Jessai Flores

What do robots have to do with humans dying? A new study from researchers at Yale and the University of Pennsylvania investigated the relationship between automation and the rising mortality rate among working-age adults in the United States.

In a paper published on Feb. 23 in the academic journal Demography, Rourke O’Brien, assistant professor of sociology at Yale, and researchers at the University of Pennsylvania sought to evaluate how automation has affected economic opportunity. This study was part of a broader research program, Opportunity for Health, aimed at examining the impact of economic opportunity on health. They found a causal relationship between increased automation and mortality in American working-age adults, and they concluded that generous social-safety net programs can soften automation’s detriments to community health.

“This work is motivated by the basic idea that health is a type of human capital, like education,” O’Brien said. “When individuals perceive greater economic opportunity, they are more likely to invest in their own health; conversely when individuals feel less hope for the future, they are less likely to invest in their health and more likely to engage in risky health behaviors.”

The rise of automation in manufacturing was a “shock” to economic opportunity in many communities, according to O’Brien. The study indicated that automation led to declines in manufacturing jobs and wages. As industrial robots displaced hundreds of thousands of manufacturing workers on factory floors, working-age adults saw an increase in mortality rates.

The team found that the causal relationship between automation and mortality operated through both “material” pathways and “despair” pathways. “Material” pathways refer to the impact of automation on current employment, wages and access to health care. The “despair” pathways refer to reduced future economic opportunities, categorized by the study as being linked to “deaths of despair.” 

“We have to realize that economic opportunity can impact health not only by impacting material outcomes — income, health insurance, access to health care — but also by shaping aspirations and expectations of the future,” O’Brien said. “It’s not just the workers who were hurt by automation, but the many ‘would be’ workers who now have fewer opportunities for upward mobility and therefore, perhaps, less reason to invest in their own health and well-being.”

The dramatic decline in the number of manufacturing jobs led to a decrease in the total number of manufacturing jobs that provided health insurance coverage. Therefore, the resulting decrease in health care access and utilization, particularly for preventative and diagnostic visits, has been linked to increased mortality for conditions such as cancer and heart disease.

The increase in working-age mortality is driven largely by increased “deaths of despair,” which refers to deaths from suicide, drug overdose and alcohol abuse. Deaths of despair have increased markedly among working-age adults in recent decades. According to O’Brien, these deaths are driving the overall decline in U.S. life expectancy that started even before the COVID-19 pandemic. One of the key arguments in the study is that a decline of manufacturing employment is likely to have a negative impact on perceived economic opportunity and future expectations, especially for residents of the industrial heartland.

“Work is a central part of our identity, an important source of meaning in our lives,” O’Brien said. “Work comes with not just economic rewards — income, health insurance, etc. — but also social status, dignity and respect. The manufacturing jobs lost by automation were relatively good jobs that afforded both economic and social benefits. When these jobs disappear, future prospects dwindle, particularly for persons without a college degree.”

The team found a more dramatic effect of automation on drug overdose mortality in areas with more prescription opioids. The risk that a displaced worker — or young person who feels little hope for the future — becomes addicted to opioids is likely to be greater in areas where the drugs are more readily available, according to O’Brien. He emphasized that declining economic opportunity was not the root cause of the opioid epidemic, but that it exacerbated the crisis.

Alexander Tsai, a researcher at Opportunity for Health, explained that if people find themselves in life circumstances where they do not have the means to accomplish their life goals, this mismatch can give way to stress, mental health problems, substance use disorder and overdose. He suggested that social and economic policy would be needed to reverse the decline in life expectancy.

“By hollowing out employment opportunities and reducing economic security,” co-author Atheendar Venkataramani GRD ’09 said, “automation made it harder for working-age adults, particularly those without a college degree, to access the middle and upper middle-class life that this sector previously afforded. That’s how automation reduced social mobility.”

The study stressed the significance and potential impact of public policy, particularly in regard to social safety net programs, such as Medicaid and unemployment benefits. Social safety net program generosity at the state level moderated the relationship between automation and mortality by blunting the social and economic hit to workers, families and their communities. Automation increased mortality more in states with less generous safety net policies, the researchers found.

Labor market policies including minimum wage rates and “right to work” laws also influenced the effects of automation, particularly in relation to suicide and drug overdose deaths. Overall, states with right to work laws and lower minimum wages had a more pronounced relationship between automation and increased working-age mortality. 

“First we must recognize that economic health and population health are inextricably intertwined,” O’Brien said. “Reversing America’s population health crisis therefore requires investing in the economic well-being of workers and their communities. Yes, that includes safety net programs that provide income support to households, but also investments in education and training. Perhaps most important is to invest in places — for the public and private sectors to work together to target new investment and jobs to communities hardest hit by manufacturing decline.”

The U.S. life expectancy is on average three years shorter than peer high-income countries.