A recent study funded by the European Research Council suggests that the adoption of industrial robots may be associated with an uneasy trade-off for American workers. More robots correlate to fewer workplace injuries but they also correlate to more deaths by drugs and alcohol, and to deteriorating mental health.
The good news is that data from Germany suggest that these trade-offs are not strictly necessary. American workers may yet find a healthy equilibrium with their robot coworkers.
Robots and Workplace Injuries in the U.S.
To analyze the relationship between workplace robots and health in the U.S., economists in the study reviewed workplace injury data from the Occupational Safety and Health Administration (OSHA), mortality and mental health data from the CDC and the National Center for Health Statistics, and data on the adoption of workplace robotics from the International Federation of Robotics (IFR).
Running a regression analysis, they found that regions with more workplace robots recorded lower rates of workplace injury. The researchers explain that these lower rates are due to robots being deployed in industrial settings, where they take heavy loads off human backs. These robots also perform repetitive tasks, for which the sheer volume of reps would eventually lead to human error and injuries.
One glaring limitation of this analysis is that the OSHA regrettably stopped collecting workplace injury data in 2011. The study therefore does not capture trends from the past decade, such as the widespread repetitive stress injuries in Amazon warehouses, where human workers must keep up with robots in the performance of repetitive tasks as these unrelenting robots serve them a constant wall of merchandise. Workers must process an item every few seconds for hours on end to keep up.
That is not to claim that recent deployments of robots have wiped out earlier gains with respect to workplace injuries. Assuming that robots have indeed offered that net benefit, there’s still bad news. Regions with more robots recorded significantly higher rates of death by drugs and alcohol, and workers reported more mentally unhealthy days. (Thankfully, the researchers did not find a correlation between workplace robots and suicide.)
MORE FOR YOU
What’s Going On in Germany?
Fortunately, the connection between robots and deteriorating mental health is not universal. The researchers ran a similar analysis for Germany, a global leader in robot adoption. The German dataset for workplace injuries is current, so the analysis could extend through 2016 — at which time the IFR robotics data ran out.
In Germany, the same basic relationship between robots and injuries held: More robots meant fewer injuries. But the relationship between robots and mental health disappeared. More workplace robots did not correlate to any change in mental health outcomes.
Why is Germany enjoying better outcomes? Following another recent study on the economics of automation in Germany, published in the Journal of the European Economic Association, the researchers explain that robots have displaced manufacturing jobs in Germany but they have generally not displaced manufacturing workers. With fewer manufacturing jobs, younger Germans have found work in other industries. Meanwhile, older Germans have not lost their existing jobs in manufacturing. Instead, they have moved into positions that have not been automated away, or they have simply retired.
Conversely, the researchers argue that in the U.S. automation has produced labor market pressures that have chipped away at workers’ mental health.
Supplementary Explanations
The explanations for the discrepancy offered in the study are plausible but questions remain. Why are young Germans able to move into new jobs while American workers struggle to adapt to automation? There are three tentative supplementary explanations.
First, additional labor union activities may contribute to the different outcomes. Labor unions increase job security, which presumably mitigates the pressures to which these researchers attribute Americans’ failing mental health. Perhaps Germany’s moderately higher levels of labor union participation have removed a stressor from German life that still burdens Americans.
Second, the German education system may be more effective at preparing young people to navigate our rapidly changing economy. For example, the German practice of sorting schoolkids into career tracks may help young people survey the economy and select a suitable path. Their education system may also better prepare working-class kids for careers outside the highly roboticized manufacturing sector.
Third, the U.S. opioid crisis may be a factor in the increased drug overdose deaths. The study shows that U.S. drug and alcohol deaths tend to cluster in regions with more workplace robots. The researchers also included fixed-effect controls that hypothetically could control for variables not explicitly identified in the study, such as regional opioid use.
However, fixed-effect controls only work if the unnamed variables remain constant — “fixed” — throughout the course of the study. The data for this study were collected during years of steadily increasing — not fixed — opioid overdose deaths. So opioid deaths due to the various other causes of the opioid crisis would not be covered well by fixed-effect controls. Before we officially pin these deaths solely on the robots, studies that consider the confounding effects of America’s struggle with opioids are needed.
Of course there’s always room for more research. As it stands, the European Research Council study provides good reasons to be mindful of the trade-offs that automation appears to bring between physical and mental health. It also suggests that we should seek to discover how to foster conditions that avoid this dilemma entirely.
