clock menu more-arrow no yes
People attend a candlelight vigil for people who died of an overdose in Staten Island, New York, on August 24, 2017.
Spencer Platt/Getty Images

“Deaths of despair”: The deadly epidemic that predated coronavirus

Mortality rates were on the rise in the US long before Covid-19. Here’s why.

In 2015, life expectancy in the wealthiest country in the world fell for the first time in decades. Then came the nearly unfathomable: Life expectancy in the US fell again in 2016 — and for a third time in a row in 2017.

It is hard to communicate just how disquieting that trend is. Outside of wars or pandemics, life expectancy across the world has been consistently rising for almost a century, and has become a hallmark of highly developed nations. The last time a three-year downturn in life expectancy happened in the US was more than a century ago when the 1918 flu pandemic wiped out anywhere between 17 million and 100 million people worldwide.

What was happening? As Princeton economists Anne Case and Angus Deaton first pointed out in a 2015 paper, working-age white men and women without four-year college degrees were dying of suicide, drug overdoses, and alcohol-related liver disease — what Case and Deaton termed “deaths of despair” — at unprecedented rates. In 2017 alone, there were 158,000 deaths of despair in the US: the equivalent of “three fully loaded Boeing 737 MAX jets falling out of the sky every day for a year.”

That death count looks eerily similar to the highest estimates of Covid-19 deaths in the United States. Deaths of despair, while less visible and less broadly disruptive than the coronavirus, amount to the equivalent of a catastrophic pandemic every single year — one that preys disproportionately on low-income and less-educated people. The difference is that, for this epidemic, there is never going to be a vaccine.

In their new book, Deaths of Despair and the Future of Capitalism, Case and Deaton try to answer a simple but profound question: Why? Why are non-college-educated whites dying of drug overdoses, alcohol poisoning, and suicide at unprecedented rates? And what does this have to do with the peculiar strain of modern American capitalism?

I spoke with Case and Deaton by phone. A lightly edited transcript of our conversation follows.

Roge Karma

The argument in the book is that deaths of despair are not just high and increasing, but they are increasing specifically for whites without a four-year college degree. How do these 158,000 deaths per year break down along educational lines, and what do the trends over time look like?

Angus Deaton

The number of 158,000 was the total number of deaths of despair in 2017; the number in the mid-1990s was closer to 69,000. And the increase in these deaths among non-Hispanic whites has been driven almost entirely by deaths among people without a four-year degree.

The data in the book show that there was no difference between the groups in the early ’90s and that it’s widened quite a lot since then.

Anne Case

When we wrote our first paper on this, we focused on people aged 45 to 54. But we’ve found it’s more helpful to think about this by birth cohort. Relative to the birth cohort of 1945, every year when people are born, the risk of dying from one of these deaths of despair is higher. That’s true for suicide, drug overdoses, and alcoholic liver disease. People born in 1960 are at higher risk than people born in 1950. People born in 1980 are much higher risk than people born in 1960, and so on. So it’s not just a baby boomer problem. Things look like they could get worse before they get better.

Roge Karma

In the book, you talk about some important labor market trends that could be driving up deaths of despair, two of which are wage stagnation and labor force participation.

I think an obvious critique is to point out that, in the years before coronavirus hit, wages and labor force participation rates were rising across the board while deaths of despair were also rising. Thus, the thinking goes, the former could not possibly be driving up the latter. What does that view get wrong?

Anne Case

It’s important to think about the long term. There’s been a long-term decline in wages for men without a four-year degree since 1979. In boom times, wages go up a bit; when hard times come again, they fall a bit. But the overall pattern has been lower and lower wages over time.

There has also been a long-term decline in attachment to the labor market. The employment-to-population ratio has seen a long-term decline for men dating back to the late ’70s. For women, the employment to population ratio rose until about 2000 for those without a college degree but has been in a long-term decline since then.

It also turns out that wage change comes with job change. People lose their jobs during a recession, and then they look around and find another job. But, on average, the jobs they are finding are not as good as the jobs they lost. Then another recession hits and the same thing happens again. So there’s not only a deep decline in the employment-to-population ratio — the people who do make it back to the labor market, on average, are not finding jobs as good as the jobs they lost.

Roge Karma

A lot of times when we discuss what’s happened to the “white working class,” we talk about forces like the 2008 financial crisis, globalization, and automation. But something you point out in the book is that all of these forces were no less powerful in Europe than they were in the US, yet the scale of the trends in both places looks very different. Can you talk a bit about the America-specificness of this deaths of despair crisis and how that helps rule out some of these common culprits?

Angus Deaton

The American peculiarity has been something that’s guided our work and that we’ve worried about from the very beginning. Globalization and automation are experienced in similar ways by other rich countries. Yet in other rich countries, you don’t see deaths of despair on anything like the same scale. So we’ve always been looking at a peculiarly American culprit.

There are several possibilities here. The one we focus on is the health care system — its cost and how it is financed — but it’s probably not the only force at work. It’s also in part because we have such a thin and fragmentary safety net compared with other rich countries.

I don’t think there’s any other country that has had declining wages for the less educated for as long as the US. There’s been some prolonged wage stagnation in Britain, Germany, and other places, but those countries have much stronger safety nets. In Britain, for instance, there’s been a huge widening of earnings at the bottom and earnings at the top, but if you take taxation and benefits into account, that’s been essentially annihilated.

Roge Karma

In addition to economic forces like wages and labor force participation, you also talk about some social forces: decreases in marriage rates, increases in out-of-wedlock births, the deterioration of social capital. How should we think about the relationship between the economic forces we just talked about and these social forces as they pertain to the rise in deaths of despair?

Angus Deaton

One thing we very strongly resist is that short-term destruction of economic opportunities is what drove deaths of despair. We know that’s not true. It has to be this long-term drip of losing opportunities and losing meaning and structure in life.

There are clearly multiple things going on. Sociologists argued for a long time that working wages are the fuel for working-class life: for marriage, for community, and all those other things. Unions were important in there, too, because unions were part of community life. The guy bowling alone on the cover of Bob Putnam’s Bowling Alone is bowling in a union hall. Now, not only is he bowling alone, but there’s no union hall. So you see the fabric of life slowly coming apart.

Social norms have also changed about whether it’s okay to get married or to have children out of wedlock. And I don’t know that we have an explanation for those social norms changing.

Anne Case

One thing to point out, though, is that, yes, social norms have changed, but it is still the case that people with a bachelor’s degree are getting married and they’re married when they have children. So one would think that if people with a bachelor’s degree — people with good jobs — are choosing to lead their lives in those ways, there’s probably a reason for it.

Roge Karma

An alternate explanation of the decaying social fabric in these communities is what we can call the “industriousness thesis” advanced by people like Charles Murray. The claim here is that the social and economic breakdown of white, rural communities is primarily attributable to a deficient culture of laziness and lack of work ethic. What does that thesis get wrong?

Anne Case

Charles Murray said the same thing about the black community back in the 1980s. But we see a lot of parallels to what happened to the black community then and what’s happened to the white working class now. In the 1970s and early ’80s, industry moved out of the inner cities, leaving a lot of men without jobs and making them unmarriageable. Marriage rates fell, out-of-wedlock childbearing went up, and they were hit by a drug epidemic.

Now we see the wheel coming around again, and this time coming for the white working class. I would say that what we really see here is that if you treat people horribly enough for long enough, bad things happen to them.

Angus Deaton

You can also show this through the standard economic tools like supply and demand. If people were getting less industrious and voluntarily pulling out of the labor force, then we would see wages going up, not down. But when you look at these graphs of labor force participation and wages both trending down together, it’s very hard to conclude anything except that it’s the supply of jobs that has gone wrong, and there’s simply less and less work for less skilled people.

Roge Karma

Let’s talk about health care. Health care debates often focus on access to health insurance and the affordability of health care for individual families. Those are both important issues, but something you talk about in the book is the total social cost of our employer-based health care system writ large and the parasitic effect it has on our economy — in ways most of us don’t even see.

Can you talk about that side of our health care system and how that relates to the economic trends we were just discussing?

Anne Case

Health care costs in the US rose from being 5 percent of GDP in 1960 to being over 18 percent of GDP in 2018, about a fifth of the entire economy. We have the most expensive health care system in the world, and it’s not delivering the best health in the world.

Our life expectancy is lower than any of our peer countries. The Swiss, who have the next most expensive health care system, live five years longer than Americans and spend 5 percent less of their GDP than what Americans spend. That adds up to a trillion dollars a year, $8,300 per family, of pure waste. It’s one-and-a-half times what we spent on the military.

That money gets sent up the income distribution to big, expensive hospitals, device manufacturers, Big Pharma, and doctors, who make up the largest single occupation in the top 1 percent of all incomes. So we have this system that’s taking money out of the pockets of regular people and sending it up the distribution to very wealthy people.

Angus Deaton

Only about [half] of the population has health care paid through employers, but the point is that the average family policy in 2019 was $21,000 per year. An individual policy was about $10,000. Those costs are just not tolerable for a firm to pay for a worker who is only worth about $30,000 or $20,000 a year. That’s one reason why so many companies have shed their jobs that are available to less educated workers.

Very few large corporations have their own janitors, food service workers, drivers, security, or call center operators anymore — they contract those jobs out. These are almost like gig jobs now: Many of them are temporary, and there is no promotion possibility. We certainly don’t claim [health care costs] are the only force at work here — globalization and automation are rightly part of the issue. But if firms can avoid paying benefits, which are a huge share of the wage for less-educated people, that’s what they’re going to do.

Local authorities are also in a big jam here. They have to pay for Medicaid and don’t have any control over Medicaid prices. Several state lawmakers told us that one of the places the money for Medicaid comes out of is state education. That means the fees go up in state universities; the number of places goes down. And that contributes to the fact that the number of people with a four-year college degree has been stagnating for some time now.

Roge Karma

A central theme throughout the book is that one of the core dividing axes in American life today in terms of life outcomes is higher education. And it seems to me that the economic impact of coronavirus tends to specifically target those who don’t have a four-year degree — whose jobs often can’t readily be done remotely.

How do you think coronavirus will impact the already stark divide in life outcomes between the college-educated and those without a college degree?

Angus Deaton

I think it will widen income inequality. But that’s been true for the past 50 years: As GDP has increased, most of it has gone to more educated people and less to people with less education. Coronavirus is going to make that worse, but saying that coronavirus will widen the gap is not the same thing as saying those people are going to die from deaths of despair. We think of those deaths as a much slower process.

Coronavirus will be one more step in destroying and undermining the lives of less-educated Americans. But we’ve gone to great pains in the book to say that short-term fluctuations, even quite severe short-term fluctuations, are not what’s causing people to die. I think that’s what’s going to happen here. It will make the long-term problem worse. It will widen income inequality between more- and less-educated people, but it’s not going to cause a mass harvesting of depths of despair among less-educated people.

Anne Case

In today’s news, we talk about what a tragedy it would be if 200,000 people died of Covid-19. Deaths of despair are in the same ballpark. The difference is that we expect [those deaths] will continue after this crisis is over.

Health Care

The Supreme Court’s very unusual new abortion orders, explained

Politics

What Democrats are still fighting over in the budget bill

Politics & Policy

The future of vaccine mandates will likely come down to Brett Kavanaugh and Amy Coney Barrett

View all stories in Politics & Policy

Sign up for the newsletter Sign up for The Weeds

Get our essential policy newsletter delivered Fridays.