Will Robots' Growing Cost-Benefit Trade-Off Make It Harder For Developing Nations To Get Rich?

The financial payback period for a welding robot in China has fallen from 5.3 years in 2010 to 1.7 years in 2015. This may stifle attempts to create a middle class based on rising household income from manufacturing wages in many countries which are on the cusp of significant economic growth. 

The question of whether technology's benefits can be optimized for entire populations is the same for the developing world as it is in the US and Europe - but if that broader distribution can not be achieved, the potential impact could be even more acute. JL

James Vincent reports in The Verge:

In the US 47 percent of jobs are at risk of automation, but in a country like Ethiopia this figure is 85 percent. (The OECD average is 57 percent.) The countries predicted to be hardest hit are in the Asian Pacific region (including Nepal, Cambodia, and Bangladesh), but nations in South America will also be affected; Africa, too.

Robots are usually discussed as a threat to jobs in developed countries, but they could have an even more destructive effect on the developing world. A recent article in The Financial Times describes how automation could destroy one of the tried-and-tested routes of economic growth, in which a country’s labor force moves from farms to factories, creating cheap goods for export. We rightly think of factory work as dangerous, monotonous, and exploitative, but that doesn’t meant it’s not better than toiling in the fields. It means more money for workers and increased urbanization, creating more jobs and an emerging middle class in turn. But robots could change this.

The problem is that if robot labor is cheaper and more reliable than human labor, why bother with the latter? The payback period for industrial robots (the time it takes for their extra costs to be paid off) is falling sharply. For a welding robot to be used in a Chinese factory, for example, the period has fallen from 5.3 years in 2010, to 1.7 years in 2015, say analysts from Citi. By 2017, they say it could be as low as 1.3 years.
the payback period for industrial robots has fallen sharply
And more robots in factories equals fewer jobs for humans. The FT quotes a joint report subtitled The Future Is Not What It Used to Be from Citi and the University of Oxford, which spells out the problem: "Today’s low-income countries will not have the same possibility of achieving rapid growth by shifting workers from farms to higher-paying factory jobs."
You might think fine, so what? Isn’t less factory work in the world a good thing? But, it’s only a good thing if factory work has never been an appealing or likely career choice for you in the first place. And, as is often pointed out, it’s capitalism that has lifted hundreds of millions of people out of poverty, with the help of processes like the farms-to-factories shift. (There’s a whole textbook to be written quibbling this statement, but let’s accept, in an roundabout fashion, that the percentage of world's population living on $1.25 a day has dramatically fallen over the last two centuries, and that capitalism had at least something to do with it.)
Of course, technology has destroyed jobs in the past with only a small effect on net unemployment (jobs are destroyed but new jobs are created), but there’s reason to think things are different this time. Past technology has improved international communication and made trade easier, but robots are obviously being designed to replace humans, not just increase their productivity. As well as automation, there are also new manufacturing processes like 3D printing, which might persuade developed companies to bring factories back to their home nation, say analysts.
"premature deindustrialization"
This robot-led slow-down is being called "premature deindustrialization," and it’s a far greater threat to developing countries. One study in the US found that 47 percent of jobs are at risk of automation, but in a country like Ethiopia, for example, this figure is 85 percent. (The OECD average is 57 percent.) The countries predicted to be hardest hit are in the Asian Pacific region (including Nepal, Cambodia, and Bangladesh), but nations in South America will also be affected; Africa, too.
Of course, this is a complex issue to navigate, and the destruction of industrial jobs isn’t going to happen overnight, and will mean different things to different countries. In China, for example, the government wants to automate more low-paid work, to counterbalance a shrinking working age population and rising pay levels. Trying to draw too many generalizations then, is unhelpful, but it is true that we’ve only just begun to consider the full, global impacts of an automated workforce.