W. Bush approved an ambitious Nasa plan to build a new generation of powerful rockets to deliver people and materials into near-Earth orbit more cheaply, but also to put a permanent manned base on the moon by 2020.

Nasa calculated that the Constellation programme would cost about $8bn a year until 2020 (the US defence budget burns through that much every five days). Maybe the cost would have risen considerably over time, but that’s not such a big deal: creating big, new technology always takes longer and costs more.

When President Obama cancelled the Constellation project in 2010, he talked about doing things in a “smarter way”, and how private enterprise would develop “space taxis” that would put people into orbit more cheaply. In reality, however, he was ending federal government support for manned space flight — though he did promise to invest a little more than a billion dollars a year in those ‘clever’ private companies.

That is not serious money: the US defence budget gets through that much every 12 hours. Lacking federal financial support, the clever companies will concentrate on doing things that make a profit. Putting people into space does not make a profit. Not in the short run, anyway, and the bean-counters are notoriously uninterested in the very long run.

The space entrepreneurs — Virgin Galactic, Northrop Grumman, Interorbital Services, XCOR, Orbital Sciences Corp. and all their rivals — make well-honed pitches about how Nasa was a bloated bureaucracy, and how private enterprise will do the same jobs more cheaply and more safely. Which may be true for launching communications satellites and the like, but is certainly not true for manned space flight and deep space exploration.

When Christopher Columbus had this idea for a new way to reach Asia, he did not talk to some Spanish fishermen about scaling up their voyages (making a profit at each stage) until eventually they would cross the entire ocean. He went to the Spanish court and got state support for his venture. Almost all of the early European voyages of discovery had state backing, because the profits were not going to flow for quite a while.