In the 1550s, English dukes had about the same life expectancy as everybody else — and sometimes fared even worse. Back then, there weren't many effective medical treatments or preventive measures for anyone, even aristocrats with money to spend.
If you play with our interactive, based on recent research led by Stanford's Raj Chetty, you can see that times have changed.
The data show American men in the top 1 percent of the income distribution live 15 years longer than men in the bottom 1 percent — a life expectancy comparable to someone from Pakistan or Sudan. The gap between US women at the extremes of the income spectrum was 10 years.
Disturbingly, in the US, Chetty found life expectancy inequality is also accelerating.
As we parsed Chetty's incredible data set further, and combed the literature on life expectancy inequality, we learned that it wasn't always this way.
There's good evidence that richer people used to have about the same life expectancy as poorer people. Moreover, the longevity divide between haves and have-nots is actually more of a gradient than a gap: The higher the income level, the longer the life, along a spectrum.
The gradient has been documented in many settings for many years around the world, and it has as much to do with factors like geography and education level as it does income.
Rich people haven't always lived longer than poor people
Data on the life expectancy of aristocrats in England has fascinated researchers for years. Before the 1700s, dukes and their families had about the same life expectancy as average Brits. So there was no gap in longevity between the rich and the rest of the population.
Then, over the next 100 years, this changed dramatically — and the rich began to pull away from the poor in terms of health.
Angus Deaton, a Princeton economist who recently won the Nobel Prize, points out that a groundswell of public-health measures and medical innovations that came online in the late 18th century meant the rich (but rarely the poor) could access, for example, inoculation against small pox, professional midwives to make childbirth less deadly, and citrus fruits that prevented scurvy.
By 1850 to 1874, the longevity gap had become quite stark: The rich lived 20 years longer than the general population.
This gap looks a lot like America's today, according to Chetty's research. As you can see in this chart on life expectancy for men (and further down, the chart on life expectancy for women) it's not just that the rich live longer than the poor in every county in every state of the nation — it's also that every step up the income ladder confers years of life. So again, there's a gradient linking income and life expectancy.
Researchers have documented this kind of gradient before.
The famous Whitehall studies of the British Civil Service established a link between the relative rank of officer and their risk of disease and death. The higher the rank, the higher the income, the better the health. Studies on Oscar winners found they live four years longer than lowly nominees. A somewhat humorous finding: The size of gravestones has been linked with life spans. The larger the stone, the longer the life.
What's special about the Chetty study is that it illustrates the gradient on a never-before-seen scale and size. He and his colleagues linked 1.4 billion federal tax returns to mortality data from the Social Security Administration from all over the US.
Though these gradients in longevity are persistent and pop up in many contexts, the data on dukes suggests they aren't inevitable.
"The absence of a gradient before 1750 shows that there is no general health benefit from status in and of itself," writes Deaton, "and that power and money are useless against the force of mortality without weapons to fight."
The longevity gradient isn't only explained by income
The cause of the longevity gradient has been the subject of much debate among researchers. And, importantly, researchers know it's a lot more complex than income. Some, like the British epidemiologist Michael Marmot, argue that the gradient emerges because of status at every rung of the social hierarchy: Health worsens as you go down the social ladder. In this view, one's relative status plays a big role in determining health and life expectancy.
Alternatively, Deaton argues for a school of thought called "the fundamental causes," which suggests that not only status, but power, money, and education, will make people healthier only when knowledge is available, as with the dukes in England.
"About 120 years ago, physicians and their kids lived the same amount of time as everybody else," Deaton said. But then, he added, "the germ theory of disease came along, and you got a gradient." This gave the educated classes knowledge about how to avoid disease. So in this view, it's not about status or even income. It doesn't matter that vaccination or cancer-screening programs exist if you can't access them, Deaton said, "especially if access is structured by geography, education, and income."
Deaton and others also argue that health helps determine status and wealth. If you're robust and healthy in body and mind, you can study and work hard, throughout a long life, and accumulate many riches along the way. Poor health in childhood in particular can compromise one's education and earnings throughout life.
Researchers still can't fully explain the gradient
This Chetty data can't answer why the gradient pops up. But as Ezra Klein points out, one of the most fascinating aspects of the study is that it upends a lot of the conventional thinking about longevity inequalities. The researchers found limited evidence to support the notion that access to health care, environmental factors, social cohesion, or local labor market conditions drive the gaps in life expectancy.
Instead, they did find that life expectancy among the poorest individuals was "significantly correlated" with health behaviors like smoking, obesity, and exercise
"Health behaviors are a key driver of differences in outcomes across areas," Chetty told me in an email. So it's not just about having a particular income — it's that being rich or poor tends to be associated with certain behaviors that influence health. Poorer people are more likely to be obese, smoke, and drink compared to their wealthier counterparts.
To Chetty's surprise, the strongest pattern in the data was the role geography played: Low-income individuals lived the longest (and had more healthy behaviors) in cities like New York and San Francisco with populations that are, on average, well educated, and high-income.
In these places, poorer people seemed to behave more like wealthier folks: They smoked less, exercised more, and were less obese. This new data, like Deaton's analysis of ducal England, suggests the longevity gap between different income groups is not inevitable.
According to Chetty, "The question is why health behaviors themselves vary so much among the low-income population — why poor Americans smoke more, eat more poorly, and exercise less in certain pockets of America."
He and his co-authors explain that what might be going on is that health-promoting public policies like anti-smoking legislation, cigarette taxes, and trans fats bans in certain cities pushed people in the direction of healthier choices and trickled down to everyone in the population.
The other explanation, said study author David Cutler, is, "Who you're around affects your health. If you're in New York City or San Francisco, where fewer people are smoking, it's just not as pleasurable to smoke."
Today's health risks are arguably more tightly linked to lifestyle choices. Americans are now living longer than ever, but they're disproportionately dying from chronic, behavior-related diseases like diabetes, heart disease, chronic obstructive pulmonary disease. Researchers have long shown that the behaviors that are linked with these diseases, like smoking, fall along a gradient that's associated with income, but also other factors (as you can see in the chart above, smoking status falls on a gradient linked to education level).
So what does this mean for policy? I asked Deaton that question. He said there's still a lot to learn. "It's like that old saying: Why did the poor behave poorly? The economists say that if people chose to smoke they chose to smoke and they like it that way. Sociologists think people have no choice — society does it."
Chetty's data suggests some of those choices and risks can be mitigated by a mixture of public-health policy and social contagion of healthy behaviors. But, as Deaton pointed out, "Until we know what [the gradient] is, we don’t know whether policy can mitigate it or not."
Since the new research also found that longevity inequality is accelerating in the US, finding out soon is more urgent than ever.