Job-killing effects of robotization may be overblown

April 10, 2021
Robotization hasn't affected the labor market in a huge way in Germany. (AP Photo/Carlos Osorio)

Robotization hasn't affected the labor market in a huge way in Germany. (AP Photo/Carlos Osorio)

Increased industrial robotization has little to no effect on the number of jobs available in the labor market, according to a German study, though researchers say employers may need to invest in retraining their employees as the economy changes.

In the study, published March 22 in the Journal of the European Economic Association, researchers used detailed administrative data from the Institute for Employment Research at the German Federal Employment Agency along with data from the International Federation of Robotics to gauge the adjustment of Germany's local labor markets to industrial robots. 

"This is a really great database on robots from the International Federation of Robotics, and where you can really see how many robots have been installed in each country, in each industry and in each year," Wolfgang Dauth, co-author of the paper and professor at the University of Würzburg, told The Academic Times.

"And so this [the IFB data] really makes the next technological change predictable," he said. "And that's how we got into this, because we saw the opportunity that this kind of research provided to us."

Germany in particular is a good case study for industrial robots, the researchers said, as Germany is more advanced in the use of robotics than other European countries and the U.S. As of 2014, Germany had 7.5 robots per 1,000 workers, outmatching the United States' 1.79 robots per 1,000 workers, according to the study. 

"Germany relies very heavily on robots, but at the same time manufacturing is still very important in Germany," Dauth said. "The share of manufacturing jobs relative to the total number of jobs in Germany is pretty high, even during the last two decades. And so we figured this makes Germany a particularly interesting laboratory because on the one hand there are many jobs that could potentially be replaced. And on the other hand, there are a lot of robots that could potentially replace those jobs."

This research adds to a growing literature investigating the impact of robots on jobs, the composition of the workforce and overall the labor market, along with other studies that show that robots have a small effect on global productivity and exacerbate the gender pay gap.

"In the general public, there's always this big fear that technology in general, and robots in particular, would kill jobs," Dauth said. "When we first started into this project, we kind of also had this intuition that we should see also that robots would lead to a decline in employment."

However, the study's key finding was that while one additional robot introduced to the labor market killed about two manufacturing jobs, this was displaced by the fact that for each additional robot, about two jobs were created in non-manufacturing industries. In sum, the effect of one additional robot increased the number of jobs by .3, but this was not significant. 

"We were pretty surprised, but also excited to see that the net effect of robots on jobs is zero," he said. 

The jobs that sprouted outside manufacturing were mostly concentrated within the business-related service industries such as information technology and business consulting, according to Dauth. But this new job growth did not come from robot-replaced employees leaving their employers.

"A hypothesis would have been that robots make people leave their firms because they replace those people," Dauth said. "But instead, employers retain their workers; they do not fire them, but they upskill them."

The data suggests that firms whose employees were displaced by robots provided training for their workers for different positions, though the data cannot show the exact content of that training, according to the researchers. 

"If you zoom into those firms, then you can see that workers that are exposed to those robots have more stable employment careers," Dauth said. "So robots even increase the probability that workers stay with that initial employer."

He added that employees are often able to improve their occupations to be better paid, more in-demand, and "more intensive on cognitive and interactive tasks" when given the opportunity to retrain.

This retraining of employees, however, may not actually be voluntary for employers, according to the researchers. This pattern of retraining employees appears "stronger in regions where labor unions are stronger," according to Dauth. 

"It's that labor laws and institutions on the labor market seem to push them [firms] towards retaining their workforce," Dauth said. "And so what you find is that technological change is not really bad for the labor market, per se. But it seems to play an important role on how institutions in the labor market support those changes and make adjustments to technological change."

Additionally, the researchers uncovered that younger people are already adapting to these labor market changes, tending to shift away from manufacturing jobs and toward business-related services jobs, as well as attending colleges and universities. 

"Young people who enter the labor market that would otherwise have entered the labor market into a manufacturing job instead now are more likely to enter the labor market and take up a job in the service industry," Dauth said.

The results of this study portray a nuanced picture of the effects of robotization on the labor market and the composition of the workforce. These results also provide important insights for policymakers and the manufacturing firms that are embracing robotization.

"For firms, they can react to technological change by retaining their workforce, by upskilling them and retraining them to do more valuable, more productive jobs," Dauth said. 

"For policymakers, they shouldn't be afraid of technological change," he said. "It does not make sense to try to reduce the pace of technological change by levying attacks on robots, for example. Instead, the change should be accompanied by incentives for firms to provide training to their workers."

The bottom line, Dauth said, is that the general fear about robots destroying jobs "is not really warranted."

"We do see that robots displace certain jobs and displace certain tasks, but on the other hand, they make the remaining tasks more productive," he said. "In the end, it's just a question of whether the workers can do those new and more productive tasks which makes a strong case for lifelong learning by having a strongly educated workforce that can change their skillset and improve their skills even while on the job."

The study, "The Adjustment of Labor Markets to Robots," published March 22 in the Journal of European Economic Association, was co-authored by Wolfgang Dauth, University of Würzburg and the Institute for Employment Research; Sebastian Findeisen, University of Konstanz; and Jens Suedekum and Nicole Woessner, DICE Heinrich-Heine-Universität Düsseldorf.

Saving
We use cookies to improve your experience on our site and to show you relevant advertising.