One year after Fukushima's nightmare, the industry's future is not bright -- for reasons of cost as much as safety.
The enormous power tucked away in the atomic nucleus, the chemist Frederick Soddy rhapsodised in 1908, could "transform a desert continent, thaw the frozen poles, and make the whole world one smiling Garden of Eden."
Militarily, that power has threatened the opposite, with its ability to make deserts out of gardens on an unparalleled scale. Idealists hoped that, in civil garb, nuclear power might redress the balance, providing a cheap, plentiful, reliable and safe source of electricity for centuries to come. But it has not, nor does it soon seem likely to.
Looking at nuclear power 26 years ago, The Economist observed that the way forward for a somewhat moribund nuclear industry was "to get plenty of nuclear plants built, and then to accumulate, year after year, a record of no deaths, no serious accidents -- and no dispute that the result is cheaper energy."
It was a fair assessment; but our conclusion that the industry was "safe as a chocolate factory" proved something of a hostage to fortune. Less than a month later one of the reactors at the Chernobyl plant in Ukraine ran out of control and exploded, killing the workers there at the time and some of those sent in to clean up afterwards, spreading contamination far and wide, leaving a swath of countryside uninhabitable and tens of thousands banished from their homes.
Then, 25 years later, when enough time had passed for some to be talking of a "nuclear renaissance," it happened again. The bureaucrats, politicians and industrialists of what has been called Japan's "nuclear village" were not unaccountable apparatchiks in a decaying authoritarian state like those that bore the guilt of Chernobyl; they had responsibilities to voters, to shareholders, to society. And still they allowed their enthusiasm for nuclear power to shelter weak regulation, safety systems that failed to work and a culpable ignorance of the tectonic risks the reactors faced, all the while blithely promulgating a myth of nuclear safety.
For nuclear to play a greater role, either it must get cheaper or other ways of generating electricity must get more expensive. In theory, the second option looks promising: the damage done to the environment by fossil fuels is currently not paid for. Putting a price on carbon emissions that recognizes the risk to the climate would drive up fossil-fuel costs.
Nuclear power would be more competitive if it were cheaper. Yet despite decades of government research-and-development programs, this does not look likely. Innovation tends to thrive where many designs can compete against each other, where newcomers can get into the game easily, where regulation is light.
Some renewable-energy technologies meet these criteria, and are getting cheaper as a result. But there is no obvious way for nuclear power to do so. Proponents say small, mass-produced reactors would avoid some of the problems of today's behemoths. But for true innovation such reactors would need a large market in which to compete against each other. Such a market does not exist.
This does not mean nuclear power will suddenly go away. Reactors bought today may end up operating into the 22nd century, and decommissioning well-regulated reactors that have been paid for when they have years to run -- as Germany did -- makes little sense.
But nuclear's promise of a global transformation is gone.