Robots, and Our Physical and Mental Health

A safe working environment remains an aspirational goal. According to the Bureau of Labor Statistics, there were 5,000 fatal and 2.8 million nonfatal workplace injuries and illnesses in 2019. Robotic automation has reduced the number of workers involved in activities associated with a high risk of injury or death. But robots cannot perceive humans entering their workspace, and humans can be distracted (ask anyone walking in New York among those looking at their phones, not the street), resulting in some ill-fated robot-human interactions. For many workers, the stress of high-risk activities has been replaced by the anxiety associated with the possibility of job loss.

The research aimed to tease out these two strands of health impacts and used OSHA’s data set, a sampling of 80,000 workers in “high-hazard” industries. It reports injury and illness rates at the level of specific establishments, and the data was correlated with deaths from the CDC National Vital Statistics data set.

• An increase of 1.34 robots/1000 workers resulted in 12 fewer injuries for 1000 full-time workers. Most of these injuries involved days away, restricted work, or transfers. There was no identified impact on more severe injuries.
• The increase in robots reduced worker exposure to physically intensive work by about 6%
• Their “back-of-the-envelope” calculation placed the savings at $1.69 billion. For context, those 2019 costs were in the range of $171 billion. The introduction of robots leads to a slight reduction in our total costs for the physical health of our workers.
• An increase of 1.34 robots/1000 workers resulted in 10% more “deaths due to drug or alcohol abuse” and a 15% increase in respondents feeling “his or her mental health was not good.”

“Overall, we interpret these findings as evidence suggesting that the labor market pressure and fears induced by robot penetration  may have detrimental effects on workers’ mental health.”

None of this should be shocking. The tradeoff in robotic efficiency is paid for by job uncertainty – lower income and hours and more unemployment. But the researchers considered another country with even more significant “robotic penetration,” Germany; the results are different here. Using a German dataset similar to OSHA's, the researcher's report

• A 5% reduction in the risk of reporting any disability
• A 4% reduction in exposure to a highly physically intensive task
• No evidence of significant effects of robot exposure on worker’s mental burden

Why the difference? The researchers suggest it was from a simultaneous shift and increase from manufacturing to service jobs. Retraining is often offered as the solution. But the replacement jobs usually pay far less than those previously held and often involve moving to another geography. Ideally, retraining involves work that provides more equivalence to what is being lost. That problem may be compounded by the diminished education of workers in fields with significant robotic automation.

Source: Industrial Robots, Workers' Safety, And Health NBER Working Paper 30180