Robots are killing people, but not the way you think

Our worst science fiction nightmares have come true! Robots are killing people, but not directly, and not in the ways you might think. While it’s true that a small number of people have been killed as a direct result of interaction with a robot, there’s also something subtler and more insidious going on.

Researchers from the Department of Sociology and the Institution for Social and Policy Studies at Yale and the Perelman School of Medicine at the University of Pennsylvania have outlined the ways in which robots are quietly causing an increase in human mortality. Their findings were published in the journal Demography.

Automation has been on the rise, particularly in manufacturing jobs, since the 1980s and we have seen a corresponding decrease in life expectancy in the United States during the same time. At present, life expectancy in the U.S. is a full three years lower than in other high-income countries. This decline is believed to be caused primarily by economic stresses.

Prior studies showed an increase in opioid overdose deaths which correlated with closures of car assembly plants, and this current study found that that the increase in automation between 1993 and 2007 led to an increase in total mortality in people between 45 and 54 years of age.

As automation increases, manufacturing jobs which have traditionally been held by humans, are now being taken up by robotic systems. The loss of economic opportunities, as a result, have driven an increase in drug overdose deaths, suicide, homicide, and cardiovascular death — also known as “deaths of despair” — according to the study.

Estimates suggest that each additional robot per 1,000 works has led to approximately eight additional deaths per 100,000 males and a further four additional deaths per 100,000 females in the 45 to 54 age group, when compared with background data. Moreover, automation could account for 12% of the overall increase in drug overdose mortality across all working age adults during the study period.

The study notes that automation itself is not wholly responsible for increased mortality in the impacted groups, but it is instead a result of the way automation impacts economic opportunities for workers. In instances where increased productivity driven by automation resulted in higher wages, the relationship disappears. But when automation displaces workers, deaths of despair rise.

The results show that a higher number of robots in manufacturing roles doesn’t increase mortality in a vacuum. Instead, it appears to be the result of a combination of factors which include availability of opioid drugs, local minimum wage rates, and the strength of social safety nets. In areas where state safety programs like Medicaid and unemployment insurance are strong, the effects of automation on mortality are mitigated, specifically as it relates to suicide and drug overdose deaths in middle-aged males.

These findings will likely become increasingly important in coming years. The study makes mention of future projections which predict that automation will double or quadruple in the next decade, necessitating innovative public policies to protect vulnerable work-age demographics who are likely to be impacted.

It’s unlikely that automation will decrease in the coming decades. In fact, as technology improves more and more jobs are likely to be subject to automation. This should, by all rights, be a good thing for society, not a cause for despair. The improvement in our collective quality of life that technology should bring can only be realized if we take steps to transition carefully and thoughtfully, ensuring the workers who have carried our economy on their backs in the past aren’t left to carry this new burden alone.

It’s a stark reminder that even when the threat of robots is more mundane than we imagine in our fiction, the solutions are the same. We win not by destroying our enemies — or our synthetic assistants — but by caring for one another.