Reshoring is the economic idea of the moment. The idea is simple. The costs saved by manufacturing goods in China will disappear as Chinese wages rise, leading manufacturing jobs to “reshore” themselves back home to the west. A rise in the renminbi would accelerate this process.

This would be a dream for Barack Obama, a politician from America’s industrial heartland. So too for David Cameron and Nick Clegg in the UK, eager to prove that we are all in it together. Manufacturing jobs have traditionally paid well, and offered good careers to men who did not excel at school. This group has been hard hit by the last 25 years of economic change. It is no surprise politicians love manufacturing jobs.

But reshoring will not happen. For a start, wages will not rise quickly in China, where 34m urban factory workers are paid an average of $2 an hour. A further 65m factory workers in town and village enterprises average just 64 cents. They would be delighted to work for $2. And 675m people are employed elsewhere in China, mainly in agriculture and at lower wages. Chinese wages will rise, but the potential supply of low-cost Chinese labour remains elastic.

If China runs out of workers willing to work for $2, low-cost producers will leave China. They will not, however, return to high-wage economies. Instead they will move to India, Bangladesh and ultimately Africa. This is history repeating itself. At the start of the 20th century Britain lost its textile industry to Japan. As wages rose, those jobs left Japan, but they did not return to Britain. They went first to Hong Kong, then to Korea, and now to China. Simple products will never be produced in developed countries in any quantity again.

Electronics is different. Here Chinese wages are well above $2, because producers want to attract the best and most reliable workers. These companies follow Henry Ford’s strategy of a century ago. He paid workers on the Model-T assembly lines $5 a day – double the prevailing wage. He needed reliable workers, willing to work relentlessly without complaining. High labour productivity allowed Ford to offer high hourly wages and achieve low unit costs. High wages at Ford then, and in Chinese electronic companies today, are a symbol of competitive success, not competitive failure.

The nexus between productivity, wages and competitiveness is critical. Chinese wages can rise without becoming uncompetitive so long as Chinese productivity keeps pace with wages. This has happened in the last 20 years, with productivity rising by 10 per cent a year on average. China is as competitive as it has ever been.

That said, Chinese labour productivity is still very low by western standards. World Bank figures show the average American industrial worker produces as much as 12 Chinese industrial workers. This means that even were reshoring to happen, it would not be a one-for-one job exchange. America would gain only one job for every eight that China loses, even taking into account the many low productivity Chinese companies producing for the domestic market. American workers are simply that effective.

The Chinese electronics sector currently employs 3m people. If the US won back a tenth of Chinese output, China would lose 300,000 jobs, but the US would gain fewer than 40,000 new jobs. America wouldn’t even notice: even if it won every single piece of manufacturing output from China’s towns and cities, unemployment would fall by just 2.75 percentage points.

Western manufacturing plants produce large amounts of output and contribute hugely to the balance of payments. But productivity is so high that they sustain very few jobs. Anyone who has visited a modern factory will be struck by the high output, and relatively few workers. Productivity is high, and still rising faster than in the service sector.

This means that manufacturing employment will continue to fall, in Britain, the US, and even in China, where it has already fallen by more than a fifth since 1996. Output has risen in all three countries, but productivity has risen faster, so that employment has fallen. There is no reason to think the next decade will be any different. Rapid productivity growth in manufacturing means that all countries must ensure that their economies deliver enough service sector jobs to return society to full employment.

The writer is an economic historian at the London School of Economics