Drug overdoses and drug-related diseases have killed nearly 400,000 people in the United States over the past decade—on par with the number of American soldiers who died during World War II. Nearly 400,000 others have committed suicide, while 250,000 died from liver disease and other alcohol-related illnesses. About a fifth of all of these deaths involved prescription painkillers or heroin. Together, these “deaths of despair” have been destructive enough to shorten the life span of a whole demographic group, namely middle-aged white people without college degrees.
Beyond its devastating effect on families and communities, there is emerging evidence that this health crisis also has political repercussions. Consider Coos County, New Hampshire, which occupies the state’s northern tip and has the highest death rate from drugs, alcohol, and suicide in all of New England. It used to be reliably Democratic: Barack Obama won the county by nearly 20 percentage points in 2008 and again in 2012. But this year Donald Trump took it easily, by nearly 10 points.
Coos County is not an isolated case. Across the country, and particularly in the industrial Midwest and New England, Trump was unexpectedly successful (as compared to Mitt Romney) in counties heavily burdened by opioid overdoses and other “deaths of despair.” That’s the conclusion of a new research paper by Shannon Monnat, a sociologist and demographer at Pennsylvania State University. It adds to previous reporting on Trump’s appeal in areas shaken by addictions and white mortality. In New England, according to Monnat’s analysis, Trump actually performed more poorly than Romney in counties experiencing low levels of mortality from drugs, alcohol, and suicide. But in the hardest-hit counties in that region, he did better than Romney by nearly 10 percent, on average. In Midwestern counties with high mortality rates, Trump bettered Romney’s averages by more than 16 percent. (The Economist also found that poor health within a county was a strong indicator of Trump’s over-performance, though that analysis considered a wider range of health metrics.)