Industrial robots are depressing wages and increasing unemployment, according to a paper published by the National Bureau of Economic Research, a private, non-profit, non-partisan research organization in America.
Written by MIT economists Daron Acemoglu and Pascual Restrepo, "Robots and Jobs: Evidence from US Labor Markets" appears only days after Treasury Secretary Steve Mnuchin dismissed the possibility of automated systems taking jobs from people, saying, "It's not even on our radar screen."
Similar to the cosmological conundrum about whether the universe will continue expanding indefinitely or collapse upon itself, the impact of automation and AI on human employment is the subject of ongoing debate about whether automated systems will create more jobs than they destroy.
Among technology advocates, there's predictable optimism. Robert D Atkinson, president and founder of the Information Technology and Innovation Foundation, has gone so far as to place a bet through the Long Now Foundation that by June of 2025 the labor force participation rate and unemployment rate, reported by the US Bureau of Labor Statistics, will respectively be above 60 per cent and below 7.5 per cent.
"The 'robots are killing our jobs' proponents miss the fact that automation lowers prices (or raises wages), which in turn spurs increased demand for goods and services, and hence labor," he states in his argument.
If Acemoglu and Restrepo are correct, however, that may not be a wise bet. The researchers analyzed how the increase in industrial robot usage between 1990 and 2007 affected US local labor markets.
These robots are fully autonomous machines that operate without human intervention, doing tasks that at some point in the past were done manually, such as welding, painting, product assembly, moving materials, and packaging.
There are presently somewhere between 1.5 and 1.75 million industrial robots operating around the globe, according to the International Federation of Robotics. The auto industry uses about 39 per cent of such robots, followed by the electronics industry (19 per cent), metal product manufacturing (9 per cent), and the plastics and chemicals industry (9 percent), according to the researchers.
Acemoglu and Restrepo found that in areas exposed to industrial robots, between 1990 and 2007, "both employment and wages decline in a robust and significant manner (compared to other less exposed areas)."
Using the researcher's preferred parameters, "one more robot per thousand workers reduces aggregate employment to population ratio by about 0.34 percentage points." In other words, every new robot added to a given commuting zone reduces employment by 5.6 workers. And the researchers project that the number of industrial robots will reach between 4 and 6 million by 2025.
If the number of industrial robots quadruples by 2025, the researchers expect 0.94 to 1.76 per cent lower employment-to-population ratio and 1.3 to 2.6 per cent lower wage growth between 2015 and 2025.
What needs to be done
Technologies such as artificial intelligence, machine learning, and robotic automation will erase 16 per cent of US jobs by 2025, research consultancy Forrester said [paywalled] last year, while adding to the job count by 9 per cent over the same period.
In an email to The Register, Daron Acemoglu suggested two policy responses: labor market adjustment assistance, which includes retraining programs for people displaced by robots; and overhauling the education system, to teach skills that are more relevant to the times.
Asked about how automation might affect jobs that involve more fluid tasks and responsibilities, Acemoglu said it was unclear how technology will affect work that isn't easily programmed, but suggested it would be unwise to dismiss automation's impact on even complex occupations.
"We don't see similar negative effects from general IT capital (eg, software and hardware)," he said. "So it's difficult to project into the future, say about the effects of artificial intelligence. But given what we know about artificial intelligence, this technology also creates ample room for major dislocations. We have to be prepared."
Acemoglu said that as with past technological innovation, robotics is creating productivity growth. "What it is not doing is to create jobs," he said.
"There might be several reasons for this:
- We are still at the beginning of this process. It took several decades for the new spinning and weaving technologies of the British Industrial Revolution to start creating jobs and higher wages.
- Many other technologies, such as computers, were displacing some workers but also complementing some other workers quite strongly (eg, computers were highly complementary to college graduates, middle managers and engineers). Robots may be less clearly complementary to existing sets of skills available in the workforce.
- We may be facing bottlenecks in other areas that are slowing down the creation of new jobs." ®